[Skip to Navigation]
Views 351
Citations 0
Comment & Response
December 9, 2019

Parent-Measured and Parent-Estimated Child’s Heights and Weights

Author Affiliations
  • 1Graduate School, Beijing University of Chinese Medicine, Beijing, China
  • 2Department of Pediatrics, China-Japan Friendship Hospital, Beijing, China
  • 3Institute of Clinical Medical Sciences, China-Japan Friendship Hospital, Beijing, China
JAMA Pediatr. 2020;174(2):205-206. doi:10.1001/jamapediatrics.2019.4719

To the Editor The study by Ohri-Vachaspati et al1 provides epidemiologic evidence of a practical consideration for the accuracy of heights and weights measured or estimated by parents as compared with the criterion-standard, professional measurement by analyzing 226 children from the New Jersey Child Health Study. The authors concluded that parent-estimated and parent-measured heights and weights in children are both effective in classifying weight status. If true, the findings of this study would be of public health importance because parent-estimated heights and weights are more feasible and less expensive in observing children growth and development than that measured. However, we are very concerned about the extrapolation and precision of this conclusion. On one hand, as stated by Weden et al,2 parent reporting error should be accounted for as underreporting of weights increased with child’s age and underreporting of heights decreased with age, especially for children aged 2 to 5 years. The results in the study by Ohri-Vachaspati et al1 need to be treated with caution, and a more detailed division of age is essential. Additionally, as revealed by the randomized clinical trial by Huybrechts et al,3 parent-reported weights were slightly more underestimated in the intervention than in the control group. On the other hand, we have noticed that sensitivity and specificity of parent-estimated values are more scattered than that of parent-measured values in classifying children’s weight status (Figure).1 To account for this demographic variance, an approach based on correlation coefficients may be more appropriated. It has been suggested that the weighted κ statistic may be calculated to determine agreement between parent-estimated and measured index body mass index status adjusted for chance.4 Alternatively, some studies adopted the Lin coefficient, where a value of ±1.0 indicates perfect concordance and a value of zero denotes nonconcordance.5 Given these concerns, we believe that the findings of this epidemiologic study are not generalizable. However, this study does underscore the need for further validations with a large sample size and careful comparison between parent-measured and parent-estimated heights and weights.