Women's Health Issues 25-2 (2015) 185–192
www.whijournal.com
Original article
Psychological Distress Among Low-Income U.S.- and ForeignBorn Women of Mexican Descent: Impact of Acculturation Venera Bekteshi, PhD, MSW, MPA, MA a,*, Qingwen Xu, PhD b, Thanh Van Tran, PhD c a
School of Social Work, University of Illinois at Urbana-Champaign, Urbana, Illinois School of Social Work, Tulane University, New Orleans, Louisiana c Graduate School of Social Work, Boston College, Chestnut Hill, Massachusetts b
Article history: Received 3 March 2014; Received in revised form 28 October 2014; Accepted 3 November 2014
a b s t r a c t Purpose: After testing the capacity of Kessler’s psychological distress (K6) scale to measure equally across low-income Mexican-born women (n ¼ 881) and U.S.-born women of Mexican descent (n ¼ 317), this study assesses the impact of acculturation on this group’s psychological distress. Methods: We employ descriptive and confirmatory factor analyses to test the cross-cultural equivalence of K6. Multivariate and logistic regression is used to test the association between acculturation and psychological distress among low-income, Mexican-American women. Results: The cross-cultural equivalence analysis shows that some of the scale’s items have the capacity to measure psychological distress equally among participants. Regression results indicate that the more acculturated these women become, the greater their psychological distress is. Conclusion: The study recommends that researchers emphasize the cross-cultural equivalence of their measures and suggests a heightened awareness among practitioners of the multidimensional impact of acculturation on clients of Mexican descent. Copyright Ó 2015 by the Jacobs Institute of Women’s Health. Published by Elsevier Inc.
Over the past 20 years, the United States has experienced one of the largest waves of immigration in its history, and foreign-born Latinos are among the fastest growing segments of the population. In 2006, this group numbered 17,690,524, reflecting an increase of 25% from 2000 (U.S. Census Bureau, 2010). According to the U.S. Census Bureau (2010) and Pew Hispanic Center (2007), the Latino population is predicted to constitute one-quarter of the U.S. population by 2050, with Mexican Americans accounting for more than half of this group. Among them, 48.4% is female (Pew Hispanic Center, 2007). Despite this demographic shift, studies focusing on the psychological distress of Latina women in the United States remain limited. Previous research shows that the shift from a predominant Mexican culture to a predominant U.S. culture, which transpires with the process of acculturation, is accompanied by a change in * Correspondence to: Venera Bekteshi, PhD, MSW, MPA, MA, Assistant Professor, University of Illinois at Urbana-Champaign School of Social Work, 1010 West Nevada Urbana, IL 61801. Phone: (618) 509-1120; Fax: (217) 244-4499. E-mail address: [email protected] (V. Bekteshi).
the psychological awareness among women of Mexican descent (Cabassa, Lester, & Zayas, 2007; Lugo Steidel & Contreras, 2003). Perceptions of mental health among Mexican-born women may be different from those of their U.S.-born counterparts, particularly because some of the essential Mexican values believed to shape beliefs about mental health among Mexican-born women may not be as influential for their U.S.-born counterparts (Rodriguez, 1998; Sabogal, Marin, Otero-Sabogal, Marin, & Perez-Stable, 1987; Lugo Steidel & Contreras, 2003; Cabassa et al., 2007). For example, a study by Cabassa and colleagues (2007) examined perceptions of depression between these two groups of women. The study found that a significantly larger proportion of U.S.-born women of Mexican descent (categorized as Americanized) viewed depression as a phenomenon related to biological, genetic, or chemical factors, while a majority of Mexican-born women associated depression with presence of interpersonal problems, such as family conflicts and lack of family support systems. Despite this difference in the cultural understanding of depression, there has been a lack of research testing simplified depression scales for crosscultural equivalency.
