Thursday, March 11, 2004
A simple method of stature prediction is as accurate as Centers for Disease Control and Prevention (CDC) growth charts when based on single height measurements, and is similar in accuracy to other methods, according to researchers for Poster Exhibit P376. This method is also independent of percentile, race, nationality and generation, they report.
The investigators propose a new, universal, age-specific method to predict adult height, termed the Height Multiplier method. The purpose of this study was to calculate Height Multipliers, validate their use for height prediction and evaluate the method's universality.
The researchers used standard growth charts, based on a diverse population, which were published by the CDC in 2000. Height multipliers (M) for boys and girls were calculated by dividing the height at skeletal maturity (Htm) by present height (Ht) (M = Htm/Ht) for each age, gender and height percentile using CDC data. These multipliers were compared to multipliers derived from 28 boys' and 24 girls' height databases. Accuracy of multipliers was tested on individual longitudinal data from 52 normal children.
The average CDC-derived multipliers were significantly different at each age for boys and girls, but within gender, different percentiles at each age were very similar. These multipliers were very similar to multipliers derived from each of the girls' and boys' databases. Median, 90 percent and standard deviation of absolute error prediction (AEP) were calculated for predictions based on the individual data from 52 children. Boys' median AEP ranged from 1.4 cm to 4.3 cm; 90 percent ranged from 1.8 cm to 8.3 cm. Girls' median AEP ranged from 0.68 cm to 4.38 cm; 90 percent AEP ranged from 1.5 cm to 10.6 cm.
Jonathan Paley, MD, of Baltimore, led the research team. Other team members, also of Baltimore, were Jonathan Talor, Anna Levin, Anil Bhave, PT, and John E. Herzenberg, MD.
|Home||2004 Academy News||2004 Academy News March 11 Index B|
Last modified 01/March/2004