Abstract

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.










Background
Keywords
Findings
Authors
What is already known on this subject?
What does this study add?

Background

Studies based on adolescents and adults have found that the use of self-reported height and weight to calculate body mass index (BMI) yields a lower prevalence of obesity than do estimates based on measured data.  Relatively few studies have examined the bias resulting from the use of parent-reported height and weight for children, and the findings have been inconsistent.

Data and methods

Data are from the 2007 to 2009 Canadian Health Measures Survey.  Parent-reported height and weight of children aged 6 to 11 (n=854) were obtained.  Subsequently, the children's height and weight were directly measured.

Results

On average, parents underestimated the height (3.3 cm) and weight (1.1 kg) of their children.  Estimates of the prevalence of obesity were significantly higher when based on parent-reported versus measured values for children aged 6 to 8; the two collection methods yielded similar estimates of obesity for children aged 9 to 11.  For children in both age groups, misclassification errors for BMI categories were substantial when based on parent-reported values.  This weakened associations between obesity and health indicators such as aerobic fitness and systolic blood pressure.  The variance explained by factors associated with the bias in parent-reported height and weight was small, particularly for height.  The use of correction equations based on variables associated with the bias resulted in a very modest reduction in misclassification errors.

Interpretation

Bias associated with parental reports of children's height and weight results in misclassification errors for obesity that affect relationships with other variables.  Efforts to establish correction equations to adjust for this bias were unsuccessful.  Direct measures are required to accurately calculate obesity estimates and their relationships with health indicators in children.

Keywords

bias, body mass index, direct measure, measurement error, misclassification, sensitivity, specificity, validity

Findings

Over the past 25 years, the prevalence of obesity among Canadian children, adolescents and adults has increased substantially, mirroring a worldwide phenomenon. Monitoring trends in obesity is essential to assess interventions aimed at preventing or reducing obesity in children. [Full Text]

Authors

Margot Shields (1-613-951-4177; margot.shields@statcan.gc.ca) is with the Health Analysis Division at Statistics Canada, Ottawa, Ontario, K1A 0T6.  Sarah Connor Gorber is with the Public Health Agency of Canada, Ottawa, Ontario.  Ian Janssen is with Queens University, Kingston, Ontario.  Mark S. Tremblay is with the Children's Hospital of Eastern Ontario Research Institute and the University of Ottawa, Ottawa, Ontario.

What is already known on this subject?

  • Studies based on adults have found that self-reported values underestimate weight and overestimate height, resulting in lower estimates of obesity than those obtained from measured data.
  • Results of the few studies of measures of height and weight among children are inconsistent, and the implications are poorly investigated.

What does this study add?

  • In the 2007 to 2009 Canadian Health Measures Survey, parents underestimated the height and weight of children aged 6 to 11, which resulted in an overestimate of body mass index (BMI) among children aged 6 to 8.
  • Use of parent-reported height and weight resulted in substantial misclassification errors in prevalence estimates by BMI category.
  • The misclassification that occurred with parent-reported values weakened associations between obesity and other variables such as aerobic fitness and systolic blood pressure.
  • Efforts to establish correction equations to adjust for the bias in parent-reported data were ineffective.