Sexual risk factors for HIV and violence among Puerto Rican women in New York City.

Sexual risk factors for HIV and violence among Puerto Rican women in New York City. This study investigated intimate partner violence (IPV) andheterosexual risk factors for HIV/AIDS and other sexually transmittedinfections (STIs) among a convenience sample of women of Puerto Ricanheritage living in New York City (NYC). Investigators sought todetermine whether those who were born in the continental United Statesor who preferred English over Spanish experience more IPV and greaterHIV risks than their counterparts who were born in the Commonwealth ofPuerto Rico or preferred to speak Spanish. HETEROSEXUAL RISK AND PUERTO RICAN WOMEN HIV continues to be a threat to Latinas in the United States and isthe fourth leading cause of death for Latinas between the ages of 35 to44 years (Centers for Disease Control and Prevention [CDC], 2008). Thosemost affected by HIV/AIDS are Latinas in their most reproductive years(ages 25 to 44), who account for 70 percent of AIDS cases and represent15 percent of all women living with HIV/AIDS. Heterosexual intercourseis the most common route of HIV transmission for Latinas (CDC, 2008). Puerto Ricans are the second largest group of Hispanics in theUnited States (Oropesa, Landale, & Greif, 2008). Puerto Ricans makeup only 9.6 percent of the Hispanic population in the United States butrepresent 58.0 percent of the AIDS cases among Latinos. Among Latinosubgroups, men and "women born in the Commonwealth of Puerto Ricohave the highest incidence of HIV; the two most dominant transmissionroutes are heterosexual contact (38 percent) and injection drug use (39percent) (CDC, 2009). Injection drug use has been found to be higheramong Puerto Ricans than any other Latino subgroup (Kang, Deren, Andia,Colon, & Robles, 2001; Montoya, Bell, Richard, Carlson, &Trevino, 1999). Among its many Hispanic ethnicities, NYC has one of the largestconcentrations of Puerto Ricans in the continental United States(Oropesa et al., 2008). Although Hispanics represented 27 percent of thepopulation in NYC, they accounted for 29 percent of newly diagnosed HIVcases in 2007 (New York City Department of Health and Mental Hygiene,2009). Of Hispanic HIV cases, 26% were among Latinas (New York CityDepartment of Health and Mental Hygiene, 2009). Frequent cyclicalmigration patterns of people of Puerto Rican heritage between PuertoRico and the continental United States have been well documented;frequent cyclical migration patterns are of concern because they canlead to higher incidence rates of HIV/AIDS in both communities (Ellis,Convay, & Bailey, 1996). Previous research among Puerto Rican women has identified severalcategories of factors that may place them at increased risk for HIV:social norms, strict gender roles, power differentials, and culturalscripts such as machismo and marianismo (Moreno & El-Bassel, 2007;Ortiz-Torres, Serrano-Garcia, & Torres-Burgos, 2000; Rafaelli &Suarez Al-Adam, 1998); structural factors, including socialmarginalization and financial hardships (Moreno, 2007; Rivera, 1994);and beliefs systems, such as having misunderstandings about how HIV/AIDSand STIs are transmitted (Ortiz-Torres et al., 2000). There is also alimited knowledge regarding the health-specific needs of the diversityof Latinas (Amaro & de la Torre, 2002). Although machismo andmarianismo are not unique to Latinos, they are present in some Latinoswith different degrees of intensity. Machismo refers to men's beingoverpowering, dictating certain sexual behaviors such as having multiplesexual partners and being more informed about sex than women. Marianismorefers to being submissive and passive, dictating some sexual behaviorssuch as being ignorant about sex and STIs and letting a male partnerdecide about all sexual matters (Moreno, 2007). IPV AND PUERTO RICAN WOMEN In the United States, Latinos represent 15.8 percent of thepopulation (U.S. Census Bureau, 2010) but account for 34 percent of thecases of IPV (Bureau of Justice Statistics, 2000).The National FamilySurvey found an even higher IPV rate among Hispanics (54 percent)compared with non-Hispanic whites (23 percent) (Strauss & Smith,1990). However, other studies have found that when the effect ofsociodemographic factors (such as income, age, and a family history ofviolence) are taken into consideration, these differences betweenHispanics and non-Hispanic whites are not statistically significant(Aldarondo, Kaufman Kantor, & Jasinski, 2002). One longitudinalstudy suggests that Hispanics are nine times more likely to reportmale-to-female violence than any other type of abuse (Field &Caetano, 2005). Only a few studies have been conducted on the difference in ratesof IPV among Latino subgroups (Aldarondo et al., 2002; Kaufman Kantor& Asdigian, 1997; Kaufman Kantor, Jasinski, & Aldarondo, 1994).In the National Alcohol and Family Violence Survey (NAFVS), KaufamnKantor et al. (1994) found that male-to-female abuse was higher amongsome Latino subgroups, with Puerto Ricans (20.4 percent) having thehighest incidence of abuse, followed by Mexican Americans (17.9 percent)and Cuban Americans (2.5 percent).Acculturation factors have beenimplicated as a possible explanation for differences in IPV among thesame ethnic groups (Caetano, Ramisetty-Mikler, Caetano, & Harris,2007). In Aldarondo et al.'s (2002) analysis of the NAFVS data set,they found that Puerto Rican women reported the highest incidence ofpartner assaults, followed by Mexican Americans.A smaller incidence ofreporting was found among immigrant groups, who might have had lowreporting rates because of their