Objective: Derangements in the GH/IGF axis are common in patients with diabetes mellitus. In insulin-dependent diabetes mellitus (IDDM), these disturbances seem to be due to a partial defect in GH action on its own receptor or via a post-receptor defect. In non-insulin-dependent diabetes mellitus (NIDDM), data are limited, and the regulation of the GH receptor (GHR) remains unclear. However, animal studies with diabetic rats demonstrated that the GHR density may be influenced by insulin disposal at the hepatocyte. With respect to this hypothesis we studied the relation between peripheral insulin status and the serum GH-binding protein (GHBP), which reflects indirectly the GHR density in the tissues. Patients with IDDM were compared to a NIDDM group as well as to a group of healthy subjects.
Design and patients: Basal blood samples for the determination of serum GHBP, GH, and IGF-I were obtained from patients with IDDM (n = 27), subjects with NIDDM (n = 112) and healthy controle (n = 42). Insulin, proinsulin, C-peptide and IGF-binding protein 1 (IGFBP-1) serum levels were used to estimate the insulin status in diabetic patients.
Results: GHBP serum levels were significantly lower in patients with IDDM than in either NIDDM or controls (P < 0.001). Conversely, the IGF-I levels were reduced in both groups of diabetics. A subgroup of hypoinsulinaemic NIDDM patients showed significantly decreased GHBP concentrations (P < 0.05) compared to the NIDDM sub-group with hyperinsulinaemia. Furthermore, GHBP levels were significantly decreased in insulin-treated patients with NIDDM compared to either non-insulin-requiring subjects or normal controls (P < 0.05). A significant direct relation was found between levels of GHBP and total insulin dose (P < 0.01) in patients with IDDM. In the NIDDM group, GHBP was correlated with proinsulin (P < 0.001), C-peptide (P < 0.01), Insulin (P < 0.05) and inversely with IGFBP-1 (P < 0.001). Multiple linear regression analysis indicated a significant contribution of proinsulin and IGFBP-1 to the variation of GHBP.
Conclusions: Decreased GHBP levels in IDDM as well as in NIDDM correlate with insulinopenia. Since the degree of insulinopenia depends on the capability of the beta-cells to secrete proinsulin, C-peptide and insulin, we hypothesize that these hormones at least partially influence the serum level of GHBP. Low GHBP levels may reflect a reduced GH receptor density and a concomitant GH insensitivity, which leads to an impaired IGF generation in insulin-deficient patients.