- Email: [email protected]
Food Reluctance of Preschool Children Attending Daycare Centers Is Associated with a Lower Body Mass Index
RESEARCH
Original Research
Food Reluctance of Preschool Children Attending Daycare Centers Is Associated with a Lower Body Mass Index Véronique Surette, MSc, RD*; Stéphanie Ward, PhD, RD*; Pascale Morin, PhD; Hassan Vatanparast, MD, PhD; Mathieu Bélanger, PhD ARTICLE INFORMATION Article history: Submitted 27 July 2016 Accepted 11 July 2017
Keywords: Child eating behaviors Childhood underweight Daycare setting Food reluctance Fussiness 2212-2672/Copyright ª 2017 by the Academy of Nutrition and Dietetics. http://dx.doi.org/10.1016/j.jand.2017.07.007 *
Certified in Canada.
ABSTRACT Background Food reluctance can present as fussiness, picky eating, slowness in eating, and high satiety responsiveness. It can be associated with inadequate weight gain during early childhood. Although a majority of preschoolers attend daycare centers, associations between their eating behaviors at daycare and their body composition have not been studied. Objective Our aim was to develop an estimate of food reluctance and to assess the relationship between food reluctance at daycare and body mass index (BMI) and waist circumference of preschoolers. Design We conducted a cross-sectional secondary analyses. Food reluctance was estimated using weighted digital plate waste analysis. Intra-rater, inter-rater, and test retest reliability and convergent validity of the food reluctance score were tested. The food reluctance score was then compared to preschool children’s BMI and waist circumference. Participants/setting Participants included 309 children aged 3 to 5 years in 24 daycare centers across the Canadian province of New Brunswick. Main outcome measures Preschool children’s waist circumference and age-adjusted BMI derived from objectively measured height and weight were analyzed. Statistical analyses performed Intraclass correlations were used to determine the reliability of the new estimate. Spearman correlation was used to compare the estimate with parental report of food reluctance. Multivariate linear regressions were used to examine the relationship between food reluctance and waist circumference and ageadjusted BMI. Results The estimated food reluctance score demonstrated excellent inter- and intrarater reliability (intraclass correlation>0.97; P<0.0001) and good testretest reliability (intraclass correlation¼0.72; P<0.0001). It also provided evidence of convergent validity through correlation with reluctance-related subscales of the Child Eating Behavior Questionnaire (r¼.53, P<0.0001). Greater demonstration of food reluctance at the daycare center was associated with a lower age-adjusted BMI (adjusted b 1.41; 95% CI .15 to 2.67), but was not associated with children’s waist circumference (adjusted b .60; 95% CI 2.06 to .86). Conclusions Signs of food reluctance can be observed in daycare and relate to lower BMI among preschoolers. J Acad Nutr Diet. 2017;-:---.
F
OOD RELUCTANCE, WHICH INCLUDES FUSSINESS, picky eating, slowness in eating, and high satiety responsiveness1-3 is a problematic eating behavior4 estimated to affect approximately one of three Canadian children aged 2 to 4 years,5 and >50% of 3- to 7-year-old Chinese children.6 Food reluctance has also been observed in the United States, where the prevalence of picky eating among children aged 3 to 4 years has been estimated to range between 15% and 22%.7-9 In contrast to children who do not demonstrate this behavior, children who are food reluctant tend to have poorer vegetable consumption and lower protein and fat intake.10,11 As such, food reluctance is often ª 2017 by the Academy of Nutrition and Dietetics.
associated with nutrition deficiencies and inadequate weight gain.12 Nutrition deficiencies at a young age put young children at higher risk of suffering from stunted growth, malnutrition, psychosocial problems, and the emergence of chronic health diseases, such as metabolic syndrome, osteoporosis, and diabetes later in life.13,14 Food reluctance in early childhood is also associated with lower, oftentimes insufficient, weight gain.12,15-17 Specifically, Canadian children who demonstrate food reluctance are twice as likely to be underweight compared with children who do not present this behavior.10 Inadequate weight gain in the preschool years increases the risk of coronary heart disease at adulthood by JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
1
RESEARCH as much as 80%.18 Despite evidence reporting these health consequences, few studies focus on preventing inadequate weight gain in the early years.18,19 Considering that the preschool years coincide with the process of establishing eating habits, and that they represent a transitional and modifiable period in life,20 these years could be optimal to implement interventions aiming to rectify problematic eating behaviors and encourage appropriate weight gain.19,21 Young children’s behaviors can largely be influenced by the environments in which they spend a lot of time.22 For example, it is well accepted that parental practices and the family environment can shape children’s eating behaviors and body composition.23,24 However, other social influences may contribute to children’s development. National survey data suggest that 54% of Canadian children aged 6 months to 5 years spend approximately 29 hours per week in daycare centers.25 This setting could therefore be an effective environment in which to intervene on preschoolers’ eating behaviors.26 Despite this, existing research on food reluctance of preschoolers is exclusively based on the familial environment in which children reside.5,11,12,15,16,23,24,27-32 Current research on the relationship between children’s eating behavior and their body composition is largely based on parent-reported assessments. Such assessments are associated with limitations, including subjectivity, because parents’ knowledge and attitudes toward health can influence their ratings of children’s eating behaviors, potentially leading to biased information.33 Also, parents whose children attend out-of-home care are generally not present when snacks or lunch are provided, further limiting the accuracy of parent-reported assessment in this population. The use of plate waste measurements can represent an alternative to counter these limitations. Plate-waste measurements involve directly estimating caloric intake, types of food, and macronutrients consumed from the difference between the weight of food served and food left at the end of a meal.34,35 This approach has been demonstrated to be a precise method for assessing diet34,35 and has been used in institutional contexts, such as schools36-38 and daycare centers,39,40 where it is feasible to measure the food consumed by a large number of individuals. Despite the current nonexistence of a method to assess food reluctance in daycare centers, plate waste data could be used to estimate food reluctance in this setting. Specifically, because not eating all or part of the food served can be a marker of eating behavior traits, such as fussiness, picky eating, eating slowly, and high satiety responsiveness, leftover data gathered in the process of plate waste data collection could be used to derive a proxy measure of food reluctance. Developing a method of estimating food reluctance in daycare centers can help researchers and health professionals gain better understanding of this eating behavior in young children. Therefore, this study had two purposes. First, we aimed to develop and test the reliability and validity of a proxy measure of food reluctance that can feasibly be applied in institutional settings. Second, we aimed to estimate the relationship between food reluctance at daycare, using the new measure, and body composition among preschoolers. Results from this study will provide information for policymakers, health professionals, daycare staff, and researchers on whether food reluctance is observed in daycare centers. 2
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
MATERIALS AND METHODS Participants Data for this project were obtained from the Healthy Start/ Départ Santé (HSDS) study; a clustered randomized controlled trial described in detail elsewhere.41 Participants retained for the current analysis were preschool children aged 3 to 5 years recruited in provincially licensed daycare centers in New Brunswick, Canada. To be considered for the study, the daycare centers had to prepare and serve lunch daily and have preschoolers attending full-time. Daycare centers were excluded if they had received an intervention focused on physical activity or nutrition in the previous 10 years, if they were licensed for fewer than 35 children, and if fewer than 20 children aged 3 to 5 attended the center. All children 3 to 5 years old attending the recruited centers on a full-time basis were eligible and invited to participate in this study (395 children were recruited from 24 centers in New Brunswick, representing a response proportion of 70%). For feasibility purposes, data collection was divided into 3 years. The Healthy Start/Départ Santé study received approval from the Health Canada, the University of Saskatchewan, and the Centre Hospitalier de l’Université de Sherbrooke ethics review boards. Ethics approval specific to the analysis presented in this article covered data collected after 2013 only. Correspondingly, analyses evaluating the reliability and validity proprieties of the new estimate of food reluctance were conducted with data obtained after the 8-month intervention in spring 2014 (n¼115). For assessing the relationship between food reluctance at daycare and body composition among preschoolers, we restricted analyses to follow-up data from centers in the control arm in June 2014 and to baseline data from the intervention and control groups in October 2014 (n¼309). This insured that the potential influence of the Healthy Start/Départ Santé intervention would not confound results. Written parental consent was obtained from all participating children.
Body Composition Measurements The children’s height (Seca 213 portable stadiometer; Seca), weight (Conair scale, CN2010CX model), and waist circumference were measured by trained research assistants following a standardized protocol.42 Two measures of height and waist circumference to the nearest 0.1 cm, and weight to the nearest 0.1 kg were obtained for each participant. If discrepancies >0.5 cm for height and waist circumference were observed between the two measures, a third measure was obtained. The average of the two closest measures was recorded. Children’s body mass index (BMI) was obtained by calculating the ratio between their weight in kilograms and their height in meters squared, and was age-adjusted as proposed by the International Obesity Task Force43 and the International Association for the Study of Obesity.44 International standards for age-adjusting waist circumference do not currently exist.45
Plate WasteBased Estimation of Food Reluctance Plate waste analysis was used to establish a proxy measure of food reluctance. This involved analyzing what children were offered and what they consumed during lunchtime on 2 consecutive weekdays. This was done during regular lunchtime for which daycare center staff were instructed not to --
2017 Volume
-
Number
-
RESEARCH alter their menu or procedures and research assistants were trained to not influence normal unfolding of lunchtime. The first step consisted of having one trained research assistant take a digital picture of each food served to children with an ASUS Memo Pad HD7 (ME173X model; Android US) and also weighing the individual food items with a digital scale that was manually calibrated (16221.5 cm; inStyle). The research assistant then took a digital picture and weighed any leftovers of those foods at the end of the meal. This was done for every serving. Pictures were taken at a distance of 62 cm and at a 45-degree angle.46 The second step consisted of calculating the amount of grams that were consumed for each of five food groups (vegetables, grains and starches, milk and milk alternatives, meats and meat alternatives, and fruits) during lunchtime. This involved subtracting the total grams leftover of a food group from the total grams served of the said food group. For this, research assistants used the weight of food recorded. When food items left on the plate were made of multiple food groups, research assistants used the pictures to estimate the proportion of food consumed to be attributed to each food group. To estimate food reluctance, a third step involved calculating the ratio between amount leftover and amount served of every food group over the 2 days of assessment. For each child, the largest proportion of food leftover over served (leftover/served) among the five food groups represented the estimated food reluctance score. For example, if a child had 30% of vegetables, 10% of grains and starches, 0% of milk and milk alternatives, 15% of meats and meat alternatives, and 10% of fruits leftover at the end of the meal, his or her estimated food reluctance score would be 30%. Higher scores were therefore estimated to represent behaviors arising from higher food reluctance, while lower scores were estimated to represent less food reluctance. For inter-rater reliability assessment, the second step of this procedure was conducted independently by two research assistants, both were registered dietitian nutritionists. One of the research assistants also repeated step two 1 month after having done it for the first time to assess intra-rater reliability.
