|Home | About | Journals | Submit | Contact Us | Français|
Under an ecodevelopmental framework, we examined lifetime segmented assimilation trajectories (diverging assimilation pathways influenced by prior life conditions) and related them to quality-of-life indicators in a diverse sample of 258 men in the Pheonix, AZ, metropolitan area.
We used a growth mixture model analysis of lifetime changes in socioeconomic status, and used acculturation to identify distinct lifetime segmented assimilation trajectory groups, which we compared on life satisfaction, exercise, and dietary behaviors. We hypothesized that lifetime assimilation change toward mainstream American culture (upward assimilation) would be associated with favorable health outcomes, and downward assimilation change with unfavorable health outcomes.
A growth mixture model latent class analysis identified 4 distinct assimilation trajectory groups. In partial support of the study hypotheses, the extreme upward assimilation trajectory group (the most successful of the assimilation pathways) exhibited the highest life satisfaction and the lowest frequency of unhealthy food consumption.
Upward segmented assimilation is associated in adulthood with certain positive health outcomes. This may be the first study to model upward and downward lifetime segmented assimilation trajectories, and to associate these with life satisfaction, exercise, and dietary behaviors.
“Acculturation” refers to a process of cultural change and adaptation that occurs across time.1–3 Despite this dynamic conception, most acculturation studies have examined acculturation at a single point in time, inferring health-related outcomes from cross-sectional comparisons across levels of acculturation.4 Unfortunately, this static approach does not examine temporal changes in the process of acculturation.
Segmented assimilation has been defined as “diverse patterns of adaptation whereby immigrant groups differentially adopt the attitudes, beliefs, and behaviors of divergent cultural groups in the United States.”3,5(p1344) More specifically, segmented assimilation is a process of cultural and economic integration into a “mainstream” society. Individuals and groups thus will differ in how effectively they succeed in their cultural and economic integration, as indicated by different assimilation trajectories (i.e., segmented assimilation).5
Segmented assimilation theory7 has identified 3 basic outcomes in this process of social integration: (1) acculturation change toward mainstream White American culture coupled with upward socioeconomic mobility (upward assimilation); (2) acculturation change and downward socioeconomic mobility into an underclass (downward assimilation); and (3) resistance to acculturation and to assimilation into the mainstream society (resistance to forced assimilation).8 Downward assimilation is predicted for groups that have low social or human capital.9–12 Within the United States, segmented assimilation has typically been examined for Latino populations, and recently it has been examined with Asian Americans.13
Differing assimilation trajectories may be associated with variations in quality of life, as indicated by differing disease risks and health-related outcomes.14–16 Among Latinos, acculturation toward mainstream American society has been associated with higher rates of alcohol, tobacco, and illegal drug use,17–20 and with higher prevalence rates of psychiatric disorders.21,22 By contrast, among Latinos, greater acculturation has also been associated with enhanced quality of life, including better employment, and access to health insurance and health care.17
One approach to the study of segmented assimilation is to examine lifetime developmental pathways of change within an ecodevelopmental framework.16 Under this framework, change trajectories can be examined for immigrants and for native-born community residents within a common “ecodevelopmental field”23 (Figure 1).
Figure 1 presents an ecodevelopmental field model of the process of segmented assimilation that depicts the sociocultural and socioeconomic trajectories described by segmented assimilation theory.24 The x-axis represents temporal effects involving 4 life milestones: (1) elementary school, (2) middle school, (3) high school, and (4) adulthood. The y-axis represents socioeconomic positions ranging from poverty to affluence. Within the US–Mexico border region, parallel dimensions along the y-axis are socioeconomic status (SES) and levels of acculturation (acculturation). In Figure 1, 3 of many lifetime developmental trajectories are illustrated as coded by high, moderate, or low resource levels, or “capital,” whether social capital (socioeconomic resources and social supports) or human capital (personal skills and capabilities). At the childhood trajectory intercept (the elementary school milestone) a person, family, or group (immigrant or native) initiates a “life journey” from an initial sociocultural position—low, moderate, or high—with the potential for moving upward or downward.
