Search

To better characterize autonomous insulin secretory behaviour in insulinoma patients and to establish diagnostic criteria with high accuracy, hyper-insulinaemic, sequentially eu- and hypoglycaemic clamp tests were performed in insulinoma patients and control subjects. Ten patients with insulinoma (benign in nine, histologically proven in nine) and 10 patients with suspected episodes of hypoglycaemia, in whom thorough clinical evaluation excluded an insulinoma, were examined. Five insulinoma patients were restudied after successful extirpation of the tumour. Suppression of C-peptide during low-dose [2 pmol kg-1 min-1 (20 mU kg-1 h-1) for 90 min, plasma insulin approximately 120 pmol L-1 (20 mUL-1)] and high-dose [8 pmol kg-1 h-1 (80 mU kg-1 h-1) for 90 min, plasma insulin approximately 450 pmol L-1 (75 mU L-1)] insulin infusion under euglycaemic conditions [plasma glucose 4.4-5.0 mmol L-1 (80-90 mg dL-1)] and during high-dose insulin infusion under hypoglycaemic conditions [glucose 2-2.2 mmol L-1 (40-45 mg dL-1)] was evaluated by radioimmunoassay (RIA). Euglycaemic hyper-insulinaemia suppressed C-peptide in control subjects (P < 0.0001), whereas in insulinoma patients apparently irregular changes in C-peptide concentrations (with spontaneous or paradoxical increments, P = 0.0006 vs. controls) were observed. The combination of hyper-insulinaemia and controlled hypoglycaemia led to a nearly complete suppression of C-peptide in normal subjects (from basal, 0.76 +/- 0.08-0.06 +/- 0.01 nmol L-1; maximum observed value 0.10 nmol L-1), which was more pronounced than at the point of discontinuation of prolonged fasting (> 48 h; 0.26 +/- 0.16 nmol L-1; P = 0.005). In insulinoma patients, C-peptide remained elevated under all conditions (P = 0.51 vs. prolonged fasting). All these findings were reversible after successful surgical removal of the insulinoma. Insulinoma patients could be identified as abnormal by (a) non-suppression of C-peptide even under hyperinsulinaemic/hypoglycaemic conditions (10 out of 10 patients) and (b) irregular increments in C-peptide under conditions that led to at least partial suppression in all normal subjects (9 out of 10 patients) and/or by an apparent shift to the left of insulin secretion relative to glucose concentrations (7 out of 10 patients). Controlled exposure to hyperinsulinaemic/hypoglycaemic conditions can help to characterize autonomous secretion in insulinoma patients and may be used as a diagnostic procedure when conventional methods yield equivocal results.

Intravenous glucagon-like peptide (GLP)-1 [7-36 amide] can normalize plasma glucose in non-insulin-dependent diabetic (NIDDM) patients. Since this is no form for routine therapeutic administration, effects of subcutaneous GLP-1 at a high dose (1.5 nmol/kg body weight) were examined. Three groups of 8, 9 and 7 patients (61 +/- 7, 61 +/- 9, 50 +/- 11 years; BMI 29.5 +/- 2.5, 26.1 +/- 2.3, 28.0 +/- 4.2 kg/m2; HbA1c 11.3 +/- 1.5, 9.9 +/- 1.0, 10.6 +/- 0.7%) were examined: after a single subcutaneous injection of 1.5 nmol/kg GLP [7-36 amide]; after repeated subcutaneous injections (0 and 120 min) in fasting patients; after a single, subcutaneous injection 30 min before a liquid test meal (amino acids 8%, and sucrose 50 g in 400 ml), all compared with a placebo. Glucose (glucose oxidase), insulin, C-peptide, GLP-1 and glucagon (specific immunoassays) were measured. Gastric emptying was assessed with the indicator-dilution method and phenol red. Repeated measures ANOVA was used for statistical analysis. GLP-1 injection led to a short-lived increment in GLP-1 concentrations (peak at 30-60 min, then return to basal levels after 90-120 min). Each GLP-1 injection stimulated insulin (insulin, C-peptide, p < 0.0001, respectively) and inhibited glucagon secretion (p < 0.0001). In fasting patients the repeated administration of GLP-1 normalized plasma glucose (5.8 +/- 0.4 mmol/l after 240 min vs 8.2 +/- 0.7 mmol/l after a single dose, p = 0.0065). With the meal, subcutaneous GLP-1 led to a complete cessation of gastric emptying for 30-45 min (p < 0.0001 statistically different from placebo) followed by emptying at a normal rate. As a consequence, integrated incremental glucose responses were reduced by 40% (p = 0.051). In conclusion, subcutaneous GLP-1 [7-36 amide] has similar effects in NIDDM patients as an intravenous infusion. Preparations with retarded release of GLP-1 would appear more suitable for therapeutic purposes because elevation of GLP-1 concentrations for 4 rather than 2 h (repeated doses) normalized fasting plasma glucose better. In the short term, there appears to be no tachyphylaxis, since insulin stimulation and glucagon suppression were similar upon repeated administrations of GLP-1 [7-36 amide]. It may be easier to influence fasting hyperglycaemia by GLP-1 than to reduce meal-related increments in glycaemia.

