Diet Quality of Preschool Aged Children in Family Child Care Homes

Diet quality of preschool-aged children in family child care homes Background: Although family child care homes (FCCH) are the second most utilized form of non-relative child care in the US, little is known about what children eat in this setting. Objectives: The primary objective is to assess the quality of meals and snacks consumed by children aged 2-5 in FCCH compared to the 2015 Dietary Guidelines for Americans (DGA 2015). The secondary, exploratory objective is to explore the association between childcare provider acculturation and dietary consumption of children in FCCH. Design: This cross-sectional study used baseline data from an ongoing cluster randomized controlled intervention trial in FCCH. Participants/setting: Family childcare providers completed a demographic survey. Preschool-aged children (n = 124; 2-5 years old) enrolled in 43 FCCH, were observed during two meals and a snack each day for two days, following a standardized protocol. Main outcome measures: Foods and beverages consumed were analyzed in the Nutrition Data System for Research 2015. Mean amounts consumed of each food group were compared to 2/3 of the daily intake recommendations established by the DGA 2015 for children 1-3 years old (1000 kcal) given that recommendations are not standard across age groups. For the exploratory analysis, acculturation was measured using a proxy measure of language spoken at home. Statistical analyses: One-sample t-tests and one-sample sign tests tested for differences between mean food group amounts consumed compared to the national guidelines. Spearman’s correlations were run to examine the association between acculturation, income and education on food groups. Results: Median and mean intakes of several foods groups did not meet recommendations. Children did not meet the recommended 0.67 cup of vegetables (Mdn = 0.20 cup, 42% of recommendation) (p<.001), 1.0 ounce of whole grains (Mdn = 0.25 ounce, 35% of recommendation) (p<.001), 9.34 grams of fiber (5.64±2.15 gram, 60% of recommendation) (p<.001) and exceeded the recommended 1.0 ounce of refined grains (1.69±0.77 cup, 169% of recommendation) (p<.001). Discretionary calories (17.14±7.02 % kcal), percent of calories from added sugars (Mdn = 6.28 % kcal) (p<.001), percent of calories from saturated fat (Mdn = 7.83 % kcal) (p<.001), and (834.86±317.33 mg) sodium (p<.001) did not exceed recommendations. There was a significant positive association with whole grain intake and acculturation, r =0.315, p <.05. Conclusion: Dietary intake of children cared for in FCCH was not consistent with national recommendations for vegetables, whole grain, and refined grain intake. Future research should continue to develop and evaluate strategies to increase fruit, vegetable and whole grain intake in this setting.


INTRODUCTION
According to recent estimates, approximately one in four preschool-aged children, ages 2 to 5, are overweight or obese. 1 Contributing to this problem are dietary patterns high in energy dense foods and beverages, and low in fruits, vegetables and whole grains. 2,3 Since food preferences develop during early childhood, [4][5][6] and contribute to healthy dietary patterns that can track into adulthood, 7 it is critical to support healthy food environments where children are spending their time.
The child care environment is an ideal setting to promote a healthy environment, 8 given that in the US, 60% of children under five years old receive non-parental care, 24% of which are enrolled in non-relative care. Many of these children are enrolled in family child care homes (FCCH), a form of childcare operated from the home of a non-relative.
Children under five years of age spend on average over 26 hours a week in this setting, 9 and are recommended to obtain 1/2 to 2/3 of their daily nutrients, depending on their age, if enrolled full-time or part-time, respectively, from meals and snacks consistent with the Dietary Guidelines for Americans (DGA) served in child care. [10][11][12] Furthermore, while the home environment plays a considerable influence on children's dietary habits, 13,14 child care settings are subject to nutritional regulations and policies, and supported by federal programs, such as the Child and Adult Food Program (CACFP). 15 Having this institutional structure could support healthy nutrition environments by targeting specific nutrition-related attitudes or practices of family child care home providers (FCCP) in professional education and training programs. 16 Of the studies that have explored what children are eating in childcare, most have focused on menu evaluations, [17][18][19][20][21] and observed foods served 20,22 and almost all have been 5 conducted in childcare centers. [23][24][25] These studies found that children in childcare centers are consuming enough vegetables and whole grains but too much saturated fat, sodium and added sugar consumption. 24 Given that in Rhode Island, at least 40% of FCCP identify as Hispanic, it is important to consider acculturation as possible contributor to the foods they choose to serve. It is well known that the healthfulness of the traditional Latino diet, which is high in legumes, fruit, 26,27 tends to deteriorate with the acculturation process due to both changes in the environment and social factors. 28,29 Some aspects of this dietary acculturation refers to the multi-dimensional process in which immigrants adopt, negotiate, and alter food attitudes and beliefs from the dominant culture that can result in changes to dietary consumption. 30 As caregivers acculturate to the US, their consumption of energy-dense snacks and sweetened beverages in children increases. 31,32 Furthermore, several studies report differences in fruit juice and sugar sweetened beverages consumed by ethnicity; Hispanic/Latino and African-American children disproportionately participate in these unhealthy dietary behaviors than their white counterparts. [33][34][35][36] Therefore, higher levels of acculturation could potentially influence increased dietary consumption of fruit juice, sugar sweetened beverages and energy-dense foods in FCCH.
A recent study with FCCP found that provider ethnicity was related to certain nutritional practices. 16 Furthermore, focus groups with Hispanic and Latino FCCP also indicate the potential influence of culture on foods served by providers to children in their care. [37][38][39] The relationship between acculturation, culture, and ethnicity is complex and 6 multifaceted, 40  We also explored the association between FCCP acculturation and dietary consumption by children enrolled in FCCH.