immigration status. Researchers have identified some factors that increase thesusceptibility of Latinos to IPV: immigration (Alderondo et al., 2002),acculturation (Caetano et al., 2007; Jasinski, 1998), socioeconomicdeprivation and stressors (Kaufman Kantor et al., 1994), cultural genderfactors such as machismo and marianismo (Moreno, 2007), and the presenceof alcohol and substance use (Caetano et al., 2007; Neff, Holamon, &Davis Schluter, I995). IPV AND HETEROSEXUAL RISK Some scholars have noted that living in an abusive situation makesit complicated for women to negotiate safe sex (Amaro, 1995; Campbell etal., 2008; Moreno, 2007; Suarez-A1-Adam, Raffaelli, & O'Leary,2000). HIV and IPV share similar risk factors and are beginning to berecognized as intersecting problems (Amaro & Raj, 2000; Campbell etal., 2008; Wyatt et al., 2002). For instance, IPV has been identified asa risk factor for having unprotected sex (Amaro, 1995), a higherincidence of STIs, and having sex with a risky partner (El Bassel etal., 1998). PLACE OF BIRTH, LANGUAGE, HETEROSEXUAL RISK, AND IPV Acculturation is the process by which migrant groups adapt theirbehaviors as they interact with the mainstream culture (Rogler, Cortes,& Magaldy, 1991). Acculturation has been linked to both HIV risk andIPV (Suarez-Al-Adam et al., 2000). Several studies have demonstratedthat acculturation acts as a buffer in the adoption of HIV protectivebehaviors, specifically for traditional Latina women (Marin, Tschann,Gomez, & Kegeles, 1993; Rojas-Guyler, Ellis, & Sanders,2005;Vargas Carmona, Romero, & Burns Loeb, 1999). For example, womenwith lower levels of acculturation have different specific HIV riskfactors (for example, low condom use, less control over sexualrelationships) than women with higher levels of acculturative factors(for example, higher number of sexual partners, riskier partners)(Flaskerud & Uman, 1996; Moreno & El-Bassel, 2007; Rojas-Guyteret al., 2005). However, Latinas who are born in the continental UnitedStates tend to perceive themselves more at-risk than their counterparts(Moreno & El-Bassel, 2007) and to use more condoms (Vargas Carmonaet al., 1999). Acculturation dynamics, such as changing gender roles andexpectations and acculturative stress, have been associated with higherincidence of IPV among Latinas (Harris, Firestone, & Vega, 2005;Jasinski, 1998). Some studies have found that the more traditional theorientation (that is, strong familism and strict gender roles) amongLatina women, the less likely they are to report IPV (Harris et al.,2005). As a construct, acculturation is complex. Birthplace and languagepreference are two core components; socioeconomic status is another; yetno one or two components can reliably serve as a proxy for theconstruct. Therefore, it is difficult to assess how much itsrelationship with IPV and risky sexual behaviors can be attributed tocomponents such as place of birth or language or to structural factorssuch as poverty. For example, low acculturation levels might reducesocioeconomic status and occupational choices, increase discriminationand stress, engender low educational attainment, and limit opportunities(Harris et al., 2005). Another possibility is that women born in thecontinental United States and those who prefer English may experiencemore acculturative stress or adopt less traditional behaviors andattitudes, either of which might contribute to more sexual risk factors. Given that women of Puerto Rican heritage have been identified ashaving high risk for both HIV and IPV, it is important to examine ingreater depth the nature of connections among risk factors. Thisdescriptive study examined the relationship of place of birth andlanguage and experience of IPV and heterosexual risk factors for HIV(STIs, num bet of partners, partner's risk factors, and condom use)in a sample of 1,003 women of Puerto Rican heritage attending anoutpatient clinic in NYC. On the basis of previous research studies thathave found a relationship between HIV risk factors and acculturation, wehypothesized that Puerto Rican women born in the continental UnitedStates and those who preferred English would have more sexual riskfactors than their counterparts. Given the associations betweenheterosexual risks for HIV and IPV, we further hypothesized that theexperience of partner abuse--irrespective of place of birth, language,and socioeconomic characteristics--would be a risk factor for HIV. METHOD Sample Recruitment and Selection Women in this study were being screened for eligibility toparticipate in Project Connect, a four-year randomized clinical trial ofa relationship-based HIV prevention program. The parent study (ProjectConnect) examined the effectiveness of a theory-driven preventionintervention adapted to low-income Latina and African American women andtheir main sexual partners (El-Bassel et al., 2003; Schiff, Witte, &El-Bassel, 2003). The institutional review board from the parentstudy's institution and the medical setting approved protocols forthe present study. This article analyzes data collected during thescreening interviews of the parent study. The women were recruited from a large, urban hospital clinic in alow-income neighborhood in the Bronx, a borough of NYC. In four of thesix areas served by this hospital, about 40 percent of the localresidents live in poverty. About 48 percent are Latino, and 35 percentare African American. Neighborhood areas served by the hospital have oneof NYC's highest rates of HIV/AIDS prevalence, two to three timeshigher than of other parts of the city (New York City Department ofHealth and Mental Hygiene, 2009). Copies of a flier were