Intimate Partner Violence Perpetration, Immigration Status, and Disparities in a Community Health Center-Based Sample of Men

SYNOPSIS

Objective

We examined disparities in male perpetration of intimate partner violence (IPV) based on immigration status.

Methods

From 2005 to 2006, 1,668 men aged 18–35 who were recruited from community health centers anonymously completed an automated, computer-assisted self-interview. Men self-reported their immigrant status (e.g., native-born, <6 years in the U.S. [recent immigrants], or ≥6 years in the U.S. [non-recent immigrants]) and IPV perpetration. We calculated differences in IPV perpetration based on immigrant status. Among immigrant men, we further examined differences in IPV perpetration based on English-speaking ability.

Results

Recent immigrants were less likely to report IPV perpetration than native-born men in the overall sample (adjusted odds ratio [AOR] = 0.60, 95% confidence interval [CI] 0.36, 1.00). However, we observed no differences in IPV perpetration between non-recent immigrants and native-born men (AOR=0.88, 95% CI 0.63, 1.23). Among immigrant men, those who were non-recent immigrants and reported limited English-speaking ability were at the highest risk for IPV perpetration, compared with recent immigrants with high English-speaking ability (AOR=7.48, 95% CI 1.92, 29.08).

Conclusions

Although immigrant men were at a lower risk as a group for IPV perpetration as compared with non-immigrants, this lower likelihood of IPV perpetration was only evident among recent immigrants. Among immigrant men, those who arrived in the U.S. more than six years ago and reported speaking English relatively poorly appeared to be at greatest risk for using violence against partners. Future research should examine the effects of fear of legal sanctions, discrimination, and changes in gender roles to clarify the present findings.

The U.S. Census Bureau reported that in 2005, 12.4% of the U.S. population was born outside of the United States—a 16.3% increase from the year 2000.¹ Among the numerous health risks faced by immigrant populations,^2–5 a growing collection of research indicates that at least some populations of immigrant women may be disproportionately affected by intimate partner violence (IPV). Recent community-based work with diverse groups of immigrant women (e.g., South Asian, Korean, and Latina women) has yielded IPV prevalence rates of 30%–60%, which is far higher than the national mean.^6–9 Unfortunately, immigrant women appear to be at high risk not only for experiencing IPV, but also for experiencing severe forms of IPV.¹⁰ Data from both New York City and Massachusetts indicate that foreign-born women are overrepresented in IPV-related homicides.^11–13

While there is growing recognition that more IPV research is needed with immigrant women to shape culturally appropriate resources,^14–17 it is also critical to simultaneously improve understanding of IPV perpetration within such communities to inform IPV prevention efforts among males. Such efforts are needed to both prevent IPV and mitigate associated deleterious health consequences (e.g., homicide, human immunodeficiency virus/acquired immunodeficiency syndrome, depression, poor child health, and substance abuse).^18–21

Unfortunately, despite documented high rates of IPV and IPV-related homicide among multiple immigrant groups and increasing numbers of foreign-born people in the U.S.,^22,23 very little data on immigrant men exist within the IPV research literature. While recent epidemiologic work has focused on pre-migration exposures and their relationship to immigrant men's IPV perpetration,²² no study has assessed potential differences in IPV perpetration based on being foreign-born vs. native-born men. Such work is necessary to expand the present state of knowledge regarding IPV perpetration and clarify how the context of being an immigrant may relate to the use of such violence against female intimate partners.

Similar to other health behaviors (e.g., smoking and substance use),^23,24 acculturation—the processes by which individuals adopt the attitudes, values, customs, beliefs, and norms of another culture²⁵—also may play an important role in immigrant men's perpetration of IPV. Many scholars posit that the process of acculturation has the potential to positively or negatively affect health behaviors, largely due to the changes and stressors that underlie this process.²⁵ While much debate exists surrounding the measurement sand conceptualization of acculturation,^25,26 the public health field commonly assesses acculturation using proxy measures, such as English-speaking ability and length of residency in the U.S.^26,27 To date, only a few IPV-related studies have examined such factors, and they have primarily focused on court-mandated batterer intervention samples.^28–30 Such work indicates that recent immigrants and men with low English-speaking ability were overrepresented in these groups (Unpublished doctoral dissertation, Kim JY. Conjugal violence in Korean American families. Chicago: University of Chicago; 1993). Research is needed to examine whether such factors relate to risk for IPV perpetration among broader samples of immigrant men.

