The prevalence of childhood obesity among racial and ethnic minority groups is high. Approximately 21% of Latino children and adolescents and 24% of African-American children and adolescents are obese compared to 14% of white children and adolescents.¹ Multiple factors contribute to the development of childhood obesity including dietary behaviors.^2-5 Health professional guidelines on the prevention and treatment of child overweight and obesity include specific dietary behaviors such as no consumption of sugar-sweetened beverages, limiting fruit juice to 4 ounces a day, limiting high energy density foods such as sweets, consuming greater than or equal to five servings of fruits and vegetables per day, and limiting the consumption of fast food or eating out (food consumed away from home).⁶ African-American and Latino children compared to their white counterparts have been shown to have dietary practices that increase the risk of obesity such as higher intake of sugar-sweetened beverages (SSB) and fast food and, lower intake of fruits and vegetables.^7-9 Few studies, however, have examined and compared the dietary practices among minority and/or low-income pediatric populations as they pertain to the professionally recommended guidelines and fewer have extended examination to the third major minority group in the United States (US)--Asian populations.¹⁰

In order for health practitioners and registered dietitians to support families in the prevention and treatment of childhood obesity, a general understanding of the dietary patterns among young children is necessary as well as an understanding of the similarities and differences across races and ethnicities. Such information can help identify suboptimal dietary patterns that may contribute to disparities in childhood obesity. Challenges like language barriers and/or specific cultural health-seeking behaviors, however, may prevent healthcare providers from effectively caring for children among diverse communities.^11,12 Language barriers are important to consider because they can affect the quality of provider-patient communication and care. In addition, language can be an indicator of acculturation which providers may need to consider in dietary counseling as English-language proficiency negatively impacts dietary practices among Latino immigrant populations.^13,14. For these reasons English-language proficiency is an important factor to consider in parent-provider interactions about a child's nutrition and dietary practices but they can be overlooked due to the increasing pace and number of recommended topics to discuss during pediatric visits.¹⁵ Therefore, understanding the dietary patterns, similarities, and differences across racial, ethnic and linguistic groups of children can help health practitioners and registered dietitians efficiently target suboptimal dietary patterns for intervention.

The purpose of this study was to focus on minority children living in California and examine a set of well-described dietary practices known to contribute to unhealthy weight development, to explore the associated socio-demographic factors with these dietary practices, and ultimately provide a starting point for practitioners to tailor nutritional guidance for diverse populations.

The data source for this study comes from the 2007 and 2009 California Health Interview Survey (CHIS). CHIS is the nation's largest state health survey sponsored by a network of public agencies and private organizations. CHIS is a population-based telephone survey designed to estimate a variety of physical, emotional, and behavioral child and adult health and healthcare indicators and is representative of California's diverse population. CHIS uses a landline and cell-phone random-digit dial approach and an oversampling of specific populations to provide statewide population estimates. A total of 9,913 children from 0-11 years of age in 2007 were sampled by the University of California, Los Angeles Center for Health Policy Research (UCLA CHPR), and another 8,945 children in 2009. Verbal parental consent and adolescent agreement are obtained by the UCLA CHPR before families participate in the CHIS survey. Interviews were conducted in five languages: English, Spanish, Chinese (Mandarin and Cantonese dialects), Vietnamese, and Korean and the translation process involved several professional translators, and multiple reviews and revisions of survey items until a consensus was achieved by both internal and external experts and translators. The translation was ultimately overseen and approved by the UCLA CHPR. Households were sampled from within 44 geographic sampling strata, each comprised of either a single county or multiple counties. Sampling weights are calculated by the UCLA CHPR to appropriately weight participants and includes a complex and iterative process that accounts for geographic strata, household level data such as ethnicity, race, age, gender, household size, and education. Details about the 2007 and 2009 CHIS methodology are described elsewhere.^16,17 The public use data files from CHIS 2007 and 2009 were merged and yielded a total of 18,858 completed interviews for children. Similar strategies of pooling data across survey years have been used to look at dietary behaviors and other health behaviors and outcomes.^18,19

The study sample included children two to eleven years old. Given the relationship between weight and dietary practices, we intended to control for Body Mass Index (BMI) in our analyses. A total of 2,956 children younger than two years of age were excluded, as BMI norms are not established for children under two. To retain as many children in the analysis and because BMI was not a primary outcome, we kept children in the sample that had missing weight or height data, using a “missing” indicator, but had complete data on all other outcome measures of interest in our a priori model. Children with missing height or weight were categorized into a “missing weight” category and included in all analyses of this study. This study was approved by the University of California, Los Angeles Institutional Review Board.