Sample
This cross-sectional study used a subset of baseline data collected between (DOCC) visually estimated the amounts of foods served and consumed by each child in FCCH, and documented a brief description of the type of foods and the quantities of foods served, dropped, traded or added during meals and snacks served in the morning and afternoon. 41 Direct observation has been shown to be a valid and reliable measure of assessing children's food and beverages in childcare. 41 Post-observation, observers clarified additional details on brands and/or cooking methods with providers. Per protocol, an observer can only accurately and reliably assess three children at a time, therefore, when more than three children were present, two observers collected the data. 41 To prevent estimation drift, staff observers conducted in-house trainings quarterly.
A registered dietitian reviewed records prior to data entry through Nutrition Data Details regarding specific food items with corresponding food pattern equivalent groups are described in the extended methods in Appendix B.

9
To compare daily recommendations found in the DGA 2015 to the recommended 2/3 amount to be consumed in childcare settings, 2/3 of the 1000 calorie Healthy Food Pattern recommendations for 1-to 3-year old age group were used as the reference standard. [10][11][12] Two-thirds of the daily nutrient recommendations was chosen, since 135 of 201 observations of children from the on-going main trial consumed at least a breakfast, lunch, and snack, which corresponds to two-thirds of meals consumed in the day (Figure   1.

10
The DGA 2015 recommends that 1/2 of grains should be whole. Total grain was calculated in addition to whole grains and refined grains. Total grain, whole grain and refined grains were measured in ounce equivalents.

Dairy
Dairy was reported in cups, and included low fat, fat free and full fat dairy sources.

Protein
Protein was calculated by summing up proteins from both animal and plant sources. Protein was measured in ounces.

Oils
Healthy fats were calculated by summing together mono-unsaturated fats (MUFA) and poly-unsaturated fats (PUFA) values together. This was reported in grams.

Added sugars
The DGA 2015 includes recommendations for no more than 10% of daily calories as added sugars. For this analysis, percent of added sugars was calculated by multiplying added sugars by total sugars in grams by their calories per gram (4 calories per gram) divided by total calories consumed multiplied by 100.

Macronutrients
Overall calories, percent of calories from saturated fat, percent of calories from added sugars and discretionary calories were determined by summing the total calories obtained from solid fat and added sugars and divided by the total caloric consumption.
Dietary fiber was reported in grams.

Acculturation
Language spoken at home at in childcare was chosen as a proxy measure for acculturation, which has been previously used in prior studies on acculturation, [42][43][44] and is known to be a strong predictor of acculturation. 41,45 Response categories were collapsed: "Spanish only" and "Other" was coded as a proxy for being less acculturated as 0, and more than one language spoken "English only", "Both, more English than Spanish", and "Both, equal amounts of time" and "Both, more Spanish than English" as 1.