Under this ecodevelopmental framework, both minority persons and persons from the mainstream culture can undergo segmented assimilation.25 Along the US–Mexico border, non-Hispanic White Americans can move culturally from their native White American culture toward Latino/Mexican culture (acculturation toward a Latino/Mexican culture) by learning to speak Spanish, making friends with Latinos, and moving from a predominantly White neighborhood into a Latino neighborhood. Similarly, Latinos can move toward mainstream White American culture (conventional acculturation) by learning English, making friends with White Americans, and moving into a White nonminority neighborhood.
These variations in segmented assimilation trajectories indicate that segmented assimilation is a bidirectional process. Moreover, downward assimilation trajectories are expected to produce illness outcomes, although some low-income, low-acculturation Latinos have exhibited unexpectedly salubrious outcomes, also known as a Hispanic or immigrant paradox.21,26 Conversely, upward assimilation toward upper socioeconomic strata is expected to produce healthy outcomes. The scientifically modeled ecodevelopmental analysis of lifetime assimilation trajectories may yield insights on the developmental processes that promote wellness or that produce health disparities.16,27,28
We modeled lifetime assimilation change by using a conditional growth mixture model (GMM).29,30 Under this model it is hypothesized that certain assimilation trajectories will be associated with an enhanced quality of life: life satisfaction, exercise behaviors, and healthy dietary behaviors. Figure 2 presents this GMM, from which we pose 2 hypotheses.
Hypothesis 1 states that lifetime acculturation change toward mainstream American culture, coupled with upward sociocultural mobility (i.e., upward assimilation) will be associated in adulthood with an enhanced quality of life, as indicated by higher life satisfaction, more frequent exercise behaviors, a more frequent consumption of healthy foods, and a less frequent consumption of unhealthy foods.
Hypothesis 2 conversely states that acculturation change involving downward sociocultural mobility (downward assimilation) will be associated with a compromised quality of life, as indicated by lower life satisfaction, less frequent exercise behaviors, less frequent consumption of healthy foods, and more frequent consumption of unhealthy foods.
From 2004 through 2008, we examined the lifetime assimilation trajectories of a diverse sample of 258 men: 140 Latino men and 85 non-Hispanic White men, and 33 men of other ethnic backgrounds, all of whom were residents of the Phoenix, Arizona, metropolitan area. This sample was obtained under 2 interrelated Corazón Life Journeys studies designed to examine lifetime acculturative and socioeconomic life journey changes in Latinos and other groups within the Southwest. We used purposive nonprobability sampling31 to obtain a diverse sample of community residents that had broad variation in levels of resilience32 and in levels of acculturation.
Members of the research team administered a 2.5-hour in-depth, structured interview that included sections on early life experiences, cultural beliefs and attitudes, mental health and health indicators, and background information. Our health-related outcomes of interest were life satisfaction, exercise behaviors, and healthy and unhealthy dietary behaviors. Under this ecodevelopmental framework, earlier life events were regarded as precursors of specific health outcomes in adulthood.16,33
A structured interview assessed life events at each of 4 life milestones: (1) the elementary school years (age 8 to 10 years), (2) the middle school years (age 11 to 13 years), (3) the high school years (age 14 to 18 years), and (4) current adulthood. For each life milestone, a memory induction procedure was used to focus attention and enhance recall accuracy. After the interview the interviewer conducted a memory rating to rule out cases of unreliable recall.
Level of education was measured by a single item with 5 levels: from 1=8th grade or less, to 5=completed college (Table 1).
At each life milestone, a single 5-level item was used to assess the socioeconomic status of the respondent’s home neighborhood. We elicited the location of the participant’s hometown and community, then asked the respondent to rate, “the type of neighborhood in which you and your family lived.” The response choices were: 1=poor: a lower-class, ghetto, or barrio neighborhood; 2=low income: a lower-middle-class neighborhood; 3=middle class: a middle-class neighborhood; 4=upper-middle income: an affluent neighborhood; and 5 = high income: a wealthy, elite neighborhood.
