|Home | About | Journals | Submit | Contact Us | Français|
Dietary intake is an important determinant of obesity and numerous chronic health conditions. A healthful diet is an essential component of chronic disease self-management. Researchers have indicated that the healthfulness of the Latino diet deteriorates during the acculturation process. However, given the many operationalizations of acculturation, conclusive evidence regarding this relationship is still lacking. This comprehensive and systematic literature review examines the relationship between acculturation and diet by examining national, quantitative, and qualitative studies involving Latinos living in the United States. Studies of diet included those that examined dietary intake using one of several validated measures (eg, food frequency questionnaire, 24-hour dietary recall, or dietary screener) and/or dietary behaviors (eg, away-from-home-eating and fat avoidance). Articles were identified through two independent searches yielding a final sample of 34 articles. Articles were abstracted by two independent reviewers and inter-rater reliability was assessed. Analyses examined the extent to which various measures of acculturation (ie, acculturation score, years in the United States, birthplace, generational status, and language use) were associated with macronutrient intake, micronutrient intake, and dietary behaviors. Several relationships were consistent irrespective of how acculturation was measured: no relationship with intake of dietary fat and percent energy from fat; the less vs more acculturated consumed more fruit, rice, beans, and less sugar and sugar-sweetened beverages. Additional observed relationships depended on the measure of acculturation used in the study. These findings suggest a differential influence of acculturation on diet, requiring greater specificity in our dietary interventions by acculturation status.
Dietary intake is an important determinant of obesity and numerous chronic health conditions among Latinos living in the United States. A healthful diet is an essential component of chronic disease self-management. Most research concludes that Latinos in the United States consume a less healthful diet compared with other racial/ethnic groups in the United States due, in part, to less access to healthful foods, food insecurity, and low socioeconomic status. One of the social mechanisms that appear to link poor health practices and risk for obesity and certain chronic health conditions are the migration and acculturation processes. The effects of migration are often examined in terms of the individual’s country of origin, age of arrival, and years living in the United States. Acculturation is a bidimensional process in which individuals may learn and/or adopt certain aspects of the dominant culture and in some cases retain most or some aspects of their culture of origin (1–3). Previous reviews on the relationship between acculturation and diet concluded that greater acculturation to the way of life in the United States is associated with less healthful dietary intake and dietary behaviors (4–6). However, the published reviews were either not inclusive of all of the studies on this topic or were focused on a specific target population (eg, individuals with type 2 diabetes) (5). In addition, there is research that indicates that being less acculturated is associated with poorer dietary habits (7–9), confirming what previous authors have suggested about the complexity of measuring the process of acculturation and its implications for understanding health and health behaviors (10,11).
This study represents a comprehensive and systematic literature review of the association between acculturation and diet among Latinos living in the United States. To reflect the varied methods in which migration and acculturation are measured, all operationalizations were examined, including single measures of migration history (eg, number of years living in the United States) to multidimensional and bidirectional measures of acculturation like the Cuellar Acculturation Rating Scale for Mexican-Americans (2). Studies of diet included those that examined dietary intake using one of several validated measures (eg, food frequency questionnaire, 24-hour dietary recall, and dietary screener) and dietary behaviors (eg, away-from-home-eating, food preparation and shopping techniques, and fat avoidance). The literature review focused exclusively on adults but included Latinos of all countries of origin in the Americas (ie, North, Central, and South America) to examine for subgroups differences by level of acculturation. However, studies were limited to populations currently residing in the United States.
Based on previous research with the target population, it was expected that greater acculturation to the US culture would have a negative influence on the Latino diet. The extent to which this was moderated by the operationalization of acculturation and Hispanic subgroup were important components of this literature review, given the inconsistent findings. The discussion provides guidance on how health care practitioners can best encourage traditional healthful dietary practices among the different Latino subgroups in the United States using the principles of the US Department of Agriculture Food Guide Pyramid as a guiding tool without sacrificing flavor, traditional foods, and cultural traditions/preferences.
