Effects of glucose-dependent insulinotropic polypeptide on gastric emptying, glycaemia and insulinaemia during critical illness: a prospective, double blind, randomised, crossover study

INTRODUCTION: Insulin is used to treat hyperglycaemia in critically ill patients but can cause hypoglycaemia, which is associated with poorer outcomes. In health glucose-dependent insulinotropic polypeptide (GIP) is a potent glucose-lowering peptide that does not cause hypoglycaemia. The objectives of this study were to determine the effects of exogenous GIP infusion on blood glucose concentrations, glucose absorption, insulinaemia and gastric emptying in critically ill patients without known diabetes. METHODS: A total of 20 ventilated patients (Median age 61 (range: 22 to 79) years, APACHE II 21.5 (17 to 26), BMI 28 (21 to 40) kg/m(2)) without known diabetes were studied on two consecutive days in a randomised, double blind, placebo controlled, cross-over fashion. Intravenous GIP (4 pmol/kg/min) or placebo (0.9% saline) was infused between T = −60 to 300 minutes. At T0, 100 ml of liquid nutrient (2 kcal/ml) containing 3-O-Methylglucose (3-OMG), 100 mcg of Octanoic acid and 20 MBq Tc-99 m Calcium Phytate, was administered via a nasogastric tube. Blood glucose and serum 3-OMG (an index of glucose absorption) concentrations were measured. Gastric emptying, insulin and glucagon levels and plasma GIP concentrations were also measured. RESULTS: While administration of GIP increased plasma GIP concentrations three- to four-fold (T = −60 23.9 (16.5 to 36.7) versus T = 0 84.2 (65.3 to 111.1); P <0.001) and plasma glucagon (iAUC(300) 4217 (1891 to 7715) versus 1232 (293 to 4545) pg/ml.300 minutes; P = 0.04), there were no effects on postprandial blood glucose (AUC(300) 2843 (2568 to 3338) versus 2819 (2550 to 3497) mmol/L.300 minutes; P = 0.86), gastric emptying (AUC(300) 15611 (10993 to 18062) versus 15660 (9694 to 22618) %.300 minutes; P = 0.61), glucose absorption (AUC(300) 50.6 (22.3 to 74.2) versus 64.3 (9.9 to 96.3) mmol/L.300 minutes; P = 0.62) or plasma insulin (AUC(300) 3945 (2280 to 6731) versus 3479 (2316 to 6081) mU/L.300 minutes; P = 0.76). CONCLUSIONS: In contrast to its profound insulinotropic effect in health, the administration of GIP at pharmacological doses does not appear to affect glycaemia, gastric emptying, glucose absorption or insulinaemia in the critically ill patient. TRIAL REGISTRATION: Australian New Zealand Clinical Trials Registry ACTRN12612000488808. Registered 3 May 2012.