Parent and Daycare Questionnaires The parent questionnaire was distributed to parents or guardians by daycare center directors, who instructed parents that they had 1 week to complete the questionnaires and return them to the daycare center, where they were placed in a concealed package and mailed back to the research team. In addition to the plate waste method, children’s eating behaviors were also measured with the Child Eating Behavior Questionnaire (CEBQ) administered to their parents or guardians.3 The CEBQ took approximately 10 minutes to complete. It is the most widely used questionnaire in this field of research.15,16,27,29-32,47 Of the eight subscales included in the CEBQ, the Satiety responsiveness (five items, eg, “My child gets full easily”), slowness in eating (four items, eg, “My child takes more than 30 minutes to finish a meal”), and food fussiness (six items, eg, “My child is difficult to please with meals) subscales were selected for the purpose of this study. Each item was scored on a five-point Likert scale (0 to 4) that ranges from “never,” “seldom,” “sometimes,” “often,” to “always.”3 It was previously observed that these three subscales are correlated and jointly represent a measure of food reluctance.3 In our sample of parents, good internal --
2017 Volume
-
Number
-
consistency was observed across the 15 items from the three subscales (Cronbach’s a¼.85, n¼57). Children and parents characteristics were collected through a self-reported validated questionnaire, administered to parents or guardians at the same time as the CEBQ.48 These characteristics included child’s date of birth and child’s sex, as well as the mother and father’s education level as assessed in the Canadian Community Health Survey.48 The questionnaire took approximately 5 minutes to complete. The urban/rural setting was determined by matching the daycare location to Statistics Canada’s metropolitan influenced zones catalogue,49 and language, French or English (proxy for culture), was recorded as the language in which the daycare was registered. These characteristics were all considered as potential cofounders as previous studies have linked them to food reluctance and to BMI and waist circumference.12,27,28,47,50
Data Analysis All analyses were conducted using IBM SPSS statistical software.51 Intraclass correlations were used to determine intrarater, inter-rater, and 2-day testretest reliability of the new proxy measure of food reluctance. Although the CEBQ is based on self-reported information, represents a subjective assessment of the behavior observed by parents, and represents eating behavior outside of daycare centers, we compared our daycare centerbased estimate of food reluctance with this score because indications of alignment between the two scores would provide support to the suggestion that they assess similar constructs.52 Because CEBQ-based food reluctance is represented on a discrete rather than continuous scale, Spearman correlation was used to compare it with our estimate of food reluctance. We conducted linear regressions models to examine the relationship between estimated food reluctance and both children’s waist circumference and their age-adjusted BMI. We sequentially computed models univariately (step 1), followed by models including all level 1 covariates (step 2; covariates included were age, sex, language, urban/rural, and parent education), and by fully adjusted models including daycare centers at a second level (step 3). Given the addition of a second level to account for potential clustering at the daycare level led to no modification of the estimates for any of the variables, we only present the more parsimonious models from steps 1 and 2.
RESULTS Properties of the Proxy Measure of Food Reluctance Data from 126 children in eight daycare centers were available for the first aim of the study. Of these, 11 children did not eat lunch at the daycare center during the 2 days of data collection. Therefore, a final sample of 115 children was available for assessing the reliability of the estimation of food reluctance. The meanstandard deviation age of children at the first data collection was 4.40.6 years (Table 1). The majority of our sample was French-speaking (72%) and from urban areas (63%). Because some daycare centers offered the option of additional servings, a total of 284 observations (plates served), from the 115 children over 2 days, were considered for the intra-rater and inter-rater reliability analyses. The mean estimated food reluctance score for these JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
3
RESEARCH Table 1. Characteristics of preschool children from the Healthy Start study included in the reliability analyses of the proxy measure of food reluctance and in the analyses of the relationship between the proxy measure of food reluctance and body composition Sample for Reliability Analyses (n[115)
Sample for the Relationship Analyses (n[309)
n
%
Mean–SDa
n
%
Mean–SD
Male
64
55.7
—
158
51.1
—
Female
51
44.3
—
151
48.9
—
English
32
27.8
—
52
16.8
—
French
83
72.2
—
257
83.2
—
Rural
43
37.4
—
87
28.2
—
Urban
72
62.6
—
222
71.8
—
None has a university degree
25
43.9
—
78
40.4
—
—
—
Characteristic Sex
Language (of the center)
Rurality (of the center)
Parents’ education At least one parent has a university degree
32
56.1
115
59.6
Age, y
—
—
4.40.6
—
—
4.10.9
Age-adjusted body mass index
—
—
20.64.2
—
—
20.73.8
Waist circumference, cm
—
—
54.85.2
—
—
53.25.5
a
SD¼standard deviation.
children was 0.38 with a standard deviation of 0.46 and median of 0.43. Results show that the estimated food reluctance score demonstrated excellent 1-month intra-rater and excellent inter-rater reliability (all intraclass correlation 0.97, Table 2). The estimated food reluctance score also demonstrated 2-day testretest reliability (intraclass correlation¼0.722; P<0.0001), as calculated by comparing the estimated food reluctance score obtained from children on their first and second days of data collection.
Table 2. Intra-rater and inter-rate reliability of the new proxy measure of food reluctance (n¼284 observations from 115 preschool children from the Healthy Start study) a
Food group
Intra-rater ICC (1-mo interval)
Inter-rater ICC
Vegetablesb
0.975***
0.982***
Grains and starches
0.994***
0.970***
Meat and meat alternatives
0.985***
0.992***
Milk and milk alternatives
0.998***
0.995***
Fruitsc
0.985***
0.998***
a
ICC¼intra-class correlation. Potatoes were included in the grains and starches category rather than the vegetables category. c Fruits do not include fruit juice. ***P<0.0001. b
4
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
Data from the 57 parents who answered the parent questionnaire were used to assess the concurrent validity of the new measure. Results indicate that the estimated food reluctance score correlated positively and moderately with the CEBQ-based parent-reported measure of food reluctance (r¼.53; P<0.001), providing evidence of concurrent validity. Our estimate of food reluctance correlated more strongly with the slowness in eating subscale (r¼0.62; P<0.001) than the food fussiness and satiety responsiveness subscales (r¼.28; P<0.2 and r¼.32; P<0.1, respectively).
Relationship between Food Reluctance and Body Composition For the second aim of the study, data from 327 children in 19 daycare centers were available. However, plate waste data could not be obtained from 18 of those children because they did not consume lunch at the daycare center during the 2 days of data collection; meaning that data from a total of 309 children were used to assess a relationship between the estimated food reluctance score and children’s body composition. They were 4.10.9 years old, and most were French-speaking (83.2%) and from urban areas (71.8%). In our sample, the mean food reluctance score was 0.40 with a standard deviation of 0.49 and a median of 0.44, and 71% of children demonstrated some food reluctance (estimated score >0%). Vegetables was the food group for which children demonstrated the most reluctance. More specifically, 54% of children who demonstrated food reluctance did so by eating relatively less of the vegetables they were served than food --
2017 Volume
-
Number
-
--
2017 Volume
-
Number
-
Univariate b from linear regressions for each separate variable to multivariate b from linear regressions including all variables. The proxy measure of food reluctance can range from 0 to 1, with higher values suggesting presentation of higher food reluctance. The ratio between amount of food leftover and amount served is calculated for five food groups. The largest of these five proportions is used as the proxy measure of food reluctance. *P<0.05. b
a
1.00 (0.63 to 2.62), 0.2
0.45 (1.02 to 1.92), 0.5 0.73 (0.55 to 2.00), 0.3 0.66 (0.05 to 1.36), 0.2 Parents’ education, 1 parent university educated
0.45 (0.40 to 1.30), 0.5
1.02 (2.98 to 0.95), 0.3
0.16 (1.24 to 1.57), 0.8
0.07 (0.49 to 0.63), 0.9 1.15 (2.86 to 0.55), 0.2 0.51 (1.08 to 0.07), 0.5
0.07 (0.55 to 0.69), 0.9
Language, French
Urban/rural setting, urban
1.24 (0.61 to 1.86), 0.04*
2.35 (1.12 to 3.58), <0.01*
0.48 (0.95 to 1.91), 0.5 0.29 (0.52 to 1.09), 0.6 0.95 (0.29 to 2.19), 0.1
2.27 (1.28 to 3.26), <0.01* 0.58 (0.49 to 1.65), 0.3 0.40 (0.49 to 1.29), 0.4
0.75 (0.06 to 1.56) 0.2 Sex, female
Age, y
0.60 (2.06 to 0.86), 0.4 0.34 (1.51 to 0.83), 0.6 1.41 (2.67 to 0.15), 0.03*
b (95% CI), P value!