posted in the outpatient clinics toencourage women to join the study. The flier described Project Connectas an intervention designed to help couples stay healthy and enhancetheir communication with each other. In addition to recruitment throughfliers, two female Project Connect staff--one African American, onePuerto Rican--approached women who came to the clinic and recruitedparticipants during the hours of 9:00 A.M. to 3:00 P.M. The Puerto Ricanrecruiter was bilingual and primarily approached Spanish-speaking women. Women who were interested completed a 15-minute face-to-facescreening interview that took place in a private room at the clinic.This interview was available in either Spanish or English, and the womenparticipating in the study were invited to choose the language to beused for the interview. As an incentive for completing the screeninginterview, participants were given a round-trip subway card, valued at$3.00. An informed consent document, which was available in English andSpanish, was verbally explained to the women. After reviewing thedocument, each woman was asked to sign it and given a copy. Once theconsent form was signed, the screening took place. A third of the womenapproached refused to be screened for the study. Women who were screenedfor the study reflected the population of patients served at thehospital in terms of age distribution, race/ethnicity, income, andemployment status. Of those screened, 1,003 identified themselves asPuerto Rican and were included in this study. Measures Sociodemographic Variables. Puerto Rican ethnicity was establishedby asking respondents who identified as being of Hispanic origin ordescent which group best describes their national origin or ancestry.Only those who chose Puerto Rican were included in this study. Toascertain birthplace, respondents indicated whether they were born inthe continental United States. For those born in the Commonwealth ofPuerto Rico, we also asked the length of time living in the UnitedStates. Language preference refers to the respondent's choice oflanguage for the screening interview (Spanish or English). Othersociodemographic indicators included age, relationship status, years ofeducation, and employment (work for pay in the past month). Categoricalvariables with multiple categories (for example, education) werecollapsed to dichotomous versions after it was ascertained that doing sodid not alter the results. Partner Abuse. Partner abuse was measured using the items from thePhysical Assault and Sexual Coercion scales of the Revised ConflictTactics Scale (Cronbach's [alpha]s = .86 and .87, respectively)(Strauss, Hamby, Boney-McCoy, & Sugarman, 1996). The interviewinstrument included questions about the experience of partner abuse inreference to prevalence, severity, and frequency of partner abuse in thepast six months and lifetime. This instrument has been previously usedwith Latinas in studies using nationally representative samples (KaufmanKantor et al., 1994; Strauss & Smith, 1990). Sexual Risk Factors. Sexual HIV risk factors were measured usingselected items of the Sexual Risk Behavior Questionnaire (SRBQ). TheSRBQ was developed by us and has been used in prior studies (El-Basselet al., 1998; Gilbert, El-Bassel, Schilling, Wada, & Bennet, 2000)with over 2,000 women and men from similar health care settings, such asemergency departments and drug and STI clinics. The selected itemsincluded number of sexual partners in lifetime, number of sexualpartners during the past year, and ever having an STI (such asgonorrhea, syphilis, chlamydia, or herpes). Because of nonnormaldistributions, the numeric variables were dichotomized at the medianinteger. Women with a main male sexual partner were asked whether or not hehad any of the following factors in the past 90 days that might placethe woman at risk for HIV infection: had sex with men or other women,contracted or exhibited symptoms of an STI (for example, pain duringurination, sores on the penis), injected drugs, and had a positive HIVdiagnosis. Interviewers showed a card listing all four risk factors;participants who responded in the affirmative were not queried as towhich specific risk factor their partner exhibited. In addition,participants who answered in the negative were asked whether or not theyworried that their partner might have had any of those risk factors. Additional items for those who had intercourse with a main partnerin the past 90 days were condom use (yes/no) and frequency of condom use(0 = never, 4 = every time) for those who had used one. A three-month(90-day) time period was used on the basis of conceptual and theoreticalarguments indicating that it provides one the most favorable balance ofreliability and validity (Jaccard, McDonald, Wan, Dittus, & Quinlan,2002). Statistical Analyses These analyses included only women who identified as ethnicallyPuerto Rican. We used descriptive statistics (t test and chi-square) tocompare the sociodemographic characteristics, women's experience ofabuse by their main partner, and sexual risk factors of women born onthe mainland United States and in the Commonwealth of Puerto Rico and ofwomen who preferred Spanish or English for the screening questionnaire. We used multiple logistic regressions to measure associations ofsociodemographic characteristics and partner abuse with sexual riskfactors. For each outcome, the model initially included as predictorsbirthplace and language; other sociodemographic characteristics--age,education, employment--and either relationship status (when the outcomepertained to the entire sample) or experience of IPV by aparticipant's main partner (when the