1049-3867/$ - see front matter Copyright Ó 2015 by the Jacobs Institute of Women’s Health. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.whi.2014.11.001
186
V. Bekteshi et al. / Women's Health Issues 25-2 (2015) 185–192
The Kessler psychological distress scale (K6), which combines a number of depression and anxiety symptoms, is widely used among researchers as an indicator of psychological distress or as a screening tool for mental illness (Drapeau, Beaulieu-Prevost, Marchand, Boyer, Preville, et al., 2010) and recently has been adopted for the Behavioral Risk Factor Surveillance System (Centers for Disease Control and Prevention [CDC], 2013). The K6 scale was deliberately constructed to contain items that function similarly across sociodemographic subgroups (Kessler et al., 2002; 2003), and the scale was adopted by the National Health Interview Survey (NHIS) to assess psychological distress across racial/ethnic groups in the United States, including Mexicans/ Mexican Americans (Bratter & Eschback, 2005). The K6 scale has been validated in at least 14 countries worldwide, including Mexico (Kessler et al, 2010). However, to our knowledge, this scale has never been tested for its capacity to measure mental health equally across Mexican- and U.S.-born women of Mexican descent. This study represents an initial effort to examine whether the K6 scale is cross-culturally equivalentdthat is, whether the items used to measure the construct are perceived identically across the two groups (Tran, 2009). It is assumed that primarily because of acculturationda process of change in cultural attitudes and behaviors as a result of an encounter between two cultures (Berry, 1997)dlow-income Mexican- and U.S.-born women of Mexican descent differ in perceptions of psychological distress. This assumption is based on the understanding that particularly for the U.S.-born second generation immigrants, acculturation takes place in a social context. For them, the outcome of acculturation depends on multiple factors including family, community, and social considerations (Rumbaut & Portes, 2001). Contact with mental health services in the United States and the stigma associated with mental illness interact with immigrants’ cultural beliefs and affect their perceptions of psychological distress. Recognizing the segmented nature of acculturation (e.g., Berry & Sabatier, 2010), researchers have noted that non-skilled (often low-income and/or unauthorized) immigrants present a unidirectional process of acculturation; that is, second- and third-generation immigrants tend to exhibit greater levels of accumulation to the host culture than previous generation members of an ethnic group. We hypothesize that Mexican- and U.S.-born women of Mexican descent differ in their ability to identify the K6 items as key aspects of their psychological distress. In this study, we seek to enable group-specific interpretation of the overall K6 score to increase our knowledge about psychological well-being among Latina women. Literature Review Acculturation, Psychological Distress, and Poverty Most of the research on the impact of acculturation on the psychological distress of Mexican women shows a positive relationship between acculturation and psychological distress: the higher the level of acculturation among this group of women, the higher their psychological distress (Kurz, Baris & Davis, 2005; Acevedo, 2000; Gordon, 2007). Studies have also confirmed that compared to less acculturated women, the more acculturated women are more likely to suffer from depression (Shatel, Smith, Colwell & Villalba, 2008), have lower life satisfaction (Shatel et al., 2008), and be more prone to substance abuse (Moracco, Hilton, Hodges & Frasier, 2005). Research attributes several factors to healthier psychological profiles among less acculturated
Mexican immigrant women. More acculturated Mexican women tend to have a higher level of awareness about discrimination. A study that examined the impact of discrimination among lowincome women of Mexican descent working at a factory found that the women who spoke English were asked to work harder and longer to help management communicate with workers who lacked English skills, but the women were not duly compensated for that assistance. Women in this study reported high psychological distress (Easter et al., 2000). In addition, less acculturated Mexican immigrant women often maintain strong networks for migration and early-stage settlement, which is seen as a positive moderating factor that diminishes the negative impact of acculturation (Vega, Gil, Zimmerman, Warheit, & Somers, 1993). Furthermore, Mexican cultural elements such as family cohesiveness and support and being family centered have been regarded as positive factors among less acculturated Mexican-born women (Hovey, 2000). Nevertheless, Mexican cultural elements may lose their buffering effect on the negative impact of acculturation where low-income women of Mexican descent are concerned. Studies have found that regardless of their level of acculturation, lowincome women of Mexican descent are likely to report higher levels of psychological distress than their high-income counterparts. (Hovey, 2000; Easter et al., 2006). Factors such as working long hours to meet their financial obligations, inability to balance the roles of mother and provider, lack of role models, and lack of support from their partners were linked with psychological distress among this group regardless of their acculturation level (Hovey, 2000; Coltrane, Parke & Adams, 2002; Gryzwacz et al., 2007). Therefore, it remains uncertain to what extent the level of acculturation and poverty interrelate and jointly affect the psychological distress of women of Mexican descent. Mexican Culture, Acculturation, and Perception of Psychological Distress Some studies indicate that less acculturated Mexican women report lower rates of psychological problems owing to different perceptions of their mental health. Several cultural values may eradicate the link as the women of Mexican descent in the United States become more acculturated, affecting their attitudes related to mental health (Rodrıguez, 1998; Sabogal et al., 1987; Lugo Steidel & Contreras, 2003). Some of the characteristics commonly identified in the Mexican culture that may shape how Mexican-born women perceive their mental health include familismo, machismo, and marianismo (Arredondo, 2002; Hinkelman, 2001). Familismo denotes a strong affiliation with family members that translates into a sense of connectedness, collectivism, and interdependence. It is manifested by active participation in activities organized with family members. Mexican-born women perceive their health as a holistic unit encompassing spiritual, physical, emotional, and familial dimensions (Vega et al., 2000). Although some dimensions of familism, such as the importance of family support, do not change with acculturation (Lugo Steidel & Contreras, 2003; Rodrıguez, 1998; Sabogal et al., 1987), Lugo Steidel and Contreras (2003) found that more acculturated individuals adhered less to the overall principles of familism, especially beliefs of familial interconnectedness and familial honor. Consequently, for some women of Mexican descent in the United States, mental health may be associated more with biological, genetic, and chemical factors than with a range of salient factors that includes family relations (Cabassa et al., 2007).