We sought to inform the current state of knowledge surrounding IPV perpetration and immigrant men by using a community health center (CHC)-based sample of men to examine (1) differences in IPV perpetration based on immigration status and (2) the relationship between acculturative factors (e.g., English-speaking ability and duration in the U.S.) and IPV perpetration among immigrant men.

METHODS

We collected data between January 2005 and December 2006 as part of an anonymous, cross-sectional study investigating risk and protective factors for men's perpetration of IPV and other forms of violence. We conducted the study in collaboration with three urban CHCs located in three neighborhoods in Boston: Roxbury, Jamaica Plain, and Dorchester. The participating CHCs provide primary care to more than 120,000 individuals per year and include a disproportionate racial/ethnic minority population, with 49% of patients identifying as black, 19% as white, and 27% as Hispanic (Unpublished report, Massachusetts Department of Public Health. Boston site-specific reports submitted to the Bureau of Family and Community Health, 2004). The percentages of foreign-born people in each neighborhood at the time of the study were 24% in Roxbury, 25% in Jamaica Plain, and 37% in Dorchester.³¹

All English-, Spanish-, or Portuguese-speaking men presenting to reception at each clinic were screened for age eligibility (i.e., 18–35 years) and recruited into the study by trained research staff who were fluent in these languages. The survey was offered in English, Spanish, and Portuguese, as these were the languages most commonly used by the populations who attended the participating CHCs. Of the 3,430 men approached for the study, 2,229 agreed to participate (65%). Reasons for nonparticipation included lack of time (58%), did not specify (41%), waiting for an appointment (2%), and other (e.g., privacy concerns, 2%). Consenting participants completed a survey using an audio computer-assisted survey instrument (ACASI), a computer-based survey tool in which participants self-administer the survey while question-and-answer choices are read aloud to them over headphones to reduce potential literacy barriers. ACASI has demonstrated increased reporting of sensitive behaviors and has been recommended specifically for research concerning violence perpetration.³² Following completion of the 30-minute survey, participants received a $20 gift card and a list of local community resources for mental and general health, including violence prevention.

A small number of men (3%) were then excluded due to data irregularities (e.g., responding with “not applicable” for all survey items). An additional 443 men reported never having had sexual intercourse (n=416) or did not provide information on this item (n=27), so they could not provide information regarding sexual violence and coercion integral to the IPV assessment. Of the remaining men, 43 did not complete the IPV assessment; thus, the present analytic sample consisted of 1,668 men. All study materials were available in English, Spanish, and Portuguese, and all protocols were approved by the Harvard School of Public Health Human Subjects Committee.

Measures

Demographic factors such as age, race/ethnicity, employment, economic status, and education were measured with single items used in the National Behavioral Risk Factor Surveillance System.³³ We measured immigrant status and length of time in the U.S. with single items developed for the current study (e.g., “Were you born in the United States?” and “How long have you lived in the United States?”). We then created a trichotomous immigrant status variable based on the response to these items: U.S.-born, lived in the U.S. for <6 years (recent immigrant), and lived in the U.S. for ≥6 years (non-recent immigrant).

While individuals from Puerto Rico are considered U.S. citizens, their migration experiences are described as comparable to those of international migrants, including limited English-speaking ability,³⁴ challenges regarding housing and mobility,^35,36 and marginalization.^37,38 Thus, consistent with existing work on immigrant experiences,^39,40 men who indicated being born outside of the U.S. and their place of birth as Puerto Rico were considered immigrants in this study. We measured English-speaking ability via a single item from the Multidimensional Acculturative Stress Inventory,⁴¹ whereby participants were asked to rate their English language competency (not well, OK, very well), and we created a dichotomous variable: low = not well/OK vs. high = very well.

We assessed past year IPV perpetration using the perpetration items from the physical assault, sexual assault, and injury subscales of the Conflict Tactics Scale 2 and sexual coercion items from the Sexual Experiences Survey.⁴² One dichotomous variable was created to assess any IPV perpetration in the past 12 months (i.e., answering “yes” to any of the items on the physical, sexual, or injury IPV assessments or “no” to all items).