The aim of the study was to investigate whether inhibition of gastric emptying of meals plays a role in the mechanism of the blood glucose-lowering action of glucagon-like peptide-1-(7-36) amide [GLP-1-(7-36) amide] in type 2 diabetes. Eight poorly controlled type 2 diabetic patients (age, 58 +/- 6 yr; body mass index, 30.0 +/- 5.2 kg/m2; hemoglobin A1c, 10.5 +/- 1.2%) were studied in the fasting state (plasma glucose, 11.1 +/- 1.1 mmol/L). A liquid meal of 400 mL containing 8% amino acids and 50 g sucrose (327 Kcal) was administered at time zero by a nasogastric tube. Gastric volume was determined by a dye dilution technique using phenol red. In randomized order, GLP-1-(7-36) amide (1.2 pmol/kg.min; Saxon Biochemicals) or placebo (0.9% NaCl with 1% human serum albumin) was infused between -30 and 240 min. In the control experiment, gastric emptying was completed within 120 min, and plasma glucose, insulin, C-peptide, GLP-1-(7-36) amide, and glucagon concentrations transiently increased. With exogenous GLP-1-(7-36) amide (plasma level, approximately 70 pmol/L), gastric volume remained constant over the period it was measured (120 min; P < 0.0001 vs. placebo), and plasma glucose fell to normal fasting values (5.4 +/- 0.7 mmol/L) within 3-4 h, whereas insulin was stimulated in most, but not all, patients, and glucagon remained at the basal level or was slightly suppressed. In conclusion, GLP-1-(7-36) amide inhibits gastric emptying in type 2 diabetic patients. Together with the stimulation of insulin and the inhibition of glucagon secretion, this effect probably contributes to the blood glucose-lowering action of GLP-1-(7-36) amide in type 2-diabetic patients when studied after meal ingestion. At the degree observed, inhibition of gastric emptying, however, must be overcome by tachyphylaxis, reduction in dose, or pharmacological interventions so as not to interfere with the therapeutic use of GLP-1-(7-36) amide in type 2 diabetic patients.

GLP-1 (7-36 amide) stimulates insulin and suppresses glucagon secretion in normal subjects and may, in pharmacological doses, normalize hyperglycaemia in type 2 diabetic patients. It is not known whether such pharmacological doses can actually lower blood glucose to hypoglycaemic levels. Therefore, in seven normal fasting subjects, GLP-1 (7-36 amide) was infused intravenously at 0.3, 0.9 and 2.7 pmol/kg per min for 30 min each. The plasma concentration of GLP-1 (7-36 amide) increased dose-dependently, but insulin secretion (insulin, C-peptide) was stimulated only marginally. Glucagon was slightly suppressed, and plasma glucose was reduced, but not into the hypoglycaemia range. In conclusion, when plasma glucose concentrations are in the normal fasting range, GLP-1 (7-36 amide) is not able to stimulate insulin secretion to a degree that causes hypoglycaemia. This should limit the risk of hypoglycaemic responses when GLP-1 (7-36 amide) is administered in pharmacological doses to reduce hyperglycaemia in type 2 diabetic patients.

Gastric inhibitory polypeptide (GIP) and glucagon-like peptide-1-(7-36) amide (GLP-1) are glucose-dependent insulinotropic gut hormones that may explain the greater insulin secretory response with oral compared to i.v. glucose (incretin effect). To study their individual and combined contributions, in eight healthy volunteers, on separate occasions, synthetic human GIP (1 pmol/kg.min) and/or GLP-1 (0.3 pmol/kg.min) or placebo were infused i.v. (-30 to 120 min), while at 0 min, a glucose infusion "isoglycemic" to the profile after an oral glucose load of 50 g/400 mL was started. After the administration of 50 g oral glucose, immunoreactive GIP rose several-fold to 337 +/- 43 pmol/L, while there was only a transient (10-30 min) and moderate increment in immunoreactive GLP-1 (from basal, 25-30, to 41 +/- 4 pmol/L). Isoglycemic i.v. glucose infusions led to smaller B-cell responses (estimated incretin effect, 41 +/- 5%). With single infusions of GIP or GLP-1 (circulating concentrations, 464 +/- 73 and 54 +/- 3 pmol/L, respectively), B-cell responses were significantly augmented compared to i.v. glucose alone and were no longer significantly different from those after oral glucose. The combination of GIP and GLP-1 led to B-cell responses that were significantly higher than those with either hormone alone (additive mode of cooperation). Plasma GIP concentrations were similar after endogenous secretion (oral glucose) and i.v. infusion, while exogenously administered GLP-1 led to plasma levels that were maintained at an elevated level for a longer period during exogenous infusion than after stimulation by oral glucose. When, in seven volunteers, a lower dose (0.15 pmol/kg.min) of GLP-1 was infused during isoglycemic glucose infusion experiments only for the duration of elevated plasma levels in the oral glucose challenges (0-30 min), a significant, but transient, increment in insulin and C-peptide concentrations was observed, which was equivalent to 26 +/- 10% of the estimated incretin effect. Therefore, in conclusion, circulating GIP seems to make a major contribution to the incretin effect after oral glucose, and GLP-1 appears to mediate a smaller proportion. GIP and GLP-1 can interact in an additive manner in normal man.

