Parental Recognition of Child Obesity

Research conducted by: Debra Etelson, Donald A. Brand, Patricia A. Patrick, and Anushree Shirali

Case study prepared by: Robert F. Houser and Georgette Baghdady

Overview
With increasing public awareness of child obesity as a major public health problem, studies are showing that it has not translated into an increased awareness of obesity in one’s own child.  Dietary patterns and weight status in childhood tend to carry into adolescence and adulthood, promoting the onset of chronic and other diseases.  A key ingredient for combating childhood obesity is parental involvement and commitment.  However, this is predicated on whether or not parents can recognize overweight and obesity in their children.

This study examined parents’ perceptions of their children’s weight status, their understanding of the health risks of obesity relative to other conditions they may perceive as health risks, and their knowledge of some healthy eating practices.  Children’s actual weight status was expressed as their body mass index (BMI) percentile, as determined by the CDC growth charts based on age and sex.  According to the CDC growth charts for children, a child with a BMI percentile less than the 5th percentile is underweight; from the 5th to less than the 85th, a child is at a healthy weight; from the 85th to less than the 95th percentile, a child is overweight; and a BMI percentile equal to or greater than the 95th percentile, a child is considered to be obese. 

A visual analog scale was used to measure parents’ perceptions of their child’s weight.  The visual analog scale consisted simply of a 10-cm straight line anchored at the left end by the label “extremely underweight” and at the right end by the label “extremely overweight.”  A parent placed a mark along the line to indicate where they perceived their child’s weight to be.  The researchers interpreted the marks as percentiles in their analysis.

Questions to Answer
Do parents recognize when their children are overweight or obese?  Do parents who make incorrect judgments about healthy food practices also make incorrect judgments about their child’s weight status?

Design Issues
This study defines a parent’s perception of their child’s BMI percentile as “accurate” if their score on a visual analog scale fell within 30 points of the child’s true BMI percentile.  This wide range defining accuracy potentially allows for misclassification of a child’s weight status among normal, overweight, and obese categories.  For example, a parent who perceives their child’s weight status as being at the 80th percentile, i.e., in the normal range, when in reality the child is obese with a BMI percentile of 98, the parent’s assessment would be considered accurate by the operational definition used in this study.  The authors explain that they chose this definition to give parents as much leeway as possible in assessing their child’s weight on the visual analog scale.

Descriptions of Variables
Variable Description
Sex The sex of the participating parent’s child
Overwt_Obese

 Whether or not a child’s body mass index (BMI) is equal to or greater than the 85th percentile for the child’s age and sex, which means that the child is either overweight (85th to less than 95th percentile) or obese (95th percentile or above)
PA_overwt Parental attitude expressing level of concern if their child were overweight, measured on a 4-point Likert Scale.  In data analysis, the four categories were condensed into two categories:
0 = “not at all” or “a little” concerned
1 = “quite” or “extremely” concerned
PA-TV Parental attitude expressing level of concern if their child watched >20 hours of TV per week, measured on a 4-point Likert Scale. In data analysis, the four categories were condensed into two categories:
0 = "not at all" or "a little" concerned
1 = "quite" or "extremely" concerned
Accurate Whether or not the parent's perception of their child's weight status was accurate. Parent's perception was considered accurate if the BMI percentile it corresponded to fell within 30 points of the child's actual BMI percentile
Juice_boxes The amount of juice that a parent thinks is healthy for their child to drink each day (a juice box contains eight ounces). We condensed the original four response categories into two categories:
0 = "1 or 2 juice boxes per day"
1 = "3 to 8 juice boxes per day"
Fast_food_meals How often a parent feels it is okay to eat at fast-food restaurants. We condensed the original four response categories into two categories: 0 = "once a month"


References

Debra Etelson, D., Brand, D. A., Patrick, P. A., Shirali, A. (2003). Childhood obesity: Do parents recognize this health risk? Obesity Research, 11, 1362-1368.

