Abstract

The stature is an anthropometric parameter that allows the assessment of the physical health status of the elderly, estimate the obesity levels, the degree of protein energy malnutrition, and other risk factors. However, in patients with mobility problems or advanced diseases is complicated to make an accurate measurement of the stature. When there are difficulties to measure the stature, a linear regression model can be used with other anthropometric variables to estimate the stature. The goal of this study is to define stature prediction equations from the knee height and the age for the elderly of Costa Rica by sex. The study used data from the longitudinal survey CRELES, representative sample of the elderly of Costa Rica. The prediction coefficients estimated with CRELES were compared with the prediction coefficients estimated by the (external) models of Chumlea, Roche, y Steinbaugh (1985) and Palloni y Guend (2005). Most of the prediction coefficients estimated with CRELES are significantly different from the coefficients of the external models; also, the prediction coefficients of CRELES generate smaller residuals and pure errors than the external models. We recommend using the new prediction equations established in this study to estimate the stature of elderly of Costa Rica with physiological and mobility problems.

Keywords: stature, stature prediction equations, knee height, anthropometry