Education
Education was also coded by collapsing response categories on the item: "less than high school diploma" and "high school or GED" as 0, and grouping "Associate's degree or equivalent" and "Bachelor's degree" as 1.

Statistical Analyses
Analyses were conducted at the FCCH group level. Spearman's correlations were run to examine the association between acculturation, income and education on legumes, vegetables, fruit, juice, whole grains, refined grains, percent of calories from discretionary calories, and percent of calories from added sugars. These foods were previously chosen in the literature on acculturation and dietary intake, and were feasibly measured in our study. 46 13

Demographics
All FCCP (100%) were female, with a mean age of 50.8 years. Family child care home providers identified as Hispanic/Latino (95%), Dominican (75%), spoke only Spanish at home outside the child care business (50%), and were not born in the US (97%), and lived for a mean of 11 years in the United States. Almost all of FCCP reported "More Spanish than English" as the language spoken to children in their care (41%) ( Table 1).
Over half of children were male (52.35%), and the average age of children were 2.9 years.

Food groups consistent with DGA 2015
Children met the recommendation for fruit and total grain consumption while in FCCH (0.5 vs. 0.7 cups, 2.0 vs. 2.0 oz equivalents, respectively) ( Table 2, Table 3 Table 3).

Exploring the association between acculturation, income, education and dietary consumption
All correlations conducted between acculturation, income, and education are reported in Table 4. There was a significant positive association with whole grain intake and acculturation, r =0.315, p <.05, as measured by the proxy measure language.
Vegetables and education level were also significantly positively correlated, r = 0.353, p <.05. and juice and education level were significantly negatively correlated r = -0.324, p <.05. Medians of significantly correlated items were then compared using Kruskal-Wallis tests. However, once adjusted for multiple comparisons with a Bonferroni correction, the results were no longer significant ( Table 5).  54 Our results also suggest that refined grains may be displacing whole grains, and could contribute to low whole grain consumption. The new CACFP guidelines could also help reduce refined grain consumption, by excluding grain based desserts from reimbursable meal pattens. 54 In addition to the possible impact of new policies, it will be important to consider interventions and training specific to improving vegetable and whole grain consumption that are consistent with federal nutritional recommendations for children in this age group. In RI, providers are required to attend one orientation with the Department of Youth, Children, and Families prior to licensure. In addition, it is recommended that they attend additional professional development trainings offered by community organizations. It is also recommended, although not required that they utilize online resources offered by the Rhode Island Department of Education, which includes topics of nutrition. Incorporating more specific information on ways to increase vegetable and whole grain consumption within these trainings, can be used to support healthy dietary habits among preschool aged children. Furthermore, it is important to continue to investigate barriers to vegetable, whole grains being served and consumed in FCCH. 55 In exploratory analysis, we found that acculturation level of FCCP was associated with dietary consumption of the children they cared for. Our study found a significant, positive association with whole grain intake and acculturation. Given the limited number of studies exploring this association in FCCP and children in their care, it is difficult to compare our findings to the literature. Although no study, to our knowledge, has reported on the effects of childcare provider acculturation on whole grain consumption by preschool-aged children, other studies have explored the relationship between generational status and whole grain consumption. For example, in a study with Mexican-American adolescents, higher whole grain consumption was associated with increasing generational status. 56 Other studies of the general population indicate that acculturation is both a protective and a risk factor to healthy dietary habits in Latino populations, associated with higher consumption of fruits and vegetables, and higher rates of energy dense food consumption, increased whole grain intake. 28,29,46,57 Our results suggest that childcare provider acculturation may play a role in specific types of healthy and unhealthy food groups consumed, and should be further explored.

DISCUSSION
We also found a positive association with vegetables and education level, and a negative association with juice and education level. Although there are no studies, to our knowledge, of FCCP educational level and dietary intake, one study with childcare centers found that that teacher? education was significantly positively associated with better knowledge of nutrition and health. 58 Given that educational level of adults has been shown to be a significant predictor of nutrition knowledge, 59 investigating the effects of education, acculturation, and nutrition knowledge on the types of healthy and unhealthy food groups consumed within early child care settings will be important.