V. Bekteshi et al. / Women's Health Issues 25-2 (2015) 185–192
Machismo characterizes much of traditional men’s behaviordthe man who proves his virility through the domination and impregnation of women is accepted as exercising his permitted authority in the typical Mexican household (Hinkelman, 2001). Marianismo is associated with Virgin Mary’s portrayal as devout, pure, noble, and giving. Both machismo and marianismo define a significant part of the behavior of Mexican women, who are expected to be submissive, self-sacrificing, dependent, sexually naive and repressed, and good housewives. Women are also considered to be morally and spiritually superior to men, and are expected to endure difficulties inflicted by men (Arredondo, 2002). Consequently, although it should not be assumed all Mexicans in Mexico subscribe to these traditional values, in agreement with Cromwell and Ruiz (1979), most Mexican women recognize these values and could be less likely to perceive the related behaviors as “psychologically distressing” (Arredondo, 2002). Scholars have indicated that the more acculturated Mexican women become, the more likely they are to adopt liberal views of gender roles (Gutman, 1996; O’Guinn, Imperia, & MacAdams, 1987). Consequently, the more likely Mexican women are to embrace Western social mores, the more aware they may become of mental health. This shift may cause them to perceive mental health differently. Because of these different cultural values, and as a result of changes owing to immigration and acculturation, it should be noted that measurement instruments for mental health might not be “cross-culturally” equivalent across the two groupsdU.S.-born women of Mexican descent, who could be regarded as more acculturated, and Mexican-born immigrant women, who are considered as less acculturated. Reflecting the growth of Latino population studies addressing their mental health have increased. However, research on lowincome women of Mexican descent are scarce despite statistics indicating that more than 25% of families of Mexican descent in the United States have annual incomes below the federal poverty line (U.S. Census, 2010). Consistent with previous research, this study recognizes the impact of acculturation and focuses on the following research questions. 1. Can K6 items that assess the psychological distress be perceived equivalently between low-income Mexican-born women and U.S.-born women of Mexican descent? Based on the available theories and empirical studies on acculturation, we hypothesize that the K6 scale is not cross-culturally equivalent between these two groups of women. 2. Is acculturation associated with the psychological distress among low-income U.S.-born and Mexican-born women of Mexican descent? We hypothesize that, when controlling for demographics, the increased level of acculturation makes the Mexican women’s odds of reporting psychological distress greater. Methods Source of Data This study uses the data from the California Health Interview Survey (CHIS) Adult Survey 2007, a project conducted by the University of California at Los Angeles Center for Health Policy Research, the California Department of Health Services, and the Public Health Institute. The CHIS collects extensive information for all age groups on health status, health conditions, healthrelated behaviors, and health insurance coverage in California.
187
A total of 43,020 participants living in California were interviewed in 2007 through a multistage sampling design by which the state was divided into 44 geographic sampling strata, 41 single-county strata, and 3 multicounty strata. Households were selected through random-digit dial, and within each household, one adult (age 18) respondent was randomly selected. For people who either did not speak English or did not speak English well enough to participate, interviews were conducted in Spanish or in one of four other languages. From the CHIS Adult Survey 2007, the present study selects 1,197 female, low-income respondents with income of less than 100% of the poverty line who claimed Mexican ethnicity; 881 participants (approximately 73%) were born in Mexico and the remaining 316 participants (27%) were born in the United States. The descriptive statistics of U.S.-born and Mexican-born women of Mexican descent are reported in Table 1. No differences were found between the two groups in terms of age, family size, family type, working status, or level of education. In general, most women in this sample are married with children, and a considerable percentage of them are fully employed. Measures In the CHIS Adult Survey 2007, the psychological distress (K6) of adults over age 18 is measured by asking adults to rate their level of agreement on six statements: 1) felt nervous during the past month (Nervous), 2) felt hopeless during the past month (Hopeless), 3) felt restless during the past month (Restless), 4) felt depressed during the past month (Depressed), 5) felt everything was an effort during the past month (Effort), and 6) felt worthless during the past month (Worthless). A Likert scale was adopted for responses to these six statements ranging 1 (“none of the time”) to 5 (“All the time”; Kessler et al., 2002; 2003). The scale was created by adding all six items, ranging from 0 to 24, and then transforming them into a dichotomous variable where 0 represents low psychological distressdranging from 0 to 12dand 1 represents high psychological distressdranging from 13 to 24 (Kessler et al., 2002; 2010). A more traditional way of measuring psychological distress is using the
Table 1 Demographic Information of Low-income Mexican- and U.S.-Born Women of Mexican Descent
Age (mean) Household income (mean) Household size (mean) Working status (%) Employed Unemployed Family type (%) Single, no children Married, no children Married with children Single with children Education (%) Elementary school At least some high school At least some college education No formal education Acculturation level (mean)* Psychological distress (mean)* *p < .05.