1.17 (2.32 to 0.02), 0.04* Food reluctance,b score
Multivariate
Waist Circumference
Univariate Multivariate Univariate
This study is the first to introduce an estimate of food reluctance based on plate waste, which is a widely used and accepted method of assessing diet.53 The new estimate of food reluctance fills a gap given previous investigations of preschoolers’ eating behaviors relied on parent-reported measures.5,11,24,27,32 Parents are generally not present at daycare centers for prolonged periods, making them imperfectly informed of what their children are eating and how they are behaving in that specific environment. This limitation further amplifies the necessity to develop an estimate of food reluctance that can be used outside of the household setting. In addition to feasibly being administered in daycare centers, the food reluctance estimate demonstrated excellent reliability. Assessment of validity for this type of measure was complicated by the absence of gold-standard measures to assess food reluctance. However, concurrent validity was suggested by demonstrating an acceptable correlation with the food-reluctance related items of the CEBQ. Given the subjective parent-reported nature of the CEBQ and the fact that it is based on eating behavior at home, it was expected that correlations between the objective and CEBQ-based measures would not be perfect. This might be related, in part, to measurement errors potentially attributable to both instruments, but most particularly to the CEBQ, as the inherent limitation associated with self-reported measures is
Variable
DISCUSSION
Age-Adjusted BMI
from the other food groups. In contrast, fruits were the food group with the highest relative proportion of consumption, contributing to only 10% of the estimated cases of food reluctance, followed by milk and milk alternatives (17%), grains and starches (18%), and meat and meat alternatives (21%). The sum of percentages exceeds 100% because some participants demonstrated the same level of food reluctance toward more than one food group. Results from the univariate models show a negative relationship between the estimated food reluctance score and age-adjusted BMI among preschoolers, meaning that greater demonstration of food reluctance at the daycare center was significantly related to lower age-adjusted BMI among preschoolers. No other variable was significantly related to BMI in univariate analyses. Age and an urban (vs rural) setting were significantly positively related to preschooler’s waist circumference, but there was no apparent association with the estimate of food reluctance (Table 3). In multivariate models, the estimated food reluctance score remained significantly related to lower age-adjusted BMI, but only age remained positively related to waist circumference. Specifically, this model suggests that for every point increase in the food reluctance score, age-adjusted BMI decreased by 1.41. For example, when comparing average children from this study (food reluctance score of 0.40) to children who do not demonstrate food reluctance (food reluctance score of 0), it is estimated that, on average, children demonstrating food reluctance have a 0.56 lower age-adjusted BMI (1.410.40). In an additional series of analyses, we also noted negative relationships between the CEBQ-based parent-reported measure of food reluctance and both age-adjusted BMI and waist circumference, but none of these analyses were statistically significant at the traditional cut point of a¼.05 (P values ranged from 0.08 to 0.7).
Table 3. Univariate and multivariate relationship between the estimate of food reluctance of preschool children from the Healthy Start study and their age-adjusted body mass index (BMI) and waist circumferencea
RESEARCH
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
5
RESEARCH the distortion of the collected information caused by the subjective opinion of each respondent.54 In studies such as this one where parents assess their child’s behaviors, variables such as parental health, family functioning, and socioeconomic status are reported as being critical factors influencing parents’ views and opinions, potentially causing reporter bias in the data collected on their child.33 An important finding of this analysis is that higher estimated food reluctance displayed at lunchtime while attending a daycare center was significantly associated with lower age-adjusted BMI. This result is consistent with two other Canadian studies examining the relationship between parent-reported eating behaviors and weight status of preschoolers.10,32 In their study, Dubois and colleagues10 reported that children who were described as being picky eaters—a demonstration of food reluctance—were two times more likely to be underweight when compared with their nonpicky eater peers. Similarly, Spence and colleagues32 reported a negative linear relationship between weight and food reluctance. Our finding also lends credence to international studies, which reported that parent-reported food reluctance could be associated with lower BMI.12,15,16,23,29,31 Our results complement these previous studies by focusing on eating behavior at daycare. By documenting the presence of potential food reluctance in this setting and demonstrating that it is related to BMI, this study provides support to the hypothesis that daycare centers represent a promising environment for healthy eating and adequate weight gainpromoting interventions. However, we did not observe a relationship between our estimate of food reluctance and waist circumference. This is in contrast with findings from Carnell and Wardle,15 who observed a negative association between a parent-reported manifestation of satiety responsiveness and children’s waist circumference. Considering that cutoffs to indicate clinical risk based on waist circumference of children do not currently exist, these authors adjusted the waist circumference value collected from their sample in reference to 1990 UK data. Given no Canadian or international standards exist, we did not adjust our waist circumference data, which possibly explains the discrepancy between these previous results and our own. It is, however, noteworthy that our multivariate analyses showed that older children had greater waist circumference. This being said, it is likely that the nonsignificant relationship we observed between estimated food reluctance and waist circumference was confounded by age. Because children in this study were in a period of important growth, age likely confounded the relationship between food reluctance and waist circumference. This highlights the need for standardization norms for children’s waist circumference. Another potential explanation for the discrepancy between previous studies and current results may be that our estimate accounted for various demonstrations of food reluctance and not just satiety responsiveness, which was measured by Carnell and Wardle. In addition to the contributions described here, the current article has implications for research and practice. Our results suggest it is feasible to use plate waste in future studies, particularly among children attending daycare centers, given data could be acquired from a large number of participants within a relatively short period of time, it could be processed relatively rapidly by trained personnel, and it possessed 6
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
excellent intra- and inter-reliability. As technology evolves, the development of mobile applications for plate waste measurement could further facilitate the collection of food consumption and eating behavior data of children in institutional settings. In addition, our finding of a significant association between estimated food reluctance displayed in daycare settings and children’s age-adjusted BMI call for experimental research to test whether daycare centers are conducive environments where interventions can be implemented to improve eating behaviors among preschoolers.
Strengths and Limitations This study included a large sample of participants, was based on objective measures, and included many steps to verify the reliability and validity of the new measure. It therefore adds to our understanding of the relationship between food reluctance and body composition among preschoolers. Still, there are limitations to consider. First, even though a significant association was reported between food reluctance and children’s age-adjusted BMI, causality cannot be established, given the cross-sectional nature of the study. Second, because our sample was selected from mostly French-speaking participants in relatively large daycare centers, our findings may not be generalizable to all Canadian preschool children attending daycare. Third, although data were collected on separate days to account for children’s variable appetite,53 those days were consecutive and we did not account for or measure appetite, limiting our ability to capture and consider additional possible variability in dietary intake. Fourth, because data were only collected at lunchtime, other food eaten at daycare was not captured. Fifth, our estimate of food reluctance represents a proxy of the behavior that does not account for appetite or food quality and does not discern different behavioral aspects of eating, including slowness in eating, picky eating, and satiety response. Last, even though research assistants were instructed to remain distant and as indiscernible as possible during data collection, it is possible that children’s food intake was distorted because they knew they were being observed.
CONCLUSIONS In summary, this study introduces a new estimate of food reluctance that showed evidence of reliability and validity in preschool children. This study also provides support to the hypothesis that food reluctance displayed at daycare is related to lower BMI among preschoolers. Future research could build on these findings to test interventions aimed at reducing preschoolers’ food reluctance in daycare centers, and explore the long-term effects of these interventions on food reluctance, body composition, and health risks later in life.
References 1.
Dovey TM, Staples PA, Gibson EL, Halford JCG. Food neophobia and “picky/fussy” eating in children: A review. Appetite. 2008;50(2-3): 181-193.
2.
Tharner A, Jansen P, Kiefte-de Jong J, et al. Toward an operative diagnosis of fussy/picky eating: A latent profile approach in a population-based cohort. Int J Behav Nutr Phys Act. 2014;11:14-24.
3.
Wardle J, Guthrie CA, Sanderson S, Rapoport L. Development of the Children’s Eating Behaviour Questionnaire. J Child Psychol Psychiatry. 2001;42(7):963-970.
--
2017 Volume
-
Number
-
RESEARCH 4.
Carruth B, Ziegler P, Gordon A, Barr S. Prevalence of picky eaters among infants and toddlers and their caregivers’ decisions about offering a new food. J Am Diet Assoc. 2004;104(1 suppl 1):S57-S64.
5.
Dubois L, Farmer A, Girard M, Peterson K. Preschool children’s eating behaviours are related to dietary adequacy and body weight. Eur J Clin Nutr. 2007;61(7):846-855.
6.
Xue Y, Zhao A, Cai L, et al. Growth and development in Chinese preschoolers with picky eating behaviour: A cross-sectional study. PLoS One. 2015;10(4):e0123664.
7.
Jacobi C, Agras W, Bryson S, Hammer L. Behavioral validation, precursors, and concomitants of picky eating in childhood. J Am Acad Child Adolesc Psychiatry. 2003;42(1):76-84.
BMI: A comparison of black Afro-Caribbean, white British and white German samples. Eur J Clin Nutr. 2013;67(2):180-184. 28.
Gubbels J, Kremers S, Stafleu A, et al. Association between parenting practices and children’s dietary intake, activity behavior and development of body mass index: The KOALA Birth Cohort Study. Int J Behav Nutr Phys Act. 2011;8:18-30.
29.
Jansen P, Roza S, Jaddoe V, et al. Children’s eating behavior, feeding practices of parents and weight problems in early childhood: Results from the population-based Generation R Study. Int J Behav Nutr Phys Act. 2012;9:130-140.
30.
Joyce J, Zimmer-Gembeck M. Parent feeding restriction and child weight. The mediating role of child disinhibited eating and the moderating role of the parenting context. Appetite. 2009;52(3): 726-734.
31.
Mallan K, Liu W, Mehta R, Daniels L, Magarey A, Battistutta D. Maternal report of young children’s eating styles. Validation of the Children’s Eating Behaviour Questionnaire in three ethnically diverse Australian samples. Appetite. 2013;64:48-55.
32.
Spence J, Carson V, Casey L, Boule N. Examining behavioural susceptibility to obesity among Canadian pre-school children: The role of eating behaviours. Int J Pediatr Obes. 2011;6(2-2):e501-e507.
8.
van der Horst K, Eldridge A, Deming D, Reidy K. Caregivers’ perceptions about picky eating: Associations with texture acceptance and food intake. FASEB J. 2014;28(1):S379.3.
9.
Mascola AJ, Bryson SW, Agras WS. Picky eating during childhood: A longitudinal study to age 11-years. Eat Behav. 2010;11(4):253-257.
10.
Dubois L, Farmer A, Girard M, Peterson K, Tatone-Tokuda F. Problem eating behaviors related to social factors and body weight in preschool children: A longitudinal study. Int J Behav Nutr Phys Act. 2007;4:9-18.
11.
