|Home | About | Journals | Submit | Contact Us | Français|
To compare total fruit and vegetable intake in cup equivalents (FVC) and its individual components among Hispanic subgroups in California.
Data are from the adult portion of the 2005 California Health Interview Survey. Hispanic/Latino subjects (n = 7954) were grouped into six subcategories (Mexican, Central American, Caribbean, Spanish American, South American, > 1 group). FVC was estimated from frequency responses about seven food categories. T-test and chi-squares were estimated to assess differences in sociodemographic characteristics across Hispanic subgroups. Multivariate linear regressions using SUDAAN were conducted to obtain means of FVC and its components by Hispanic subgroups controlling for confounders.
Hispanic subgroups did not differ in their intake of total FVC (mean cups = 3.4 and 2.9 for men and women, respectively). Small but significant differences (p<0.01) were found across Hispanic subgroups in individual FVC components (green salad (women only), cooked dried beans and non-fried white potatoes) after adjusting for potential sociodemographic and acculturation confounders.
Hispanic FVC intake did not meet the national recommendation, although their reported intake is higher compared to other race/ethnicity groups. The public health message remains the same: to increase FVC. Examination of intake for subgroups of Hispanics may enhance the utility of dietary information for surveillance, program and message design, and intervention and evaluation.
Hispanics as a group is the fastest growing demographic group in the United States (US); it is estimated that by 2050, 25% of the population in the US will be Hispanic (1). The Office of Management and Budget uses the terms ‘Hispanic’, ‘Latino’ or ‘of Spanish Origin’ interchangeably; the classification is composed of any person of Cuban, Mexican, Puerto Rican, South or Central American or other Spanish culture or origin (2). This group, however, is not a homogeneous one; it includes people from different racial and socioeconomic backgrounds and cultural influences. Health behavior practices and morbidity and mortality incidences are also different (3–11). Furthermore, Hispanic subgroups may differ in immigration history and rate of acculturation to the US diet (12,13). However, data about dietary intake for Hispanic subgroups in the US are scarce; only a few studies have compared dietary intake among some Hispanic subgroups (6,12,14–17).
Prior research has noted that consuming a diet high in fruits and vegetables is associated with lower risk of chronic diseases including cardiovascular disease, diabetes and certain types of cancers (18). In previous surveys, Hispanics have reported higher intake of fruits and vegetables than other ethnic groups (19,20). However, none have examined potential differences in fruit and vegetable consumption across subgroups of Hispanics. This study uses data from the 2005 California Health Interview Survey (CHIS 2005) to examine potential differences in total and individual components of fruit and vegetable intake among Hispanic subgroups. The results will broaden our understanding of common and distinct dietary patterns, and inform design of public health nutrition programs for specific subpopulations within this heterogeneous group.
Data are from the adult portion of the CHIS 2005, a biennial, population-based random digit dial telephone survey of California’s population conducted between July 2005 and April 2006. The CHIS is based at the University of California at Los Angeles’ (UCLA) Center for Health Policy Research; it is conducted in collaboration with the California Department of Public Health, the California Department of Health Care Services and the Public Health Institute. The CHIS was designed to produce reliable estimates of various health parameters, including diet (questions on diet were partially funded by the National Cancer Institute (NCI)). The CHIS study protocols were approved by the UCLA and the NCI Institutional Review Boards.
The CHIS surveys are the largest health surveys ever conducted in any state. The total number of completed interviews among adults for the CHIS 2005 was 43,020; a response rate of 26.9% (49.8% screener completion rate times 54.0% adult interview completion) is comparable to other surveys conducted in California in the 2005–2006 time frame (21). CHIS 2005 methodology and sample design are described in detail elsewhere (22) (http://www.ucla.chis.edu).
One of the advantages of the CHIS 2005 is its ability to capture the rich diversity of the California population, including a large sample of Hispanic respondents (n=8036 in total). In order to recruit English and non-English speakers, the interviews were conducted in five different languages. About 10% of adult interviews were conducted in languages other than English. Of those self-identified as Hispanics/Latinos, 38.9% of the interviews were conducted in Spanish.