Metabolic potencies of the destetrapeptide insulin analogues des-(B27-B30)-insulin, des-(B27-B30)-insulin-B26-amide, [ThrB26] des-(B27-B30)-insulin-B26-amide and [GluB26] des-(B27-B30)-insulin-B26-amide were studied in anaesthetized adult rats and in primary cultures of rat hepatocytes and compared with that of the native hormone. Hypoglycaemic effects following intravenous bolus injection of insulin or analogues were similar, as were the stimulatory actions on total body glucose disposal during euglycaemic clamping. In these latter studies a maximal stimulation in the range 16-20 mg glucose/kg per hour was observed and identical half-maximally effective serum concentrations for all peptides of about 1 pmol/ml were obtained. Analogue actions on individual peripheral tissues estimated by the uptake of 2-deoxyglucose were not different from those of insulin. In hepatocyte cultures the stimulatory action of destetrapeptide analogues on glycogenesis and on aminoisobutyric acid transport was indistinguishable from that of native insulin, with identical half-maximally effective concentrations. These data demonstrate that des-(B27-B30)-insulin and related destetrapeptide analogues have high biological activity. Since the truncated non-amidated analogue appeared to be monomeric in solution, this peptide could be a candidate for an insulin preparation potentially showing rapid absorption from subcutaneous tissue.

In type-2 diabetes, the overall incretin effect is reduced. The present investigation was designed to compare insulinotropic actions of exogenous incretin hormones (gastric inhibitory peptide [GIP] and glucagon-like peptide 1 [GLP-1] [7-36 amide]) in nine type-2 diabetic patients (fasting plasma glucose 7.8 mmol/liter; hemoglobin A1c 6.3 +/- 0.6%) and in nine age- and weight-matched normal subjects. Synthetic human GIP (0.8 and 2.4 pmol/kg.min over 1 h each), GLP-1 [7-36 amide] (0.4 and 1.2 pmol/kg.min over 1 h each), and placebo were administered under hyperglycemic clamp conditions (8.75 mmol/liter) in separate experiments. Plasma GIP and GLP-1 [7-36 amide] concentrations (radioimmunoassay) were comparable to those after oral glucose with the low, and clearly supraphysiological with the high infusion rates. Both GIP and GLP-1 [7-36 amide] dose-dependently augmented insulin secretion (insulin, C-peptide) in both groups (P < 0.05). With GIP, the maximum effect in type-2 diabetic patients was significantly lower (by 54%; P < 0.05) than in normal subjects. With GLP-1 [7-36 amide] type-2 diabetic patients reached 71% of the increments in C-peptide of normal subjects (difference not significant). Glucagon was lowered during hyperglycemic clamps in normal subjects, but not in type-2 diabetic patients, and further by GLP-1 [7-36 amide] in both groups (P < 0.05), but not by GIP. In conclusion, in mild type-2 diabetes, GLP-1 [7-36 amide], in contrast to GIP, retains much of its insulinotropic activity. It also lowers glucagon concentrations.

Insulin secretion is stimulated better by oral than by intravenous glucose (incretin effect). The contribution of the autonomic nervous system to the incretin effect after oral glucose in humans is unclear. We therefore examined nine type 1 diabetic (insulin-dependent) patients with end-stage nephropathy, studied after combined heterotopic pancreas and kidney transplantation, and 7 non-diabetic kidney recipients (matched for creatinine clearance and immunosuppressive medication). The release of gastric inhibitory polypeptide (GIP) and glucagon-like peptide 1 (GLP-1) immunoreactivity and B cell secretory responses (IR insulin and C-peptide) to oral (50 g) and "isoglycaemic" intravenous glucose (identical glycaemic profile) were measured by radioimmunoassay. The difference in B cell responses between the two tests represents the contribution of the enteroinsular axis to the response after oral glucose (incretin effect). Insulin responses after the oral glucose challenge were similar in the two patient groups despite systemic venous drainage of the pancreas graft in the pancreas-kidney-transplanted group. In both groups GIP and GLP-1 increased after oral but not after intravenous glucose, and B cell secretory responses were significantly smaller (by 55.2 +/- 7.7% and 46.5 +/- 12.5%, respectively) with "isoglycaemic" intravenous glucose infusions. The lack of reduction in the incretin effect in pancreas-kidney-transplanted patients, whose functioning pancreas is denervated, indicates a lesser role for the nervous system and a more important contribution of circulating incretin hormones in mediating the enteroinsular axis in man.

Subcutaneous injections of soluble human insulin and of the monomeric insulin analogue des-(B26-B30)-insulin-B25-amide were given to fasted rabbits, pigs and healthy man. The time course of blood glucose decline and the rate of appearance of the hormones in the peripheral circulation were studied for doses of 0.35 and 0.7 nmol/kg. In rabbits an identical time course with early glucose nadirs between 45 and 60 min and hormone peaks at about 20 min were observed for both hormones. Similar results were obtained for insulin and the analogue in pigs. When injected into humans, slightly earlier peaks of the analogue at 30 to 45 min were measured compared to insulin peaks at 60 min. In addition, a trend towards faster decline in blood glucose could be demonstrated after analogue injection. In contrast to monomeric insulin analogues produced by recombinant DNA technology, s.c. injections of des-(B26-B30)-insulin-B25-amide do not appear to meet mealtime insulin requirements better than soluble insulin.