Pinhas-Hamiel O, Bar-Zvi E, Boyko V, Reichman B, Lerner-Geva L. Prevalence of overweight in kindergarten children in the centre of Israel—Association with lifestyle habits. Child Care Health Dev. 2009;35(2):147-152.
33.
Collishaw S, Goodman R, Ford T, Rabe-Hesketh S, Pickles A. How far are associations between child, family and community factors and child psychopathology informant-specific and informant-general? J Child Psychol Psychiatry. 2009;50(5):571-580.
12.
Ekstein S, Laniado D, Glick B. Does picky eating affect weight-forlength measurements in young children? Clin Pediatr (Phila). 2010;49(3):217-220.
34.
13.
Halterman J, Kaczorowski J, Aligne C, Auinger P, Szilagyi P. Iron deficiency and cognitive achievement among school-aged children and adolescents in the United States. Pediatrics. 2001;107(6):13811386.
Wolper C, Heshka S, Heymsfield S. Measuring food intake: An overview. In: Allison D, ed. Handbook of Assessment Measures for Eating Behaviors and Weight-Related Problems. Thousand Oaks, CA: Sage Publishing; 1995:215-240.
35.
Jacko C, Dellava J, Ensle K, Hoffman D. Use of the plate-waste method to measure food intake in children. J Ext. 2007;45(6):6RIB7.
36.
14.
Liuba P, Pesonen E. Infection and early atherosclerosis: Does the evidence support causation? Acta Paediatr. 2005;94(6):643-651.
Blakeway S, Knickrehm M. Nutrition education in the Little Rock School Lunch Program. J Am Diet Assoc. 1978;72(4):389-391.
37.
15.
Carnell S, Wardle J. Appetite and adiposity in children: Evidence for a behavioral susceptibility theory of obesity. Am J Clin Nutr. 2008;88(1):22-29.
Whatley J, Donnelly J, Jacobsen D, Hill J, Carlson M. Energy and macronutrient consumption of elementary school children served modified lower fat and sodium lunches or standard higher fat and sodium lunches. J Am Coll Nutr. 1996;15(6):602-607.
16.
Mackenbach J, Tiemeier H, Ende J, et al. Relation of emotional and behavioral problems with body mass index in preschool children: The Generation R study. J Dev Behav Pediatr. 2012;33(8):641-648.
38.
Lee H, Lee K, Shanklin C. Elementary students’ food consumption at lunch does not meet recommended dietary allowance for energy, iron, and vitamin A. J Am Diet Assoc. 2001;101(9):1060-1063.
17.
Mallan KM, Nambiar S, Magarey AM, Daniels LA. Satiety responsiveness in toddlerhood predicts energy intake and weight status at four years of age. Appetite. 2014;74:79-85.
39.
Branen L, Fletcher J. Effects of restrictive and self-selected feeding on preschool children’s food intake and waste at snacktime. J Nutr Educ. 1994;26(6):273-277.
18.
Eriksson J, Forsén T, Tuomilehto J, Osmond C, Barker D. Early growth and coronary heart disease in later life: Longitudinal study. Br Med J. 2001;322(7292):949-953.
40.
Ireton C, Guthrie H. Modification of vegetable-eating behavior in preschool children. J Nutr Educ. 1972;4(3):100-103.
41.
19.
Kuhl E, Clifford L, Stark L. Obesity in preschoolers: Behavioral correlates and directions for treatment. Obesity. 2012;20(1):3-29.
20.
Anzman S, Rollins B, Birch L. Parental influence on children’s early eating environments and obesity risk: Implications for prevention. Int J Obes. 2010;34(7):1116-1124.
Bélanger M, Humbert L, Vatanparast H, et al. A multilevel intervention to increase physical activity and improve healthy eating and physical literacy among young children (ages 3-5) attending early childcare centres: The Healthy Start-Départ Santé cluster randomised controlled trial study protocol. BMC Public Health. 2016;16: 313-322.
21.
Birch L, Ventura A. Preventing childhood obesity: What works? Int J Obes. 2009;33(suppl 1):S74-S81.
42.
22.
Bronfenbrenner U, Morris P. The ecology of developmental processes. In: Damon W, Lerner RM, eds. Handbook of Child Psychology: Volume 1: Theoritical Models of Human Development. 5th ed. Hoboken, NJ: John Wiley & Sons; 1998:993-1028.
National Health and Nutrition Examination Survey. Body Composition Procedures Manual. http://www.cdc.gov/nchs/data/nhanes/ nhanes_03_04/BC.pdf. Revised January 2004. Accessed January 1, 2014.
43.
Cole T, Bellizzi M, Flegal K, Dietz W. Establishing a standard definition for child overweight and obesity worldwide: International survey. Br Med J. 2000;320(7244):1240-1243.
23.
Hittner J, Faith M. Typology of emergent eating patterns in early childhood. Eat Behav. 2011;12(4):242-248.
44.
24.
Johannsen D, Johannsen N, Specker B. Influence of parents’ eating behaviors and child feeding practices on children’s weight status. Obesity. 2006;14(3):431-439.
Lobstein T, Baur L, Uauy R; IASO International Obesity TaskForce. Obesity in children and young people: A crisis in public health. Obes Rev. 2004;5(suppl 1):4-104.
45.
Fredriks A, van Buuren S, Fekkes M, Verloove-Vanhorick S, Wit J. Are age references for waist circumference, hip circumference and waisthip ratio in Dutch children useful in clinical practice? Eur J Pediatr. 2005;164(4):216-222.
46.
Williamson D, Allen H, Martin P, Alfonso A, Gerald B, Hunt A. Comparison of digital photography to weighed and visual estimation of portion sizes. J Am Diet Assoc. 2003;103(9):1139-1145.
47.
Powers S, Chamberlin L, van Schaick K, Sherman S, Whitaker R. Maternal feeding strategies, child eating behaviors, and child BMI in low-income African-American preschoolers. Obesity. 2006;14(11): 2026-2033.
25.
26.
27.
--
Sinha M, Bleakney A. Spotlight on Canadians: Results from the General Social Survey. Receiving care at home. http://www.statcan. gc.ca/pub/89-652-x/89-652-x2014002-eng.htm. Published June 2014. Accessed November 15, 2014. Larson N, Ward D, Neelon S, Story M. What role can child-care settings play in obesity prevention? A review of the evidence and call for research efforts. J Am Diet Assoc. 2011;111(9):1343-1362. Blissett J, Bennett C. Cultural differences in parental feeding practices and children’s eating behaviours and their relationships with child
2017 Volume
-
Number
-
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
7
RESEARCH 48.
Statistics Canada. Canadian Community Health Survey (CCHS): Annual Component2010 Questionnaire. http://www23.statcan.gc. ca/imdb-bmdi/instrument/3226_Q1_V7-eng.pdf. Published July 2010. Accessed September 29, 2013.
49.
Community Information Database. Metropolitan Influence Zones (MIZ) typology 2006. http://map.cid-bdc.ca/#s¼2006;i¼comtype. miz;sly¼can_sdr_DR;sid¼689;v¼map1;l¼en;z¼2149095,322536, 269304,188240. Published 2006. Accessed September 5, 2013.
50.
Roberts K, Shields M, de Groh M, Aziz A, Gilbert J. Overweight and obesity in children and adolescents: Results from the 2009 to 2011 Canadian Health Measures Survey. Health Rep. 2012;23(3):37-41.
51.
IBM SPSS Statistical Software [computer program]. Version 22.0. Armonk, NY: IBM Corp; 2013.
52.
McIntire S, Miller L. Foundations of Psychological Testing. 2nd edition. Thousand Oaks, CA: Sage Publishing; 2005.
53.
Yon B, Johnson R, Stickle T. School children’s consumption of lowercalorie flavored milk: A plate waste study. J Acad Nutr Diet. 2012;112(1):132-136.
54.
Barker C, Pistrang N, Elliott R. Research Methods in Clinical Psychology: An Introduction for Students and Practitioners. 2nd ed. West Sussex, UK: John Wiley & Sons; 2002.
AUTHOR INFORMATION V. Surette is a registered dietician and territory manager, Nestlé Canada, Montreal, Quebec; at the time of the study, she was a graduate student, Faculty of Medicine and Health Sciences, Université de Sherbrooke, Sherbrooke, Canada. S. Ward is assistant professor of nutrition, Université de Moncton, Moncton, Canada; at the time of the study, she was a PhD candidate, Faculty of Medicine and Health Sciences, Université de Sherbrooke, Sherbrooke, Canada. P. Morin is an associate professor, Faculty of Physical Activity Sciences, Université de Sherbrooke, Sherbrooke, Canada. H. Vatanparast is an associate professor, University of Saskatchewan, School of Public Health, Saskatoon, Canada. M. Bélanger is an associate professor and director of research, Université de Sherbrooke, Centre de formation médicale du Nouveau-Brunswick, Moncton, Canada, and an epidemiologist, Réseau de santé Vitalité, Bathurst, Canada. Address correspondence to: Mathieu Bélanger, PhD, Université de Sherbrooke, Centre de formation médicale du Nouveau-Brunswick, 100, rue des Aboiteaux, Moncton, New Brunswick, Canada, E1A 3E9. E-mail: [email protected]
STATEMENT OF POTENTIAL CONFLICT OF INTEREST No potential conflict of interest was reported by the authors.
FUNDING/SUPPORT Healthy Start/Départ Santé is financially supported by grants from the Public Health Agency of Canada, the Consortium national de formation en santé, the Heart and Stroke Foundation of Canada and the Community Initiative Fund from the Government of Saskatchewan. Véronique Surette was funded through a Faculté de médecine et des sciences de la santé de l’Université de Sherbrooke Graduate Scholarship and Stéphanie Ward was funded through a Canadian Institutes of Health Research Charles Best Canada Graduate Scholarships Doctoral Award and a Gérard-Eugène-Plante Doctoral Scholarship.
ACKNOWLEDGEMENTS The authors gratefully acknowledge the important contributions of Anne Leis, PhD, principal investigator; Gabrielle Lepage-Lavoie, coordinator; the Réseau santé en français de la Saskatchewan; the daycare centers; and the participants in the Healthy Start/Départ Santé study.