The CHIS asked respondents about their racial and ethnic background. Those who self-reported being Latino or Hispanic were then asked about their specific country of origin. For the purposes of this study, the authors grouped countries of origin by geographic regions as follows: Mexican (Mexican, Mexican American, Chicano), Central American (Salvadoran, Guatemalan, Costa Rican, Honduran, Nicaraguan, Panamanian), Caribbean (Puerto Rican, Cuban, other Caribbean origin), Spanish-American (from Spain or Spanish origin), South American (Colombian, Argentinean, Peruvian and other South American origin); and more than one of the other five Hispanic or Latino origin groups. Eighty-two Hispanic respondents who could not be categorized into one of the six groups were excluded from Hispanic subtype analyses (Hispanic group n=7954). Geographic regions tend to share similar socio-political history and cultures. Some social scientists may argue that Hispanics should be categorized by their immigration status and not by geographic region of origin (i.e. Cubans and Mexicans share immigration status, so they should be grouped together, and Puerto Ricans should be grouped with other American citizens such as Mexican American and Chicano) (23). In the CHIS, the authors found no statistical differences between the groups of Puerto Ricans and Cubans, and the groups of Mexicans, Mexican-Americans and Chicanos in sociodemographic variables of interest (education, income, BMI, percentage of smokers). Therefore, Puerto Ricans and Cubans were grouped together, as were Mexicans, Mexican-Americans and Chicanos.
Fruit and vegetable intake in cup equivalents (FVC) was estimated from responses about frequency of consumption of seven categories of fruits and vegetables (fruit; 100% fruit juice; green leafy or lettuce salad; cooked dried beans; French fries, home fries or hash browns; other potatoes; and other vegetables), using the following procedures. The reported frequency category for each individual food was converted to mean daily number of times consumed. Because portion size was not asked, external estimates of median portions (in cup equivalents) by gender and 10-year age group were applied. Generally, a cup equivalent of fruits is defined as 1 cup of fruit or 100% fruit juice or ½ cup of dried fruit; a cup equivalent of vegetables is defined as 1 cup of raw or cooked vegetables or vegetable juice or 2 cups of raw leafy greens (24). For each component, frequency of intake was multiplied by age-sex specific portion size. In order to estimate total FVC, these component estimates were summed, and regression coefficients (developed externally to reflect the relationship between the screener and 24-hour dietary recall), were applied (25). Estimates of intake for individual components and total fruits and vegetables are given in cup equivalents.
were included in the analyses either as categorical or as continuous variables. These included: age; gender; marital status (married, never married or other); annual household income; hours worked per week for those working outside the home; food security status (high food security, marginal food security, low food security, very low food security); percentage at or under 300% federal poverty level; years of education completed; smoking status (current smoker vs. not current smoker); body mass index (BMI = self-reported weight (kg)/self-reported height (m2)); days in poor health per month; physical activity (self-reported vigorous leisure activity for at least 3 days per week and 20 minutes per day or moderate leisure activity for at least 5 days per week and 30 minutes per day; moderate and vigorous activities were defined respectively as activities that make you breathe somewhat harder or much harder than normal); and number of children in household, as well as self-reported history of high cholesterol, heart disease, hypertension, diabetes and colo-rectal cancer. Variables indicating receipt of social services (i.e. Food Stamps, Social Security, Temporary Assistance to Needy Families and California Work Opportunity and Responsibility to Kids (TANF/CALWORKS), and Women, Infants, Children (WIC) Supplemental Nutrition Program) were also included, as well as information regarding acculturation (born in U.S. vs. not; number of years living in the US; language of interview (Spanish, English); and English proficiency (speak only English; speak English very well/well; do not speak English well or at all)).
Descriptive statistics were estimated to assess differences in socio-demographic characteristics and dietary intake across race/ethnicity and by Hispanic subgroup. SUDAAN (Survey Data Analysis, Research Triangle Institute, Research Park Triangle, NC) was used with jackknife replicate weights to account for the complex sampling design in computing variance estimates. For comparisons with national data, the authors also analyzed mean total FVC from the National Health and Nutrition Examination Survey which used 24-hour dietary recalls rather than a screener (NHANES 2003–2004 presented in Table 1). These data were also analyzed in SUDAAN using jackknife replicate weights. All analyses were weighted to obtain California (CHIS) or national (NHANES) estimates. Table 1 shows age-adjusted means of total FVC from NHANES 03–04 and CHIS by race/ethnicity. P-values were obtained using pair-wise comparisons between Hispanics and other ethnic groups.
Sociodemographic and health characteristics among the Hispanic subgroups are presented in Table 2. Because FVC consumption is related to both gender and age, further analyses tested for significant differences in selected characteristics (those that are continuous rather than categorical in scaling) among the groups after adjusting for gender and age. All variables that remained statistically significantly different across Hispanic subgroups were tested one by one in their association with FVC or its components. Variables that showed a significant association (p<0.05) with FVC and ethnic subgroups were included in the final multivariate regression models to obtain mean FVC and its components.