After ingestion of glucose both GIP (gastric inhibitory polypeptide, glucose-dependent insulinotropic polypeptide) and GLP-1(7-36amide) (glucagon-like polypeptide-1, 7-36amide) may play a physiological role in augmenting insulin release. Their insulinotropic effect was compared in isolated rat islets after 24-h maintenance in tissue culture (11 mmol l-1 glucose). Ten islets per vial were then incubated in Krebs-Ringer-Hepes buffer for 30 min; insulin was measured radioimmunologically. Both hormones were always compared in the same experiment. At 16.7 mmol l-1 glucose both GIP and GLP-1(7-36amide) 2 x 10(-10) mol l-1 significantly increased insulin release; 10(-10) mol l-1 of either hormone had no significant effect. The response at 10(-9) and 10(-8) mol l-1 was similar for both; at 4 x 10(-10) mol l-1 GLP-1(7-36amide), however, was clearly more effective than GIP. At low glucose (2.8 or 5.0 mol l-1) no significant differences were found. A concentration of 10(-8) mol l-1 of both hormones was slightly stimulatory. At 8.3 mmol l-1 glucose, 10(-9) mol l-1 GLP-1(7-36amide) was 60% more effective than GIP (4.8 +/- 0.4 vs. 3.0 +/- 0.4, n = 13, P less than 0.005), the response to 10(-8) mol l-1 was similar. These data show comparable effects of high concentrations of GIP and GLP-1(7-36amide) on glucose-induced insulin release; at presumably physiological concentrations, however, GLP-1(7-36amide) was clearly more effective. The combination of the two peptides was not more than additive, suggesting that they act via the same final mechanism.

To estimate the contribution of postprandial cholecystokinin (CCK) responses to circulating insulin concentrations and insulin secretion, a specific CCK receptor antagonist (loxiglumide; 10 mg/kg body weight/h) or saline were infused intravenously in normal volunteers, beginning 90 min before insulin secretion was stimulated on separate occasions by the intraduodenal administrations of glucose, glucose and protein, and glucose plus protein with the admixture of pancreatin. The release of CCK (radioimmunoassay) was stimulated by the protein component of the nutrients from basal 2.4 +/- 0.4 to 8.0 +/- 1.2 pmol/l. CCK plasma levels were significantly higher with loxiglumide (p < 0.05). Glucose-dependent insulinotropic polypeptide (GIP) was also released by all nutrient mixtures. Loxiglumide significantly inhibited the amount of bilirubin and pancreatic enzymes recovered from duodenal aspirates. In contrast, in none of the experiments, C-peptide increments and hence insulin secretion rates were altered by loxiglumide. With glucose and protein as intraduodenal stimulus (no pancreatin added), the plasma amino acids rose significantly less (by approximately 50% of the control experiment) and the increment in insulin (but not C-peptide) concentrations was significantly reduced by loxiglumide. This is most likely explained by a change in insulin metabolic clearance. This effect cannot be a primary action of CCK because there was no similar effect of loxiglumide with the same intraduodenal stimulus plus added pancreatin. Pancreatic enzymes reduced maldigestion secondary to loxiglumide effects on pancreatic exocrine secretion: The increment in circulating amino acid concentrations was similar with and without loxiglumide. In conclusion, CCK does not alter insulin secretion and, therefore, is not an incretin hormone in man. Blocking CCK actions on the exocrine pancreas by loxiglumide, however, can secondarily cause reductions in postprandial insulin profiles by altering insulin clearance. These changes are possibly related to reductions in circulating amino acid concentrations.

Release of immature secretory granules rich in incompletely processed proinsulin has been proposed to explain the relative hyperproinsulinemia in type 2 diabetic and insulinoma patients because of a constant secretory drive resulting from hyperglycemia and autonomous secretion, respectively. To test this hypothesis, insulin secretion was stimulated by a combination of hyperglycemia (11 mmol/L clamp), intravenous (i.v.) tolbutamide (1 g), and i.v. glucagon (initial bolus 10 micrograms/kg body weight, maintenance infusion 2 micrograms/kg body weight per hour) for 3 h. Circulating IR-insulin and IR-C-peptide concentrations increased 89-fold and 14-fold over basal values, respectively, but IR-proinsulin concentrations increased only ninefold over basal values. Estimation of the amount of insulin secreted (based on deconvolution analysis of plasma C-peptide values) showed that approximately 76 +/- 21 U were secreted during the stimulation period. This amount is a significant proportion of pancreatic insulin content in normal humans. In molar terms, IR-proinsulin (integrated incremental response multiplied by metabolic clearance rate of proinsulin) relative to IR-C-peptide (= insulin) secretion (deconvolution analysis) was estimated to be equal or even lower than the known proportion in islets (0.22 +/- 0.05%). Thus, using a near-maximal stimulation of insulin secretion maintained long enough to cause release of amounts of insulin approaching the estimated pancreatic content, no preferential release of proinsulin was observed in normal humans. Therefore, the hyperproinsulinemia of type 2 diabetes and in insulinoma patients may be caused by additional defects in the proinsulin to insulin conversion process.