The acculturation measure consisted of a 5-item scale used in prior acculturation research.34,35 These 5 acculturation items assessed: (1) self-reported Spanish–English skills in speaking (from 1=Spanish only to 5=English only); (2) self-reported Spanish–English skills in reading (from 1=Spanish only to 5=English only); (3) best friends (from 1=almost all Latinos or other ethnic minority persons to 5=almost all White Americans); (4) neighbors (from 1= almost all Latinos or other ethnic minority persons to 5=almost all White Americans); and (5) the language of television (or radio) shows watched or listened to (from 1=Spanish only to 5=English only). Higher acculturation scale scores indicated higher levels of acculturation to the “mainstream” White American culture. For this sample, the acculturation scale’s Cronbach α reliability coefficient was 0.87 for the elementary school, middle school, and high school milestones, and it was 0.80 for acculturation at the adult milestone.
Items from the Lifestyle Survey Health Assessment Inventory assessed specifically operationalized health behaviors and psychological aspects of quality of life.35 These exercise and food frequency items and scales were developed under several prior lifestyle studies.34,36,37 These units of dietary and exercise behavior were not designed to measure dietary calories or metabolic energy expenditure. They were designed to measure discrete operationalized behaviors and behavior change by using common units of lifestyle activity.35,36
The Life Satisfaction Scale consists of 10 items that assess an overall sense of happiness in relation to one’s current life satisfaction.35 All items elicit responses to the question, “How satisfied are you with having….” Two example items ask about life satisfaction with “the ability to overcome life’s problems,” and “clear life goals and a direction in life.” Items were rated on a dimension ranging from1=not at all satisfied to 5=extremely satisfied. For this sample, the Cronbach coefficient α was 0.92.
The Exercise Scale is a 4-item behavioral scale of frequencies for 4 specific types of exercise behaviors measured in units of exercise per week.35 Units consist of activity episodes lasting 15 minutes or more for: (1) light aerobic exercise (e.g., walking), (2) stretching exercises to increase flexibility, (3) moderate aerobic exercise (e.g., jogging, swimming), and (4) weight-bearing exercises (e.g., push-ups, weight lifting to build muscles). For this sample, these behavior frequency units exhibited additive properties, yielding a Cronbach coefficient α of 0.71.
The Healthy Dietary Behaviors Scale (fruits, vegetables, and fiber) is a 5-item scale that measures frequency in units per week in consuming healthy foods: (1) a salad or raw vegetable; (2) a fruit (e.g., apple, orange, pear, peach); (3) a serving of whole-grain wheat breads or cereal (e.g., bran, not refined white dough); (4) a portion of dark green and leafy vegetables (e.g., broccoli, spinach); and (5) a portion of white meat (e.g., 6 oz or more of fish, chicken).35 For this sample, these behavior frequency units exhibited additive properties, yielding a Cronbach coefficient α of 0.70.
The Unhealthy Dietary Behaviors Scale (fat and high-sugar foods) is a 7-item scale that measures frequency in units per week in consuming unhealthy foods: (1) a food cooked or fried in lard or containing grease (e.g., french fries, potato chips); (2) butter, mayonnaise, or oil added to food; (3) a sweet snack (candy bar, chocolate, jelly beans); (4) a soft drink containing sugar (e.g., Coke, Pepsi); (5) a pastry (e.g., a doughnut, slice of cake, or pie); (6) ice cream or a dessert (a cone, a bowl); (7) a serving of red meat (e.g., 6 oz or more of beef, pork, hamburger, hot dog, steak).35 For this sample, these behavior frequency units also exhibited additive properties, yielding a Cronbach coefficient α of 0.72.
We used a conditional GMM to identify the optimal number of assimilation latent classes based on intercept and slope latent factors for socioeconomic status and for level of acculturation, along with the conditional covariates of immigration and education (Figure 2).29,30 In this model the socioeconomic intercept and slope latent factors were identified from the 4 neighborhood socioeconomic variables, as assessed at each of the 4 life milestones. Similarly, the acculturation intercept and slope latent factors were identified from the 4 levels of acculturation, as assessed at each of the 4 life milestones. The intercept loadings were fixed at 1, and the slope loadings were fixed at 0,1, and 2, respectively, as proscribed by Bollen and Curran.29 Each adult loading was freed (which is indicated within the MPlus program as an asterisk) because it was a parameter to be estimated as a result of variability across persons in the period of time from high school to adulthood milestones.