Using methods developed by Cooper (12), articles were identified for inclusion using three approaches: database literature searches, reviewing the references of articles selected for the review (also known as backward searching), and examining published reviews on the topic. Consistent with the multiple operationalization approach recommended for literature reviews (12), an initial search was conducted by the first author in Medline and PsycInfo using the following key terms for acculturation and diet: acculturation, culture, migration, years in United States, country of origin, generation status, nutrition, dietary intake, dietary behaviors, food frequency, food recall, food habits, food preparation, food preferences, and food shopping. A second search, conducted by the health sciences librarian at San Diego State University, was inclusive of several additional databases: CINAHL, Academic Search Premiere, Sociological Abstracts, and Web of Science. The searches were then limited to studies conducted with Latinos/Hispanics using the following key terms: Latino/a, Hispanic, Spanish-speaking, Chicano/a, Spanish, Mexican American, Puerto Rican, Latin immigrant, Central American, and South American. Literature searches were downloaded into EndNote (version 9.0, 2000, Thomson Scientific, Philadelphia, PA) to eliminate duplicates and to facilitate the review process. Both literature searches were conducted in June and July 2007 to ensure inclusion of the latest research.
Articles were included in the review if the abstract, title, or key words indicated that the authors examined or were interested in the influence of acculturation on diet. In a few cases, this was not clear. These articles were included to ensure that no possible study was omitted. Articles had to be published in a peer-reviewed journal between 1965 and the present (1980 to the present in Web of Science) and available in Spanish or English. Articles were excluded from the review process if they addressed some aspect of health or health behaviors not specific to diet (eg, obesity or diabetes), if the publication was a dissertation or nonpeer-reviewed publication (eg, newspaper article or report), or did not contain sufficient information for evaluation purposes (eg, conference abstracts).
Using Cooper’s Method-Description Approach (12), a data coding sheet was created to capture information pertinent to this review. Reviewers coded all information available in each article pertaining to the study’s implementation and evaluation. When discrepancies were observed between information presented in the text vs a table or figure, final coding was based on information provided in the table or figure. Similarly, studies with potential methodologic flaws and/or threats to internal validity were included in the review process to eliminate the need for reviewers to make judgments about study design during the review process. This decision also provided the opportunity to examine the relationship between study design issues and outcomes.
Each study’s sample size was noted, and its participants were characterized based on their sex, age and age range, income, employment, education, body mass index (BMI) or weight status, and country of origin/Latino ethnic subgroup.
Characteristics of a study included the study objectives, primary outcomes, the study design specific to the acculturation–diet analyses (cross-sectional or longitudinal), the geographic region of the study, and analytic approach.
Measures of migration and acculturation included generation status, language of assessment, years in the United States, age at arrival to the United States, and acculturation score, the latter inclusive of scales that measured acculturation on a unidimensional/unidirectional scale (eg, language used at home) to multidimensional and bidirectional scales (2).
Measures of diet included dietary intake assessed using 24-hour dietary recall methodology, a food frequency questionnaire, or a dietary screener; dietary behaviors included but were not limited to away-from-home eating, food preparation and shopping techniques, and fat avoidance; and dietary stages of change.
Several strategies were used to ensure accurate coding of information onto the coding sheets. First, two reviewers reached consensus on the appropriate use of the coding sheet. This process involved discussions about each variable; testing the coding sheet with two articles, comparing results, addressing discrepancies, and revising the coding sheet as needed until adequate inter-rater reliability estimates were obtained between the raters. All articles were then double-coded for verification purposes, with discrepancies addressed using a consensus approach (12). Inter-rater reliability on a subsample of the articles yielded results that ranged from 0.70 to 0.99, with more errors observed on coding of the acculturation variable used in primary analyses, and fewer errors observed on demographic variables. All discrepancies were resolved and consensus reached.
Results of the literature search are outlined in Figure 1. Two separate searches and backward searches from identified articles yielded 54 articles describing the acculturation–diet relationship. Five (10%) studies were excluded from the review process because they themselves were review articles (4–6,10,11). The other studies were excluded because they examined the outcomes of a poor diet (eg, obesity, diabetes, or hypertension) rather than on diet itself or they did not systematically measure acculturation (13–23). This yielded a final sample of 34 articles that underwent further examination. Although not included in this review, one study was notable in its examination of the relationship between acculturation, diet, and low birthweight; Cobas and colleagues (23) found that acculturation had a negative effect on dietary intake, and dietary intake had a direct effect on low birthweight status.