FVC was examined in the original scale as well as the transformed scale (using the square root in order to approximate normality); no notable differences were seen in the means of FVC, so cup equivalents on the original scale are presented in the tables for easier interpretation. Overall p-values for Hispanic groups were obtained and significant differences noted (p<0.05). Individual fruit and vegetable components that varied significantly across Hispanic subgroups (p<0.05) were further examined in pair-wise comparisons. Because of multiple comparisons, the Bonferroni correction was used to detect statistically significant differences (p<0.005).
Table 1 shows the means of total FVC from CHIS 2005. For comparisons with national data, Table 1 also presents results from NHANES 2003–2004. Due to sampling design, the Hispanic group of NHANES is predominantly Mexican-Americans, while Hispanics in CHIS included all Hispanic groups. Results from both surveys are generally consistent: Among women in both surveys, Hispanics reported higher mean total FVC than non-Hispanic Whites (CHIS p<0.001; NHANES p<0.01) or Blacks (CHIS p<0.05; NHANES p<0.001). Among men, in CHIS, there were no significant differences by race/ethnicity. Among men in NHANES, while there were no differences in reported intakes between non-Hispanic Whites and Blacks, and between non-Hispanic Whites and Mexican Americans, Mexican-American men reported higher (p<0.01) FVC consumption than non-Hispanic Black men.
Using CHIS 2005 data, there were significant differences (p<0.05) in demographic characteristics (education, income, below federal poverty level, high food security, number of hours worked per week, number of children in household, born outside the US, number of years living in the US for those foreign-born, and English proficiency); behavioral characteristics (smoking, BMI); use of income assistance programs (Food Stamps, WIC); and self-reported health conditions (days of poor physical health, diabetes) among Hispanic subgroups (Table 2). These relationships in ratio scaled variables remained after gender and age adjustment. Respondents of Mexican and Central American origins had lower education and income, as compared to the other groups; the subgroup of Mexican origin had the largest number of children in the household and the highest BMI. South American and Central American reported longer working hours per week. Caribbean respondents reported the most days in poor physical health and the largest number of years living in the US (not surprising, since the majority of respondents in this group were Puerto Rican, and Puerto Rico is considered as part of the US). Groups also differed (p<0.05) in prevalence of smoking and reported history of diabetes. Those of Mexican and Central American origins were more likely to be at or under the 300% federal poverty level, and report receiving WIC benefits. Respondents of Mexican and Central American origins were more likely (p<0.05) to have been born outside the US and report that they did not speak English well or at all.
Final models for analysis of differences in FVC among Hispanic subgroups were stratified by gender and controlled for potential confounders (age, income, education, BMI, smoking, born in the US). Further adjustment for English proficiency, language of interview or number of years lived in the US did not alter the results (Tables 3A and Table 3B). There were no differences in total FVC among Hispanic subgroups (p=0.86 for females and p=0.24 for males) (Table 3A and 3B). However, an examination of the components of FVC suggests differences by gender and across these subgroups, even after adjusting for potential confounders (age, income, education, BMI, smoking and born in the US). South American women reported higher intake of other white potatoes (significant differences among all subgroups) and green salad (significant difference compared to Mexican and Central Americans) and a lower intake of cooked dried beans (significant differences among all subgroups). Among men, only two components of total fruit and vegetables remained significantly different across subgroups after adjusting for potential confounders: Mexican men reported higher consumption of cooked dried beans than all other groups except respondents of multiple origin; and South American men reported higher intake of other white potatoes than did Mexican, Spanish, and multiple origin men.
Consuming a diet high in fruits and vegetables is associated with lower risk of chronic diseases including cardiovascular disease, diabetes and certain types of cancers (18). Results from this study found higher intake of total fruit and vegetable cup equivalents for Hispanic men and women, as compared to non-Hispanic Whites and Blacks. These findings are consistent with national results from NHANES 2003–2004 using 24-hour dietary recalls. Previous data from the 2000 National Health Interview Survey (NHIS) also found that Latino men and women reported significantly higher servings of fruits and vegetables than non-Hispanic White and Black counterparts (20). Although total fruit and vegetable consumption for Hispanic subjects is somewhat higher than other race/ethnicity groups in the US population, it is lower than the recommended intake of fruits and vegetables for Americans (a total of 4.5 cups of fruit and vegetables per day for a 2,000-calorie intake) (26). Identifying differences in intake across the Hispanic subgroups may help public health programs mold nutrition education messages to increase fruit and vegetable intake in this population.