 

Links

Etelson et al. article
BMI percentiles for children


Exercises
  1. What is the operational definition of "accurate parental perception" of a child's weight status used in this study?
  2. The following contingency table shows cell counts for the sex of the child versus whether or not the parent perceived their child's weight status accurately. Perform a Pearson Chi Square test based on p<0.05.

    Sex of child

    		 Accurate parental perception Inaccurate parental perception Total
    Female 24 19 43
    Male 17 23 40
    Total 41 42 83

    1. What percentage of parents accurately perceived their child's weight status?
    2. Were parents more likely to perceive accurately a daughter or son's weight status? Report your findings with relevant percentages and p value.
  3. The following contingency table shows cell counts for the actual weight status of a child versus parents' perception of their child's weight status for healthy weight versus overweight or obese categories. Perform a Pearson Chi Square test based on p<0.05.

    Child’s actual weight status category Accurate parental perception Inaccurate parental perception Total
    Not overweight or obese 34 11 45
    Overweight or obese 7 31 38
    Total 41 42 83

    1. What percentage of children were overweight or obese in this sample?
    2. Were parents of overweight or obese children more or less likely to perceive their child's weight status accurately compared with parents of children who were not overweight or obese?
  4. Was the visual analog scale able to detect a significant difference in parental perceptions between children who were truly not overweight or obese versus children who were truly overweight or obese? Perform an independent samples t test based on p<0.05 with the values in the following table and interpret your findings.

    Actual weight status category of child Perceived Percentile using the marks by parents on the visual analog scale
      Mean SD n
    Not overweight or obese (less than 85th percentile) 44.75556 12.99153 45
    Overweight or obese (equal to or greater than 85th percentile) 54.78947 12.17698 38

  5. Did parents who reported they would be concerned if their child watched >20 hours of TV per week also report that they would be concerned if their child were overweight?
    1. Perform a Pearson Chi Square test based on p<0.05 with the cell counts shown below. Report your findings with percentages and p value.
    2. Do your results suggest that parents are aware of a link between hours spent watching TV and a child's weight status? Why or why not? What factors might mediate such a link?

        Would NOT be concerned if child were overweight   Would be concerned if child were overweight   Total
      NOT Concerned if child were to watch >20 hours of TV per week   12 16 28
      Concerned if child were to watch >20 hours of TV per week   6 49 55
      Total   18 65 83

  6. Explore whether or not parents’ judgments about healthy food intake patterns are related to their judgments about their children’s weight status.  Perform Pearson Chi Square tests based on p<0.05 with the cell counts shown below in the two contingency tables.
    1. Juice intake frequency.
      Juice boxes per day   Accurate parental perception Inaccurate parental perception Total
      Healthy for their child to drink 1 or 2 juice boxes per day   27 29 56
      Healthy for their child to drink 3 to 8 juice boxes per day   13 12 25
      Total   40 41 81

    2. Fast-food meals frequency.
      Fast-food meals frequency   Accurate parental perception Inaccurate parental perception Total
      Okay to eat at fast food restaurants once a month   14 16 30
      Okay to eat at fast food restaurants one to five times a week   25 25 50
      Total   39 41 80

    3. Do parents who make incorrect assessments about healthy food intake patterns also make incorrect assessments about child weight status? Interpret the results you obtained in parts a. and b.
  7. What are some possible reasons why some of the respondents refused to answer the fast food and juice intake frequency questions?
  8. The ages of the children in this study ranged from 4 to 8 years (mean 5.8, SD 1.4). Why might parents of children in this age range have difficulty recognizing that their child is overweight or obese?
  9. Examine the regression output below, which was generated using the STATA statistical software. The STATA command is bolded.

    1. Write the simple linear regression equation predicting a parent's perceived percentile of their child's weight status from the child's actual percentile.
    2. What is the percent variation in parents' perceived percentiles explained by the children's actual percentiles?
    3. Do the data points show homoscedasticity or heteroscedasticity about the regression line?
    4. Does the regression output show some degree of parental recognition in regard to their children's weight status? Why or why not?