Development of regression equations for estimating height and weight using body segments in Argentine children

RUIZ BRÜNNER, MARIA M.; BUTLER, CHARLENE; CUESTAS, EDUARDO

NUTRITION

ELSEVIER SCIENCE INC

Año: 2019 vol. 57 p. 122 - 126

0899-9007

Objectives: Body weight and height measurements are essential in children for assessing growth and nutrition, for the calculation of medication doses, and for the effectiveness of medical interventions. When direct measurements cannot be made, segmental measures can be used to estimate weight and height. The equations available to estimate height and weight, however, are limited. The aim of this study was to use segmental measures to develop equations for use in pediatric clinical practice. Methods: A cross-sectional study design was used to collect data from 861 healthy children (484 females and 377 males) ages 2 to 18 y to develop equations for estimating weight and height from midarm circumference (MAC) and knee?heel height (KH), respectively. A multi-linear regression model was used to develop the equations. Results: The high correlation between MAC and the actual weight and KH and height indicates strong agreement. Four equations were developed to estimate weight and height using segmental measures. 1. To estimate weight from MAC for females: W = 2.37 × MAC + 1.64 × age (y) ? 28.28. 2. To estimate weight for males: W = 2.54 × MAC + 1.82 × age (y) ? 32.73. 3. To estimate height from KH for females: H = 2.88 × KH + 0.15. 4. To estimate height from KH for males: H = 2.73 × KH + 0.21. Conclusions: MAC and KH can be used for estimation equations for weight and height with a very good predictive power. Sex and age were significant covariates in estimating weight. To predict height, only sex was needed to fit the model.