 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
ARTICLE |
a Weight and Eating Disorders Program, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
b Department of Child and Adolescent Psychiatry, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania
c Bureau of Health Risk Reduction, Division of Chronic Disease Prevention and Adult Health, New York State Department of Health, Albany, New York
d Research Institute, Bassett Healthcare, Cooperstown, New York
e Division of Nutrition, New York State Department of Health, Albany, New York
f Department of Epidemiology and Biostatistics, School of Public Health, University at Albany, State University of New York, Albany, New York
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
PATIENTS AND METHODS. A sample of parents or guardians of children aged 1 to 4 years who attended 1 of 49 Special Supplemental Nutrition Program for Women, Infants, and Children agencies in New York State were surveyed in 1999 or 2000 (N = 2801). The survey addressed children's dietary intake, parental feeding practices, and parental exposure to nutritional counseling messages to increase fruit, vegetable, and low-fat milk intakes. Each child's height and weight were measured approximately every 6 months for up to 48 months. A prospective cohort design was used in which survey variables were the predictors and the outcome was change in children's adiposity, defined as change in age- and gender-standardized BMI per month (ie, BMI z-score slope).
RESULTS. Controlling for gender and ethnicity, the relationship between juice intake and adiposity gain depended on children's initial overweight status. Among children who were initially either at risk for overweight or overweight, increased fruit juice intake was associated with excess adiposity gain, whereas parental offerings of whole fruits were associated with reduced adiposity gain. Each additional daily serving of fruit juice was associated with an excess adiposity gain of 0.009 SD per month. Feeding restriction was greater among parents whose children were initially at risk for overweight or overweight compared with those at a healthy weight. Parental exposure to nutritional messages was not associated with reduced child adiposity gain.
CONCLUSION. This study supports the Institute of Medicine recommendations to reduce fruit juice intake as a strategy for overweight prevention in high-risk children.
Key Words: obesity • child nutrition • feeding behavior • primary prevention • ethnic groups • fruit
Abbreviations: WIC—Special Supplemental Nutrition Program for Women, Infants, and Children • IOM—Institute of Medicine
Childhood overweight prevention is a national priority because of its rising prevalence,1 comorbidities,2 and persistence into adulthood.3 Modifying children's intake of fruit juice, as well as fruits and vegetables, may be viable targets for overweight prevention,4 although data linking these variables to child BMI (kg/m2) are inconsistent.5 Dennison et al6 showed that low- to middle-class preschool-aged children who consumed >12 oz/day of fruit juice were significantly more likely to have a BMI above either the 75th or 90th percentile than children who consumed <12 oz/day. This association was subsequently related to increased intake of apple, but not orange or grape, juice.7 Tanasescu et al8 also showed higher fruit juice intake among overweight compared with nonoverweight preschool-aged children. Skinner et al,9 however, found no significant association between fruit juice intake and child overweight, although trends were similar to those reported by Dennison et al.6 Fruit, vegetable, and fruit juice intake were not associated with 3-year changes in BMI z score in a study of more than 14000 youth,10 and among moderate- to high-income white children, fruit juice intake decreased between 2 and 6 years of age and was not associated prospectively with childhood overweight.11 Finally, the association between juice intake and weight gain may be limited to children who are at risk for overweight or are already overweight.12
Parental feeding styles are another potential target for overweight prevention, because restrictive feeding practices have been linked to increased children's food intake and to weight status.13 However, with few exceptions,14 these studies were conducted in predominantly middle-class, white families. No study has examined whether parental feeding attitudes and practices predict adiposity gain in children from low-income families. Any association between feeding restriction and child weight status may also depend on the child's initial weight status. The association between restriction and children's adiposity gain was documented in children who had a family history of obesity but not in children having a family history of thinness.15
In our study we tested the association between children's diet and beverage intake, parental feeding restriction and attitudes, and changes in children's adiposity over 18 to 48 months. We studied an ethnically diverse, low-income sample of children enrolled in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC). Children from low-income families are at increased risk for overweight compared with children from middle-class or more affluent families,16–18 possibly attributable, in part, to limited resources to purchase healthier but more costly foods.19,20 We predicted that increased children's fruit juice intake and parental restriction of children's eating would be associated with increased adiposity gain. We also tested whether any observed association depended on the children's baseline overweight status.12,15 As a secondary aim, we explored whether parental exposure to WIC counseling to increase fruit, vegetable, or low-fat milk intake was associated with reduced children's adiposity gain.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The institutional review boards of the Mary Imogene Bassett Hospital and the New York State Department of Health approved this study. WIC staff removed all identifying information from the surveys before data entry.