Tables 1 and and22 provide detailed information on the relationship between diet and the various operationalization(s) of acculturation. Given our specific interest in the relationship between diet and acculturation among Latinos, non-Latino subsamples were excluded. The following information is presented in the tables: sample characteristics (eg, number, sex, age, income, and education); ethnicity, acculturation, and final acculturation score(s) used in primary analyses (eg, ethnic subgroup, generation status, language of assessment, years in the United States, age at first arrival, and acculturation); operationalization(s) of diet; and results. Information on employment status and BMI were abstracted but excluded from the Tables given that <50% of studies reported these characteristics. Information in the Tables is organized first by year of publication and then alphabetically by first author, the former to show progression in the science.
Table 1 presents the results based on five studies using Hispanic Health and Nutrition Examination Survey and National Health and Nutrition Examination Survey data, representing the most generalizable findings available (24–28). Irrespective of the samples included and the operationalization of acculturation used, being less acculturated was associated with more healthful levels of nutrient consumption. For example, Dixon and colleagues (24) found that a higher percentage of Mexican-born men and women, compared with US-born English and Spanish-speaking Mexican/Mexican-American men and women, were more likely to meet dietary guidelines for total fat, saturated fat, fiber, and potassium, as well as consume the recommended dietary allowance of vitamin C, vitamin B-6, folate, calcium, and magnesium. Based on reported consumption in the past month, Mexican-born men and women consumed more fruit, vegetables, fruit juice, grains and legumes, as well as fewer salty snacks, desserts, and added fats than US-born Mexican/Mexican-American men and women. The only exception was observed among Cuban Americans where no relationships were observed between acculturation and diet (24,26). Two studies also demonstrated no relationship between acculturation and total energy among women (25,27).
An examination of the 24 site-specific quantitative studies identified several important findings related to the study design, Latino subgroup, measurement of acculturation and diet, and their relationship.
Most studies (71%) were conducted in the Southwest region of the United States, including 13 in California (see Figure 2). As such, it is difficult to draw generalizable conclusions about the relationship between acculturation and diet given that acculturation is influenced, in part, by contextual factors (eg, density of the Latino population in a given neighborhood, history of migration into the region, and influence of social network members who are also Latino) (29–31), as is diet (32–34).
Fifty percent of the studies collected data from both men and women. This is a much larger percentage of studies than anticipated given the perceived dearth of studies on Latino men’s lifestyle behaviors. All but six studies (35–40) involved primarily a young adult to middle-age population, consistent with the age distribution of the Latino population in the United States (41). However, this also points to the lack research among Hispanic elders, a concern given the continued growth of this population (42), and the fact that they are more likely to be living in poverty (43). Markers of socioeconomic status suggest that researchers have primarily focused their efforts on a lower socioeconomic population. For example, among studies that reported education based on the percentage of sample who did not complete high school, the lowest number was 29% (58) and the highest number was 83% (53). Although not included in the tables given the limited findings, seven studies reported that the mean BMI ranged from 27 to 29.7, indicating that the average BMI was in the overweight range (35,36,40,46–49). In addition, two studies reported percent overweight/obese at 59% (45) to 77% (37).
Four studies did not report the percent of individuals representing different Latino subgroups, although three of these studies involved residents of California who are, according to US census data, primarily Mexican/Mexican American (49–51) and one involved residents of New York and Connecticut who are predominantly Puerto Rican (52). Three studies reported country of origin in global terms (foreign- vs US-born) (35,53,54). Eleven studies involved primarily Mexicans/Mexican Americans (36,37,40,44–47,55–58) or Puerto Ricans (48,59), precluding the possibility of examining Latino subgroup differences. Of the four remaining studies that included multiple Latino subgroups, analyses did not examine the relationship between acculturation and diet by subgroup (38,39,60,61).
All studies published in the 1990s used either the Cuellar and colleagues (62) or the Marin and Gamba (3) scales to measure acculturation. By the year 2000, studies began to examine several dimensions of acculturation including birthplace, years in the United States and age of arrival to the United States. However, a less desirable trend during this era was the use of project-specific scales to measure acculturation, limiting our ability to draw comparisons across studies.