outcome pertained only toparticipants with a main partner); and interaction terms for birthplaceand language crossed with each of the other predictors. The final modelfor each outcome dropped nonsignificant interactions but retainedbirthplace and language and each of the other sociodemographicpredictors so as to control for possible confounding, because each wasassociated with birthplace or language or (in most cases) both.Theanalyses were performed using SAS (version 8.02). RESULTS Place of Birth, Language, Sociodemographic Characteristics, and IPV The sample of 1,003 was 50 percent born in the continental UnitedStates and 50 percent born in the Commonwealth of Puerto Rico; 39percent preferred a Spanish interview, whereas 61 percent preferredEnglish. Sociodemographic characteristics of these women and reportedfrequency of experiencing IPV, according to birthplace and languagepreference, are displayed in Table 1. Commonwealth-born women (versus continental U.S.-born women),especially those who preferred Spanish (versus English), were older thantheir counterparts and less likely to have completed high school or beemployed/Women interviewed in English were more Likely to have a mainmale sexual partner than those interviewed in Spanish. More women whopreferred English had been abused by their main partner, but differencesby birthplace were not significant. Birth Place, Language, and Sexual Risk Factors The frequencies of reported sexual risk factors for Puerto Ricanwomen according to birthplace and language preference are displayed inTable 2. Significant differences were found in most of these variables. More of the women who preferred English and more of those born inthe continental United States reported having five or more lifetimesexual partners, two or more sexual partners in the past year, and ahistory of STIs. Also, more women who preferred English and were born onthe U.S, mainland used condoms with their main partner; however, amongthose who used condoms at all, more Spanish speakers used them everytime they had sex. With regard to partner risk factors, more women whopreferred English and more born in the continental United States knew orworried that their main sexual partner had risk factors for HIV.Differences by birthplace were not significant with regard towomen's knowing that their partner had risk factors for HIV,although slightly more Commonwealth-born women worried that theirpartner had such risk factors. Multivariate Logistic Regression The results of logistic regressions of the associations among thetwo measures of place of birth and language, sociodemographiccharacteristics, and IPV and each of the six main sexual risk factorsare displayed in Tables 3 and 4. Multivariate analyses revealed an interaction between birthplaceand age as predictors of multiple partners in the past year and havinghad an STI. Regarding women with more than one partner as compared withwomen with one or fewer partners, among those born on the U.S. mainland,there was no significant age difference, but among those born in theCommonwealth of Puerto Rico (who were, on average, older), women withmore than one partner were younger than women with one or fewerpartners. Regarding women who had an STI as compared with women who hadnever had an STI, among those born in the continental United States,women who had had an STI were older, but among those born in PuertoRico, women who had had an STI were comparable in age to those who hadnot. Among women with a main partner, multivariate analyses revealedthat those who preferred English and those who had experienced IPV weremore likely to report condom use with the main partner in the past 90days and to have knowledge of or worry about partner risk factors. Therewas also an interaction between education level and acculturation aspredictors of condom use. Among women with less than a high schooleducation, those born on the U.S. mainland and those who preferredEnglish were more likely than their counterparts to use a condom, Amongwomen with a high school degree or more, however, there was nodifference in condom use by birthplace, and women who preferred Englishwere far more likely to use a condom. DISCUSSION Acculturation as a construct--and the proxy variables used tomeasure it--has been controversial in the literature because of itscomplexity and imperfection (Hunt, Schneider, & Comer, 2004); Lara,Gamboa, Kahramanian, Morales, & Hayes-Bautista, 2005). In thisstudy, we strove to understand the differences among groups of women ofPuerto Rican heritage (born in the continental United States or in theCommonwealth of Puerto Rico), assuming that Puerto Ricans born on theU.S. mainland tend to be different and might suffer more fromacculturative stress or adopt a less traditional perspective with regardto behaviors and attitudes. This might contribute to greater sexual riskand influence the experience of IPV as well. For virtually all IPV and sexual risk factors, women who were bornin the Commonwealth of Puerto Rico and those who preferred Spanish fortheir interview (that is, those who were less acculturated) had lowerrisk than their counterparts who were born in the continental UnitedStates or preferred English. It is possible that the lower rates of IPVreported by women preferring Spanish can be attributed to social factorssuch as stigma of abuse or norms concerning privacy (Harris et al.,2005). Also, Commonwealth-born women were significantly older thanothers in the sample (revealed in multivariate results as an interactionbetween birthplace and age). A limitation of this study is our inabilityto segregate effects of age from effects of being