Objective: Many GH deficient (GHD) patients have impaired glucose tolerance and GH substitution in these patients has caused deleterious effects on glucose tolerance with hyperinsulinaemia. This further impairment of glucose tolerance might be due to an unphysiologically high dose of GH. Whether such a deterioration can be avoided by an optimal GH replacement dose is not known. In most previous studies, the GH dose was calculated according to body weight or body surface area and not adjusted according to the serum IGF-I response.
Design: The study was of open design and investigations were performed before the start of GH substitution and after nine months of treatment. The GH dose was adjusted according to the response in serum IGF-I, and in patients with sub-normal serum IGF-I levels (all but two) we aimed for a serum IGF-I level in the middle of the normal range. The median GH dose at the end of the study was 0.14 IU/kg/week.
Patients: Ten patients, eight males and two females, with childhood onset GHD were examined. Their median age was 27 years (range 21-28).
Measurements: Overnight and 24-h fasting levels of glucose, insulin and IGFBP-1 were measured. Directly after the 24-h fast an oral glucose tolerance test (OGTT), with measurements of glucose, insulin and IGFBP-1 was performed. An intravenous glucose tolerance test (IVGTT) was performed after overnight fasting. Body composition was measured with bio-impedance analysis (BIA) and quality of life was assessed using a self-rating questionnaire, Qol-AGHDA.
Results: After GH treatment, there were no significant changes in glucose tolerance, measured by overnight and 24-h fasting levels of glucose, insulin and IGFBP-1, an oral glucose tolerance test (after 24-h fasting) and an intravenous glucose tolerance test (after overnight fasting). Percentage fat mass and BMI correlated negatively with both the 24 h fasting IGFBP-1 levels and the IGFBP-1 responses after the OGTT. All patients decreased their percentage of fat mass measured by BIA [median -2.9%; range -1.0-(-6.6); P = 0.005]. The administered GH dose correlated negatively with the relative change in whole body resistance (r = -0.66; P = 0.04). All, but one of the patients improved their quality of life score after GH therapy.
Conclusions: In a group of young patients with childhood onset GH deficiency, 9 months of treatment with a low GH dose (median 0.14 IU/kg/week) caused no significant deterioration of glucose tolerance. The strong negative associations between BMI or percentage fat mass and IGFBP-1 suggest that serum IGFBP-1 is more closely related than insulin to body composition in GH deficient patients. It is important to consider which critical endpoints should determine the GH dose. We would suggest that, apart for normalizing the serum IGF-I level, another main endpoint should be normalization of, or at least avoidance of any deterioration in glucose tolerance.