Document Type



Serial predictions of adult stature are needed in the clinical management of many children. The changes in these predictions might reflect effects of intervention and, at least in part, the changes that occur between serial predictions for normal children. The mean differences between predictions for normal children made 1, 2-, or 5 years apart are near zero; at most ages the standard deviations of these changes approximate 1.2, 1.5 and 2.0 cm respectively. These findings are important in evaluating the effects of diseases or intervention and in determining the sample sizes necessary to demonstrate significant effects in studies based upon alterations in predicted adult statures.