Analysis

We generated frequencies for demographics, immigrant status, and IPV perpetration. We used Chi-square analyses to assess bivariate associations between demographics with immigrant status and IPV perpetration (significance at p<0.05). Following the rationale of Miettinen and Cook, we simultaneously entered all demographic covariates considered in this study for adjusted logistic regression, as they have all been proposed or recognized as correlates of IPV perpetration.⁴³

We analyzed the immigrant-only subsample using adjusted logistic regression to obtain estimates of odds ratios (ORs) and 95% confidence intervals (CIs) for the significance of English-speaking ability and duration in the U.S. in relation to IPV perpetration.

RESULTS

Sample demographics: total sample

As shown in Table 1, among the 1,668 men who participated in the study, 73.4% were native-born, 9.4% were recent immigrants, and 17.1% were non-recent immigrants. Immigrants were more likely than native-born men to be Hispanic; non-immigrants were more likely to be white or black.

Table 1.

Bivariate associations between demographics and immigrant status, and past year IPV perpetration (n=1,668)^a among men attending Boston-based community health centers from January 2005 to December 2006

^aNumbers <1,668 are due to missing values.

^bColumn percentages

^cRow percentages

IPV = intimate partner violence

GED = general educational development

Prevalence of past year IPV perpetration and bivariate associations with demographics: total sample

Approximately one-quarter of men (24.8%) reported perpetrating any IPV in the past 12 months (Table 1). Perpetration also differed by age, with men aged 22–26 years being most likely to report perpetration of IPV in the past year (29.8%). Men who were divorced/separated (40.3%) were more likely to report any IPV perpetration in the past 12 months as compared with married men (22.0%), men who were never married (23.5%), or men with other intimate partner status (e.g., cohabiting [17.2%]). Male participants with biological children reported more past year IPV perpetration (32.1%) than those without biological children (18.6%). We observed no significant associations for IPV and income, race/ethnicity, educational attainment, or employment status.

Regression analyses of association between immigrant status and past year IPV perpetration: total sample

As shown in Table 2, prevalence of past year IPV differed based on immigration status. One in six men (16.6%) who were recent immigrants reported perpetrating IPV in the past year. More than one in five men who were non-recent immigrants (23.9%), and more than one in four who were native-born (26.0%) reported such IPV perpetration. Adjusted logistic regression revealed that, compared with native-born men, recent immigrant men were less likely to report IPV perpetration in the past 12 months (adjusted OR [AOR] = 0.60, 95% CI 0.36, 1.00). However, we observed no significant difference in IPV prevalence among non-recent immigrants and non-immigrants (AOR=0.88, 95% CI 0.63, 1.23).

Table 2.

Crude and adjusted logistic regression for associations between immigrant status and past year IPV perpetration among men attending Boston-based community health centers from January 2005 to December 2006: total sample

^aA recent immigrant has lived <6 years in the U.S. A non-recent immigrant has lived ≥6 years in the U.S.

^bAdjusted for age, education, race/ethnicity, occupation, having kids, and marital status

IPV = intimate partner violence

CI = confidence interval

English-speaking ability, duration in the U.S., and IPV perpetration: immigrant men

More than one-quarter (25.8%) of immigrant men who reported low English-speaking ability reported perpetrating IPV in the past year (Table 3). These men were more than twice as likely to report IPV perpetration as compared with immigrant men with high English-speaking ability (AOR=2.67, 95% CI 1.43, 4.97). As observed with the entire sample of men, being in the U.S. for fewer than six years was associated with being protective against past year IPV perpetration (AOR=0.47, 95% CI 0.24, 0.91) relative to having lived for a longer period of time in the U.S.

Table 3.

Adjusted logistic regression for associations between English-speaking ability and past year IPV perpetration among an immigrant subsample of men attending Boston-based community health centers from January 2005 to December 2006

^aA recent immigrant has lived <6 years in the U.S. A non-recent immigrant has lived ≥6 years in the U.S.

