Computer simulation of final height after growth-promoting therapy

Abstract

This short communication discusses the possible influence of final height prediction on the estimation of the growth response to therapy. These influences are illustrated in terms of height at 8 and 18 years of age for 28 prepubertally short boys. On the basis of several treatment simulations and assumptions, we conclude that the mean residual final height; that is, the mean of individual differences between final height and predicted final height, is a robust outcome measure in growth-promoting studies. However, the treatment residual final height (the difference between final height and predicted final height following growth-promoting treatment) was only partially correlated (r2 = 0.27) with the gain due to treatment, but was highly correlated (r2 = 0.75) with the residual final height; thus, growth-response estimations made on the basis of treatment residual final height may be more dependent on the predicted interval, or confidence interval, of final height than on the effect of growth-promoting treatment.