A strength of this review is the finding related to dietary assessment method used. In 18 of24 (75%) studies, dietary intake was measured using a food frequency questionnaire or a 24-hour dietary recall. The remaining studies examined dietary behaviors, including fat avoidance (50,51,61) and fat/fiber-related behaviors (47,54,55).
Table 3 summarizes the results regarding the relationship between acculturation and diet. Across several studies, the evidence is fairly conclusive that dietary fat and percent energy from fat are not associated with any measure of acculturation. The one exception observed was in a study by Monroe and colleagues (40), which found that second-generation respondents consumed more monounsaturated fat than first-generation Mexican immigrants. This may be explained by the additional pattern of findings: more acculturated individuals consume more fast food, fatty snacks, and added fats than less acculturated individuals; however, they were also more likely to engage in fat avoidance behaviors. Compounding this is evidence that less acculturated individuals were more likely to consume whole milk and more likely to fry food with lard and other meat fat than their more acculturated counterparts. There is also strong evidence that acculturation is not associated with dietary cholesterol intake. One study provides striking evidence that contradicts the rest of the literature (40). The relationship between acculturation and dairy and meat depend on the measure of acculturation.
The evidence regarding fiber intake and acculturation is equally mixed. Fruit, rice, and beans are negatively associated with acculturation (less acculturated eat more fruit/rice/beans than more acculturated individuals), as are reports of total fruit and vegetable consumption. However, there is mixed evidence on the relationship between acculturation and fiber, and this relationship varies by measurement of acculturation.
One’s acculturation score was not associated with total energy consumed; however, being born in the United States and living more years in the United States were associated with less energy intake. Nevertheless, those who spoke English and had lived in the United States for more years consumed more sugar, including sugar-sweetened beverages. For the most part, acculturation was not associated with micronutrient intake. The only variable that seemed to differentiate individuals was generation status/birthplace; foreign-born individuals consumed more vitamins A, C, and E; folate; calcium; and zinc compared with their counterparts.
Three studies are not included in Table 3 because they do not contain acculturation and/or dietary variables that are comparable with other studies (39,47,53). However, the results are consistent with those contained in Table 3. For example, diet quality was negatively associated with age of arrival and years in the United States (53); pregnant women who arrived to live in the United States at an earlier age and who lived in the United States for fewer years had a better quality diet than their counterparts. Lin and colleagues (39) determined that people who consumed fruit and cereal were more acculturated than those who consumed rice in the past 3 months. In a study conducted with Mexican/Mexican-American women in San Diego (47), more acculturated women were more likely to engage in several healthful dietary behaviors; however, they also were more likely to eat out for lunch and dinner, and to eat at fast-food restaurants than the less acculturated individuals.
Five qualitative studies were identified that provided support for the findings in this systematic review (63–67). Most of these studies were conducted within the past 6 years and involved focus groups and interviews with members of the target population. In a study involving 18 middle-aged Latina women originating from Mexico, Central America, and South America and now living in Mississippi, Gray and colleagues (65) found that 61% of women who completed the interview said that they generally ate more healthfully in their country of origin. Since arriving in the United States reductions were reported in fish, seafood, bread, and corn. Increases were reported in vegetables, pork, chicken, hamburgers, and pizza. No changes were reported in milk, beef, rice, potatoes, beans, and eggs. In a second study involving 23 Honduran women living in Louisiana (63) and a third study involving 117 Hispanic migrant workers in Pennsylvania (67), increases were noted in consumption of hamburger meat and vegetables.
The five qualitative studies also shed light on possible mechanism for the diet–acculturation relationship. Purchasing of fast food occurred more frequently in the United States (63,65,67) despite reports that fast-food restaurants were perceived to be equally accessible in their home countries (65). In Honduras, 52% of the women reported never eating fast food, whereas 100% reported consuming fast food, including nearly 50% who reported consuming fast food every weekend (63). Lack of time due to work obligations explained these findings (63). Three quotes illustrate these findings:
“When one works outside the home, yes, you go to eat hamburgers, fried chicken, and fast food” (63).
“In Mexico, (women) did not have to work, but here you have to work. Because of this, sometimes you have to buy easy things to give to the children. You arrive from work tired. In Mexico, you attended to your kids more. Here there is not time” (65).
