LATINO FAMILY MENTAL HEALTH: EXPLORING THE ROLE OF DISCRIMINATION AND FAMILISMO

The purpose of this study was to examine the role of discrimination and familismo on internalizing mental health symptoms among two generations of Latinos, youth and their parents, residing in the Southwest region of the United States. Data from the Latino Acculturation and Health Project was used to determine the direct and moderation effects of discrimination and familismo on internalizing mental health symptoms. The sample included 150 Latino youth–parent dyads who were immigrants or U.S. born. Descriptive results indicate that youth had significantly higher scores on the familismo scale whereas parents reported higher levels of perceived discrimination. Regression analyses results revealed direct effects of familismo and perceived discrimination on internalizing mental health symptoms. Implications for practice are discussed.

The Latino community is the largest and fastest growing minority population in the U.S. (U.S. Bureau of the Census, 2004). Many Latinos are in socially vulnerable positions. Structural factors such as anti-immigrant policies and poverty, and the inequities and discriminatory practices that accompany these factors impact Latino mental health and well-being. For example, the Latino population has been identified as a high risk group for depression and anxiety connected to poverty, poor housing conditions, and rigid work demands (Magana & Hovey, 2003); and poor mental health related to discrimination (Araujo & Borrell, 2006; Ramos, Jaccard, & Guilamo-Ramos, 2003).

Anti-immigrant policies, particularly those proposed and implemented in the Southwest region of the United States (Rubio-Goldsmith, Romero, Rubio-Goldsmith, Escobedo, & Khoury, 2009; Takei, Saenz, & Li, 2009) can accelerate the stress and fear experienced by immigrant populations thus potentially impacting the Latino community’s mental health and overall wellbeing. Yet, the Latino community’s value of familismo, their strong ties to immediate and extended family members, can be protective of their wellbeing (Parsai, Voisine, Marsiglia, Kulis, & Nieri, 2009). “Evidence suggests that features of familismo such as pride, belonging, and obligation members of the family continue to be distinctive attributes across generations regardless of the length of time one has resided in the U.S.” (Santiago-Rivera, 2003). As the family is such a critical aspect in Latinos lives and there is a high reliance on the family for material and emotional support and help (Marin & Marin, 1991), it is crucial to understand how familismo influences Latinos mental health. The purpose of this study is to examine the role of discrimination and familismo on internalizing mental health symptoms among a sample of Latino youth and their parents, residing in a large metropolitan area of the Southwest region of the United States.

LITERATURE REVIEW

The Latino population accounts for 41.8 million or 14% of the people living in the U.S. (U.S. Bureau of the Census, 2004). For the last 400 years, people of Latino ancestry have been part of the lands identified today as the U.S. More recent Latino immigrants continue to experience the challenges of integrating into the host society (Ellis & Gunnar, 2009). Although many recent immigrants are documented, it is estimated that seven million Latino immigrants are undocumented (Organista, 2007). Ten percent of all children in the United States live in mixed status households where one parent or household member is undocumented (Community for Hispanic Children and Families, 2004; Kanaiaupuni, 2000). Arizona is one of 11 states with the most rapid growth in the undocumented population, now representing 40 to 49% of all immigrants (Passel, Capps, & Fix, 2004).

Recent immigration policies such as proposition 187 in California and SB1175 in Arizona have heightened the anti-immigrant sentiment in the Southwest region of the United States (Massey, 2009). Such policies increase the barriers to accessing needed services as undocumented individuals’ fear being detected by immigration authorities and subsequently being deported (Kullgren, 2003). Concomitantly, the Latino community experiences higher levels of perceived and actual discrimination (Hovey, Rojas, Kain, & Magaña, 2000). There is substantial evidence supporting the negative effects of discrimination on individuals mental health (Araújo Dawson, 2009; Gee, Ryan, Laflamme, & Holt, 2006; Moradi & Risco, 2006; Umaña-Taylor & Updegraff, 2007; Yip, Gee, & Takeuchi, 2008).

A plethora of studies have found that “discrimination is associated with multiple indicators of poorer physical health and, especially, mental health” (Williams, Neighbors, & Jackson, 2008, p. s29). In a sample of adults, discrimination was identified as a contributing factor to lower scores on the Mental Component Summary (MCS12), a measure of overall psychological wellbeing (Gee, Ryan, Laflamme, & Holt, 2006). Several factors have been identified as moderating the relationship between discrimination and mental health status. Gee and colleagues (2006) found that the length of time residing in the United States moderated the relationship between mental health and discrimination with Latino individuals who have resided in United States longer experiencing more negative effects. Among Mexican origin participants, the effects of perceived discrimination on depression were greater if participants were born in the United States, female, highly acculturated (as measured by language behaviors), and educated in both Mexico and the United States (Finch, Kolody, & Vega, 2000). Similarly, acculturation was identified as moderating the relationship between discrimination and stress levels among a sample of Dominican women (Araujo & Dawson, 2009). Individuals with lower levels of social support experience more harmful effects of discrimination on health (Finch & Vega, 2003).