Even though all people of Hispanic or Latino descent are often grouped together in surveys, the Hispanic population is heterogeneous, composed of persons of Hispanic, Latino and Spanish origin (2). Previous studies that have compared dietary intake by ethnicity have focused mostly on Mexican-Americans (13,19,27–30), and do not represent well over one-third of the US Hispanic population of other countries of origin (1). To date, there are only a few studies that have examined dietary intake among Hispanic subgroups and compared differences. Studies that have looked at particular dietary intakes between selected Hispanic groups, suggest different eating behaviors by country of origin (3,6,12,15,16).
The current study is the first to compare fruit and vegetable intake among five Hispanic subgroups in a representative sample. Although no differences were found in total FVC across Californian Hispanic subgroups, there were significant differences in the estimated intake of the individual fruit and vegetable components (beans and other potatoes for both genders, and salad for women), even after considering potential sociodemographic and acculturation confounders. These findings suggest that among different Hispanic subgroups, there are preferences for certain types of fruits and vegetables that go beyond sociodemographic predictors or acculturation factors. However, with the very short food questionnaire, it was not possible to examine potential differences in individual type of fruit and vegetable in detail.
The Hispanic population in the US varies widely by region of origin and, thereby, by race, culture and health behavior practices, among others (4,5,23,31,32). Previous studies have examined how subpopulations of this heterogeneous group differ in their sociodemographic characteristics, and determinants and predictors of disease and mortality (4,6,9–11,13,33,34). The various Hispanic subgroups may also differ in their level of acculturation (English proficiency, years lived in the US, language of interview, etc). In separate studies, Gordon-Larsen et al.(12), and Lin et al. (35) reported different effects of acculturation on dietary patterns by Hispanic subgroups, specifically, adolescent Puerto Ricans, Cubans and Mexicans, and elderly Dominicans and Puerto Ricans respectively. The present study found that Hispanic subgroups differed in sociodemographic variables and various characteristics that suggest acculturation; however, these differences did not explain all of the variation found in fruit and vegetable intake within certain subgroups. A plausible explanation is that food is still related to culture from the country of origin among certain groups. In a careful examination of cultural diversity of diet among Hispanics, Sanjur noted that dietary patterns are bound to the local cultures and sociopolitical histories. Foods that remain more ‘static’ in diets tend to be staple foods, such as maize in Central America, and corn and beans among Mexicans and Mexican-American diets (17). However, it is still not well understood which factors foster dietary change away from traditional dietary patterns (17,35), whether this change may lead to healthier or unhealthier outcomes, and how that may differ by Hispanic subgroup. The current study identified small, but significant differences in intake of various components of total fruit and vegetable consumption. Although the main public health message is still that all race/ethnicity groups should increase intake of total fruit and vegetables, this information on small but significant differences among Hispanic subgroups may be useful to local public health programs in California that want to target the fruit and vegetable consumption through the food culture of specific Hispanic subgroups in their communities.
There are several limitations to this study, including that all measures of health behavior and sociodemographic variables were self-reported, which may lead to some misclassification and bias. In addition, the overall response rate for CHIS 2005 suggests a low response; however, it is comparable to the response rates in other national telephone surveys (36). An important limitation of this study is the use of a short dietary assessment instrument, instead of a more detailed measure, such as multiple 24-hour dietary recalls, to estimate fruit and vegetable intake. The screener only asked about frequency of consumption of seven items; portion sizes were estimated using external data. An analytical procedure which calibrated screener-type responses against 24-hour recalls was used to provide a better approximation of fruit and vegetable intake. However, race/ethnicity was not incorporated in the calibration. Despite this limitation, screener estimates of differences among Hispanics, non-Hispanic whites and non-Hispanic blacks in CHIS 2005 were generally consistent with national estimates (NHANES 2003–2004) that used multiple 24-hour recalls to collect dietary data. Another limitation of the screener is the lack of information on specific fruits and vegetables that would allow the authors to examine consumption in accordance to the Dietary Guidelines’ specific recommendations (26). For example, the USDA Food Guide recommends consumption of 2 cups per week of orange vegetables; 3 cups per week of dark green vegetables and 3 cups per week of dry beans; the screener does not ask about orange and dark green vegetables separately.
The CHIS dataset is unique in that it is able to capture the rich diversity of the California Hispanic population, providing a major strength to this study. The large number of Hispanic respondents in the CHIS adult dataset allowed the authors to estimate differences in fruit and vegetable intake among Hispanic subgroups with considerable confidence. This is the first such study in this growing segment of the US population. The results are generalizable to the state of California but also suggest patterns that may pertain to other regions of the US.
