The Accuracy of Height Prediction Equations in Greek Patients: A Cross-Sectional Study.

Abstract

In clinical settings, standing height measurement is often difficult to perform due to patients' inability to stand upright. Height prediction equations derived from measurements of the length of other body segments have been published; however, they are not readily applicable to all populations since ethnic differences affect the relationship between standing height and body segment length. This cross-sectional study aimed to examine the accuracy of height prediction using the Malnutrition Universal Screening Tool (MUST) height predictive equations among Greek patients and to develop new, nationally representative equations. The study population consisted of 1198 Greek adult outpatients able to stand upright without assistance and without medical conditions that affected their height. Standing height, ulna length, knee height and demi-span measurements were obtained from 599 males and 599 females. Patients were stratified into age groups of <55 and ≥55 years, <60 and ≥60 years and <65 and ≥65 years according to the categories indicated by the MUST for height prediction from alternative measurements. There were positive correlations between standing height and ulna length and knee height and demi-span length (p < 0.001) in both sexes and all age categories. A strong correlation was observed between the measured and predicted standing height using ulna length (rho = 0.870, p < 0.001), knee height (rho = 0.923, p < 0.001) and demi-span length (rho = 0.906, p < 0.001). The average difference between the MUST indicative equations' height predictions from alternative measurements and actual height was -3.04 (-3.32, -2.76), -1.21 (-1.43, -0.988) and 2.16 (1.92, 2.41), respectively. New height prediction equations for Greek patients were identified, with the predicted values closer to the measured standing heights than those predicted with the MUST indicative equations for height prediction from alternative measurements.