Plasma glucose, immunoreactive insulin and C-peptide concentrations were compared in nine pancreas-kidney-transplanted patients (systemic venous drainage) and in ten non-diabetic kidney-transplanted patients with similar kidney function. In the basal state, C-peptide (insulin secretion) was similar, but immunoreactive insulin was higher and glucose concentrations were slightly, but significantly lower in pancrease-transplanted patients. After 50 g oral glucose, the plasma glucose and IR-insulin profiles were similar in both groups. The circumvention of first-pass hepatic insulin extraction (decreased endogenous insulin clearance) was compensated for by a significant reduction in insulin secretion (C-peptide; p = 0.036). In conclusion, hyperinsulinaemia in pancreas-transplanted patients with systemic venous drainage is significant only in the basal state. Insulin delivered into the portal and peripheral circulation, when leading to similar insulin profiles, maintains comparable degrees of glucose tolerance.

The acute metabolic actions of insulin-like growth factor I were studied in anaesthetized adult rats and its potency was compared to that of insulin. Following an i.v. bolus injection of insulin-like growth factor I a dose-dependent decrease of blood glucose and serum non-esterified fatty acid concentrations was noted with a potency of about 2% that of insulin. Stimulation of total body glucose disposal during euglycaemic clamping required approximately 50 times higher insulin-like growth factor I serum concentrations to achieve an identical half-maximal response. A similar difference in potency was observed for the stimulatory action on 2-deoxyglucose uptake and on glycogen formation in skeletal muscle. Lipogenesis in epididymal fat pads was increased dose-dependently by both hormones requiring approximately 30 times higher half-maximally effective serum concentrations of insulin-like growth factor I. These data demonstrate that insulin-like growth factor I exerted acute insulin-like metabolic actions in vivo with low potency. These effects were probably mediated via insulin receptors. A preferential stimulation of glucose metabolism in skeletal muscle was not observed.

The absorbable deoxynojirimycin derivative emiglitate (BAY o 1248) is a potent competitive inhibitor of small intestinal alpha-glucosidases in man. In two similar randomized, placebo-controlled, double blind investigations, the efficacy, duration of action and tolerability of single doses of 10, 20 and 40 mg emiglitate have been assessed in healthy male volunteers after repeated sucrose or maize-starch loads at 08.00, 12.00 and 17.00 h. Even at the lowest dose used, emiglitate almost abolished the glycaemic (-88%) and hormonal responses after the first sucrose meal, simultaneously evoking significant hydrogen evolution (mean peak H2-concentration greater than 100 ppm), which was not related to the dose, and which induced unacceptable symptoms of carbohydrate malabsorption, i.e. at the dosages tested, the inhibition of glycaemic and hormonal responses was at the expense of intolerable gastrointestinal adverse effects. Flattening of postprandial responses of blood glucose, serum insulin and gastric inhibitory polypeptide was still apparent after a second sucrose load 4 h later, demonstrating long-lasting inhibition of alpha-glucosidase activity. After starch, the dose dependency of inhibition emerged more clearly than after sucrose, i.e. the reduction was less pronounced. However, emiglitate led to significant reduction of the glycaemic and hormonal rises after both the first and second starch meals. Symptoms of carbohydrate malabsorption were absent after 10 mg and were negligible with 20 mg or 40 mg emiglitate. Breath hydrogen concentration increased gradually, indicating slight but significant carbohydrate malabsorption after the highest dose of the alpha-glucosidase inhibitor. The results show that a single morning dose of 20-40 mg emiglitate might be useful in the control of postprandial hyperglycaemia after breakfast and lunch.(ABSTRACT TRUNCATED AT 250 WORDS)

Metabolic potency of des-(B26-B30)-insulin-B25-amide, [TyrB25]des- (B26-B30)-insulin-B25-amide and [HisB25]des-(B26-B30)-insulin-B25-amide was studied in anaesthetized rats. Compared to insulin, full potency for des-(B26-B30)-insulin-B25-amide and an enhanced potency for both substituted analogues has been described previously on rat adipocytes in vitro. Hypoglycaemic effects following i.v. injection of all of these analogues were almost identical to those of native insulin with a half-maximal effective dose of approximately 3 nmol.kg-1. Stimulation of glucose metabolism during euglycaemic hyperinsulin-/analogueaemic clamp studies was indistinguishable from that of the native hormone with a maximal stimulation of approximately 19 mg.kg-1.min-1 and half-maximal effective hormone concentrations of approximately 1 pmol.ml-1. Analogue action on individual peripheral tissues estimated by the uptake of 2-deoxyglucose as well as stimulation of lipogenesis in epididymal fat was not different to that of insulin. These data demonstrate that C-terminal amidation of des-(B26-B30)-insulin results in a shortened molecule with full in vivo metabolic potency. When substituting phenylalanine in position B25 by tyrosine or histidine, the insulin-identical potency is preserved.

Sucrose (100 g) loading tests were performed in 10 healthy volunteers before and during the intake of an alpha-glucosidehydrolase inhibitor (acarbose) for 8 weeks (3 X 200 mg daily) and serum levels of glucose, immunoreactive insulin (IRI), and immunoreactive gastric inhibitory polypeptide (IR-GIP) were measured. The addition of 200 mg of acarbose to the sucrose load attenuated the sucrose-induced glycaemia and IRI response and completely abolished the IR-GIP release. The volunteers complained about meteorism and abdominal pain during the intake of the inhibitor. These side effects became less marked at the end of the study. The attenuation of complaints cannot be explained by a decreasing sucrase inhibition, since the increase of glucose, IRI, and IR-GIP after sucrose loading at the beginning and after 4 and 8 weeks was equally impaired by acarbose.