Analyses were performed using Stata 12 (Stata Corp., College Station, TX) and the replicate weights provided by UCLA CHPR to account for the complex survey design of CHIS and to yield state estimates. The 2007 data and weights provided by CHIS were merged with the data and weights from 2009. This pooled data was used for all of the analyses of this study. All dietary outcomes were dichotomized based on professional dietary guidelines or evidence supported by the literature on dietary behaviors associated with unhealthy weight status development. Descriptive statistics of the independent variables were completed using the sample population and then by each racial/ethnic and linguistic group. Chi-square tests were completed to examine the bivariate associations between dietary behaviors and children's racial/ethnic and linguistic group and children's weight status. Separate multivariable logistic regression models of each dietary outcome were used to examine factors significantly associated with each dietary behavior after adjustment for relevant covariates.

This study highlights the racial and ethnic differences in dietary behaviors among California children as well as several sociodemographic factors related to specific dietary behaviors. In multivariate regressions, substantial differences in fruit juice, fruit, vegetable, sugar-sweetened beverages, sweets, and fast food consumption were found among the major racial and ethnic groups of children. This is the first study to our knowledge that examines a comprehensive set of dietary practices, known to be important determinants of healthy weight development for white children and the general pediatric population⁶, among the three major minority groups of California. Several dietary practices are worth noting as they have not previously been well described and may serve as a starting point for practitioners and registered dietitians to explore dietary practices with families as well as guide researchers who work with diverse populations.

Children of Asian families interviewed in non-English, suggestive of households with lower levels of US acculturation, were not found to have higher odds of consuming fast food than Whites, while most other minority groups had significantly higher odds of fast food consumption. One other study has shown an association of increased fast food intake with higher levels of acculturation among Asian children.³⁹ Thus, it appears that lower intake of fast food consumption among Asian-American children may be more likely in less acculturated households. There is a scarcity of research on Asian-American families to explain how this healthy dietary practice is attained. Beliefs about the ills of fast food among Asian-Americans seem to be similar to those of other minority groups.⁴⁰ The higher parental educational attainment and income brackets of Asian families do not explain these differences—in fact, fast food consumption tends to be greater in some higher educational and income brackets in our sample, suggesting a more complex process. Additional research is needed to explore what factors may be contributing to this healthy dietary practice. All other minority children in this study, however, had significantly higher odds of fast food consumption than Whites and over half of all California children, regardless of racial or ethnic group, reported eating fast food at least once a week. This high prevalence of fast food consumption among children mirrors national estimates⁹ and highlights the need to address this dietary practice across all groups of young children.

Children of Asian families interviewed in both English and non-English, however, were also found to have higher odds of consuming less than two vegetable servings in the previous day compared to Whites. There were higher odds of consuming less than two fruit and vegetables servings among Asians interviewed in a non-English language. The findings of lower vegetable and fruit consumption among Asians may be somewhat counterintuitive, as several Asian cultures have native food patterns that are high in vegetables, ^41,42 however, contributing factors may include the acculturation effect of a high-fat and low fruit and vegetable American diet on Asian dietary practices,^10,43,44 difficulty finding traditional fruits and vegetables that are native to Asian countries, and cultural and native patterns where low fruit consumption may be typical.⁴⁰ Because Asian-American children have been understudied in obesity prevention and management, it is difficult to determine whether this finding holds true for Asian-American children outside of California and the extent to which traditional beliefs may be at the root of fruit and vegetable consumption patterns. Our findings, however, suggest that one potential mechanism to address the lower fruit and vegetable consumption among Asian children might be for physicians to talk to families about nutrition and healthy eating. In our study, children had lower odds of consuming less than two fruit and vegetables servings if parents reported that their child's physician had discussed nutrition during their child's last routine physical exam visit. These findings are consistent with the literature that indicates parents welcome counseling on healthy foods from physicians, and that healthcare professionals in general play an important role in counseling and changing lifestyle behaviors.^45-48 Additional studies are needed, though, to evaluate whether healthcare professional nutrition counseling has a differential effect on children from ethnically diverse backgrounds.