The concept of segmented assimilation suggests that assimilation is a product of 2 interrelated factors: socioeconomic mobility (upward or downward) and acculturative integration into a host society (i.e., toward or away from American “mainstream” culture). Accordingly, we developed a 2-factor socioeconomic–acculturative conditional GMM that operationalizes intercept and change latent growth factors to model temporal assimilation effects, as constrained by 2 covariates: immigration status (native-born, immigrant) and level of education. We used MPlus version 4.1 (Muthen and Muthen, Los Angeles, CA) to define this GMM and to conduct the latent class analyses.
The primary study objective was to use this GMM to identify an optimal number of latent classes that represent distinct assimilation trajectory groups. We anticipated that at least 1 of these trajectory groups would define an upward assimilation trajectory, and that at least 1 would define a downward segmented assimilation trajectory.
From this GMM model, a 4-class optimal solution was identified on the basis of the minimization of the relative fit statistics: the Akike Information Criterion, the Bayesian Information Criterion, and the Adjusted Bayesian Information Criterion23 (Table 2). For these log-likelihood fit statistics, Hanson et al. indicated that “smaller values denote better models,” that is, the best fit or optimal solution.23(p205)
We then labeled these classes: (1) group 1, extreme upward assimilation; (2) group 2, extreme downward assimilation; (3) group 3, moderate upward assimilation; and (4) group 4, moderate downward assimilation (Figure 3). These assimilation trajectory groups exhibited the following demographic characteristics. For group 1 (n=53), the mean age was 43.73 years (SD=12.66), with 83% US native-born (17% immigrant), and with an ethnicity profile of 56.6% Latino, 30.2% non-Hispanic White, and 13.2% other. For group 2 (n=21), the mean age was 35.42 years (SD=7.16), with 95.2% US native-born (4.8% immigrant), and with an ethnicity profile of 42.9% Latino, 28.6% non-Hispanic White, and 28.6% other. For group 3 (n=130), the mean age was 37.82 years (SD=11.16), with 78.5% US native-born (21.5% immigrant), and with an ethnicity profile of 56.2% Latino, 33.1% non-Hispanic White, and 10.8% other. For group 4 (n=54), the mean age was 38.85 years (SD=8.93), with 83.3% US native-born (and, thus, 16.7% immigrant), and with an ethnicity profile of 51.9% Latino, 37.0% non-Hispanic White, and 11.1% other. It is noteworthy that none of these groups consisted of a single ethnic or nativity group; ethnicity and nativity were distributed throughout these trajectory groups.
Because of the relatively small sample sizes for some of the trajectory groups, we could not conduct a multiple group structural equation model analysis. Instead, we conducted 4 one-way analyses of variance to examine group mean differences on the 4 health outcomes of interest: life satisfaction, exercise, healthy diet, and unhealthy diet (Table 3).
One of these analyses of variance showed an omnibus group difference across assimilation trajectory groups for the Life Satisfaction Scale (F3,254 =2.99; P <.05). Posthoc contrasts revealed that group 1, the extreme upward assimilation group, had a significantly higher level of life satisfaction compared with group 4, the moderate downward assimilation group. Also, the omnibus test for the Unhealthy Diet Scale was significant (F3,254 =4.40; P <.01). Here, group 1, the extreme upward assimilation group, exhibited a lower frequency of unhealthy food consumption relative to group 2, the extreme downward assimilation group, and also relative to group 3, the moderate upward assimilation group.
Regarding assimilation group changes across the life milestones, the extreme upward assimilation trajectory group (group 1) exhibited a steady linear increase in acculturation across time. By contrast, the extreme downward assimilation group (group 2) exhibited a net downward acculturation trajectory away from mainstream American culture. Also, the moderate upward assimilation trajectory group (group 3) exhibited no acculturative change until the transition from high school to adulthood, then exhibited an increase toward mainstream American culture. Finally, the moderate downward assimilation trajectory group (group 4) exhibited a stable level of acculturation from childhood through adolescence, whereas from high school to adulthood it exhibited a decrease in acculturation—in other words, an increasing orientation toward Latino/Mexican culture.