8
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
--
2017 Volume
-
Number
-
Original Research
Food Reluctance of Preschool Children Attending Daycare Centers Is Associated with a Lower Body Mass Index Véronique Surette, MSc, RD*; Stéphanie Ward, PhD, RD*; Pascale Morin, PhD; Hassan Vatanparast, MD, PhD; Mathieu Bélanger, PhD ARTICLE INFORMATION Article history: Submitted 27 July 2016 Accepted 11 July 2017
Keywords: Child eating behaviors Childhood underweight Daycare setting Food reluctance Fussiness 2212-2672/Copyright ª 2017 by the Academy of Nutrition and Dietetics. http://dx.doi.org/10.1016/j.jand.2017.07.007 *
Certified in Canada.
ABSTRACT Background Food reluctance can present as fussiness, picky eating, slowness in eating, and high satiety responsiveness. It can be associated with inadequate weight gain during early childhood. Although a majority of preschoolers attend daycare centers, associations between their eating behaviors at daycare and their body composition have not been studied. Objective Our aim was to develop an estimate of food reluctance and to assess the relationship between food reluctance at daycare and body mass index (BMI) and waist circumference of preschoolers. Design We conducted a cross-sectional secondary analyses. Food reluctance was estimated using weighted digital plate waste analysis. Intra-rater, inter-rater, and test retest reliability and convergent validity of the food reluctance score were tested. The food reluctance score was then compared to preschool children’s BMI and waist circumference. Participants/setting Participants included 309 children aged 3 to 5 years in 24 daycare centers across the Canadian province of New Brunswick. Main outcome measures Preschool children’s waist circumference and age-adjusted BMI derived from objectively measured height and weight were analyzed. Statistical analyses performed Intraclass correlations were used to determine the reliability of the new estimate. Spearman correlation was used to compare the estimate with parental report of food reluctance. Multivariate linear regressions were used to examine the relationship between food reluctance and waist circumference and ageadjusted BMI. Results The estimated food reluctance score demonstrated excellent inter- and intrarater reliability (intraclass correlation>0.97; P<0.0001) and good testretest reliability (intraclass correlation¼0.72; P<0.0001). It also provided evidence of convergent validity through correlation with reluctance-related subscales of the Child Eating Behavior Questionnaire (r¼.53, P<0.0001). Greater demonstration of food reluctance at the daycare center was associated with a lower age-adjusted BMI (adjusted b 1.41; 95% CI .15 to 2.67), but was not associated with children’s waist circumference (adjusted b .60; 95% CI 2.06 to .86). Conclusions Signs of food reluctance can be observed in daycare and relate to lower BMI among preschoolers. J Acad Nutr Diet. 2017;-:---.
F
OOD RELUCTANCE, WHICH INCLUDES FUSSINESS, picky eating, slowness in eating, and high satiety responsiveness1-3 is a problematic eating behavior4 estimated to affect approximately one of three Canadian children aged 2 to 4 years,5 and >50% of 3- to 7-year-old Chinese children.6 Food reluctance has also been observed in the United States, where the prevalence of picky eating among children aged 3 to 4 years has been estimated to range between 15% and 22%.7-9 In contrast to children who do not demonstrate this behavior, children who are food reluctant tend to have poorer vegetable consumption and lower protein and fat intake.10,11 As such, food reluctance is often ª 2017 by the Academy of Nutrition and Dietetics.
associated with nutrition deficiencies and inadequate weight gain.12 Nutrition deficiencies at a young age put young children at higher risk of suffering from stunted growth, malnutrition, psychosocial problems, and the emergence of chronic health diseases, such as metabolic syndrome, osteoporosis, and diabetes later in life.13,14 Food reluctance in early childhood is also associated with lower, oftentimes insufficient, weight gain.12,15-17 Specifically, Canadian children who demonstrate food reluctance are twice as likely to be underweight compared with children who do not present this behavior.10 Inadequate weight gain in the preschool years increases the risk of coronary heart disease at adulthood by JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
1
RESEARCH as much as 80%.18 Despite evidence reporting these health consequences, few studies focus on preventing inadequate weight gain in the early years.18,19 Considering that the preschool years coincide with the process of establishing eating habits, and that they represent a transitional and modifiable period in life,20 these years could be optimal to implement interventions aiming to rectify problematic eating behaviors and encourage appropriate weight gain.19,21 Young children’s behaviors can largely be influenced by the environments in which they spend a lot of time.22 For example, it is well accepted that parental practices and the family environment can shape children’s eating behaviors and body composition.23,24 However, other social influences may contribute to children’s development. National survey data suggest that 54% of Canadian children aged 6 months to 5 years spend approximately 29 hours per week in daycare centers.25 This setting could therefore be an effective environment in which to intervene on preschoolers’ eating behaviors.26 Despite this, existing research on food reluctance of preschoolers is exclusively based on the familial environment in which children reside.5,11,12,15,16,23,24,27-32 Current research on the relationship between children’s eating behavior and their body composition is largely based on parent-reported assessments. Such assessments are associated with limitations, including subjectivity, because parents’ knowledge and attitudes toward health can influence their ratings of children’s eating behaviors, potentially leading to biased information.33 Also, parents whose children attend out-of-home care are generally not present when snacks or lunch are provided, further limiting the accuracy of parent-reported assessment in this population. The use of plate waste measurements can represent an alternative to counter these limitations. Plate-waste measurements involve directly estimating caloric intake, types of food, and macronutrients consumed from the difference between the weight of food served and food left at the end of a meal.34,35 This approach has been demonstrated to be a precise method for assessing diet34,35 and has been used in institutional contexts, such as schools36-38 and daycare centers,39,40 where it is feasible to measure the food consumed by a large number of individuals. Despite the current nonexistence of a method to assess food reluctance in daycare centers, plate waste data could be used to estimate food reluctance in this setting. Specifically, because not eating all or part of the food served can be a marker of eating behavior traits, such as fussiness, picky eating, eating slowly, and high satiety responsiveness, leftover data gathered in the process of plate waste data collection could be used to derive a proxy measure of food reluctance. Developing a method of estimating food reluctance in daycare centers can help researchers and health professionals gain better understanding of this eating behavior in young children. Therefore, this study had two purposes. First, we aimed to develop and test the reliability and validity of a proxy measure of food reluctance that can feasibly be applied in institutional settings. Second, we aimed to estimate the relationship between food reluctance at daycare, using the new measure, and body composition among preschoolers. Results from this study will provide information for policymakers, health professionals, daycare staff, and researchers on whether food reluctance is observed in daycare centers. 2
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
MATERIALS AND METHODS Participants Data for this project were obtained from the Healthy Start/ Départ Santé (HSDS) study; a clustered randomized controlled trial described in detail elsewhere.41 Participants retained for the current analysis were preschool children aged 3 to 5 years recruited in provincially licensed daycare centers in New Brunswick, Canada. To be considered for the study, the daycare centers had to prepare and serve lunch daily and have preschoolers attending full-time. Daycare centers were excluded if they had received an intervention focused on physical activity or nutrition in the previous 10 years, if they were licensed for fewer than 35 children, and if fewer than 20 children aged 3 to 5 attended the center. All children 3 to 5 years old attending the recruited centers on a full-time basis were eligible and invited to participate in this study (395 children were recruited from 24 centers in New Brunswick, representing a response proportion of 70%). For feasibility purposes, data collection was divided into 3 years. The Healthy Start/Départ Santé study received approval from the Health Canada, the University of Saskatchewan, and the Centre Hospitalier de l’Université de Sherbrooke ethics review boards. Ethics approval specific to the analysis presented in this article covered data collected after 2013 only. Correspondingly, analyses evaluating the reliability and validity proprieties of the new estimate of food reluctance were conducted with data obtained after the 8-month intervention in spring 2014 (n¼115). For assessing the relationship between food reluctance at daycare and body composition among preschoolers, we restricted analyses to follow-up data from centers in the control arm in June 2014 and to baseline data from the intervention and control groups in October 2014 (n¼309). This insured that the potential influence of the Healthy Start/Départ Santé intervention would not confound results. Written parental consent was obtained from all participating children.
Body Composition Measurements The children’s height (Seca 213 portable stadiometer; Seca), weight (Conair scale, CN2010CX model), and waist circumference were measured by trained research assistants following a standardized protocol.42 Two measures of height and waist circumference to the nearest 0.1 cm, and weight to the nearest 0.1 kg were obtained for each participant. If discrepancies >0.5 cm for height and waist circumference were observed between the two measures, a third measure was obtained. The average of the two closest measures was recorded. Children’s body mass index (BMI) was obtained by calculating the ratio between their weight in kilograms and their height in meters squared, and was age-adjusted as proposed by the International Obesity Task Force43 and the International Association for the Study of Obesity.44 International standards for age-adjusting waist circumference do not currently exist.45
Plate WasteBased Estimation of Food Reluctance Plate waste analysis was used to establish a proxy measure of food reluctance. This involved analyzing what children were offered and what they consumed during lunchtime on 2 consecutive weekdays. This was done during regular lunchtime for which daycare center staff were instructed not to --
2017 Volume
-
Number
-
RESEARCH alter their menu or procedures and research assistants were trained to not influence normal unfolding of lunchtime. The first step consisted of having one trained research assistant take a digital picture of each food served to children with an ASUS Memo Pad HD7 (ME173X model; Android US) and also weighing the individual food items with a digital scale that was manually calibrated (16221.5 cm; inStyle). The research assistant then took a digital picture and weighed any leftovers of those foods at the end of the meal. This was done for every serving. Pictures were taken at a distance of 62 cm and at a 45-degree angle.46 The second step consisted of calculating the amount of grams that were consumed for each of five food groups (vegetables, grains and starches, milk and milk alternatives, meats and meat alternatives, and fruits) during lunchtime. This involved subtracting the total grams leftover of a food group from the total grams served of the said food group. For this, research assistants used the weight of food recorded. When food items left on the plate were made of multiple food groups, research assistants used the pictures to estimate the proportion of food consumed to be attributed to each food group. To estimate food reluctance, a third step involved calculating the ratio between amount leftover and amount served of every food group over the 2 days of assessment. For each child, the largest proportion of food leftover over served (leftover/served) among the five food groups represented the estimated food reluctance score. For example, if a child had 30% of vegetables, 10% of grains and starches, 0% of milk and milk alternatives, 15% of meats and meat alternatives, and 10% of fruits leftover at the end of the meal, his or her estimated food reluctance score would be 30%. Higher scores were therefore estimated to represent behaviors arising from higher food reluctance, while lower scores were estimated to represent less food reluctance. For inter-rater reliability assessment, the second step of this procedure was conducted independently by two research assistants, both were registered dietitian nutritionists. One of the research assistants also repeated step two 1 month after having done it for the first time to assess intra-rater reliability.