Latino and African-American children had higher odds of consuming more than the recommended amount of daily fruit juice and sugar-sweetened beverages than Whites. These unhealthy dietary practices have been well described in the literature.^7,30,49 Latino children, however, had lower odds of consuming sweets than Whites. Several national studies show an overall increase in the consumption of energy-dense, nutrient-poor foods such as sweets among children,^50,51 but it is unclear whether all groups of children are increasing their consumption of sweets at similar rates. Our results suggest that the documented increase in consumption of sweets over the last several years may not be driven by Latino children, and relative avoidance of sweets in fact may be a common healthy weight behavior among Latino families. This possibility is supported by many qualitative studies that illustrate that Latino parents overwhelmingly describe limiting junk food such as chips and candy as important for the prevention of childhood obesity, but are less likely to mention non-carbonated sweetened drinks and high fat snacks, that can also lead to unhealthy weight development.^52,53 Whether Latino children consume lower rates of sweets compared to other groups of children, however, is an important area for future research particularly because protective factors that can buffer against childhood obesity among Latinos are less described or absent from the literature.

White children have been shown to have lower rates of sugar-sweetened beverage, fruit juice and fast food consumption, and higher rates of fruit and vegetable intake when compared to minority groups of children. Many of these patterns were also found in this analysis. The consumption of sweets, however, did not follow this healthy dietary pattern. Instead, we found higher odds of consumption of sweets among whites when compared to Latino children. In addition, unlike many of the other unhealthy dietary practices, the consumption of sweets was positively related to family income and did not vary by parental educational attainment.

This finding indicates that parents in general may consider sweets to be qualitatively different from sugar-sweetened beverages and fruit juices. This dietary practice requires further study, particularly the context in which sweets are provided to children—sweets given as reward, for instance, can increase a child's risk of overweight due to poor self-regulation of appetite, problematic eating attitudes, and an increased preference for and consumption of sweet food.^54-56

This study had several limitations. First, the study sample was created from 2007 and 2009 data, and we may not have captured variation between years. This approach however was necessary to have adequate sample size for the comparison of the major racial and ethnic groups of California children. Analyses of each individual year (albeit underpowered) only showed secular trends of a decrease in sugar-sweetened beverage consumption which has been supported by other publications.⁵⁷ No other secular trends were noted. Second, dietary practices were collected using caregiver report with caregivers defining the specific foods and servings, and did not include direct observations of dietary intake or specifications from the interviewer related to age-specific serving sizes. As a result, there may be reporting bias and under or over estimation of each food item. The survey is also administered in 5 different languages and although rigorous steps and strategies are incorporated by CHLA CHPR to ensure optimal translation of each question, there is always the possibility that caregivers incorrectly understood the specific food related questions thereby introducing bias into our results. It should also be noted that Asians and Latinos are not homogenous groups and therefore aggregating each of them into one group may have biased the findings. In addition, height and weight were based on caregiver report, which may bias our estimates; however, many national studies continue to use this method of survey data collection.⁵⁸ There was a significant amount of missing height and weight data particularly for non-English interviewed Latinos and Asians that could have biased our estimates. The large amount of missing data for these populations may be due to low health literacy levels or confusion regarding height and weight metrics as many Latin American and Asian countries use the International System of Units (i.e. meters and kilograms). Children with missing weight or height data, however, were flagged with an indicator and not omitted from the analysis, in order to avoid exclusion bias. Lastly, time spent in the US was not included in this analysis, which can play an important role in dietary practices; however, English language fluency (interview language) was included which is often also used as a proxy for acculturation.

There are significant racial and ethnic differences in dietary practices among California children, and each group has practices that are protective and risk factors of unhealthy weight development; our study, in particular, provides new information on Asian children. Thus, all groups of children exhibit different problematic dietary practices and additional variations may exist within specific ethnic groups based on linguistic preferences. Healthcare providers, and more specifically registered dietitians, can use these results for broad guidance regarding dietary screening and as starting points for more tailored assessments and counseling for specific ethnic and racial groups.