Limitations and Strengths
This study was not without limitations and results should be interpreted in context of social and environmental factors. The cross-sectional design of our study limits the ability to draw causal inferences. In addition, our sample was primarily composed of

Introduction
According to the socio-ecological framework for health promotion, 1 biological, social, and environmental factors influence the development of childhood obesity.
Contributing in part to this epidemic is the consumption of energy dense, low-fiber, and high fat foods. 2 Since food preferences are established in early childhood, 3

Childhood Obesity -A Public Health Concern
Childhood obesity is a major public health crisis. According to National Health and Nutrition Examination Survey (NHANES) data from 2011-2014, approximately one in five preschool-aged (2-5 years old) children are overweight or obese in the United States (US). 4 This is concerning, given that childhood overweight and obesity often tracks into adulthood, 5 and increases the risk for chronic diseases like type 2 diabetes mellitus, cardiovascular disease, and some cancers. 6 To prevent the development of such diseases, the United States Preventive Task force recommends screening for obesity in children older than six years of age and adolescents. 7 Therefore, the Center for Disease Control and Prevention (CDC) has developed screening tools (e.g. indicators of adiposity) to diagnose childhood obesity. 8 32 Childhood (ages 2 to 19 years old) overweight and obesity is measured through body mass index (BMI), a measure of weight adjusted by height, which correlates to body fat. 9 Body mass index can also predict future adiposity, morbidity, and mortality in children. 10 In adults, absolute measures of BMI are used, where a BMI range of 18.5 to 24.9 indicates a healthy weight. 11 However, due to periods of rapid growth and development in children, weight and height change with age and differ by sex. 12 Therefore, to determine overweight and obesity in children, percentiles specific to age and sex are used. 12

Child Care Settings
In 2012, almost 60% of children ages 3 to 5-years were enrolled in some type of non-parental care with 56% enrolled in center-based child care, and 24% enrolled in nonrelative home-based child care, also known as a family child care home (FCCH). 24 Previously, research on the nutrition environment and child diet quality has focused in the home, rather than in childcare. 30  Caregivers are considered the nutritional gatekeepers of a child's diets. 42 In FCCH, child care providers play a major role in shaping children's health behaviors by influencing a child's dietary intake by controlling the availability and access to healthy or unhealthy foods, meal structure, feeding, and mealtime social practices. 43

37
The CACFP is a federally funded program for reimbursing the cost of meals and snacks offered in child care programs. 45 Child care centers and FCCH that meet eligibility requirements are able to participate in a tiered reimbursement program. 45 The program functions as a food safety net for low-income individuals and vulnerable populations with greater nutritional risk. 45 Participation is contingent on compliance with federal nutrition standards defined by meal-patterns based off of the DGA. 45

Caregiver dietary socio-demographics and diet
Although literature is well established between socio-demographic factors and diet in adults, 46 little is known about the relationship between caregiver sociodemographics and the diet of children in their care. A study by Erinosho et al. 47 reported that caregivers' demographic characteristics, such as increasing caregiver age and education was associated with more healthful dietary intake in children 3-5 years of age and Latino subgroup was associated with less healthful dietary intakes by children 3-5 years of age. 48 Acculturation refers to the process by which cultural exchange occurs between an immigrant and their host culture, which can result in a change in many behaviors, including diet. 49,50 Although no standards have been established for measuring 38 acculturation, some measures include length of residence in host country, language spoken, location of birth, and also ethnic identification. 51  To quantify dietary intake in child-care centers without relying on provider report or intrusive weighing methods, 64 Ball et al. 55

developed the Diet Observation in Child
Care (DOCC) as a validated protocol for observing dietary intakes in child care settings.

Results from a larger study utilizing the DOCC protocol indicate that children in North
Carolina child care centers were not meeting DGA in 2005, whereby consumption of whole grain and dark green and orange vegetable was limited and consumption of energy dense snacks and condiments was high. 60 By developing the DOCC protocol, Ball et al. 55 was able to perform analyses on foods served and consumed, accounting for observed child intake of nutrients or food groups (foods consumed) rather than observing what was served.
In a recent study using the DOCC protocol, Schwartz et al. 56  For the main trial, a total of 132 FCCPs will be randomized into 66 matched pairs. These FCCPs will be randomized into the nutrition and physical activity intervention group or early literacy and reading readiness control group. A monthly topic was chosen and FCCP received a newsletter/video with information related to their topic.