^bAdjusted for age, education, race/ethnicity, occupation, having kids, marital status, and country of origin

IPV = intimate partner violence

CI = confidence interval

Post-hoc analysis: stratified analysis of English-speaking ability, past year IPV perpetration, and years in the U.S.: immigrant men

Because the data were not adequately powered to conduct a formal statistical interaction test to examine if the effect of English-speaking ability on past year IPV perpetration varied by duration in the U.S., we stratified data to investigate if such a pattern were present (Table 4). Given the significant risk for past year IPV perpetration observed for immigrant men with lesser English-speaking ability, and the association with a protective effect observed for being in the U.S. less than six years, it was expected that the highest percentage of past year IPV perpetration would be observed among immigrant men who reported low English-speaking ability and being in the U.S. for at least six years, and that the lowest risk for IPV perpetration would be among recent immigrant men with high English-speaking ability. We created a four-level categorical variable: low English-speaking ability and recent immigrant, low English-speaking ability and non-recent immigrant, high English-speaking ability and non-recent immigrant, and high English-speaking ability and recent immigrant (i.e., referent group).

Table 4.

Stratified analysis of any IPV perpetration in past year, English-speaking ability, and duration in the U.S. among an immigrant subsample of men attending Boston-based community health centers from January 2005 to December 2006

^aA recent immigrant has lived <6 years in the U.S. A non-recent immigrant has lived ≥6 years in the U.S.

^bAdjusted for age, education, race/ethnicity, occupation, having kids, marital status, and country of origin

IPV = intimate partner violence

CI = confidence interval

As expected, adjusted logistic regression analyses indicated that non-recent immigrants with low English-speaking ability were at the greatest risk for perpetrating IPV (AOR=7.48, 95% CI 1.92, 29.08) as compared with the referent group. We observed a significantly greater likelihood of IPV perpetration among recent immigrant men with low English-speaking ability (AOR=4.02, 95% CI 1.05, 15.42) as compared with the referent group.

DISCUSSION

This U.S.-based study of native-born and immigrant men attending urban CHCs suggests that recent immigrants were less likely to perpetrate IPV against a female partner in the past year. However, this association with a protective effect does not seem to persist once immigrants have lived in the U.S. for an extended period of time (e.g., six years). The observed association with a protective effect of recent immigration (fewer than six years) on IPV perpetration and the attenuated effect for longer-term immigrants is consistent with the larger body of literature on immigrant health and length of stay. Such studies indicate that with greater time in the U.S., there is a tendency to converge to the health behaviors of native populations.^24,44,45 Similar patterns have also been noted in investigations examining incarceration rates and engagement in acts of general violence,⁴⁶ as well as specific health-risk behaviors including tobacco use, diet,⁴⁷ and substance abuse.²⁴ As with other behaviors studied, it is unclear which factors pertaining to the social context of being an immigrant may contribute to lower IPV perpetration upon initial arrival, followed by increased IPV perpetration with longer time spent in the U.S.

One possible explanation for less reported IPV perpetration among recent immigrants is a greater likelihood of foreign-born people to view IPV as illegal, as documented by a statewide study conducted in 2006.⁴⁸ As immigrants first arrive in the U.S., they may fear breaking laws or engaging in acts that may draw negative attention to themselves or their communities,¹⁴ particularly in the current context of rising anti- immigrant sentiment⁴⁹ and the threat of deportation. What is not known, however, is whether viewing IPV as illegal is actually predictive of lower IPV perpetration. It may also be possible that immigrant men were less likely to disclose perpetration out of fear of legal consequences, despite ACASI and assurance that all responses would remain anonymous. Notably, similar fears have been documented among immigrant women as factors for not reporting IPV victimization to service providers.⁵⁰ Over time, fears associated with legal ramifications and stigma from mainstream society may become less salient, thus yielding increased reporting of IPV perpetration among foreign-born men. Future research must be devoted to examining the effects of fear of legal authorities, length of residence, IPV perpetration, and reporting biases among this population.

Changes in gender roles with increased time in the U.S. may also be associated with IPV perpetration. With greater acculturation of immigrant women, increased risk of IPV victimization from a male immigrant partner has been noted in qualitative work.⁵¹ Furthermore, recent work with Chinese male immigrants has indicated the importance of post-migration power dynamics in influencing attitudes toward IPV.²⁸ Currently, it is not clear whether such a factor would be more salient among non-recent vs. recent immigrants (i.e., whether such effects accumulate over time). More research is needed on these issues.