Parent and Daycare Questionnaires The parent questionnaire was distributed to parents or guardians by daycare center directors, who instructed parents that they had 1 week to complete the questionnaires and return them to the daycare center, where they were placed in a concealed package and mailed back to the research team. In addition to the plate waste method, children’s eating behaviors were also measured with the Child Eating Behavior Questionnaire (CEBQ) administered to their parents or guardians.3 The CEBQ took approximately 10 minutes to complete. It is the most widely used questionnaire in this field of research.15,16,27,29-32,47 Of the eight subscales included in the CEBQ, the Satiety responsiveness (five items, eg, “My child gets full easily”), slowness in eating (four items, eg, “My child takes more than 30 minutes to finish a meal”), and food fussiness (six items, eg, “My child is difficult to please with meals) subscales were selected for the purpose of this study. Each item was scored on a five-point Likert scale (0 to 4) that ranges from “never,” “seldom,” “sometimes,” “often,” to “always.”3 It was previously observed that these three subscales are correlated and jointly represent a measure of food reluctance.3 In our sample of parents, good internal --
2017 Volume
-
Number
-
consistency was observed across the 15 items from the three subscales (Cronbach’s a¼.85, n¼57). Children and parents characteristics were collected through a self-reported validated questionnaire, administered to parents or guardians at the same time as the CEBQ.48 These characteristics included child’s date of birth and child’s sex, as well as the mother and father’s education level as assessed in the Canadian Community Health Survey.48 The questionnaire took approximately 5 minutes to complete. The urban/rural setting was determined by matching the daycare location to Statistics Canada’s metropolitan influenced zones catalogue,49 and language, French or English (proxy for culture), was recorded as the language in which the daycare was registered. These characteristics were all considered as potential cofounders as previous studies have linked them to food reluctance and to BMI and waist circumference.12,27,28,47,50
Data Analysis All analyses were conducted using IBM SPSS statistical software.51 Intraclass correlations were used to determine intrarater, inter-rater, and 2-day testretest reliability of the new proxy measure of food reluctance. Although the CEBQ is based on self-reported information, represents a subjective assessment of the behavior observed by parents, and represents eating behavior outside of daycare centers, we compared our daycare centerbased estimate of food reluctance with this score because indications of alignment between the two scores would provide support to the suggestion that they assess similar constructs.52 Because CEBQ-based food reluctance is represented on a discrete rather than continuous scale, Spearman correlation was used to compare it with our estimate of food reluctance. We conducted linear regressions models to examine the relationship between estimated food reluctance and both children’s waist circumference and their age-adjusted BMI. We sequentially computed models univariately (step 1), followed by models including all level 1 covariates (step 2; covariates included were age, sex, language, urban/rural, and parent education), and by fully adjusted models including daycare centers at a second level (step 3). Given the addition of a second level to account for potential clustering at the daycare level led to no modification of the estimates for any of the variables, we only present the more parsimonious models from steps 1 and 2.
RESULTS Properties of the Proxy Measure of Food Reluctance Data from 126 children in eight daycare centers were available for the first aim of the study. Of these, 11 children did not eat lunch at the daycare center during the 2 days of data collection. Therefore, a final sample of 115 children was available for assessing the reliability of the estimation of food reluctance. The meanstandard deviation age of children at the first data collection was 4.40.6 years (Table 1). The majority of our sample was French-speaking (72%) and from urban areas (63%). Because some daycare centers offered the option of additional servings, a total of 284 observations (plates served), from the 115 children over 2 days, were considered for the intra-rater and inter-rater reliability analyses. The mean estimated food reluctance score for these JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
3
RESEARCH Table 1. Characteristics of preschool children from the Healthy Start study included in the reliability analyses of the proxy measure of food reluctance and in the analyses of the relationship between the proxy measure of food reluctance and body composition Sample for Reliability Analyses (n[115)
Sample for the Relationship Analyses (n[309)
n
%
Mean–SDa
n
%
Mean–SD
Male
64
55.7
—
158
51.1
—
Female
51
44.3
—
151
48.9
—
English
32
27.8
—
52
16.8
—
French
83
72.2
—
257
83.2
—
Rural
43
37.4
—
87
28.2
—
Urban
72
62.6
—
222
71.8
—
None has a university degree
25
43.9
—
78
40.4
—
—
—
Characteristic Sex
Language (of the center)
Rurality (of the center)
Parents’ education At least one parent has a university degree
32
56.1
115
59.6
Age, y
—
—
4.40.6
—
—
4.10.9
Age-adjusted body mass index
—
—
20.64.2
—
—
20.73.8
Waist circumference, cm
—
—
54.85.2
—
—
53.25.5
a
SD¼standard deviation.
children was 0.38 with a standard deviation of 0.46 and median of 0.43. Results show that the estimated food reluctance score demonstrated excellent 1-month intra-rater and excellent inter-rater reliability (all intraclass correlation 0.97, Table 2). The estimated food reluctance score also demonstrated 2-day testretest reliability (intraclass correlation¼0.722; P<0.0001), as calculated by comparing the estimated food reluctance score obtained from children on their first and second days of data collection.
Table 2. Intra-rater and inter-rate reliability of the new proxy measure of food reluctance (n¼284 observations from 115 preschool children from the Healthy Start study) a
Food group
Intra-rater ICC (1-mo interval)
Inter-rater ICC
Vegetablesb
0.975***
0.982***
Grains and starches
0.994***
0.970***
Meat and meat alternatives
0.985***
0.992***
Milk and milk alternatives
0.998***
0.995***
Fruitsc
0.985***
0.998***
a
ICC¼intra-class correlation. Potatoes were included in the grains and starches category rather than the vegetables category. c Fruits do not include fruit juice. ***P<0.0001. b
4
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
Data from the 57 parents who answered the parent questionnaire were used to assess the concurrent validity of the new measure. Results indicate that the estimated food reluctance score correlated positively and moderately with the CEBQ-based parent-reported measure of food reluctance (r¼.53; P<0.001), providing evidence of concurrent validity. Our estimate of food reluctance correlated more strongly with the slowness in eating subscale (r¼0.62; P<0.001) than the food fussiness and satiety responsiveness subscales (r¼.28; P<0.2 and r¼.32; P<0.1, respectively).
Relationship between Food Reluctance and Body Composition For the second aim of the study, data from 327 children in 19 daycare centers were available. However, plate waste data could not be obtained from 18 of those children because they did not consume lunch at the daycare center during the 2 days of data collection; meaning that data from a total of 309 children were used to assess a relationship between the estimated food reluctance score and children’s body composition. They were 4.10.9 years old, and most were French-speaking (83.2%) and from urban areas (71.8%). In our sample, the mean food reluctance score was 0.40 with a standard deviation of 0.49 and a median of 0.44, and 71% of children demonstrated some food reluctance (estimated score >0%). Vegetables was the food group for which children demonstrated the most reluctance. More specifically, 54% of children who demonstrated food reluctance did so by eating relatively less of the vegetables they were served than food --
2017 Volume
-
Number
-
--
2017 Volume
-
Number
-
Univariate b from linear regressions for each separate variable to multivariate b from linear regressions including all variables. The proxy measure of food reluctance can range from 0 to 1, with higher values suggesting presentation of higher food reluctance. The ratio between amount of food leftover and amount served is calculated for five food groups. The largest of these five proportions is used as the proxy measure of food reluctance. *P<0.05. b
a
1.00 (0.63 to 2.62), 0.2
0.45 (1.02 to 1.92), 0.5 0.73 (0.55 to 2.00), 0.3 0.66 (0.05 to 1.36), 0.2 Parents’ education, 1 parent university educated
0.45 (0.40 to 1.30), 0.5
1.02 (2.98 to 0.95), 0.3
0.16 (1.24 to 1.57), 0.8
0.07 (0.49 to 0.63), 0.9 1.15 (2.86 to 0.55), 0.2 0.51 (1.08 to 0.07), 0.5
0.07 (0.55 to 0.69), 0.9
Language, French
Urban/rural setting, urban
1.24 (0.61 to 1.86), 0.04*
2.35 (1.12 to 3.58), <0.01*
0.48 (0.95 to 1.91), 0.5 0.29 (0.52 to 1.09), 0.6 0.95 (0.29 to 2.19), 0.1
2.27 (1.28 to 3.26), <0.01* 0.58 (0.49 to 1.65), 0.3 0.40 (0.49 to 1.29), 0.4
0.75 (0.06 to 1.56) 0.2 Sex, female
Age, y
0.60 (2.06 to 0.86), 0.4 0.34 (1.51 to 0.83), 0.6 1.41 (2.67 to 0.15), 0.03*
b (95% CI), P value!