U.S. Born (n ¼ 316)
Mexican Born (n ¼ 881)
41.7 (SD ¼ 19.83) 11,642 (SD ¼ 6,384) 3.56 (SD ¼ 2.03)
41.61 (SD ¼ 14.68) 10,776 (SD ¼ 6011) 4.1 (SD ¼ 1.03)
54.6 41.1
51.9 49.1
28.1 6.9 31.9 33.1
30.5 10.0 36.9 22.6
38.5 45.7 9.5
34.4 44.4 15.0
6.0 7.58 (SD ¼ 1.70) 24.34 (SD ¼ 4.81)
5.9 5.65 (SD ¼ 1.46) 25.00 (SD ¼ 4.41)
188
V. Bekteshi et al. / Women's Health Issues 25-2 (2015) 185–192
Statistical Methods
Table 2 Corrected Item-Total Correlation and Cronbach’s a Items*
U.S.-Born Women
Mexican-Born Women
Nervous Restless Effort Hopeless Depressed Worthless Cronbach’s a
0.59 0.64 0.57 0.67 0.69 0.64 0.85
0.58 0.57 0.54 0.70 0.67 0.66 0.83
*
All inter-item correlation coefficients are significant at .05.
K6 score of 13 or greater (Forman-Hoffman et al., 2014; Kessler, 2002); however, consistent with previous research (Kessler et al., 2010) looking at the distribution of variable we decided to lower the cutoff score by 1.0 point to capture more people in this category. Nonetheless, in our analysis of the impact of acculturation on psychological distress, we use both cutoff points (See Table 7 for detailed results). Use of English (i.e., language acculturation) is adopted in this study as a proxy to measure the level of acculturation, owing to the limitations of the CHIS Adult Survey data. Although measures of acculturation are usually constructed multidimensionally (e.g., llar scale; Cue llar, Harris & Jasso, 1980), it is noted that the the Cue use of English remains the most important element that captures the most variation in acculturation (Alegria et al., 2009; Vega et al., 1993). Scales of acculturation based on language use, similar to ours, have been shown to be reliable and valid measure of acculturation (Deyo, Diehl, Hazuda & Stern, 1985). Use of English in this study is measured by asking participants to rate their level of agreement with three statements: 1) language used at home, 2) language used with friends, and 3) language used in the media/TV/radio. A Likert-type scale was adopted for the responses to these statements: 1 (“language other than English”), 2 (“English and another language”), and 3 (“English only”). The items were then added, with higher results reflecting higher level of acculturation. The acculturation scale ranges from 4 to 12. The interitem correlation is strong (Cronbach’s a ¼ .79). Table 3 Within Group Confirmatory Factor Analysis with One Factor Model for Six Items of K6 Scale Items
Nervous Restless Effort Hopeless Depressed Worthless
U.S.-Born Women of Mexican Descent (n ¼ 317)
Mexico-Born Immigrant Women (n ¼ 881)
ML* (d)
CSS (d)
ML (d)
CSS (d)
1.00 1.28 1.12 1.41 1.27 1.31
0.56 0.67 0.50 0.82 0.76 0.83
1.00 1.16 1.04 1.08 1.20 1.01
0.62 0.72 0.67 0.70 0.80 0.72
(0.14) (0.15) (0.14) (0.13) (0.13)
(0.69) (0.54) (0.75) (0.33) (0.43) (0.31)
(0.11) (0.11) (0.11) (0.11) (0.12)
(0.61) (0.48) (0.55) (0.51) (0.36) (0.49)
Goodness of Fit Statistics Minimum fit function c2 (DF) RMSEAz NNFIz GFIz AGFIz
14.30 (9) (p ¼ .11*)y .046 .99 .98 .96
17.50 (9) (p ¼ .042)y 0.056 0.99 0.98 0.96
* ML stands for maximum likelihood estimates; error theta-delta; CSS stands for completely standardized solution. These factor loadings are statistically significant at .05. y The p-value of this test should be greater than .05. z The root-mean-square error of approximation (RMSEA) with a value of less than 0.05 indicates good fit; non-normed fit index (NNFI), goodness of fit indices (GFI), and adjusted GFI (AGFI) with a value close to 0.95 indicate good fit.