1. The insulinogenic factor of the gastrointestinal mucosa named "incretin" is only one part of the complex enteroinsular axis. --2. Of the chemically defined gastrointestinal hormones GIP is the strongest incretin candidate. --3. Because of the dual function of GIP as gastrone and insulinotropic substance several safeguards against GIP-mediated insulin hypoglycaemia exist. --4. No pathological condition has yet been found which is causally related to hyper- or hyposecretion of GIP. However, an exaggerated GIP response (usually secondary to the disease) may participate in the pathogenesis of hyperinsulinaemia of patients with obesity and duodenal ulcer. --5. The injection of GIP antibodies only partially abolishes the incretin effect. Therefore, GIP, although important, is not the only incretin.

The effect of glucose and insulin on fat- and glucose-induced gastric inhibitory polypeptide (GIP) release has been studied in insulin-dependent juvenile-type diabetics. Blood glucose and serum immunoreactive GIP (IR-GIP) were measured after an oral load of 100 g glucose or 100 g fat was given and during an infusion of one of the following: saline, glucose, glucose plus insulin, or insulin. The infusion of insulin alone (in the presence of elevated glucose levels) or together with glucose significantly suppressed the IR-GIP rise after fat ingestion, but it did not alter the GIP response to oral glucose. Intravenous infusion of glucose had a slight but significant inhibitory effect on fat-stimulated increase of IR-GIP, which cannot be related to endogenous insulin release in these insulin-deficient diabetics. It is suggested that an insulin-mediated increase of glucose utilization in the GIP cell interferes only with increased GIP secretion stimulated by the utilization of fatty acids but not of glucose. This could explain the existence of a negative feedback control between insulin and GIP secretion for fat but not for glucose-induced GIP release.

The action of watery rat gut extracts on glucose-induced insulin release in anaesthetized rats was examined before and after removal of GIP by immunoadsorption. Infusions of GIP-containing rat gut extracts nearly doubled the insulin release induced by intravenous glucose (1 g X kg -1 X h -1). Peak insulin secretion was 98 +/- 11 mU/l (mean +/- SEM) after intravenous glucose and increased to 178 +/- 16 mU/l following infusion of glucose plus gut extract (p less than 0.005). After injection of GIP antiserum in sufficient amounts to neutralize the GIP activity in the gut extract preparation, the additional insulin release due to the gut extract was reduced by only 30%. After complete removal of GIP from gut extracts by immuno-absorption, more than 50% of the incretin effect remained. These data suggest that the insulinotropic activity of rat gut extracts can only be partially related to GIP. The existence of additional insulinotropic gut factors which may also be released following oral glucose is postulated.

The rate of proinsulin/insulin turnover has been studied in human insulinoma tissue from ten patients. During the incubation of tumor tissue the ratio of immunoreactive insulin (IRI) release to content was significantly higher than that from isolated human pancreatic islets. The tumor cell cytoplasm (100,000 g supernatant S-100) contained approximately 45% of the IRI. Endogenous proinsulin and insulin were released throughout the incubation. Newly synthesized proinsulin was detected in the 15 min pulse microsomal, secretory granule, S-100 and media samples. These results suggest that defective hormonal storage and release mechanisms are operative in human insulinoma cells resulting in a higher turnover of proinsulin and insulin compared with pancreatic islet tissue.

The influence of impaired digestion on nutrient induced release of gastric inhibitory polypeptide (GIP) and insulin have been investigated in patients with chronic pancreatitis. All patients had massive steatorrhea (> 25 g/24 h), and glucose intolerance. A standard liquid test meal comprising fat and glucose were ingested with or without pancreatic enzyme substitution (9.0 g pancreatin). In the presence of pancreatin the response of serum levels of GIP to the test meal was significantly enhanced (81.2 vs 194.5 microgram/l X 180 min). Concurrently, the insulin response was augmented (3.4 vs 6.4 U/l X 180 min), resulting in improved glucose tolerance. Addition of pancreatin also significantly augmented the GIP response to oral fat (100 g), but not to oral glucose (100 g). In patients with pancreatogenic steatorrhea the insulin response to an IV glucose infusion (0.7 g/kg/h for 90 min) was augmented by oral fat only after addition of 9.0 g pancreatin to the fat load (3.5 vs 7.3 U/l X 180 min). After restoration of the GIP response to fat by pancreatin, the inhibitory effect of IV glucose on fat-induced GIP increase was restored. These data indicate that the GIP response to a mixed meal or triglycerides is dependent on the absorption of nutrients. In patients with chronic pancreatitis improvement of pancreatogenic insufficiency reverses the impaired GIP response, restores the incretin effect of fat, and improves glucose tolerance.