“If we’re somewhere and they’re hungry … how can I not? You know, we can stop off at [fast-food restaurant] and get a cheeseburger and french fries. It’s like a dollar” (66).
Whether an immigrant consumes fewer fruits and vegetables after emigrating to the United States appears to depend, in part, on availability in their home countries and changes in income status following immigration. Fresh foods were perceived as more readily available in their home countries than in the United States (67). For example, Gray and colleagues (65) found that 67% (n=12) of women said they bought at least some of their fresh food from street markets and 61% (n=11) said that food was fresher and more natural in their home country compared with the United States. These fresh foods also were perceived as tastier in their home countries than in the United States (65–67). However, availability appears to vary by country of origin. In Honduras, women reported that there was little access to fresh vegetables (63). Compounding this issue was access to familiar products in the United States. Changes were attributed to the unavailability of familiar food products in United States (65,67), and increases in the availability of unfamiliar food products, specifically fresh fruits and vegetables (64). In focus groups with 79 Latina women living in Arizona, the women “saw an assortment of vegetables and fruits in their grocery stores but did not know how to prepare them” (64). They were less likely to purchase unfamiliar products because they were concerned about wasting money. Finally, the dietary acculturation process of their children appears to be a major source of influence in the parents’ dietary behaviors and intake. Family dietary choices are affected by the children’s desire to eat out and requests for “Americanized” food (64–67).
This review sought to examine the relationship between acculturation and diet among Latinos living in the United States. Researchers are growing increasingly interested in understanding the influence of acculturation on health behaviors and health outcomes (11,68). Some researchers are trying to disentangle which acculturation variable is most relevant to health (54); others argue that acculturation may not be the most appropriate variable to consider in understanding Latino health disparities (31). This review attempted to bring some clarity to the issue by examining the relationship between acculturation and diet using multiple operationalizations of acculturation and rigorous methods for capturing diet.
Several relationships were consistent irrespective of how acculturation was measured. First, there was no relationship between acculturation and dietary fat intake or percent energy from fat, despite evidence that fat-related behaviors seem to differ between those who are less or more acculturated. Those who are less acculturated consume more whole milk and use fat in food preparation, whereas the more acculturated consume more fast food, snacks, and added fats. Second, less vs more acculturated individuals consumed more fruit, rice, and beans. Third, less acculturated individuals consumed less sugar and sugar-sweetened beverages than more acculturated individuals. Some of our findings support the conclusions drawn by Perez-Escamilla and Putnik (5), whereas other findings are directly opposed. They indicated that “The process of acculturation among Latinos is associated with suboptimal dietary choices, including lack of breast-feeding, low intake of fruits and vegetables, and a higher consumption of fats and artificial drinks containing high levels of refined sugar” (p 867). Our conclusions also differ somewhat from the conclusions drawn by Satia-Abouta and colleagues (6) who wrote that “Overall, most of the studies found some statistically significant associations of levels of acculturation with diet. Unfortunately there was no consistent direction of effect between level of acculturation and dietary intake” (p 1116). The latter review involved only nine articles making it difficult to tease out this complicated relationship. In a more general review on the topic, Lara and colleagues (11) noted that “More acculturated Latinos (those who are highly acculturated) are more likely to engage in substance abuse and undesirable dietary behaviors and experience worse birth outcomes compared with their less acculturated counterparts.” (p 374). When one considers behaviors such as fast-food consumption, sugar-sweetened beverage consumption, and fruit consumption, our findings support those of Lara’s; however, the less acculturated individuals also used meat fat to prepare foods. Overall these findings suggest a differential influence of acculturation on diet, requiring greater specificity in our dietary interventions by acculturation status.