born in theCommonwealth of Puerto Rico or preferring Spanish. A weaknesses ofexplanations based on acculturation is that acculturation alone does nottake into consideration the complex and differential social andpolitical contexts in which immigrant groups live (Arcia, Skinner,Bailey, & Correa, 2001) and the different levels of women'sempowerment, regardless of language, place of birth, and length of timein the United States. Limitations of the data also prevented us fromassessing other dimensions of acculturation such as stress, adaptation,and attitudes. These issues pose formidable challenges for researchers,and more studies are needed that probe their complexities. Our results show that birth in the continental United States andEnglish language preference appear to be harbingers of greater risk forIPV, risky sexual practices, and risky partners. Although the greatersexual risk may be somewhat controlled by condom use, only inconsistentcondom use was reported by the women who preferred English. The findingthat among women who used condoms at all with their main partner,Spanish-speaking women were more likely to use condoms consistentlycontradicts most previous research. However, because of the small numberof participants in this particular analysis, generalization is notwarranted unless the finding is replicated. The present study has several weaknesses regarding samplingstrategy. The sample was a convenience sample and lacked randomization.There were also many opportunities for selection bias. Participants wererecruited primarily during the day time, probably targeting more womenwho do not work during the day and might have significantly differentsociodemographic profiles in areas such as education levels andacculturation than do women who do work during the day. It is alsopossible that less acculturated women tend to attend health clinics suchas the one where recruitment was conducted for this study. As mentionedearlier, the sample's sociodemographic characteristics reflectedthose of the clinic, which has a high poverty rate, a large Latinopopulation, and the highest HIV incidence in the country. In addition,the sample had a high refusal rate, which was probably a result of thesensitive content of the study. IMPLICATIONS FOR INTERVENTION Acculturation remains a paradox for Latinas in that healthyprotective behaviors decline with acculturation (Abraido-Lanza, Florez,& Florez, 2005). Our findings suggest that this paradox isparticularly complex for Puerto Rican women in relation to sexual riskbehaviors and IPV. Moreover, the conceptualization of acculturation as aconstruct used in research requires further refinement. For bothreasons, further studies are needed to shed light on this topic. Behavioral interventions remain the only way to control the spreadof HIV (Crepaz et al., 2006), and it is important to have effectiveinterventions to eliminate racial and ethnic disparities (Amaro & dela Torre, 2002). Findings from the present study highlight the need forspecific prevention messages that take into consideration differencesamong Puerto Rican women that are based on place of birth and languagepreference. Community-based programs, service providers, and researchersoften struggle with the challenge of how to tailor specific preventionmessages and interventions to specific ethnic groups and subgroups.Findings from the present study can be used to tailor interventionmessages because they reveal different risk factors. For example, giventhat Spanish-speaking Puerto Rican women in this study did not perceivethemselves to be at risk for HIV, prevention strategies for them mightfocus on promoting awareness of risk and accuracy of risk assessment. Partner abuse increases women's vulnerability to HIV becausewomen often have difficulties negotiating safer sex and getting accessto health services, preventive programs, and treatment care(Gonzalez-Guarda, Peragallo, Urrutia, Vasquez, & Mitrani, 2008;Moreno, 2007). Our findings reveal that exposure to IPV creates uniqueHIV risk factors for Puerto Rican women "who speak English. It isimperative to integrate these considerations into prevention messagesand programs for Puerto Rican women who live in abusive situations. Original manuscript received June 18, 2010 Final revision received August 23, 2010 Accepted September 3, 2010 REFERENCES Abraido-Lanza, A., Florez, C., & Florez, K. (2005). Do healthybehaviors decline with greater acculturation? Implications for theLatino mortality paradox. Social Science & Medicine, 61, 1243-1255. Aldarondo, E., Kaufman Kantor, G., & Jasinski,J. L. (2002). Arisk marker analysis of wife assault in Latino families. ViolenceAgainst Women, 8, 429-254. Amaro, H. (1995). Love, sex, and power: Considering women'srealities in H1V prevention. American Psychologist, 50, 437-447. Amaro, H., & de la Torre, A. (2002). Public health needs andscientific opportunities in research on Latinos. American Journal ofPublic Health, 92, 525-529. Amaro, H., & Raj, A. (2000). On the margin: Power andwomen's HIV risk deduction strategies. Sex Roles, 42, 723-749. Arcia, E., Skinner, M., Bailey, D., & Correa, J. (2001). Modelsof acculturation and health behaviors among Latino immigrants to the US.Social Science & Medicine, 53, 41-53. Bureau of Justice Statistics. (2000). Female victims of violentcrimes. Washington, DC: U.S. Department of Justice. Caetano, R., Ramisetty-Mikler, S., Caetano, P.A., & Harris, J.R. (2007). Acculturation stress, drinking and intimate partner violenceamong Hispanic couples in the U.S.Journal of Interpersonal Violence, 22,1431-1447. Campbell, J. C., Baty, M. L., Ghandour, R. M., Stockman, J. K.,Francisco, L., & Wagman, J. (2008).The