Among adolescents a relationship between discrimination and mental health has also been established. Researchers have focused on developmental issues related to identity formation, self-esteem, and parent–child relationships. Umaña-Taylor and Updegraff (2007) found that various aspects of the self (including self-esteem, ethnic identity, and cultural orientation) protect or enhance the risks associated with discrimination. For example, as adolescents reported more discrimination they reported lower self-esteem and more depressive symptoms. Among male adolescents higher levels of orientation toward mainstream culture were related to a positive relationship between discrimination and depressive symptoms suggesting that a strong orientation toward mainstream culture may heighten the negative effects of discrimination (Umaña-Taylor & Updegraff, 2007). Umaña-Taylor and Updegraff also found that high levels of involvement in Latino culture served as a protective factor minimizing the negative effects of discrimination on youth’s development. Similarly, Smokowski and Bacallao (2007) found that perceived discrimination and parent adolescent conflict were significant predictors of internalizing and externalizing symptoms. As evidenced by multiple studies the role of discrimination must be considered when conceptualizing interventions and treatment plans for the Latino population (Moradi & Risco, 2006).

Recent immigrants tend to have better mental health status as compared to U.S.-born Latinos. This finding is commonly referred to as the epidemiological or immigrant paradox as immigrants tend to have better outcomes, although they often experience greater hardships than non immigrants of similar socioeconomic characteristics (Johnson & Marchi, 2009). For example, when compared with recent immigrants (less than 13 years), immigrants with longer residency in the United States (more than 13 years), and U.S.-born Mexicans scored worse on multiple measures of mental health (Vega et al., 1998). Results indicate that immigrants who have resided in the U.S. longer (13 years or more) were at an increased risk of experiencing a range of mental health problems and substance/alcohol abuse. Similarly, Mexican mothers and their male partners experiencing poverty-related hardships were found to have healthier lifestyles than their U.S. born counterparts (Mull, Agran, Winn, & Anderson, 2001). Mexican mothers were less likely to use drugs, alcohol, or experience mental health disorders compared to Mexican American and White mothers. The epidemiological paradox has been attributed to a protective or buffering effect of traditional cultural values and practices (Escobar, 1998; Vega et al., 1998). Latino families are often described as close knit with extended family networks that offer a great deal of support (Escobar, 1998; Finch & Vega, 2003).

Familismo, a Latino cultural value, refers to the importance of strong family loyalty, closeness, and getting along with and contributing to the wellbeing of the nuclear family, extended family, and kinship networks (Cauce & Domenech-Rodriguez, 2000; Guilamo-Ramos et al., 2007). The strong ties between family members have been attributed to helping newly immigrated individuals adjust and confront social inequities in the United States (Baca Zinn, 1994). However, Latino families may face challenges to maintaining strong support networks after immigration and in coping with the changes in values due to acculturation (Aranda & Knight, 1997). Evidence suggests that familismo is a protective factor for Latino families as this cultural value, for example, has been linked to positive health outcomes including lower levels of substance and drug abuse (Gil, Wagner, & Vega, 2000; Unger et al., 2002), increase likelihood of seeking out mammogram exams (Suarez, 1994), and decreased likelihood of child maltreatment (Coohey, 2001).

THEORETICAL FRAMEWORK

This study is informed by the ecological perspective (Bronfenbrenner, 1979). The ecological perspective suggests that multiple factors at multiple systemic levels intersect to influence individuals’ wellbeing. The ecological perspective lends itself to the analysis of structural factors that impact Latino families’ wellbeing as well as the strengths associated with Latino culture (Hancock, 2005). Bronfenbrenner conceptualized the context in which one develops, or the ecological environment, as a set of nested structures including micro-, meso-, exo-, and macrosystems (Eamon, 2001). The microsystem involves immediate interactions with one’s parents, people residing in one’s home, and peers. Consistent with the Latino cultural value of familismo the family structure and ties among family members promote and influence Latino family wellbeing. The mesosystem involves interactions among two or more microsystems, for example, children’s interactions with their parents may influence their interactions with their peers. The exosystem involves the process between two or more settings where only one setting involves the developing person. For example, the type of formal and informal sources of support that parents have may influence a child indirectly. The macrosystem includes policies, opportunity structures, material resources that promote or hinder development and wellbeing. Assessing Latino immigrant families’ macro-level dynamics is a necessary component of culturally competent assessments and interventions (Hancock, 2005). The ecological perspective indicates that individual’s perception of their social environment can have significant effects on their wellbeing (Bronfenbrenner, 1979).