Eligibility criteria
For this analysis, we included FCCP (n = 43) who had complete baseline demographic surveys conducted in-person and over the phone, and those that had

Measures
Details regarding specific food items with corresponding food pattern equivalent groups are described in the extended methods in Appendix B.

Dietary Guidelines for Americans 2015
To compare daily recommendations found in the DGA 2015 to the recommended 2/3 amount to be consumed in child care settings, 2/3 of the 1000 calorie Healthy Food

Grains
The DGA 2015 recommends that 1/2 of grains should be whole. Total grain was calculated in addition to whole grains and refined grains. Total grain, whole grain and refined grains were measured in ounce equivalents. Values were obtained from raw NDSR output 04.

Dairy
Dairy was calculated in cups, and included milkwhole, milkreduced fat, milk low fat and fat free, milknondairy and ready to drink, flavored milkreduced fat, ready to drink flavored milklow fat and fat free, sweetened flavored milk beverage powder with non-fat dry milk, artificially sweetened flavored milk beverage powder with non-fat dry milk, cheesefull fat, cheeselow fat and fat free, cheesenon-dairy, yogurtsweetened with whole milk, yogurtsweetened low fat, yogurtsweetened fat free, yogurtartificially sweetened low fat, yogurtartificially sweetened fat free, yogurtnondairy, dairy based sweetened meal replacement/supplement, dairy based artificially sweetened meal replacement/supplement. Values were obtained from NDSR output 09.

Protein
Protein was calculated by summing up proteins from both animal and plant sources. Protein was measured in ounces. This included beef, lean beef, veal, lean veal, lamb, lean lamb, fresh pork, lean fresh pork, cured pork, lean cured, game, poultry, lean poultry, fishfresh and smoked, lean fishfresh and smoked, shellfish, cold cuts and sausage, lean cold cuts and sausage, organ meats, eggs, egg substitute, nuts and seeds, nut and seed butters, meat alternatives. Values were obtained from NDSR output 09.

Oils
Healthy oils group was measured by summing mono-unsaturated fatty acids (MUFA) and poly-unsaturated fatty acids (PUFA). The MUFA and PUFA values were obtained from NDSR output 04 and reported in grams.

Sodium
Sodium was reported in milligrams and obtained from the raw NDSR output 04.

Dietary Fiber
Dietary fiber was reported in grams and obtained from the raw NDSR output 04.

Macronutrients
Overall calories, percent of calories from saturated fat, percent of calories from added sugars and discretionary calories were determined by summing the total calories obtained from total solid fat and added sugars by total sugars and divided by the total calorie consumption multiplied by 100. Percent of saturated fat calories were obtained from the raw NDSR output 04. Percentage of added sugars calories were obtained by multiplying added sugars by total sugars by 4 calories per gram divided by total calories and multiplied by 100. Both of these values were obtained from NDSR output 04.

Acculturation
Language spoken at home at in childcare was chosen as a proxy measure for acculturation, which has been previously used in prior studies on acculturation, and is known to be a strong predictor of acculturation. Response categories were collapsed: "Spanish only" and "Other" was coded as a proxy for being less acculturated as 0, and more than one language spoken "English only", "Both, more English than Spanish", and "Both, equal amounts of time" and "Both, more Spanish than English" as 1.

Education
Education was also coded by collapsing response categories on the item: " less than high school diploma" and "high school or GED" as 0, and grouping "Associate's degree or equivalent" and "Bachelor's degree" as 1.

Statistical Analyses
Analyses were conducted at the FCCH group level. All descriptive statistics on food group variables were conducted by FCCH, and not by individual child. Normality was assessed visually and by examining skewness and kurtosis. For normal variables, a one-sample t-test was used to compare means of food group variables compared to DGA recommendations for FCCH. A one-sample sign test was run to compare medians of nonnormally distributed food group variables compared to the DGA recommendations for FCCH. The primary aim was sufficiently powered at β = .80 with a sample size of 22 FCCH. Since analyses were not adjusted for multiple comparisons, the statistically significant threshold was set at p <0.01.
Spearman's correlations were run to examine the association between acculturation, income and education on legumes, vegetables, fruit, juice, whole grains, refined grains, 57 percent of calories from discretionary calories, and percent of calories from added sugars.