1.17 (2.32 to 0.02), 0.04* Food reluctance,b score
Multivariate
Waist Circumference
Univariate Multivariate Univariate
This study is the first to introduce an estimate of food reluctance based on plate waste, which is a widely used and accepted method of assessing diet.53 The new estimate of food reluctance fills a gap given previous investigations of preschoolers’ eating behaviors relied on parent-reported measures.5,11,24,27,32 Parents are generally not present at daycare centers for prolonged periods, making them imperfectly informed of what their children are eating and how they are behaving in that specific environment. This limitation further amplifies the necessity to develop an estimate of food reluctance that can be used outside of the household setting. In addition to feasibly being administered in daycare centers, the food reluctance estimate demonstrated excellent reliability. Assessment of validity for this type of measure was complicated by the absence of gold-standard measures to assess food reluctance. However, concurrent validity was suggested by demonstrating an acceptable correlation with the food-reluctance related items of the CEBQ. Given the subjective parent-reported nature of the CEBQ and the fact that it is based on eating behavior at home, it was expected that correlations between the objective and CEBQ-based measures would not be perfect. This might be related, in part, to measurement errors potentially attributable to both instruments, but most particularly to the CEBQ, as the inherent limitation associated with self-reported measures is
Variable
DISCUSSION
Age-Adjusted BMI
from the other food groups. In contrast, fruits were the food group with the highest relative proportion of consumption, contributing to only 10% of the estimated cases of food reluctance, followed by milk and milk alternatives (17%), grains and starches (18%), and meat and meat alternatives (21%). The sum of percentages exceeds 100% because some participants demonstrated the same level of food reluctance toward more than one food group. Results from the univariate models show a negative relationship between the estimated food reluctance score and age-adjusted BMI among preschoolers, meaning that greater demonstration of food reluctance at the daycare center was significantly related to lower age-adjusted BMI among preschoolers. No other variable was significantly related to BMI in univariate analyses. Age and an urban (vs rural) setting were significantly positively related to preschooler’s waist circumference, but there was no apparent association with the estimate of food reluctance (Table 3). In multivariate models, the estimated food reluctance score remained significantly related to lower age-adjusted BMI, but only age remained positively related to waist circumference. Specifically, this model suggests that for every point increase in the food reluctance score, age-adjusted BMI decreased by 1.41. For example, when comparing average children from this study (food reluctance score of 0.40) to children who do not demonstrate food reluctance (food reluctance score of 0), it is estimated that, on average, children demonstrating food reluctance have a 0.56 lower age-adjusted BMI (1.410.40). In an additional series of analyses, we also noted negative relationships between the CEBQ-based parent-reported measure of food reluctance and both age-adjusted BMI and waist circumference, but none of these analyses were statistically significant at the traditional cut point of a¼.05 (P values ranged from 0.08 to 0.7).
Table 3. Univariate and multivariate relationship between the estimate of food reluctance of preschool children from the Healthy Start study and their age-adjusted body mass index (BMI) and waist circumferencea
RESEARCH
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
5
RESEARCH the distortion of the collected information caused by the subjective opinion of each respondent.54 In studies such as this one where parents assess their child’s behaviors, variables such as parental health, family functioning, and socioeconomic status are reported as being critical factors influencing parents’ views and opinions, potentially causing reporter bias in the data collected on their child.33 An important finding of this analysis is that higher estimated food reluctance displayed at lunchtime while attending a daycare center was significantly associated with lower age-adjusted BMI. This result is consistent with two other Canadian studies examining the relationship between parent-reported eating behaviors and weight status of preschoolers.10,32 In their study, Dubois and colleagues10 reported that children who were described as being picky eaters—a demonstration of food reluctance—were two times more likely to be underweight when compared with their nonpicky eater peers. Similarly, Spence and colleagues32 reported a negative linear relationship between weight and food reluctance. Our finding also lends credence to international studies, which reported that parent-reported food reluctance could be associated with lower BMI.12,15,16,23,29,31 Our results complement these previous studies by focusing on eating behavior at daycare. By documenting the presence of potential food reluctance in this setting and demonstrating that it is related to BMI, this study provides support to the hypothesis that daycare centers represent a promising environment for healthy eating and adequate weight gainpromoting interventions. However, we did not observe a relationship between our estimate of food reluctance and waist circumference. This is in contrast with findings from Carnell and Wardle,15 who observed a negative association between a parent-reported manifestation of satiety responsiveness and children’s waist circumference. Considering that cutoffs to indicate clinical risk based on waist circumference of children do not currently exist, these authors adjusted the waist circumference value collected from their sample in reference to 1990 UK data. Given no Canadian or international standards exist, we did not adjust our waist circumference data, which possibly explains the discrepancy between these previous results and our own. It is, however, noteworthy that our multivariate analyses showed that older children had greater waist circumference. This being said, it is likely that the nonsignificant relationship we observed between estimated food reluctance and waist circumference was confounded by age. Because children in this study were in a period of important growth, age likely confounded the relationship between food reluctance and waist circumference. This highlights the need for standardization norms for children’s waist circumference. Another potential explanation for the discrepancy between previous studies and current results may be that our estimate accounted for various demonstrations of food reluctance and not just satiety responsiveness, which was measured by Carnell and Wardle. In addition to the contributions described here, the current article has implications for research and practice. Our results suggest it is feasible to use plate waste in future studies, particularly among children attending daycare centers, given data could be acquired from a large number of participants within a relatively short period of time, it could be processed relatively rapidly by trained personnel, and it possessed 6
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
excellent intra- and inter-reliability. As technology evolves, the development of mobile applications for plate waste measurement could further facilitate the collection of food consumption and eating behavior data of children in institutional settings. In addition, our finding of a significant association between estimated food reluctance displayed in daycare settings and children’s age-adjusted BMI call for experimental research to test whether daycare centers are conducive environments where interventions can be implemented to improve eating behaviors among preschoolers.
Strengths and Limitations This study included a large sample of participants, was based on objective measures, and included many steps to verify the reliability and validity of the new measure. It therefore adds to our understanding of the relationship between food reluctance and body composition among preschoolers. Still, there are limitations to consider. First, even though a significant association was reported between food reluctance and children’s age-adjusted BMI, causality cannot be established, given the cross-sectional nature of the study. Second, because our sample was selected from mostly French-speaking participants in relatively large daycare centers, our findings may not be generalizable to all Canadian preschool children attending daycare. Third, although data were collected on separate days to account for children’s variable appetite,53 those days were consecutive and we did not account for or measure appetite, limiting our ability to capture and consider additional possible variability in dietary intake. Fourth, because data were only collected at lunchtime, other food eaten at daycare was not captured. Fifth, our estimate of food reluctance represents a proxy of the behavior that does not account for appetite or food quality and does not discern different behavioral aspects of eating, including slowness in eating, picky eating, and satiety response. Last, even though research assistants were instructed to remain distant and as indiscernible as possible during data collection, it is possible that children’s food intake was distorted because they knew they were being observed.
CONCLUSIONS In summary, this study introduces a new estimate of food reluctance that showed evidence of reliability and validity in preschool children. This study also provides support to the hypothesis that food reluctance displayed at daycare is related to lower BMI among preschoolers. Future research could build on these findings to test interventions aimed at reducing preschoolers’ food reluctance in daycare centers, and explore the long-term effects of these interventions on food reluctance, body composition, and health risks later in life.
References 1.
Dovey TM, Staples PA, Gibson EL, Halford JCG. Food neophobia and “picky/fussy” eating in children: A review. Appetite. 2008;50(2-3): 181-193.
2.
Tharner A, Jansen P, Kiefte-de Jong J, et al. Toward an operative diagnosis of fussy/picky eating: A latent profile approach in a population-based cohort. Int J Behav Nutr Phys Act. 2014;11:14-24.
3.
Wardle J, Guthrie CA, Sanderson S, Rapoport L. Development of the Children’s Eating Behaviour Questionnaire. J Child Psychol Psychiatry. 2001;42(7):963-970.
--
2017 Volume
-
Number
-
RESEARCH 4.
Carruth B, Ziegler P, Gordon A, Barr S. Prevalence of picky eaters among infants and toddlers and their caregivers’ decisions about offering a new food. J Am Diet Assoc. 2004;104(1 suppl 1):S57-S64.
5.
Dubois L, Farmer A, Girard M, Peterson K. Preschool children’s eating behaviours are related to dietary adequacy and body weight. Eur J Clin Nutr. 2007;61(7):846-855.
6.
Xue Y, Zhao A, Cai L, et al. Growth and development in Chinese preschoolers with picky eating behaviour: A cross-sectional study. PLoS One. 2015;10(4):e0123664.
7.
Jacobi C, Agras W, Bryson S, Hammer L. Behavioral validation, precursors, and concomitants of picky eating in childhood. J Am Acad Child Adolesc Psychiatry. 2003;42(1):76-84.
BMI: A comparison of black Afro-Caribbean, white British and white German samples. Eur J Clin Nutr. 2013;67(2):180-184. 28.
Gubbels J, Kremers S, Stafleu A, et al. Association between parenting practices and children’s dietary intake, activity behavior and development of body mass index: The KOALA Birth Cohort Study. Int J Behav Nutr Phys Act. 2011;8:18-30.
29.
Jansen P, Roza S, Jaddoe V, et al. Children’s eating behavior, feeding practices of parents and weight problems in early childhood: Results from the population-based Generation R Study. Int J Behav Nutr Phys Act. 2012;9:130-140.
30.
Joyce J, Zimmer-Gembeck M. Parent feeding restriction and child weight. The mediating role of child disinhibited eating and the moderating role of the parenting context. Appetite. 2009;52(3): 726-734.
31.
Mallan K, Liu W, Mehta R, Daniels L, Magarey A, Battistutta D. Maternal report of young children’s eating styles. Validation of the Children’s Eating Behaviour Questionnaire in three ethnically diverse Australian samples. Appetite. 2013;64:48-55.
32.
Spence J, Carson V, Casey L, Boule N. Examining behavioural susceptibility to obesity among Canadian pre-school children: The role of eating behaviours. Int J Pediatr Obes. 2011;6(2-2):e501-e507.
8.
van der Horst K, Eldridge A, Deming D, Reidy K. Caregivers’ perceptions about picky eating: Associations with texture acceptance and food intake. FASEB J. 2014;28(1):S379.3.
9.
Mascola AJ, Bryson SW, Agras WS. Picky eating during childhood: A longitudinal study to age 11-years. Eat Behav. 2010;11(4):253-257.
10.
Dubois L, Farmer A, Girard M, Peterson K, Tatone-Tokuda F. Problem eating behaviors related to social factors and body weight in preschool children: A longitudinal study. Int J Behav Nutr Phys Act. 2007;4:9-18.
11.