Measurement equivalence is defined as the degree to which “research instruments used to collect data for the defined variables bear the same meanings and psychometric properties” (Tran, 2009, p. 67). To test the cross-cultural equivalence of the K6 scale, the study applies an analysis guided by a widely referred framework of cross-cultural validation (e.g., IngersollDayton, 2011; Luyt, 2011; Tran, 2009; Tran, Chan & Nguyen, 2011). This comprehensive technique consists of several layers of analysis, including descriptive and confirmatory factor analysis (CFA; for a detailed explanation of this framework, see Tran, 2009). The major goal of this extensive analysis is to detect differences in results between two groups of women, not to establish the validity of the K6 scale to assess psychological distress. Descriptive statistics, including kurtosis, skewness, corrected item–total correlation, and Cronbach’s a, are used to compare the group difference in terms of item distribution, distance, and closeness. Corrected item–total correlation tells how well a particular item correlates with all other items of a scale; Cronbach’s a is a summary statistic of how well the items of a scale tie together. CFA tests variance structures and factor patterns. Tran (2009) indicates that key statistics to accepting or rejecting the hypothesisdcross-cultural equivalencedare c2 (>0.05), root mean square error of approximation (RMSEA;
www.whijournal.com
Original article
Psychological Distress Among Low-Income U.S.- and ForeignBorn Women of Mexican Descent: Impact of Acculturation Venera Bekteshi, PhD, MSW, MPA, MA a,*, Qingwen Xu, PhD b, Thanh Van Tran, PhD c a
School of Social Work, University of Illinois at Urbana-Champaign, Urbana, Illinois School of Social Work, Tulane University, New Orleans, Louisiana c Graduate School of Social Work, Boston College, Chestnut Hill, Massachusetts b
Article history: Received 3 March 2014; Received in revised form 28 October 2014; Accepted 3 November 2014
a b s t r a c t Purpose: After testing the capacity of Kessler’s psychological distress (K6) scale to measure equally across low-income Mexican-born women (n ¼ 881) and U.S.-born women of Mexican descent (n ¼ 317), this study assesses the impact of acculturation on this group’s psychological distress. Methods: We employ descriptive and confirmatory factor analyses to test the cross-cultural equivalence of K6. Multivariate and logistic regression is used to test the association between acculturation and psychological distress among low-income, Mexican-American women. Results: The cross-cultural equivalence analysis shows that some of the scale’s items have the capacity to measure psychological distress equally among participants. Regression results indicate that the more acculturated these women become, the greater their psychological distress is. Conclusion: The study recommends that researchers emphasize the cross-cultural equivalence of their measures and suggests a heightened awareness among practitioners of the multidimensional impact of acculturation on clients of Mexican descent. Copyright Ó 2015 by the Jacobs Institute of Women’s Health. Published by Elsevier Inc.
Over the past 20 years, the United States has experienced one of the largest waves of immigration in its history, and foreign-born Latinos are among the fastest growing segments of the population. In 2006, this group numbered 17,690,524, reflecting an increase of 25% from 2000 (U.S. Census Bureau, 2010). According to the U.S. Census Bureau (2010) and Pew Hispanic Center (2007), the Latino population is predicted to constitute one-quarter of the U.S. population by 2050, with Mexican Americans accounting for more than half of this group. Among them, 48.4% is female (Pew Hispanic Center, 2007). Despite this demographic shift, studies focusing on the psychological distress of Latina women in the United States remain limited. Previous research shows that the shift from a predominant Mexican culture to a predominant U.S. culture, which transpires with the process of acculturation, is accompanied by a change in * Correspondence to: Venera Bekteshi, PhD, MSW, MPA, MA, Assistant Professor, University of Illinois at Urbana-Champaign School of Social Work, 1010 West Nevada Urbana, IL 61801. Phone: (618) 509-1120; Fax: (217) 244-4499. E-mail address: [email protected] (V. Bekteshi).
the psychological awareness among women of Mexican descent (Cabassa, Lester, & Zayas, 2007; Lugo Steidel & Contreras, 2003). Perceptions of mental health among Mexican-born women may be different from those of their U.S.-born counterparts, particularly because some of the essential Mexican values believed to shape beliefs about mental health among Mexican-born women may not be as influential for their U.S.-born counterparts (Rodriguez, 1998; Sabogal, Marin, Otero-Sabogal, Marin, & Perez-Stable, 1987; Lugo Steidel & Contreras, 2003; Cabassa et al., 2007). For example, a study by Cabassa and colleagues (2007) examined perceptions of depression between these two groups of women. The study found that a significantly larger proportion of U.S.-born women of Mexican descent (categorized as Americanized) viewed depression as a phenomenon related to biological, genetic, or chemical factors, while a majority of Mexican-born women associated depression with presence of interpersonal problems, such as family conflicts and lack of family support systems. Despite this difference in the cultural understanding of depression, there has been a lack of research testing simplified depression scales for crosscultural equivalency.