The effect of a dialysate exchange with both 1.5 and 4.25% glucose solutions on plasma levels of glucose, insulin, gastric inhibitory polypeptide (GIP), and glucagon has been investigated in 5 continuous ambulatory peritoneal dialysis (CAPD) patients. Only in the case of the 4.25% solution did plasma glucose levels rise above 100 mg/dl. 4 of the 5 patients responded to this change with a marked insulin secretion. Employing the 1.5% solution, plasma glucose remained stable and only a slight insulin stimulation was observed in 2 patients. It is concluded that provided the 4.25% dialysates are used only occasionally, there will be no continuous stimulation of the pancreatic beta-cells due to absorption of glucose from the dialysate alone during CAPD treatment. GIP levels are highly elevated in CAPD patients. A dialysate exchange with either a 1.5 or a 4.25% glucose solution had no effect on this gastrointestinal hormone. Hyperglucagonemia was also observed in this collective. An initial suppression of glucagon levels occurred in 4 of the patients after a 4.25% dialysate exchange. The 5th patient demonstrated an initial rise followed by a later decrease in glucagon, a response similar to that reported in adult onset diabetes after an oral glucose tolerance test.

A radioimmunoassay, using an antiserum that is specific for human proinsulin, has been used to study the response of serum proinsulin to low (25 g) and high (75 g) oral glucose loads in non-obese patients with non-insulin-dependent diabetes mellitus (NIDDM). Diabetic patients were treated by diet only (N = 8) or were receiving oral anti-hyperglycemic agents (N = 8) and therapy was not interrupted during the study. In the fasted state, proinsulin concentrations were higher (P less than 0.05) in the drug-treated patients (31 +/- 3 pmol/l (SEM)) compared with age- and weight-matched healthy subjects (22 +/- 2 pmol/l; N = 10), but concentrations in the diet-treated patients 25 +/- 3 pmol/l) were not significantly different. Following 25 g and 75 g glucose loads, the rises in serum immunoreactive insulin and C-peptide concentrations in both groups of diabetic patients were impaired and delayed relative to those in the control subjects. The responses of serum proinsulin, however, were not significantly different in the NIDDM patients compared with controls at any time point up to 180 min except in the case of drug-treated patients receiving 25 g of glucose who had elevated (P less than 0.05) proinsulin concentrations at 150 min and 180 min after ingestion. It is concluded that NIDDM is not associated with an exaggerated release of proinsulin in response to glucose compared with healthy subjects, but the islets have maintained the ability to release proinsulin better than the ability to release insulin.

Intraduodenal phenylalanine administration (333 mg/min over 60 min) released endogenous cholecystokinin in healthy young subjects as demonstrated radioimmunologically and by intraduodenal bilirubin and pancreatic enzyme output. Concomitantly, there was only a small increase over basal in circulating immunoreactive-insulin and immunoreactive-C-peptide concentrations. In healthy volunteers intraduodenal infusions of saline (10 ml/min), glucose (333 mg/min) or phenylalanine (333 mg/min) were performed for 60 min when plasma glucose was clamped at approximately 8 mmol/l. Phenylalanine enhanced immunoreactive-insulin and immunoreactive-C-peptide responses three-fold more than did the same amount of glucose. Immuno-reactive gastric inhibitory polypeptide responses were small and not different after glucose and phenylalanine administration. Immunoreactive cholecystokinin was significantly stimulated to 9.4 +/- 1.4 pmol/l only by intraduodenal phenylalanine. Plasma phenylalanine concentrations increased into the supraphysiological range (approximately 1.5 mmol/l). Intravenous infusions of phenylalanine achieving plasma concentrations of 1.2 mmol/l stimulated insulin secretion at elevated plasma glucose concentrations (approximately 8 mmol/l clamp experiments), but had no effect at basal plasma glucose concentrations. A small increase in cholecystokinin also was observed. Intravenous infusions of synthetic sulphated cholecystokinin-8 leading to plasma concentrations in the upper postprandial range (8-12 pmol/l) did not augment the immunoreactive-insulin or immunoreactive-C-peptide levels during hyperglycaemic clamp experiments, in the absence or presence of elevated plasma phenylalanine concentrations. It is concluded that the augmentation of the glucose-induced insulin release by intraduodenal administration of phenylalanine cannot be related to cholecystokinin release, but rather is explained by the combined effects of elevated glucose and phenylalanine concentrations. In man, cholecystokinin does not augment insulin secretion caused by moderate hyperglycaemia, elevations of phenylalanine concentrations, or combinations thereof.

Gastric Inhibitory Polypeptide (GIP; 1 or 10 mug/ml) potentiated glucose-induced (8 or 16.6 mM) insulin (IRI) release from isolated rat pancreatic islets. Basal release was unaffected. The threshold concentration of glucose necessary for GIP to modulate IRI release was between 6 and 8 mM. GIP had no effect on IRI release from islets submitted to a maximal glucose stimulus (25 mM).

The present study was designed to establish an in vitro perfused porcine pancreas preparation as a model for testing the effect of organ protective solutions on stimulated pancreatic endocrine and exocrine secretion. The pancreas was prepared and perfused for 10 min with Euro Collins solution, thereafter it was stored in the cold (4 degrees C) for various times. After 3-h and 6-h ischemia pancreatic insulin release in response to glucose was not significantly affected. After 12-h ischemia reduced pancreatic insulin secretin, increased perfusion pressure, and increased amylase and lipase release indicated pancreatic damage. Complete pancreatic dysfunction was seen after 24-h and 48-h ischemia with massive increase in perfusion pressure and low insulin secretion which did not follow a glucose-dependent release pattern, while amylase and lipase concentrations in the perfusion medium increased. Stimulated exocrine pancreatic secretion was significantly decreased already after 3-h ischemia and completely lost after 12 h.