intersection of intimatepartner violence against women and HIV/A1DS: A review. InternationalJournal of Injury Control and Safety Promotion, 15, 221-231. Centers for Disease Control and Prevention. (2008). Cases of HIVinfection and AIDS in the United States by race/ethnicity, 2002-2006.HIV/AIDS Surveillance Supplemental Report, 13(1). Retrieved fromhttp://www.cdc.gov/hiv/surveillance/resources/ reports/2008supp_vo113no1/ Centers for Disease Control and Prevention. (2009, August). HIVamong Hispanics/Latinos [Fact sheet]. Retrieved fromhttp://www.cdc.gov/hiv/hispanics/ resources/factsheets/hispanic.htm Crepaz, N., Lyles, C., Wolitski, R., Passin, W., Rama, S.,Herbest,J., et al. (2006). Do prevention interventions reduce HIV riskbehaviours among people living with HIV? A metanalytic review ofcontrolled trials. AIDS, 20; 143-157. El-Bassel, N., Gilbert, L, Krishman, S, Schilling, R. E, Gaetha,T., & Purpura, S. (1998). Partner violence and HIV-risk behaviorsamong women in an inner-city emergency department. Violence and Victims,13, 377-393. El-Bassel, N., Witte, S. S., Gilbert, L., Wu, E., Chang, M., Hill,J, & Steinglass, P. (2003). The efficacy of a relationship basedHIV/STI prevention for heterosexual couples. American Journal of PublicHealth, 93, 963-969. Ellis, M., Convay, D., & Bailey, A. (1996).The circularmigration of Puerto Rican women: Towards a gendered explanation.International Migration/Migraciones Internationales, 34, 31-64. Field, C.A., & Caetano, R. (2005). Longitudinal modelpredicting mutual partner violence among white, black, and Hispaniccouples in the United States general population. Violence and Victims,20, 499-511. Flaskerud, J. H., & Uman, G. (1996).Acculturation and itseffects on self-esteem among immigrant Latina women. BehavioralMedicine, 22, 123-133. Gilbert, L., El-Bassel, N., Schilling, R., Wada, T., & Bennet,B. (2000). Partner violence and sexual risk behavior among women inmethadone treatment. AIDS and Behavior, 4, 261-269. Gonzalez-Guarda, R., Peragallo, N., Urrutia, M., Vasquez, E., &Mitrani, V. (2008). HIV risks, substance abuse, and intimate partners.Journal of the Association of Nurses in AIDS Care, 19, 252-266. Harris, R., Firestone, J., & Vega, W. (2005). The interactionof country of origin, acculturation and gender role ideology on wifeabuse. Social Science Quarterly, 86, 463-483. Hunt, L. M., Schneider, S., & Comer, B. (2004). Should"acculturation" be a variable in health research? A criticalreview of research on US Hispanics. Social Science & Medicine, 59,973-986. Jaccard, J., McDonald, R.,Wan, C. K., Dittus, P., & Quinlan, S.(2002). The accuracy of self-reports of condom use and sexualbehavior.Journal of Applied Social Psychology, 32, 1863-1903. Jasinski, J. L. (1998). The role of acculturation in wife assault.Hispanic Journal of Behavioral Sciences, 20, 175-191. Kang, S.Y., Deren, S., Andia, J., Colon, H., & Robles, R.(2001). Gender differences in HIV risk behaviors among Puerto Rican druginjectors by awareness of HIV seropositive status. AIDS & Behavior,5, 241-249. Kaufman Kantor, G., & Asdigian, N. L. (1997). Genderdifferences in alcohol-related spousal aggression. In R.W. Wilsnack& S. C. Wilsnack (Eds.), Gender and alcohol: Individual and socialperspectives (pp. 312-334). Piscataway, NJ: Rutgers Center of Alcohol Studies. Kaufman Kantor,G., Jasinski, J., & Aldarondo, E. (1994). Sociocultural status andincidence of marital violence in Hispanic families. Violence andVictims, 9, 207-222. Lara, M., Gamboa, C., Kahramanian, M.Y., Morales, L., &Hayes-Bautista, D. (2005). Acculturation and health in the UnitedStates: A review of the literature and its sociopolitical context.Annual Review of Public Health, 26, 367-397. Marin, B.V.,Tschann,J. M., Gomez, A. C., & Kegeles, S. M.(1993). Acculturation and gender differences in sexual attitudes andbehaviors: Hispanic vs. nonHispanic white unmarried adults. AmericanJournal of Public Health, 83, 1759-1761. Montoya, I. D., Bell, D. C., Richard, A.J., Carlson, J.W., &Trevino, R.A. (1999). Estimated HIV risk among Hispanics in a nationalsample of drug users. Journal of Acquired Immune Deficiency Syndrome andHuman Retrovirology, 11, 83-87. Moreno, C. L. (2007).The relationship between culture, gender,structural factors, abuse, trauma, and HIV/ AIDS for Latinas.Qualitative Health Research, 17(3), 1-13. Moreno, C. L,, & El-Bassel, N. (2007). Dominican and PuertoRican women in partnerships and their sexual risk behaviors. HispanicJournal of Behavioral Sciences, 29, 336-348. Neff, J. A., Holamon, B., & Davis Schluter, T. (1995). Spousalviolence among Anglos, blacks, and Mexican Americans: The role ofdemographic variables, psychosocial predictors, and alcohol consumption.Journal of Family Violence, 10, 1-21. New York City Department of Health and Mental Hygiene. (2009). HIVepidemiology and field services semiannual report: January 1,2008-December 31, 2008 4(2). Retrieved from http://www.nyc,gov/html doh/downloads/pdf/dires/ dires-2009-report-semi2. pdf Oropesa, R. S., Landale, N. S., & Greif, M.J. (2008). FromPuerto Rican to pan-ethnic in New York City. Ethnic & RacialStudies, 31, 1315-1339. Ortiz-Torres, B., Serrano-Garcia, I., & Torres-Burgos, N.(2000). Subverting culture: Promoting HIV/AIDS prevention among PuertoRican and Dominican women. American Journal of Community Psychology, 28,859-881. Raffaelli, M., & Suarez Al-Adam, M. (1998). Reconsidering theHIV/AIDS prevention needs in Latino women in the United States. In N. L.Roth & L. K. Fuller (Eds.), Women and AIDS: Negotiating saferpractices, care, and representation (pp. 7-41). Binghamton, NY: HaworthPress. Rivera, J. (1994). Domestic violence against Latinas by Latinomales:An analysis of race, national origin, and gender differentials.Boston College Third World Law Journal, 14, 231-257. Rogler, L. H., Cortes, D. E., & Malgady, R.