1049-3867/$ - see front matter Copyright Ó 2015 by the Jacobs Institute of Women’s Health. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.whi.2014.11.001
186
V. Bekteshi et al. / Women's Health Issues 25-2 (2015) 185–192
The Kessler psychological distress scale (K6), which combines a number of depression and anxiety symptoms, is widely used among researchers as an indicator of psychological distress or as a screening tool for mental illness (Drapeau, Beaulieu-Prevost, Marchand, Boyer, Preville, et al., 2010) and recently has been adopted for the Behavioral Risk Factor Surveillance System (Centers for Disease Control and Prevention [CDC], 2013). The K6 scale was deliberately constructed to contain items that function similarly across sociodemographic subgroups (Kessler et al., 2002; 2003), and the scale was adopted by the National Health Interview Survey (NHIS) to assess psychological distress across racial/ethnic groups in the United States, including Mexicans/ Mexican Americans (Bratter & Eschback, 2005). The K6 scale has been validated in at least 14 countries worldwide, including Mexico (Kessler et al, 2010). However, to our knowledge, this scale has never been tested for its capacity to measure mental health equally across Mexican- and U.S.-born women of Mexican descent. This study represents an initial effort to examine whether the K6 scale is cross-culturally equivalentdthat is, whether the items used to measure the construct are perceived identically across the two groups (Tran, 2009). It is assumed that primarily because of acculturationda process of change in cultural attitudes and behaviors as a result of an encounter between two cultures (Berry, 1997)dlow-income Mexican- and U.S.-born women of Mexican descent differ in perceptions of psychological distress. This assumption is based on the understanding that particularly for the U.S.-born second generation immigrants, acculturation takes place in a social context. For them, the outcome of acculturation depends on multiple factors including family, community, and social considerations (Rumbaut & Portes, 2001). Contact with mental health services in the United States and the stigma associated with mental illness interact with immigrants’ cultural beliefs and affect their perceptions of psychological distress. Recognizing the segmented nature of acculturation (e.g., Berry & Sabatier, 2010), researchers have noted that non-skilled (often low-income and/or unauthorized) immigrants present a unidirectional process of acculturation; that is, second- and third-generation immigrants tend to exhibit greater levels of accumulation to the host culture than previous generation members of an ethnic group. We hypothesize that Mexican- and U.S.-born women of Mexican descent differ in their ability to identify the K6 items as key aspects of their psychological distress. In this study, we seek to enable group-specific interpretation of the overall K6 score to increase our knowledge about psychological well-being among Latina women. Literature Review Acculturation, Psychological Distress, and Poverty Most of the research on the impact of acculturation on the psychological distress of Mexican women shows a positive relationship between acculturation and psychological distress: the higher the level of acculturation among this group of women, the higher their psychological distress (Kurz, Baris & Davis, 2005; Acevedo, 2000; Gordon, 2007). Studies have also confirmed that compared to less acculturated women, the more acculturated women are more likely to suffer from depression (Shatel, Smith, Colwell & Villalba, 2008), have lower life satisfaction (Shatel et al., 2008), and be more prone to substance abuse (Moracco, Hilton, Hodges & Frasier, 2005). Research attributes several factors to healthier psychological profiles among less acculturated
Mexican immigrant women. More acculturated Mexican women tend to have a higher level of awareness about discrimination. A study that examined the impact of discrimination among lowincome women of Mexican descent working at a factory found that the women who spoke English were asked to work harder and longer to help management communicate with workers who lacked English skills, but the women were not duly compensated for that assistance. Women in this study reported high psychological distress (Easter et al., 2000). In addition, less acculturated Mexican immigrant women often maintain strong networks for migration and early-stage settlement, which is seen as a positive moderating factor that diminishes the negative impact of acculturation (Vega, Gil, Zimmerman, Warheit, & Somers, 1993). Furthermore, Mexican cultural elements such as family cohesiveness and support and being family centered have been regarded as positive factors among less acculturated Mexican-born women (Hovey, 2000). Nevertheless, Mexican cultural elements may lose their buffering effect on the negative impact of acculturation where low-income women of Mexican descent are concerned. Studies have found that regardless of their level of acculturation, lowincome women of Mexican descent are likely to report higher levels of psychological distress than their high-income counterparts. (Hovey, 2000; Easter et al., 2006). Factors such as working long hours to meet their financial obligations, inability to balance the roles of mother and provider, lack of role models, and lack of support from their partners were linked with psychological distress among this group regardless of their acculturation level (Hovey, 2000; Coltrane, Parke & Adams, 2002; Gryzwacz et al., 2007). Therefore, it remains uncertain to what extent the level of acculturation and poverty interrelate and jointly affect the psychological distress of women of Mexican descent. Mexican Culture, Acculturation, and Perception of Psychological Distress Some studies indicate that less acculturated Mexican women report lower rates of psychological problems owing to different perceptions of their mental health. Several cultural values may eradicate the link as the women of Mexican descent in the United States become more acculturated, affecting their attitudes related to mental health (Rodrıguez, 1998; Sabogal et al., 1987; Lugo Steidel & Contreras, 2003). Some of the characteristics commonly identified in the Mexican culture that may shape how Mexican-born women perceive their mental health include familismo, machismo, and marianismo (Arredondo, 2002; Hinkelman, 2001). Familismo denotes a strong affiliation with family members that translates into a sense of connectedness, collectivism, and interdependence. It is manifested by active participation in activities organized with family members. Mexican-born women perceive their health as a holistic unit encompassing spiritual, physical, emotional, and familial dimensions (Vega et al., 2000). Although some dimensions of familism, such as the importance of family support, do not change with acculturation (Lugo Steidel & Contreras, 2003; Rodrıguez, 1998; Sabogal et al., 1987), Lugo Steidel and Contreras (2003) found that more acculturated individuals adhered less to the overall principles of familism, especially beliefs of familial interconnectedness and familial honor. Consequently, for some women of Mexican descent in the United States, mental health may be associated more with biological, genetic, and chemical factors than with a range of salient factors that includes family relations (Cabassa et al., 2007).
