One of the major problems facing girls and women with Turner syndrome (TS) is extreme short stature, their mean adult height (Ht) of 142 cm (4'88”) being 20 cm (8“) below the mean adult Ht of women without TS. Treatment with recombinant human growth hormone (GH) increases the adult Ht of women with TS. However, it has long been recognized that there is variability in the magnitude of individuals' responses to GH. A number of factors may contribute to this variability, including underlying genetic factors beyond TS itself, age at initiation of GH and GH dose.

In a randomized, multicenter, long-term trial of GH in 232 patients with TS the Ht of the 99 patients whose bone age (BA) was ≥14 y at last observation(approx. final Ht), was 148.7 ± 6.1 cm after 5.4 ± 1.8 y of GH started at a mean age of 10.9 ± 2.3 y. The patients' mean Ht increased by 1.3 ± 0.6 standard deviation scores (SDS) from baseline, relative to Lyon standards; 80% of patients had Ht > Lyon median and 33% had Ht > 152 cm (5'), compared with only 50% and 5%, respectively, of patients in the Lyon reference population. The dose-response relationship to GH was addressed by comparing mean change in Ht SDS between patient groups who received 0.27 mg/kg/wk (n= 39) or 0.36 mg/kg/wk (n= 60). In a linear regression model of change from baseline Ht SDS of patients with BA ≥14 y at time of analysis, the significant explanatory variables were age at GH initiation (p<0.001) and GH dose (p=0.028). The mean loss in final Ht SDS due to delaying GH by one year was 0.15 ± 0.02 SDS, and the mean gain in Ht SDS from using the higher GH dose was 0.25 ± 0.11 SDS. Midparental Ht was not a significant explanatory variable of change in Ht SDS.

To avoid subjecting poorly-responsive patients to ineffective therapy it would be valuable to have a method for predicting long-term response to GH. Thus we evaluated the relationship between growth velocity (GV) after 1 year of GH, and long-term response to therapy. A linear regression model shows a clear positive relationship (p<0.001) between mean GV after 1 year of GH and mean change in Ht SDS from baseline to time of analysis at BA ≥14 y. However, the variability in this relationship is great enough that the regression equation (using all information on age at initiation, Ht SDS at baseline, GH dose, and GV at 1y) cannot, for an individual patient, adequately predict outcome.

These findings indicate a dose-response effect of GH in patients with TS and highlight the importance of initiating GH therapy as early as possible, in order to optimize gain in final height.