(1991).Acculturation and mental health status among Hispanics:Convergence and new directions for research. Hispanic Journal ofBehavioral Sciences, 46, 585-597. Rojas-Guyler, L., Ellis, N., &Sanders, S. (2005). Acculturation, health protective sexualcommunication, and HIV/AIDS risk behaviors among Hispanic women in alarge midwestern city. Health Education & Behavior, 32, 767-779. Schiff, M., Witte, S. S., & El-Bassel, N. (2003). Clientsatisfaction and perceived helping aspects of an HIV/ AIDS preventiveintervention for urban couples. Research on Social Work Practice, 13,468-492. Strauss, M.A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B.(1996).The Revised Conflict Tactics Scale (CTS2): Development andpreliminary psychometric data. Journal of Family Issues, 17, 283-316. Strauss, M.A., & Smith, C. (1990).Violence in Hispanic familiesin the United States: Incidence rates and structural interpretations. InM. A. Strauss & R.J. Gelles (Eds.), Physical violence in Americanfamilies: Risk factors and adaptations to violence in 8,145families (pp.341-367). New Brunswick, NJ:Transaction. Suarez-Al-Adam, M., Raffaelli, M., & O'Leary, A. (2000).Influence of abuse and partner hypermasculinity on the sexual behaviorof Latinas. AIDS Education and Prevention, 12, 263-274. U.S. Census Bureau. (2010). State & county quick facts.Retrieved from http://quickfacts.census.gov/qfd/ states/00000.html Vargas Carmona, J., Romero, G., & Burns Loeb, T. (1999). Theimpact of HIV status and acculturation on Latinas' sexual risktaking. Cultural Diversity & Ethnic Minority Psychology, 5, 209-221. Wyatt, G. E., Myers, H., Williams, J. K., Ramirez Kitchen, C.,Loeb, T., Vargas-Camona, J., et al. (2002). Does a history of traumacontribute to HIV risk for women of color? Implications for preventionand policy. American Journal of Public Health, 92, 660-665. Claudia L. Moreno, PhD, is associate professor, Graduate School ofSocial Service, Fordham University, 113 West 60th Street, New York, NY10024; e-mail: cmoreno@fordham.edu. Allison C. Morrill, PhD, JD, is president and founder ofCapacities, Watertown, MA. Nabila El-Bassel, DSW, MSW, is professor,School of Social Work, Columbia University, New York. The study reportedhere was conducted at Columbia University and was supported by NationalInstitute of Mental Health grant MH57145, awarded to Nabila El-Bassel.Table 1: Sociodemographic Characteristics of Participants and ReportedFrequencies of Experiencing Intimate Partner Violence,by Language Preference and Birthplace (N = 1,003) Spanish English Preferred (a) Preferred (b) Total (c)Characteristic M SD M SD M SD Age (years): 18-73P.R. born (d) 45.96 0.57 38.77 0.80 43.58 12.23U.S. born (c) 35.26 1.42 32.78 0.49 33.04 8.75Total (c) 44.51 12.74 34.41 9.38 38.32 11.87 Years in United States 0-64P.R. born (d) 25.04 15.26 28.84 13.03 26.30 14.66 Effect of Effect of Language BirthplaceCharacteristic t (df) p t (df) p Age (years): 18-73P.R. born (d)U.S. born (c)Total (c) 13.5 (652) <.0001 15.7(910) <.0001 Years in United States 0-64P.R. born (d) -2.9 (374) .004 Spanish English Preferred (a) Preferred (b) Total (c) % n % n % n High School education (f)P.R. born (d) 37 107 43 57 39 164U.S. born (e) 55 24 54 220 54 244Total (c) 39 131 52 277 47 408 Employee (d)P.R. born (d) 7 23 13 21 9 44U.S. born (e) 13 7 28 123 26 130Total (c) 8 30 23 144 17 174 Main male sexual partnerP.R. born (d) 47 158 65 108 53 266U.S. born (e) 45 24 76 339 73 363Total (c) 47 182 73 447 63 629 Effect of Effect of Language Birthplace [chi square] [chi square] (1) p (1) p High School education (f)P.R. born (d)U.S. born (e)Total (c) 13.4 (g) .0002 21.8 (g) <.0001 Employee (d)P.R. born (d)U.S. born (e)Total (c) 40.9 (h) <.0001 51.9 (h) <.0001 Main male sexual partnerP.R. born (d)U.S. born (e)Total (c) 68.2 (h) <.0001 41.2 (h) <.0001 Spanish EnglishWoman with Preferred (a) Preferred (b) Total (j)a Main Partner % n % n % n History of IPV: EverP.R. born (k) 8 13 29 31 16 44U.S. born (l) 21 5 19 66 20 71Total (m) 10 18 22 97 18 115 Effect of Effect ofWoman with Language Birthplacea Main Partner [chi square] [chi square] (1) p (1) p History of IPV: EverP.R. born (k)U.S. born (l)Total (m) 12.2 (n) .0005 1.0 (a) 0.32Note: P.R. born = born in the Commonwealth of Puerto Rico; U.S.born = born in the contintental United States; IPV = intimatepartner violence. (a) n = 389. (b) n = 614. (c) n = 1,003. (d) n= 503. (e) n = 500. (f) n = 874. (g) n = 874. (h) N = 1,002. (i)n = 183. (j) n = 447. (k) n = 267, (l) n = 263. (m) n = 620. (n)N = 630.Table 2: Frequencies of Reported Sexual Risk Factors,by Language Preference and Birthplace (N = 1,003) Spanish English Preferred (a) Preferred (b) Total (c)Characteristic % n % n % n Partners in lifetime: >4P.R. born (d) 18 58 44 68 26 126U.S. born (e) 33 17 59 255 56 272Total (c) 20 75 55 323 41 398 Partners in past year: >1P.R. born (d) 8 25 24 40 13 65U.S. born (e) 11 6 29 130 27 136Total (c) 8 31 28 170 20 201 History of STI: EverP.R. born (d) 9 31 32 54 17 85U.S. born (e) 15 8 31 140 30 148Total (c) 10 39 32 194 23 233 Effect of Effect of Language Birthplace [chi square] [chi square]Characteristic (1) p (1) p Partners in lifetime: >4P.R. born (d)U.S. born (e)Total (c) 117.5 (f) <.0001 89.5 (f) <.0001 Partners in past year: >1P.R. born (d)U.S. born (e)Total (c) 57.5 (g) <.0001 31.7 (g) <.0001 History of STI: EverP.R. born (d)U.S. born (e)Total (c) 61.8 (g) <.0001 22.5 (g) <.0001 Spanish