V. Bekteshi et al. / Women's Health Issues 25-2 (2015) 185–192
Machismo characterizes much of traditional men’s behaviordthe man who proves his virility through the domination and impregnation of women is accepted as exercising his permitted authority in the typical Mexican household (Hinkelman, 2001). Marianismo is associated with Virgin Mary’s portrayal as devout, pure, noble, and giving. Both machismo and marianismo define a significant part of the behavior of Mexican women, who are expected to be submissive, self-sacrificing, dependent, sexually naive and repressed, and good housewives. Women are also considered to be morally and spiritually superior to men, and are expected to endure difficulties inflicted by men (Arredondo, 2002). Consequently, although it should not be assumed all Mexicans in Mexico subscribe to these traditional values, in agreement with Cromwell and Ruiz (1979), most Mexican women recognize these values and could be less likely to perceive the related behaviors as “psychologically distressing” (Arredondo, 2002). Scholars have indicated that the more acculturated Mexican women become, the more likely they are to adopt liberal views of gender roles (Gutman, 1996; O’Guinn, Imperia, & MacAdams, 1987). Consequently, the more likely Mexican women are to embrace Western social mores, the more aware they may become of mental health. This shift may cause them to perceive mental health differently. Because of these different cultural values, and as a result of changes owing to immigration and acculturation, it should be noted that measurement instruments for mental health might not be “cross-culturally” equivalent across the two groupsdU.S.-born women of Mexican descent, who could be regarded as more acculturated, and Mexican-born immigrant women, who are considered as less acculturated. Reflecting the growth of Latino population studies addressing their mental health have increased. However, research on lowincome women of Mexican descent are scarce despite statistics indicating that more than 25% of families of Mexican descent in the United States have annual incomes below the federal poverty line (U.S. Census, 2010). Consistent with previous research, this study recognizes the impact of acculturation and focuses on the following research questions. 1. Can K6 items that assess the psychological distress be perceived equivalently between low-income Mexican-born women and U.S.-born women of Mexican descent? Based on the available theories and empirical studies on acculturation, we hypothesize that the K6 scale is not cross-culturally equivalent between these two groups of women. 2. Is acculturation associated with the psychological distress among low-income U.S.-born and Mexican-born women of Mexican descent? We hypothesize that, when controlling for demographics, the increased level of acculturation makes the Mexican women’s odds of reporting psychological distress greater. Methods Source of Data This study uses the data from the California Health Interview Survey (CHIS) Adult Survey 2007, a project conducted by the University of California at Los Angeles Center for Health Policy Research, the California Department of Health Services, and the Public Health Institute. The CHIS collects extensive information for all age groups on health status, health conditions, healthrelated behaviors, and health insurance coverage in California.
187
A total of 43,020 participants living in California were interviewed in 2007 through a multistage sampling design by which the state was divided into 44 geographic sampling strata, 41 single-county strata, and 3 multicounty strata. Households were selected through random-digit dial, and within each household, one adult (age 18) respondent was randomly selected. For people who either did not speak English or did not speak English well enough to participate, interviews were conducted in Spanish or in one of four other languages. From the CHIS Adult Survey 2007, the present study selects 1,197 female, low-income respondents with income of less than 100% of the poverty line who claimed Mexican ethnicity; 881 participants (approximately 73%) were born in Mexico and the remaining 316 participants (27%) were born in the United States. The descriptive statistics of U.S.-born and Mexican-born women of Mexican descent are reported in Table 1. No differences were found between the two groups in terms of age, family size, family type, working status, or level of education. In general, most women in this sample are married with children, and a considerable percentage of them are fully employed. Measures In the CHIS Adult Survey 2007, the psychological distress (K6) of adults over age 18 is measured by asking adults to rate their level of agreement on six statements: 1) felt nervous during the past month (Nervous), 2) felt hopeless during the past month (Hopeless), 3) felt restless during the past month (Restless), 4) felt depressed during the past month (Depressed), 5) felt everything was an effort during the past month (Effort), and 6) felt worthless during the past month (Worthless). A Likert scale was adopted for responses to these six statements ranging 1 (“none of the time”) to 5 (“All the time”; Kessler et al., 2002; 2003). The scale was created by adding all six items, ranging from 0 to 24, and then transforming them into a dichotomous variable where 0 represents low psychological distressdranging from 0 to 12dand 1 represents high psychological distressdranging from 13 to 24 (Kessler et al., 2002; 2010). A more traditional way of measuring psychological distress is using the
Table 1 Demographic Information of Low-income Mexican- and U.S.-Born Women of Mexican Descent
Age (mean) Household income (mean) Household size (mean) Working status (%) Employed Unemployed Family type (%) Single, no children Married, no children Married with children Single with children Education (%) Elementary school At least some high school At least some college education No formal education Acculturation level (mean)* Psychological distress (mean)* *p < .05.