Pinhas-Hamiel O, Bar-Zvi E, Boyko V, Reichman B, Lerner-Geva L. Prevalence of overweight in kindergarten children in the centre of Israel—Association with lifestyle habits. Child Care Health Dev. 2009;35(2):147-152.
33.
Collishaw S, Goodman R, Ford T, Rabe-Hesketh S, Pickles A. How far are associations between child, family and community factors and child psychopathology informant-specific and informant-general? J Child Psychol Psychiatry. 2009;50(5):571-580.
12.
Ekstein S, Laniado D, Glick B. Does picky eating affect weight-forlength measurements in young children? Clin Pediatr (Phila). 2010;49(3):217-220.
34.
13.
Halterman J, Kaczorowski J, Aligne C, Auinger P, Szilagyi P. Iron deficiency and cognitive achievement among school-aged children and adolescents in the United States. Pediatrics. 2001;107(6):13811386.
Wolper C, Heshka S, Heymsfield S. Measuring food intake: An overview. In: Allison D, ed. Handbook of Assessment Measures for Eating Behaviors and Weight-Related Problems. Thousand Oaks, CA: Sage Publishing; 1995:215-240.
35.
Jacko C, Dellava J, Ensle K, Hoffman D. Use of the plate-waste method to measure food intake in children. J Ext. 2007;45(6):6RIB7.
36.
14.
Liuba P, Pesonen E. Infection and early atherosclerosis: Does the evidence support causation? Acta Paediatr. 2005;94(6):643-651.
Blakeway S, Knickrehm M. Nutrition education in the Little Rock School Lunch Program. J Am Diet Assoc. 1978;72(4):389-391.
37.
15.
Carnell S, Wardle J. Appetite and adiposity in children: Evidence for a behavioral susceptibility theory of obesity. Am J Clin Nutr. 2008;88(1):22-29.
Whatley J, Donnelly J, Jacobsen D, Hill J, Carlson M. Energy and macronutrient consumption of elementary school children served modified lower fat and sodium lunches or standard higher fat and sodium lunches. J Am Coll Nutr. 1996;15(6):602-607.
16.
Mackenbach J, Tiemeier H, Ende J, et al. Relation of emotional and behavioral problems with body mass index in preschool children: The Generation R study. J Dev Behav Pediatr. 2012;33(8):641-648.
38.
Lee H, Lee K, Shanklin C. Elementary students’ food consumption at lunch does not meet recommended dietary allowance for energy, iron, and vitamin A. J Am Diet Assoc. 2001;101(9):1060-1063.
17.
Mallan KM, Nambiar S, Magarey AM, Daniels LA. Satiety responsiveness in toddlerhood predicts energy intake and weight status at four years of age. Appetite. 2014;74:79-85.
39.
Branen L, Fletcher J. Effects of restrictive and self-selected feeding on preschool children’s food intake and waste at snacktime. J Nutr Educ. 1994;26(6):273-277.
18.
Eriksson J, Forsén T, Tuomilehto J, Osmond C, Barker D. Early growth and coronary heart disease in later life: Longitudinal study. Br Med J. 2001;322(7292):949-953.
40.
Ireton C, Guthrie H. Modification of vegetable-eating behavior in preschool children. J Nutr Educ. 1972;4(3):100-103.
41.
19.
Kuhl E, Clifford L, Stark L. Obesity in preschoolers: Behavioral correlates and directions for treatment. Obesity. 2012;20(1):3-29.
20.
Anzman S, Rollins B, Birch L. Parental influence on children’s early eating environments and obesity risk: Implications for prevention. Int J Obes. 2010;34(7):1116-1124.
Bélanger M, Humbert L, Vatanparast H, et al. A multilevel intervention to increase physical activity and improve healthy eating and physical literacy among young children (ages 3-5) attending early childcare centres: The Healthy Start-Départ Santé cluster randomised controlled trial study protocol. BMC Public Health. 2016;16: 313-322.
21.
Birch L, Ventura A. Preventing childhood obesity: What works? Int J Obes. 2009;33(suppl 1):S74-S81.
42.
22.
Bronfenbrenner U, Morris P. The ecology of developmental processes. In: Damon W, Lerner RM, eds. Handbook of Child Psychology: Volume 1: Theoritical Models of Human Development. 5th ed. Hoboken, NJ: John Wiley & Sons; 1998:993-1028.
National Health and Nutrition Examination Survey. Body Composition Procedures Manual. http://www.cdc.gov/nchs/data/nhanes/ nhanes_03_04/BC.pdf. Revised January 2004. Accessed January 1, 2014.
43.
Cole T, Bellizzi M, Flegal K, Dietz W. Establishing a standard definition for child overweight and obesity worldwide: International survey. Br Med J. 2000;320(7244):1240-1243.
23.
Hittner J, Faith M. Typology of emergent eating patterns in early childhood. Eat Behav. 2011;12(4):242-248.
44.
24.
Johannsen D, Johannsen N, Specker B. Influence of parents’ eating behaviors and child feeding practices on children’s weight status. Obesity. 2006;14(3):431-439.
Lobstein T, Baur L, Uauy R; IASO International Obesity TaskForce. Obesity in children and young people: A crisis in public health. Obes Rev. 2004;5(suppl 1):4-104.
45.
Fredriks A, van Buuren S, Fekkes M, Verloove-Vanhorick S, Wit J. Are age references for waist circumference, hip circumference and waisthip ratio in Dutch children useful in clinical practice? Eur J Pediatr. 2005;164(4):216-222.
46.
Williamson D, Allen H, Martin P, Alfonso A, Gerald B, Hunt A. Comparison of digital photography to weighed and visual estimation of portion sizes. J Am Diet Assoc. 2003;103(9):1139-1145.
47.
Powers S, Chamberlin L, van Schaick K, Sherman S, Whitaker R. Maternal feeding strategies, child eating behaviors, and child BMI in low-income African-American preschoolers. Obesity. 2006;14(11): 2026-2033.
25.
26.
27.
--
Sinha M, Bleakney A. Spotlight on Canadians: Results from the General Social Survey. Receiving care at home. http://www.statcan. gc.ca/pub/89-652-x/89-652-x2014002-eng.htm. Published June 2014. Accessed November 15, 2014. Larson N, Ward D, Neelon S, Story M. What role can child-care settings play in obesity prevention? A review of the evidence and call for research efforts. J Am Diet Assoc. 2011;111(9):1343-1362. Blissett J, Bennett C. Cultural differences in parental feeding practices and children’s eating behaviours and their relationships with child
2017 Volume
-
Number
-
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
7
RESEARCH 48.
Statistics Canada. Canadian Community Health Survey (CCHS): Annual Component2010 Questionnaire. http://www23.statcan.gc. ca/imdb-bmdi/instrument/3226_Q1_V7-eng.pdf. Published July 2010. Accessed September 29, 2013.
49.
Community Information Database. Metropolitan Influence Zones (MIZ) typology 2006. http://map.cid-bdc.ca/#s¼2006;i¼comtype. miz;sly¼can_sdr_DR;sid¼689;v¼map1;l¼en;z¼2149095,322536, 269304,188240. Published 2006. Accessed September 5, 2013.
50.
Roberts K, Shields M, de Groh M, Aziz A, Gilbert J. Overweight and obesity in children and adolescents: Results from the 2009 to 2011 Canadian Health Measures Survey. Health Rep. 2012;23(3):37-41.
51.
IBM SPSS Statistical Software [computer program]. Version 22.0. Armonk, NY: IBM Corp; 2013.
52.
McIntire S, Miller L. Foundations of Psychological Testing. 2nd edition. Thousand Oaks, CA: Sage Publishing; 2005.
53.
Yon B, Johnson R, Stickle T. School children’s consumption of lowercalorie flavored milk: A plate waste study. J Acad Nutr Diet. 2012;112(1):132-136.
54.
Barker C, Pistrang N, Elliott R. Research Methods in Clinical Psychology: An Introduction for Students and Practitioners. 2nd ed. West Sussex, UK: John Wiley & Sons; 2002.
AUTHOR INFORMATION V. Surette is a registered dietician and territory manager, Nestlé Canada, Montreal, Quebec; at the time of the study, she was a graduate student, Faculty of Medicine and Health Sciences, Université de Sherbrooke, Sherbrooke, Canada. S. Ward is assistant professor of nutrition, Université de Moncton, Moncton, Canada; at the time of the study, she was a PhD candidate, Faculty of Medicine and Health Sciences, Université de Sherbrooke, Sherbrooke, Canada. P. Morin is an associate professor, Faculty of Physical Activity Sciences, Université de Sherbrooke, Sherbrooke, Canada. H. Vatanparast is an associate professor, University of Saskatchewan, School of Public Health, Saskatoon, Canada. M. Bélanger is an associate professor and director of research, Université de Sherbrooke, Centre de formation médicale du Nouveau-Brunswick, Moncton, Canada, and an epidemiologist, Réseau de santé Vitalité, Bathurst, Canada. Address correspondence to: Mathieu Bélanger, PhD, Université de Sherbrooke, Centre de formation médicale du Nouveau-Brunswick, 100, rue des Aboiteaux, Moncton, New Brunswick, Canada, E1A 3E9. E-mail: [email protected]
STATEMENT OF POTENTIAL CONFLICT OF INTEREST No potential conflict of interest was reported by the authors.
FUNDING/SUPPORT Healthy Start/Départ Santé is financially supported by grants from the Public Health Agency of Canada, the Consortium national de formation en santé, the Heart and Stroke Foundation of Canada and the Community Initiative Fund from the Government of Saskatchewan. Véronique Surette was funded through a Faculté de médecine et des sciences de la santé de l’Université de Sherbrooke Graduate Scholarship and Stéphanie Ward was funded through a Canadian Institutes of Health Research Charles Best Canada Graduate Scholarships Doctoral Award and a Gérard-Eugène-Plante Doctoral Scholarship.
ACKNOWLEDGEMENTS The authors gratefully acknowledge the important contributions of Anne Leis, PhD, principal investigator; Gabrielle Lepage-Lavoie, coordinator; the Réseau santé en français de la Saskatchewan; the daycare centers; and the participants in the Healthy Start/Départ Santé study.
8
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
--
2017 Volume
-
Number
-