Growth parameters and predictors of growth in short children with and without growth hormone (GH) deficiency treated with human GH: a randomized controlled study.

Seventy-seven prepubertal short children with heights below the third centile for age and gender were divided into three groups according to their peak GH response to clonidine and insulin provocation. Group (I) included 30 children with peak GH response < 7 micrograms/l, group (II) included 19 children, with GH peak response between 7 and 10 micrograms/l, and group (III) included 24 children with GH peak > 10 micrograms/l. Each group was divided into two subgroups, a and b. Subgroups (I)b, (II)a and (III)b were treated daily for 1 year with subcutaneous recombinant human growth hormone (GH) 15 U/m2/week, and group (I)a was treated with GH (30 U/m2/week). Before initiation of treatment, the chronological age, the height standard deviation score (HtSDS), and the bone age delay did not differ among the study subgroups. The height growth velocity (GV) and insulin-like growth factor-I concentrations were significantly higher in group (III), with normal GH response to provocation, compared to those for group (I) with GH deficiency. All the children had normal thyroid function and normal glucose tolerance. CT examination of the hypothalamic-pituitary area revealed a picture of empty sella (either partial or complete) in 35 per cent of the children in group (I) and 21 per cent of children in group (II). After 1 year of GH therapy, the HtSDS, GV, and IGF-I concentrations increased significantly in the four subgroups treated with GH compared to their pretreatment values and to their controls. All the children in group (I) were responders (increment in GV of 2 cm2/year above the pretreatment GV), of the nine subjects treated in group (II)a, one child was a non-responder and of the 12 children in group (III)a three children were non-responders. GV was non-significantly higher in group (I)a (30 U/m2/week) v. group (I)b (15 U/m2/week). GV of children in groups (I)b, with abnormal GH response to provocation, was significantly higher than GV of children in group (III)a. Bone age advanced by less than 1 year in the treated groups (0.84 +/- 0.14 years) v. the untreated groups (0.73 +/- 0.3 years). None of the children had impaired glucose tolerance or abnormal thyroid function after 1 year of GH therapy. In all the treated children, GV after 1 year of GH treatment was correlated significantly with the pretreatment GV (r = -0.63, P < 0.01), peak GH response to provocation (r = -0.59, P < 0.01), IGF-I concentration (r = -0.54, P < 0.01), and positively with the GH dose (r = 0.589, P < 0.01). In group (III) children, with normal GH reserve, GV correlated significantly with the pretreatment GV (r = -0.48, P < 0.01) and negatively with the GH peak response to provocation (r = -0.25, P < 0.05). In conclusion, GH therapy improved GV of children growing along or parallel to the 3rd centile, irrespective of their GH response to provocation, without untoward effect on skeletal maturation, thyroid function or glucose tolerance.