The proximal process or interactions with others and various structures occur within the context of multiple environments and over time (Bronfenbrenner & Ceci, 1994). For example, the proximal process may include learning appropriate ways to behave, learning to read and write, and accessing appropriate resources. The proximal processes are influenced by micro interactions such as parent–child connections. However, other environments can also influence proximal processes, for example, one’s community (exosystem) and remote environments such as federal laws (macrosystem). The proximal process informs one’s experience of multiple environments.

Anti-immigrant sentiment and immigration policies in the macro environment can influence children and parents indirectly through their community environment or parents’ work environment, for example. Discrimination may replace Latino families’ hope for a better future with a sense of thwarted social mobility and marginalization (Finch et al., 2000; Hancock, 2005). Moreover, a person’s perception of discrimination is informed by the interactions, or the proximal process, with multiple environments including anti-immigrant sentiment and policies in the macro environment. The aims of this study were to (a) assess for the direct effects of discrimination and familismo on internalizing symptoms while controlling for demographics and immigration status, and (b) identify if there is an interaction effect between discrimination and familismo on internalizing symptoms of mental health among Latino families. The overall hypothesis leading the study was that discrimination and familismo had direct but opposite effects on internalizing symptoms among youth and their parents after controlling for demographics. The secondary hypothesis of the study was that there was an interaction effect between discrimination and familismo on internalizing symptoms of mental health among Latino families such that high levels of familismo will reduce the harmful effects of discrimination. The contribution of this study lies in its analysis of the impact of discrimination and the retention of traditional values such as familismo, on the mental health of two generations within the same household. In addition, the sample consists of Latino U.S.-born and immigrant families residing in a state with nativist (Cohen-Marks, Nuño, & Sanchez, 2009) immigration attitudes and policies.

METHOD

This study was completed using data from the Southwest subsample of the Latino Acculturation Health Project (LAHP) dataset. The study’s protocol and bilingual measures were approved by the Arizona State University Institutional Review Board. The Southwest LAHP dataset consists of mental health, acculturation, and demographic data on 150 families residing in the United States–Mexico border region.

Sampling and Procedures

Families were recruited at multiple sites including English as a Second Language (ESL) classes, community centers, local churches, and community fairs in a large metropolitan area. Criteria for inclusion in the study was self-identifying as Latino/a, agreeing to participate in paper & pencil questionnaires every 6 months for a total of four times (a span of about 2.5 years), and being a parent of an adolescent 14–18 years of age who would also agree to participate in the study. Participants were asked in which country and city they were born. Although the target population was Latinos in general, due to the demographics in the city in which the study was conducted, all participants recruited were of Mexican descent with the exception of six who were born in Central America. Parent–child dyads were interviewed in their homes separately (total sample N = 300). Questionnaires were available in both Spanish and English. Participants could choose to answer the questions on their own or to have interviewers read the questions to them. This article is informed by baseline data.

Measures

The dependent variable, internalizing mental health, is measured using the internalizing score in the Youth Self-Report (YSR) for youth and Center for Epidemiologic Studies Depression Scale (CES-D) for parents.

The YSR (Achenbach & Rescorla, 2001) is a standardized measure used to assess adolescents’ emotional and behavioral problems. The YSR has 112 items related to academic performance, social competency, family and peer relationships, and maladaptive behaviors. Each item is scored using a 3-point scale (0 = not true; 1 = somewhat or sometimes true; 2 = very true or often true) with higher scores indicating more problem behaviors. The questions are computed into three scores, internalizing, externalizing, and total score. For the purpose of this study only the internalizing scale was used. The internalizing scores consist of scales measuring anxious and withdrawn depression symptoms and somatic complaints. T scores of 60 and above on the internalizing scale are indicative of clinical/borderline emotional and behavioral problems (Achenbach, 1991). Raw scores were used for the moderation analysis. The reliability for this measure was good (α = .87).

The CES-D (Radloff, 1977) was completed by the parent participants. The CES-D is a widely used scale to screen for depression symptomatology in the general population. Short versions of the CES-D have been previously used in other studies with results suggesting that shorter forms are reliable with Mexican immigrant populations and no measurement precision relative to the full CES-D version is lost (Grzywacz, Hovey, Seligman, Arcury, & Quandt, 2006). The instrument was pilot tested with a small group of Latino parents and youth. To reduce participant burden only 12 of the original 20 questions were used in the LAHP study. The most reliable items were retained (α = .82). The following are sample questions: “I was bothered by things that usually don’t bother me”; “I felt that I was just as good as other people”; and “I was happy.” Participants are asked to report the frequency for each statement within the past week using a 4-point scale (0 = less than once day, 1 = 1–2 days, 2 = 3–4 days, and 3 = 5–7 days). Scores range between zero and 36 with higher scores indicating more depressive symptomatology. The clinical cutoff for the original 20-item CES-D was 16, and for this study it was recalculated to 12.7.

