We used theWilcoxon signed-rank test as a non-parametric test, to
compare repeated measurements in the same adolescent, or observed
and predicted height outcomes.
We used the Pearson Correlation to test the linear relationship
between two variables.
We fitted a multivariate linear model to the data to estimate the
proper effect of rhGH dose, residual growth rate at onset of rhGH treatment,
and knee score, on height gain (HG) or height increment (HI) in
response to rhGH. Statistics and estimations of effect given in Table 3
are adjusted for the others whenever appropriate. Other covariables
were not included in the model since they showed no significant relationship
with HG or HI. All statistics and linear model were computed
using R 2.10.1. Results are expressed as mean ± SD.