In accord with hypothesis 1, the extreme upward assimilation trajectory group (group 1) exhibited the highest level of life satisfaction. Conversely, for hypothesis 2, the extreme downward assimilation group (group 2) exhibited the lowest level of life satisfaction. Thus, hypotheses 1 and 2 received support regarding life satisfaction. Some supportive evidence involving a lower frequency of unhealthy eating was also observed for the extreme upward assimilation trajectory group (group 1), although no effect was observed for exercise or for healthy eating. In summary, upward socioeconomic and acculturative mobility (extreme upward assimilation; group 1) were associated with enhanced life satisfaction and with 1 adaptive health behavior—a lower frequency of unhealthy food consumption, suggesting efforts to avoid unhealthy foods. This may be the first study that assesses and models the temporal process of lifetime assimilative change, and relates these trajectories to selected indicators of quality of life.
The use of retrospective recall to identify patterns of socioeconomic status and acculturation at early life milestones might be a study limitation, as these indicators are subject to memory distortion. To reduce memory distortion, we elicited responses under a structured interview and with a memory induction procedure. Also, we recognize that a retrospective study is not as potent as a long-term prospective study. Nonetheless, the present results based on this retrospective methodology offer insights regarding variations in segmented assimilation and their health-related outcomes, whereby under a conventional long-term prospective study attaining such results would require data gathering over a period of 20 years or more.
Another limitation of this study is the relatively small sample size as applied to the analysis of growth mixture and latent class models, and for the identification of a stable model that required multiple iterations and bootstrap draws.38 Future studies would benefit from having a larger sample that includes female participants, especially because women have exhibited more pronounced patterns of acculturative change.39 The absence of female representation in this sample is another limitation, whereby these findings can only be generalized to men having similar characteristics.
This study presented an ecodevelopmental field model of socioeconomic and acculturation changes to identify distinct lifetime segmented assimilation trajectory groups, as related to life satisfaction, exercise, and dietary behaviors. These findings suggest that certain health outcomes in adulthood are influenced in predictable ways by lifetime segmented assimilation trajectories.
This study also suggests that “culture counts,”40 whereby lifetime trajectories of socioeconomic and acculturative change matter as influences on certain health-related outcomes.33,41,42 Many immigrants and persons in racial/ethnic minority groups will develop aspirations for success in pursuit of the American Dream. Some succeed, whereas others fail and drift downward in socioeconomic status and in separation from mainstream society. Moreover, White Americans raised within affluent environments can also suffer economic downturns, drug addiction, and other losses that result in descending assimilation trajectories.
Developing human and social capital early in life may promote upward assimilation trajectories and contribute to healthy outcomes later in life.43,44 This study also reveals that the transition from late adolescence to young adulthood is an especially critical period, presenting both risks and opportunities. Accordingly, culturally relevant health promotion and disease prevention interventions designed for young adults at this crucial life stage may focus on building human and social capital via life-skills training, strengthening social networks, and building community support systems.45 A general aim is to facilitate transitions into upward assimilation trajectories to promote life satisfaction and healthy behaviors in efforts to reduce or eliminate health disparities.16,42
This article was supported by the National Center on Minority Health and Health Disparities (grant P20 MD002316-010003, Felipe González Castro, principal investigator, and grant P20 MD002316-01, Flavio F. Marsiglia, principal investigator). The content is solely the responsibility of the authors and does not necessarily represent the official view of the National Center on Minority Health and Health Disparities or the National Institutes of Health. We also thank the Institute for Mental Health Research, Phoenix, Arizona, for its support of one of the Corazón studies.
Reprints can be ordered at http://www.ajph.org by clicking the “Reprints/Eprints” link.
Human Participant Protection
This study was approved by the institutional review board of Arizona State University. Also, a federal certificate of confidentiality was obtained to protect the confidentiality of information from drug-using respondents.
ContributorsF. G. Castro originated this study and conducted the data analyses and major writing. F.F. Marsiglia provided conceptual and editorial support. S. Kulis provided methodological and substantive feedback and writing. J.G. Kellison provided editorial feedback and some writing.
Felipe González Castro, Department of Psychology, Arizona State University, Tempe.
Flavio F. Marsiglia, School of Social Work at Arizona State University, and the Southwest Interdisciplinary Research Center, Phoenix, AZ.
Stephen Kulis, Southwest Interdisciplinary Research Center and the School of Social and Family Dynamics, Arizona State University.
Joshua G. Kellison, Graduate student in the clinical training program of the Department of Psychology, Arizona State University.