The predictor perceived discrimination was measured with a 3-item scale. This scale was previously used by Finch et al. (2000) with a sample of Mexicans and Mexican Americans. Scores on this scale range from 3 to 15 with higher scores indicating greater perceived discrimination. The following is a sample question: “You are treated unfairly because you are Latino.” The reliability of the perceived discrimination scale was acceptable (for youth α = .74, for adults α = .77).

The measure for the moderator, familismo, consisted of six items previously used by Gil and colleagues (2000). The scale is measured using a 4-point scale (1 = strongly disagree to 4 = strongly agree). Scores on the familismo scale can range from 6 to 24 with higher scores indicating a greater tie to the value of familismo. The following are sample questions: “Family members respect one another?” and “Family members feel loyal to other family members?” One of the original seven items from the scale was excluded to increase the scale’s reliability. In this study, the scale’s reliability is good (for youth, α = .86; for parents, α = .90).

Demographic variables (gender, age, level of education, marital status, and income) were used to describe the sample and as control variables in the moderation analysis. Gender was coded with male as the reference category. Consistent with the literature on Latino mental health, we also included length of time residing in the United States as a control variable. We differentiate between U.S. born and immigrants and also took into account the length of time immigrant youth and parents have resided in the United States. There has been much variation in the literature regarding how the cutoff points for number of years living in the United States are determined. We followed the procedures used by Finch and Vega (2003) as their study involves a sample of Latinos in the western region of the United States. Upon close analysis of the distribution of the data for parents we found three cutoff points (0–7 years, 8–15 years, and 16 and more years) that split the sample roughly the same. Similarly, we found that two groups (1–5 years and 6 and more years) for the youth sample of immigrants. Moreover, the variable for length of time was coded as follows: (1) for parents, whole life or U.S. born, 0–7 years, 8–15 years, and 16 and more years, with U.S. born as the reference category; and (2) for youth, whole life or U.S. born, 1–5 years, and 6 and more years, with U.S. born as the reference category. Similar to Finch and Vega (2003), we are using length of time in the United States as a proxy for acculturation.

Moderation Analysis

Hierarchical regression analyses were completed to determine the moderator effect of the relationship between familismo and perceived discrimination on internalizing mental health (as measured by the CES-D for parents and the CBCL internalizing scale for youth). The regression analysis was completed using three blocks. The first block consisted of the control variables. For youth, Block 1 included gender, age, and length of time in the United States. For parents, Block 1 included gender, age, level of education, income, and length of time in the United States. In the second block, the predictor, perceived discrimination, and moderator, familismo, were entered to identify a main effect. A main effect was deemed present when after controlling for the effects of the control variables, discrimination and/or familismo were significant predictors of internalized symptoms. Finally, the third block consisted of the interaction term, Perceived Discrimination × Familismo. The predictors and interaction term were mean-centered prior to completing the analyses. A moderation effect was deemed to exist under the following conditions: (a) the coefficient for the interaction term was statistically significant, and (b) the interaction term significantly increased the amount of variance explained in the dependent variable (Cohen & Cohen, 1983).

RESULTS

Descriptive Analyses

One hundred fifty youth–parent dyads participated in this study. A majority of the parent participants were female (n = 141, 94%), married (n = 109, 72.6%), and immigrants (n = 131, 87.3%) with a mean age of 40 (SD = 6.73). It was expected that a high percentage of parents would be married or in a committed relationship as many are immigrant and Latino immigrants tend to have higher rates of two-parent households compared to other groups (Passel & Cohn, 2009). The adult sample in this study is on average a few years older than adult Latino samples found in other similar studies (Finch & Vega, 2003; Gee et al., 2006; Guarnaccia et al., 2007; Moradi & Risco, 2006). There was much variation in the length of time parents had resided in the United States if they were immigrants (see Table 1 ). The mean household income was 24,191 (SD = 15, 447). The mean household income in this study is lower than the median household income of 28,820 for Latinos nationwide (Pew Hispanic Center, 2010). Parents’ level of education ranged from no schooling (1.3%) to college graduate (9.9%) with approximately 36% reporting some high school education and 23% reporting that they were high school graduates. The sample in this study had slightly higher levels of education compared to the study completed by Gee and colleagues (2006). The mean age for youth participants was 15.5 (SD = 1.25). Youth were female (n = 90, 60%), U.S. born (n = 83, 55.7%), in high school (n = 104, 69.3%), and resided with two parents (n = 116, 77.3%). The demographics are summarized in Tables 1 and ​ and2 2 .

Table 1