Survey
At each participating WIC agency, questionnaires written in both English and Spanish were distributed to eligible parents. Self-administered questionnaires were distributed between July and December 1999 and May and December 2000. Records with missing race/ethnicity were excluded (n = 33), yielding a baseline survey sample of 1797. The parent self-reported race/ethnicity (Hispanic, non-Hispanic black, non-Hispanic white, Asian, Native American, multicultural, or other), relationship to the child, and last grade of school he or she completed.
In the 1999 and 2000 surveys, the parent reported the usual number of servings per day of milk, fruit juice, fruit, and vegetables (excluding carrots, potatoes, or salad) that the child consumes. The fruit and vegetable questions were adapted from the Behavioral Risk Factor Surveillance System questionnaire.21 When the survey for this study was piloted, parents rarely reported that their child consumed any salad; therefore, this question was omitted. A single serving was defined as cup of milk or juice and cup of fruit, vegetables, potatoes, or carrots for this age group (1–5 years). The parent also reported the type(s) of milk the child usually drinks (whole, 2%, 1%, and/or skim milk). In addition, parents completed 4 questions pertaining to their feeding style: The first question asked, "How often do you limit how much this child eats?" Response options, scored on a 1-to-5 scale, were "never," "rarely," "sometimes," "usually," and "always." The second question asked, "Do you agree with the statement ‘Children need to finish dinner before dessert’?" Response options, scored on a 1-to-5 scale, were "strongly disagree," "disagree," "neutral," "agree," and "strongly agree." Questions 3 and 4, respectively, asked, "Have you tried offering this child more fruit to eat?" and "Have you tried offering this child more vegetables to eat?" Response options were "yes" or "no."
For the year 2000 survey only, parents were asked the following 3 questions about their exposure to WIC nutrition counseling services: "Has anyone at WIC discussed serving this child low-fat milk?"; "Has anyone at WIC discussed serving this child more fruits?"; and "Has anyone at WIC discussed serving this child more vegetables?" Response options were "yes" or "no." Data for these 3 questions were only available for about half of the sample and were used for our exploratory research question.
Anthropometric Measurements
Each child's most recent height or length, weight, and date of measurement were abstracted from his or her WIC chart at the time of the survey. Children are required to have their height and weight measured every 6 months as part of the WIC certification process. WIC staff measured each child, frequently on the same day as the survey was completed. Weight was measured by using a standard balance-beam scale. Height for children >36 inches tall or >2 years of age was measured by using a wall-mounted stadiometer. Otherwise, child length was measured by using a recumbent board; child length was converted to child height by subtracting 0.5 cm.
Additional data were obtained for each study child beginning in December 2001 through September 2002. WIC staff abstracted height, weight, and measurement data from paper charts and/or computer records subsequent to the intake survey date. These data were linked to the baseline data.
BMI was computed for children >2 years of age as weight in kilograms divided by the square of the height in meters (kg/m2). Age- and gender-specific BMI percentiles and BMI z scores, as well as weight-for-height z scores, were computed from the 2000 Centers for Disease Control and Prevention growth charts for the United States.22 Weight-for-height z score was computed to control for children's initial weight status in regression analyses.
To more accurately predict change in adiposity, we included only children who had at least 3 measures after the age of 2 years and both a height and weight measurement at baseline (N = 971). Specifically, 382 children had BMI calculated for 3 time points, 297 for 4 time points, 205 for 5 time points, 72 for 6 time points, 12 for 7 times points, 1 for 8 time points, and 2 for 9 time points.
Statistical Analysis
Data were analyzed by using SAS 8.2 (SAS Institute, Inc, Cary, NC). Proportions or means ± SD characterized the study sample. 
2 statistics, t tests, or analysis of variance assessed differences in characteristics between participants included in the study sample and those excluded and were used to test differences in study variables by child race/ethnicity, gender, and baseline overweight status (defined later in this article).
Certain predictor variable scores were recoded for data analyses: race/ethnicity was collapsed into Hispanic, non-Hispanic black, non-Hispanic white, and other. Educational attainment was collapsed into "less than high school," "high school or general equivalency diploma," or "more than high school." Type of milk was collapsed into a dichotomous variable indicating high-fat (whole or 2%) or low-fat (1% or skim) milk.