EnglishWoman with Preferred (a) Preferred (b) Total (j)a Main Partner % n % n % n Used condom with main partner of past 90 days: At all (k)P.R. born (l) 9 11 25 23 16 34U.S. born (e) 17 4 26 75 25 79Total (m) 10 15 25 98 21 113 Effect of Effect ofWoman with Language Birthplacea Main Partner [chi square] [chi square] (1) p (1) p Used condom with main partner of past 90 days: At all (k)P.R. born (l)U.S. born (e)Total (m) 15.2 (n) <.0001 6.8 (n) .01Woman with a Spanish EnglishMain Partner Preferred (o) Preferred (p) Total (q)Who Useda Condom % n % n % n Used condom with main partner of past 90 days: Every time (r)P.R. born (s) 64 7 39 9 47 16U.S. born (t) 75 3 27 20 29 23Total (c) 67 10 30 29 36 41 Effect of Effect ofWoman with a Language BirthplaceMain PartnerWho Used [chi square] [chi square]a Condom (1) p (1) p Used condom with main partner of past 90 days: Every time (r)P.R. born (s)U.S. born (t)Total (c) 7.7 (u) .005 3.2 (u) 0.07 Spanish EnglishWoman with Preferred (v) Preferred (w) Total (m)a Main Partner % n % n % n Know partner has risk factorP.R. born (x) 8 10 27 25 16 31U.S. born (y) 13 3 14 42 14 45Total (m) 9 13 17 67 15 80 Effect of Effect ofWoman with Language Birthplacea Main Partner [chi square] [chi square] (1) p (1) p Know partner has risk factorP.R. born (x)U.S. born (y)Total (m) 6.4 (n) .01 0.3 (n) 0.56 Spanish EnglishWoman with Preferred (z) Preferred (aa) Total (bb)a Main Partner % n % n % n Worry partner has risk factorP.R. born (cc) 12 14 31 21 19 35U.S. born (dd) 10 2 20 74 28 76Total (bb) 12 16 30 95 24 111 Effect of Effect ofWoman with Language Birthplacea Main Partner [chi square] [chi square] (1) p (1) p Worry partner has risk factorP.R. born (cc)U.S. born (dd)Total (bb) 17.1 (ee) .0001 4.5 (ee) 0.03Note: P.R. born = born in the Commonwealth of Puerto Rico;U.S. born = born in the continental United States;STI = sexually transmitted infection; IPV = intimate partner violence.(a) n = 389. (b) n = 614. (c) n = 1,003. (d) n = 503.(e) n = 500. (f) N = 967. (g) N = 1,002. (h) n = 183.(i) n = 447. (j) n = 620. (k) Of the 534 who had sex. (l) n = 218.(m) n = 534. (n) N = 534. (o) n =15. (p) n = 97. (q) n = 112. (r) Ofthe 113 who used a condom at all in past 90 days. (s) n = 34.(t) n = 500. (u) N= 112. (v) n = 149. (w) n = 385. (x) n = 218.(y) n = 316. (z) n = 137. (aa) n = 319. (bb) n = 456. (cc) n = 183.(dd) n = 273. (ee) N = 456.Table 3: Logistic Regression: Birthplace, Language, and OtherSociodemographics Characteristics as Predictors of Sexual RiskFactor for Entire Sample (N = 1,003) Partners in Lifetime: >4 Para.Characteristic Est. pAge nsHigh school education nsEmployed nsRelationship status * nsBirthplace: U.S. -.29 .002 Birthplace x Age --Prefer English -.57 <.0001 Partners in Lifetime: >4 OR CIBirthplace: U.S. 0.56 0.39-0.81 Age: U.S. born Age: P.R. bornPrefer English 0.32 0.22-0.48 Partners in Past Year: >1 Para.Characteristic Est. pAge -.03 .0006High school education nsEmployed nsRelationship status * nsBitthplace: U.S. 0.89 .01 Birthplace x Age -.03 .005Prefer English -.58 <.0001 Partners in Past Year: >1 OR CIBirthplace: U.S. Age: U.S. born 1.00 0.97-1.02 Age: P.R. born 0.94 0.91-0.96Prefer English 0.31 0.19-0.53 History of STI Para.Characteristic Est. pAge nsHigh school education nsEmployed .31 .01Relationship status * nsBitthplace: U.S. .66 Birthplace x Age -.02 .02Prefer English -.67 <.0001 History of STI OR CIBirthplace: U.S. Age: U.S. born 1.02 1.00-1.05 Age: P.R. born 0.98 0.96-1.01Prefer English 0.26 0.16-0.43Notes: Para. Est. = parameter estimate; p = p of Wald [chisquare] OR = odds ratio; Ci = confidence interval. A dash in acell indicates an interaction that was not significant in theinitial full model and was dropped from the final model.(a) Has main partner.Table 4: Logistic Regression: Acculturation, OtherSociodemographic Characteristics, and Partner Abuse asPredictors of Sexual Risk Factors for Women with a MainPartner (N = 620) Used Condom with Main Partner: Past 90 Days Para. pCharacteristic Est.Age nsHigh school education nsEmployed nsBirthplace: U.S. ns Birthplace x Education -.39 .01Prefer English -.66 .002Language x Education .56 .01IPV .33 .05 Used Condom with Main Partner: Past 90 Days OR CIBirthplace: U.S. Birthplace: Less than 0.37 0.14-0.94 high school education Birthplace: High school 1.87 0.86-4.06 graduatePrefer English Birthplace: Less than 0.81 0.29-2.29 high school education Birthplace: Less than 0.09 0.02-0.32 high school education Knows Partner Has a Risk Factor Para.Characteristic Est. pAge nsHigh school education .31 .03Employed nsBirthplace: U.S. ns Birthplace x Education --Prefer English -.75 .0005Language x Education --IPV -.33 .03 Knows Partner Has a Risk Factor OR CIBirthplace: U.S. ns Birthplace: Less than -- high school education Birthplace: High school -- graduatePrefer English 0.23 0.10-0.52 Birthplace: Less than -- high school education Birthplace: Less than -- high school education Worries Partner Has a Risk Factor ParsCharacteristic Est. pAgeHigh school education .29 .02Employed .44 .01Birthplace: U.S. ns Birthplace x Education --Prefer English -.64 .001Language x Education --IPV -.52 .0005 Worries Partner Has a Risk Factor OR CIBirthplace: U.S. ns Birthplace: Less than -- high school education Birthplace: High school -- graduatePrefer English 0.28 0.13-0.60 Birthplace: Less than -- high school education Birthplace: Less than -- high school educationNotes: Para. Est. = parameter estimate; p = p of Wald [chisquare]; OR = odds ratio; CI = confidence interval. A dash in acell indicates an interaction that was not significant in theinitial full model and was dropped from the final model.