The main public health message remains to increase FVC among all ethnic and Hispanic subgroups, but the small differences in FVC found among Hispanic subgroups can have valuable implications for the design and implementation of this public health message. Results can be particularly useful in the design of culturally-sensitive public health programs that promote fruit and vegetable consumption among local Hispanic subgroups. The findings can be relevant to identify barriers among various subgroups to increase intake of fruits and vegetables and to identify how different groups respond to various public health messages. Future epidemiologic studies that wish to examine disease outcomes on dietary data about specific fruits and vegetables among various Hispanic groups could also find these results useful. Recognition of the heterogeneity of Hispanics also implies the need for surveillance of the diet among Hispanic subgroups.
In California, Hispanics reported higher consumption of fruits and vegetables than non-Hispanic Whites and Blacks, although their intake still does not meet the national recommendation. There were no significant differences in total fruit and vegetable intake among different Hispanic subgroups. However, different Hispanic subgroups reported differences in intake of specific fruits and vegetables. As Hispanics are the largest minority in the US, it is increasingly important to examine the health behaviors of this heterogeneous group more carefully. Future research of Hispanics should consider design and adequate sample size in order to capture any potential diversity in health behaviors. Also, public health programs that wish to promote fruit and vegetable consumption among Hispanics should consider the heterogeneous nature of this population.
The authors would like to thank Susan Foerster, MPH, RD, from the Network for a Healthy California, who provided assistance and support during this project.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Uriyoán Colón-Ramos, Formerly, CRTA post-doc fellow, Health Promotion Research Branch, Behavioral Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, 6130 Executive Blvd / EPN 4074 MSC 7335, Bethesda, MD 20892-7335, Currently: Email: ude.llenroc@31cu.
Frances E. Thompson, Epidemiologist, Risk Factor Monitoring and Methods Branch, Applied Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, 6130 Executive Blvd / EPN 4095A MSC 7344, Bethesda, MD 20892-7344, Phone: 301-435-4410, Fax: 301-435-3710, Email: vog.hin.liam@fospmoht.
Amy Lazarus Yaroch, Formerly, Behavioral Scientist, Program Director, Health Promotion Research Branch, Behavioral Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, 6130 Executive Blvd / EPN 4074 MSC 7335, Bethesda, MD 20892-7335, Currently: Director, Center for Human Nutrition, Rm. 1024, 505 Durham Research Plaza, Omaha, NE 68105-1313, Phone: 402-559-5500, Fax: 402-559-7302, Email: ude.cmnu@hcoraya..
Richard P. Moser, Research Psychologist, Office of the Associate Director, Behavioral Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, 6130 Executive Blvd., MSC 7326, Executive Plaza North, Room 4052, Bethesda, MD 20892-7326, Phone: 301-496-0273, Fax: 301- 435-7547, Email: vog.hin.liam@rresom.
Timothy S. McNeel, Information Management Services, Inc., 12501 Prosperity Dr., Suite 200, Silver Spring, MD 20904, Phone: 301-680-9770, Fax: 301-680-8304, Email: moc.bewsmi@tleencm..
Kevin W. Dodd, Biometry Research Group, Division of Cancer Prevention, National Cancer Institute, National Institutes of Health, Department of Health and Human Services, 6130 Executive Blvd Room 3129 MSC 7362, Bethesda, MD 20892-7362, Phone: 301-496-7461, Fax: 301-402-0816, Email: vog.hin.liam@kddod.
Audie A. Atienza, National Cancer Institute, Division of Cancer Control and Population Sciences, Behavioral Research Program, Health Promotion, Research Branch, 6130 Executive Blvd, EPN 4082, Bethesda, MD 20892-7335, Phone: 301-402-8426, Fax: 301-480-2087, Email: vog.hin.liam@aazneita.
Sharon B. Sugerman, Research Scientist II, California Department of Public Health Cancer Prevention and Nutrition Section Research and Evaluation Unit, MS-7204, PO Box 997377, Sacramento, CA 95899-7377, Phone: 916- 449-5406, Fax: 916- 449-5415, Email: vog.ac.hpdc@namreguS.norahS.
Linda Nebeling, Chief, Health Promotion Research Branch, Behavioral Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, 6130 Executive Blvd, EPN 4060, Bethesda, MD 20892-7326, Phone: 301- 435-2841, Fax: 301- 435-7547, Email: vog.hin.liam@lnileben.