Comparison of measured and parents' reported height and weight in children and adolescents

The objectives of this study were to (i) compare parent-reported height and weight to measured height and weight in children between ages 2 and 17 years, (ii) investigate correlations between magnitude of error of parent-reported data or refusal to estimate height and weight with gender, race/ethnicity, child's age, and age-specific BMI z-score, and (iii) determine sensitivity and specificity of identifying obese youth based on parent-reported data. The authors studied 1,430 consecutive outpatients between ages 2 and 17 years at an outpatient orthopedic clinic. At the initial visit, parents completed a questionnaire including their child's height and weight; height and weight were then measured. Mean height error was very small, with slight overestimation in boys and underestimation in girls. Mean weight error increased with age (P < 0.001), and girls had a larger mean weight error (1.29 kg, 95% confidence interval (CI): 0.65, 1.45) than boys (0.85 kg, 95% CI: 0.8: 0.58, 1.12). Mean weight error also increased with age-specific BMI z-score (r = 0.32, P < 0.001). Correlation between weight error and age-specific BMI z-score was higher among black children (r = 0.45, P < 0.001) than among Hispanic children (r = 0.37, P < 0.001) and was lowest among white children (r = 0.29, P < 0.001). Refusal or inability to estimate weight did not correlate with age-specific BMI z-score. Twenty-one percent of children who were obese would not be identified by using parent-reported data to calculate the BMI.