The main outcome variable of interest, change in adiposity, was defined as change in age- and gender-standardized BMI per month (ie, BMI z-score slope). This change score was estimated individually by linear regression for each child.23,24 Specifically, BMI z score was regressed on child's age in months at the time of assessment to derive the BMI z-score slope. Multiple regression models tested the prospective association between children's food and beverage intake, as well as parental feeding styles and attitudes variables, with change in children's adiposity. Because the number of data points used to compute change in BMI z score differed across children, we used the log of precision of the estimated slope (ie, 1/variance of BMI z-score slope) as a weighting factor. Baseline weight-for-height z score was statistically controlled in regression models to adjust for children's initial weight status. Initial weight-for-height z score was controlled, rather than initial BMI z score, because some children were <2 years of age at their first assessment. Weight for height and BMI are highly correlated with similar sensitivity and specificity for predicting overweight in preschool-aged children.25
Weight-for-height z score was also used to categorize children with respect to "overweight status" at initial assessment. This was a dummy-coded variable for which children were scored either at 0 (ie, weight-for-height z score <85th percentile) or 1 (ie, weight-for-height z score 
85th percentile). The former category was considered "neither at risk for overweight nor overweight" at baseline, whereas the latter category was considered "at risk for overweight or already overweight" (referred to as at risk/overweight) at baseline.
The primary regression model included as predictors all child food- and beverage- intake variables, all parenting behavior variables, and initial child weight-for-height z score. Next, an interaction model tested whether any significant predictor from the primary regression model interacted with child baseline overweight status. This model included the same predictors from the primary model (although the dummy-coded overweight status variable replaced the continuous weight-for-height z-score variable) plus appropriate 2-way interaction terms between baseline overweight status and any significant predictor from the primary regression model. A significant interaction indicated that the association between the predictor variable and adiposity change depended on child baseline overweight status. The interaction model may also be considered a moderator analysis. We followed up significant interactions with stratified regression analyses testing children who were (n = 655) versus those who were not (n = 316) at risk/overweight at baseline.
To test the secondary research question, additional regression models added the 3 questions concerning parental exposure to WIC discussions about increasing children's fruit, vegetable, or low-fat milk intake. Alpha was set at .05 (2-tailed) for all analyses.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
Parental Feeding Practices
Reported parental child-feeding practices and attitudes differed by race/ethnicity (Table 3). Compared with the other groups, parents of non-Hispanic white children reported being less restrictive and were less inclined to agree that children should be encouraged to finish eating dinner before dessert (P < .05 for all comparisons; Fig 1).
|
Predictors of Change in Child BMI z Score
In the primary multiple-regression model, increased juice intake was associated with excess adiposity gain (P < .01); each incremental serving per day was associated with an additional BMI z-score gain of 0.005 SD per month (see Table 4). No other predictor was significant. In the interaction model, the main effect of juice intake was no longer statistically significant (P = .79), although the overweight status-by-juice intake interaction was significant (P = .01). These results, including the significant interaction (P = .01), did not change when modeling a random-effect term for WIC site.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Among high-risk children, excess juice consumption may promote adiposity gain through passive overconsumption of total energy intake. Children generally maintain a consistent level of energy intake,26,27 although there is variability in "caloric compensation" ability.27 Moreover, beverage intakes seem to be less precisely regulated than solid foods in humans28–31 and animals.32–36 In young children, energy imbalances as subtle as 125.82 kJ/day can promote overweight,37 which could be easily achieved by juice intake given that a standard 6- to 8-oz. fruit juice box has 
418.4 kJ. Indeed, overconsumption of sugar-sweetened beverages has been associated with excess weight gain in studies with adults38 and children.39
At the same time, parental offering of whole fruit was associated with reduced adiposity gain among initially at-risk/overweight children. This pattern is consistent with a substitution phenomenon,40 such that increased fruit intake may have displaced juice intake. That fruits and vegetables can displace intake of energy-dense foods has been shown previously in children,4,41 although it has not been demonstrated that increasing healthy foods can displace fruit juice intake per se. If true, such findings would support American Academy of Pediatrics' recommendations that whole fruits be used as an alternative for juice intake.42 Controlled, clinical trials are needed to test the influence of reduced juice intake on weight-gain prevention and to determine whether reinforcing healthier foods or water can displace fruit juice intake.
It is noteworthy that there was a cross-sectional association between low-fat milk intake and child overweight status but no longitudinal association. This suggests that the prescription of low-fat milk intake by staff is a reaction to the child rather than a contributing factor to excess adiposity gain. Regarding parenting styles, parents of at-risk/overweight children reported greater feeding restriction than parents of healthy-weight children. This finding is consistent with a review13 that noted that restriction of children's eating was the only feeding domain consistently linked to increased child food intake and weight status. This has not been documented previously in children from low-income families, and there was a trend for restriction to be associated with greater adiposity gain in children who were initially at risk/overweight (P = .09) but not in other children (P = .64) in this sample. Similarly, we recently found that the association between feeding restriction and child BMI gain was only seen in children who had a familial history of obesity.15 Future studies examining the effects of feeding styles on children's adiposity gain should evaluate the moderating role of child weight status or familial obesity history.