U.S. Born (n ¼ 316)
Mexican Born (n ¼ 881)
41.7 (SD ¼ 19.83) 11,642 (SD ¼ 6,384) 3.56 (SD ¼ 2.03)
41.61 (SD ¼ 14.68) 10,776 (SD ¼ 6011) 4.1 (SD ¼ 1.03)
54.6 41.1
51.9 49.1
28.1 6.9 31.9 33.1
30.5 10.0 36.9 22.6
38.5 45.7 9.5
34.4 44.4 15.0
6.0 7.58 (SD ¼ 1.70) 24.34 (SD ¼ 4.81)
5.9 5.65 (SD ¼ 1.46) 25.00 (SD ¼ 4.41)
188
V. Bekteshi et al. / Women's Health Issues 25-2 (2015) 185–192
Statistical Methods
Table 2 Corrected Item-Total Correlation and Cronbach’s a Items*
U.S.-Born Women
Mexican-Born Women
Nervous Restless Effort Hopeless Depressed Worthless Cronbach’s a
0.59 0.64 0.57 0.67 0.69 0.64 0.85
0.58 0.57 0.54 0.70 0.67 0.66 0.83
*
All inter-item correlation coefficients are significant at .05.
K6 score of 13 or greater (Forman-Hoffman et al., 2014; Kessler, 2002); however, consistent with previous research (Kessler et al., 2010) looking at the distribution of variable we decided to lower the cutoff score by 1.0 point to capture more people in this category. Nonetheless, in our analysis of the impact of acculturation on psychological distress, we use both cutoff points (See Table 7 for detailed results). Use of English (i.e., language acculturation) is adopted in this study as a proxy to measure the level of acculturation, owing to the limitations of the CHIS Adult Survey data. Although measures of acculturation are usually constructed multidimensionally (e.g., llar scale; Cue llar, Harris & Jasso, 1980), it is noted that the the Cue use of English remains the most important element that captures the most variation in acculturation (Alegria et al., 2009; Vega et al., 1993). Scales of acculturation based on language use, similar to ours, have been shown to be reliable and valid measure of acculturation (Deyo, Diehl, Hazuda & Stern, 1985). Use of English in this study is measured by asking participants to rate their level of agreement with three statements: 1) language used at home, 2) language used with friends, and 3) language used in the media/TV/radio. A Likert-type scale was adopted for the responses to these statements: 1 (“language other than English”), 2 (“English and another language”), and 3 (“English only”). The items were then added, with higher results reflecting higher level of acculturation. The acculturation scale ranges from 4 to 12. The interitem correlation is strong (Cronbach’s a ¼ .79). Table 3 Within Group Confirmatory Factor Analysis with One Factor Model for Six Items of K6 Scale Items
Nervous Restless Effort Hopeless Depressed Worthless
U.S.-Born Women of Mexican Descent (n ¼ 317)
Mexico-Born Immigrant Women (n ¼ 881)
ML* (d)
CSS (d)
ML (d)
CSS (d)
1.00 1.28 1.12 1.41 1.27 1.31
0.56 0.67 0.50 0.82 0.76 0.83
1.00 1.16 1.04 1.08 1.20 1.01
0.62 0.72 0.67 0.70 0.80 0.72
(0.14) (0.15) (0.14) (0.13) (0.13)
(0.69) (0.54) (0.75) (0.33) (0.43) (0.31)
(0.11) (0.11) (0.11) (0.11) (0.12)
(0.61) (0.48) (0.55) (0.51) (0.36) (0.49)
Goodness of Fit Statistics Minimum fit function c2 (DF) RMSEAz NNFIz GFIz AGFIz
14.30 (9) (p ¼ .11*)y .046 .99 .98 .96
17.50 (9) (p ¼ .042)y 0.056 0.99 0.98 0.96
* ML stands for maximum likelihood estimates; error theta-delta; CSS stands for completely standardized solution. These factor loadings are statistically significant at .05. y The p-value of this test should be greater than .05. z The root-mean-square error of approximation (RMSEA) with a value of less than 0.05 indicates good fit; non-normed fit index (NNFI), goodness of fit indices (GFI), and adjusted GFI (AGFI) with a value close to 0.95 indicate good fit.
Measurement equivalence is defined as the degree to which “research instruments used to collect data for the defined variables bear the same meanings and psychometric properties” (Tran, 2009, p. 67). To test the cross-cultural equivalence of the K6 scale, the study applies an analysis guided by a widely referred framework of cross-cultural validation (e.g., IngersollDayton, 2011; Luyt, 2011; Tran, 2009; Tran, Chan & Nguyen, 2011). This comprehensive technique consists of several layers of analysis, including descriptive and confirmatory factor analysis (CFA; for a detailed explanation of this framework, see Tran, 2009). The major goal of this extensive analysis is to detect differences in results between two groups of women, not to establish the validity of the K6 scale to assess psychological distress. Descriptive statistics, including kurtosis, skewness, corrected item–total correlation, and Cronbach’s a, are used to compare the group difference in terms of item distribution, distance, and closeness. Corrected item–total correlation tells how well a particular item correlates with all other items of a scale; Cronbach’s a is a summary statistic of how well the items of a scale tie together. CFA tests variance structures and factor patterns. Tran (2009) indicates that key statistics to accepting or rejecting the hypothesisdcross-cultural equivalencedare c2 (>0.05), root mean square error of approximation (RMSEA;
