Objectives: Recent animal studies have suggested that insulin-like growth factor binding protein (IGFBP)-1 may regulate the insulin-like actions of the circulating IGFs. In man, IGFBP-1 levels change rapidly with nutritional status and are inversely related to changes in insulin. In-vitro studies have shown that both insulin and glucose independently regulate IGFBP-1 secretion in an inverse manner. A rapid rise of serum IGFBP-1 levels following insulin-induced hypoglycaemia suggested that glucose or glucose availability, rather than insulin, may be the major regulator of IGFBP-1.
Design: Three separate experiments both in patients and in normal volunteers were designed to examine the possibility that in these extreme circumstances glucose rather than insulin was the predominant regulator of IGFBP-1.
Methods: Insulin tolerance tests (ITT) were performed during the routine assessment of pituitary function in seven patients, four female and three male (mean age +/- SEM 36.8 +/- 6.3 years, range 20.7-69.3 years) with pituitary disease. Hypoglycaemic glucose clamp studies (insulin 2 mU/kg/min for 180 minutes) were performed in five normal volunteers, two female and three male (mean age 33.6 +/- 2.2 years, range 23.5-42.0 years). A three-part infusion study was performed in five volunteers, three female and two male (mean age 22.9 +/- 0.9 years, range 20.8-25.0 years) who received for 45 minutes on three occasions separated by at least 7 days either octreotide (long-acting somatostatin analogue) (1 microgram/min), adrenaline (3 micrograms/min) or control.
Measurements: Serum levels of IGFBP-1, insulin, glucose, C-peptide and cortisol were measured at varying intervals during the three studies by radioimmunoassay (RIA).
Results: Symptomatic hypoglycaemia (1.0 +/- 0.1 mmol/l) occurred at 30 minutes in all patients during the ITT. Serum IGFBP-1 levels rose from 28 +/- 7 to 86 +/- 15 micrograms/l at 180 minutes. During the hypoglycaemic glucose clamp study plasma glucose fell from 4.8 +/- 0.3 to 2.2 +/- 0.3 mmol/l. In contrast to the response observed during ITT, IGFBP-1 levels fell from 22 +/- 6 to 10 +/- 1 microgram/l by 180 minutes. During the octreotide infusion study there was no change in plasma glucose and plasma insulin levels fell from 5.8 +/- 1.9 to < 2.0 mU/l. Serum IGFBP-1 levels rose from 21 +/- 2 to 68 +/- 5 micrograms/l by 180 minutes. There was no change in IGFBP-1 during either the adrenaline infusion or the control study. The rise in IGFBP-1 following the octreotide infusion (68 +/- 5 micrograms/l) was similar to that in the patients undergoing the ITT (86 +/- 15 micrograms/l) (P = 0.3).
Conclusion: The rapid rise of serum IGFBP-1 levels induced by acute hypoglycaemia could be reproduced in euglycaemic conditions with octreotide when insulin secretion was suppressed, whereas IGFBP-1 levels did not rise with hypoglycaemia induced by a prolonged insulin infusion. These findings suggest that the surprising rise of IGFBP-1 levels observed during ITT is not secondary to changes in glucose. The rapid removal from the portal circulation of endogenous insulin with its inhibitory effect on IGFBP-1 secretion therefore appears to be the likely cause for the rapid rise of IGFBP-1 following an ITT. This conclusion supports the hypothesis that IGFBP-1 may inhibit the insulin-like actions of 'free' IGF when insulin secretion is low and so directly link the availability and hence actions of IGFs to acute but temporary changes in nutritional status.