Beyond legal views and gender roles, experiences with discrimination in the U.S. may also contribute to decreased protective association of recent immigration with IPV perpetration observed in the current study. While discrimination against both recent and non-recent immigrants has been documented,^52,53 the negative effect of perceived discrimination may emerge as stronger among immigrants with longer duration in the U.S.^54,55

One study indicated that immigrants' perceived discrimination was also associated with aggressive behavior;⁵⁶ and recent work with minority men has documented a significant relationship between racial discrimination and IPV perpetration.⁵⁷ Thus, it may be that the observed protective association of being a recent immigrant may erode over time as cumulative experiences with discrimination manifest as stress and violent behavior.⁵⁶ The potential effects of discrimination on IPV perpetration may be particularly important given increased concern regarding hate crimes and mistreatment of immigrants in recent years.^49,53 More work is needed to elucidate potential pathways among discrimination, length of stay, and IPV perpetration.

The complexities of the effects of recent immigration on IPV perpetration were further highlighted as we examined the immigrant-only subsample. When simultaneously considering recency of immigration and English-speaking ability, recent immigration continued to be associated with lower likelihood of IPV perpetration. However, low English-speaking ability yielded an increased likelihood of reporting this behavior—a finding that is consistent with prior work that has documented an overrepresentation of immigrant men with low levels of English-speaking ability in batterer intervention programs. Furthermore, the combination of low English-speaking ability and longer duration in the U.S. appeared to confer the highest risk for IPV perpetration, in comparison with immigrant men with high English-speaking ability and recent arrival in the U.S.

It is not clear why this combination was associated with the greatest risk for IPV perpetration among immigrant men. It may be that despite having resided in the U.S. for longer periods of time, these men may be less integrated with mainstream U.S. society due to segmented assimilation (i.e., residing in predominantly immigrant communities) and, thus, may not have increased their English language competency. Therefore, they may be less able to access benefits and resources that would be more readily obtainable with increased English language acquisition and greater integration, as suggested by studies examining other health behaviors.^58,59 Immigrant men with limited English-speaking skills may also face more discrimination than those with greater fluency in English. Segmented assimilation and residential segregation have been highlighted as factors that may play an important role in influencing the health of immigrant communities. However, more work is needed to examine and clarify these issues, as they may pertain to IPV perpetration.²⁵

Limitations

This study had several limitations in addition to those already discussed. The cross-sectional nature of the study limited our ability to infer causality. Another important limitation was our reliance on a non- probability sample of men drawn from CHCs. Additionally, while the current study's use of three languages made research participation more accessible to immigrants who were not fluent in English, it is important to note that immigrants fluent in languages other than English, Spanish, or Portuguese were not eligible for this study. Furthermore, as no consensus has been reached regarding definitions for recent and non-recent immigrants, it is possible that variations in these definitions may yield different findings.

Another limitation was the lack of information on the specific immigration/visa status of the men in the current sample (e.g., undocumented status). It is possible that undocumented men are less likely to engage in and/or report such illegal behaviors due to fear of authorities. It should also be noted that duration in the U.S. and English-speaking ability may both be proxies for other underlying acculturation and assimilation processes not captured by data in the current study. In addition to consideration of discrimination experiences, access to services, and views of IPV as a legal activity, it is also important to consider other aspects of the immigrant context (e.g., social support, power dynamics with in-laws, and lack of culturally appropriate services) to further explain the observed phenomena.

CONCLUSIONS

This study represents an initial step toward understanding the complexities surrounding immigrant-based differences in IPV perpetration, an issue that is severely understudied in the public health arena. While additional work is needed, preliminary recommendations for practice can be proposed based on these findings. First and foremost, CHCs serving large populations of immigrant men may be an effective and appropriate venue for reaching immigrant men regarding IPV perpetration issues. In addition to activities regarding men's health, CHCs may also consider conducting outreach to immigrant men on IPV issues, particularly as they struggle with issues pertaining to acculturation and assimilation (e.g., English as a second language classes). Additionally, batterer intervention programs must continue to work to ensure that men who speak languages other than English are effectively reached.

Future research should integrate the use of more representative samples and longitudinal designs with qualitative work, to obtain a better understanding of mechanisms surrounding immigration, length of residency, English-speaking ability, and IPV perpetration. More comprehensive measures of acculturation are also needed. Such work can help inform the development of tailored programs to prevent both IPV perpetration among immigrant men and the related negative health and social impacts associated with this grave public health concern.