A number of ethnicity differences were found in child and parent behaviors. As in earlier studies,43,44 mothers of white children reported less feeding restriction than did other mothers. We also observed ethnicity differences in children's dietary intake patterns, a relatively understudied issue. Kant45 reported that elevated intake of "low nutrient dense" foods was more common among non-Hispanic white than black or Hispanic children in the National Health and Nutrition Examination Survey cohort. There may be important sociocultural influences on children's food intake that have been overlooked because of the lack of validated assessment instruments.46,47
Nutrition education is an important component of WIC, and the New York State WIC program is currently training local WIC agency staff in client-centered nutrition education techniques and the "FITWIC" healthy lifestyle initiative.48 However, reported parental discussions with WIC staff about fruits, vegetables, and low-fat milk were not associated with change in children's adiposity in this study, a finding consistent with the literature.49 The foods and beverages provided to WIC participants may exert a more powerful influence on dietary intake than the information alone.
At the time of this study, federal guidelines allowed for up to 288 fl oz per month (9.5 fl oz per day) of fruit juice to be dispensed for children aged 1 to 5 years,50 which is twice the 4 to 6 fl oz per day guideline for children aged 1 to 6 years recommended by the American Academy of Pediatrics.42 This also exceeded the 1 serving per day maximum recommended in the Dietary Guidelines for Americans.51 Juice intake is greater among infants and children enrolled in WIC compared with those who are not.52 The findings of this study support the recent recommendations by the Institute of Medicine (IOM)53 and the proposed revisions by the US Department of Agriculture that the WIC food packages reduce the maximum amount of fruit juice dispensed and add fruits and vegetables.54
Our study has several limitations. First, parental report of children's dietary intake is subject to reporting inaccuracies.55 Parental feeding questions were assessed by only single-item questions for which validity data were not available. Second, data on total energy intake were not available, which would have been an informative covariate in regression analyses.10 Third, measures of physical activity were not included. Fourth, subjects in our study were families who were eligible for the WIC program and chose to participate; findings, therefore, cannot be generalized to those who, although eligible, choose not to participate. Also, the immigration status of the survey respondents was unknown, although immigration status has no impact on WIC eligibility.
These limitations are balanced against the study strengths, which include the large sample size, racial/ethnic diversity of the cohort, and the long-term prospective follow-up of adiposity.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Address correspondence to Myles S. Faith, PhD, Weight and Eating Disorders Program, University of Pennsylvania School of Medicine, 3535 Market St, 3rd Floor, Philadelphia, PA 19104. E-mail: mfaith{at}mail.med.upenn.edu
Views expressed in this report are not necessarily those of the New York State Department of Health or the US Department of Agriculture.
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
246.10 supplemental foods. Available at: www.fns.usda.gov/wic/lawsandregulations/WICRegulations-7CFR246.pdf. Accessed September 6, 2006This article has been cited by other articles:
 |
|
 |
|
  C. E. O'Neil and T. A. Nicklas A Review of the Relationship Between 100% Fruit Juice Consumption and Weight in Children and Adolescents American Journal of Lifestyle Medicine, July 1, 2008; 2(4): 315 - 354. [Abstract] [PDF]
|
|
|
|
 |
|
 T. A. Nicklas, C. E. O'Neil, and R. Kleinman Association Between 100% Juice Consumption and Nutrient Intake and Weight of Children Aged 2 to 11 Years Arch Pediatr Adolesc Med, June 1, 2008; 162(6): 557 - 565. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. C. Wang, S. N. Bleich, and S. L. Gortmaker Increasing Caloric Contribution From Sugar-Sweetened Beverages and 100% Fruit Juices Among US Children and Adolescents, 1988-2004 Pediatrics, June 1, 2008; 121(6): e1604 - e1614. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Stinson Clarity in Juice Classification Pediatrics, September 1, 2007; 120(3): 696 - 697. [Full Text] [PDF]
|
|
|
|
 |
|
 T. V. Kral, A. J Stunkard, R. I Berkowitz, V. A Stallings, D. D Brown, and M. S Faith Daily food intake in relation to dietary energy density in the free-living environment: a prospective analysis of children born at different risk of obesity Am. J. Clinical Nutrition, July 1, 2007; 86(1): 41 - 47. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. K. Newby Are Dietary Intakes and Eating Behaviors Related to Childhood Obesity? A Comprehensive Review of the Evidence J. Law Med. Ethics, March 1, 2007; 35(1): 35 - 60. [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