The results of this review must be considered in light of its limitations. Although some studies examined multiple cohorts to assess the relationship over time, none of the studies were longitudinal in design severely limiting conclusions that can be drawn about the relationship between acculturation and diet. Second, Lara and colleagues (11) indicated the need to examine whether the relationship between acculturation and diet differs by Latino subgroup. Unfortunately, few studies provided sufficient information to draw any meaningful conclusions. A current research initiative, led by the National Heart, Lung, and Blood Institute, addresses these two concerns. The Hispanic Community Health Study (http://www.cscc.unc.edu/hchs/) will collect data from 16,000 Latinos living in the United States, including their dietary intake using rigorous methods of dietary data collection (ie, 24-hour dietary recalls, a food propensity questionnaire, and biomarkers); it is longitudinal in nature. Selected measures of acculturation and migration history are being collected, although limited given concern for participant burden. All Latino subgroups will be represented with the study sites located in San Diego, Chicago, Miami, and the Bronx. The four groupings of Latinos are Cubans (mostly in Miami), Puerto Ricans and Dominicans (mostly in the Bronx), Mexican Americans (mostly in San Diego and Chicago) and Central/South Americans (mostly in Chicago and Miami).
The strength of this review is that it is more comprehensive (eg, studies range from national studies using National Health and Nutrition Examination Survey data to qualitative studies involving small samples) and focused (adults only; dietary intake or behaviors and not attitudes and beliefs; and within-group analysis vs non-Hispanic whites as a healthy reference group) than previous reviews given the study’s inclusion and exclusion criteria. For example, Perez-Escamilla and Putnik (5) reviewed 16 studies; 11 of their studies were included in this review (duplicates), four were excluded because they involved an adolescent population, and one was excluded because the focus was on breastfeeding. Satia-Abouta and colleagues (6) identified nine studies involving Hispanics, all of which were duplicates in this review. In their comprehensive review of acculturation and a variety of health behaviors, Lara and colleagues (11) reviewed 10 studies on diet; eight were duplicates, one involved adolescents, and one was new (11). The Benavides-Vaello review (4) provided the greatest number of previously unidentified studies: five of the 11 studies were new, three were duplicates, and three involved adolescents.
Future studies should examine this relationship in other geographic regions of the United States and with a more socioeconomically diverse Latino population. The fact that most of these studies took place in California is striking. Given the growing body of evidence of the influence of the neighborhood environment on the acculturation process and dietary intake, more research is clearer needed in emerging Latino communities (eg, Georgia and North Carolina). Second, acculturation is associated with improvements in socioeconomic status, including more education, better jobs, and more income. Although the poor are at significantly greater health risk due to disparities in the environment (69,70), access to quality care, and lower incomes, this review highlights the need for more research on the health of middle-class Latinos. The food marketing industry is aware of this subpopulation. Our public health efforts should not be far behind.
Culturally competent care targeting healthful lifestyles, whether for prevention or management of a chronic condition like diabetes, needs to recognize the commonalities across Latino subgroups, as well as those aspects that set them apart. Consistent with other collectivistic societies, Mexicans, Puerto Ricans, and Cubans alike emphasize the family to a greater extent than non-Hispanics in the United States. This does not seem to change with the acculturation process, although social networks may become less dense over time and generations. Families influence our food environment and what we eat (32). Second, by definition, Latinos share a fairly common language, a bond that extends from interpersonal to mass media communication. It affects the ability to communicate and receive information from health care providers. Initiatives to develop linguistically appropriate interventions, as well as improve the language skills of health care providers, have the potential to affect positive changes in health. Messages to less acculturated Latinos may need to stress maintenance of healthful lifestyle behaviors such as eating recommended qualities of vegetables, portion control to reduce energy intake, and changes to food preparation practices; the more acculturated may benefit from messages that stress moderation of fast food, sugar-sweetened beverages, and other away-from-home foods. In the United States, the common language shared by all Latino subgroups helps to bridge two cultures—the one that is retained and built upon in the United States based on one’s country of origin and the dominant culture of the individual’s community. Practitioners need to build into this common language the need for healthful eating habits.
This review was funded by the Grain Foods Foundation. Additional support was provided to the first author by the National Cancer Institute, grant no. R21 CA120929-01, and the American Cancer Society, grant no. RSGPB 113653.
GUADALUPE X. AYALA, Division of Health Promotion, Graduate School of Public Health.
BARBARA BAQUERO, Joint Doctoral Program in Behavioral Science, Center for Behavioral and Community Health, San Diego State University, San Diego, CA.
SYLVIA KLINGER, Hispanic Food Communications, Inc, Hinsdale, IL.