Experimental and clinical work over the last 6 years has confirmed and broadened, but also challenged, the incretin concept. The nervous component of the entero-insular axis is still poorly defined, especially the peptidergic nerves, of which several contain insulinotropic regulatory peptides. The incretin effect is preserved after complete denervation of the porcine pancreas. Type 2 (non insulin-dependent) diabetic patients have a significantly decreased incretin effect. GIP (gastric inhibitory polypeptide; glucose dependent insulin releasing peptide) remains the strongest incretin factor. Its secretion depends on the absorption of nutrients. However, the correlation between the GIP response and disturbances of the entero-insular axis in some gastrointestinal diseases and, in particular, Type 2 diabetes, is poor. Furthermore, physiological concentrations of exogenous GIP do not produce fully the incretin effect and injection of GIP antibodies does not abolish the incretin effect. This suggests the existence of additional humoral incretin factors. On the other hand, GIP seems to have direct metabolic effects independent of its insulinotropic activity. The incretin effect of oral glucose is smaller if plasma levels of C-peptide rather than insulin are measured. However, decreased hepatic extraction of insulin after glucose ingestion only accounts partially for the incretin effect. GIP is unlikely to be the gut factor which regulates hepatic insulin extraction.

In the entero-insular axis, humoral factors predominate over neural factors. Gastric inhibitory polypeptide (GIP) is the most important incretin candidate. Release of GIP depends on nutrient absorption, and therefore secondary disturbances of GIP secretion occur in a number of gastrointestinal and metabolic diseases. Alterations of GIP secretion are not necessarily followed by alterations of insulin secretion. Disturbances of the entero-insular axis are best evaluated by estimating the incretin effect (i.e. comparing the insulin response to oral glucose with the insulin response to an isoglycaemic i.v. glucose infusion). From the poor correlation between disturbances of the entero-insular axis and GIP abnormalities it can even be concluded that other intestinal factors with incretin activity exist. The extent of the incretin effect may be overestimated if only serum insulin levels are measured. Unknown gut factors seem to exist which inhibit hepatic insulin extraction and thus alter serum insulin levels without stimulating insulin secretion.

The effect of insulin on fasting levels of immunoreactive gastric inhibitory polypeptide (IR-GIP) has been examined in insulin-dependent, juvenile-type diabetics who were well-controlled with two doses of an intermediate insulin. After withdrawal of the evening insulin injection the fasting blood glucose and serum IR-GIP levels were elevated and decreased significantly following intravenous insulin towards normal values. There was a significant positive correlation between levels of blood glucose and serum IR-GIP before and during insulin application. It is suggested that fasting serum GIP levels increase in case of insulin deficiency because basal GIP secretion is suppressed by normal insulin levels.

This study investigated early alterations of glucose metabolism in idiopathic haemochromatosis. Circulating concentrations of glucose, insulin, C-peptide, glucagon, and gastric inhibitory polypeptide (GIP) were measured after a 100-g oral glucose load in 10 men with idiopathic haemochromatosis in the non-cirrhotic stage of the disease. All had normal glucose tolerance and normal body weight. Ten matched healthy subjects were studied as controls. Insulin concentrations increased to significantly higher levels in patients with idiopathic haemochromatosis than in the control subjects from 30 to 180 min after the glucose load (p less than or equal to 0.01), while fasting insulin concentrations were not significantly different (p greater than 0.05). Concentrations of glucose, glucagon, C-peptide, and GIP were not significantly different at any time (p greater than 0.05). Thus, patients with idiopathic haemochromatosis show hyperinsulinaemia and hence insulin resistance without impaired glucose tolerance in the non-cirrhotic stage. Since pancreatic insulin secretion (C-peptide), glucagon secretion, and the entero-insulinar axis (GIP) are not impaired in these non-cirrhotic patients with idiopathic haemochromatosis, iron accumulation in the hepatocytes may be responsible for the impaired insulin effect and may cause impaired hepatic insulin extraction.

Glucagon-like peptide-1 and glucagon-like peptide-2 are encoded by the m-RNA of pancreatic preproglucagon. They show high conservation in different species and substantial sequence homology to glucagon. Because no definite biological activity of these peptides has been reported, we investigated the effect of synthetic C-terminally amidated glucagon-like peptide-1 [1-36] and synthetic human glucagon-like peptide-2 [1-34] with a free C-terminus on insulin release from isolated precultured rat pancreatic islets in the presence of glucose. Glucagon-like peptide-1 stimulates insulin release at 10 and 16.7 mmol/l glucose in a dose-dependent manner. Significant stimulation starts at 2.5 nmol/l in the presence of 10 mmol/l glucose and near maximal release is observed at 250 nmol/l, with approximately 100% increase over basal at both glucose concentrations. The peptide reaches 63% of the maximal stimulatory effect of glucagon. No stimulation occurs in the presence of 2.8 mmol/l glucose. Glucagon-like peptide-2 has no effect on insulin secretion at any glucose concentration tested. It is concluded that glucagon-like peptide-1, in contrast to glucagon-like peptide-2, exhibits a glucose-dependent insulinotropic action on isolated rat pancreatic islets similar to that of glucagon and gastric inhibitory polypeptide.

The effect of insulin on fat-induced gastric inhibitory polypeptide (GIP) release has been studied in seven healthy volunteers during euglycemic blood glucose clamping. In the first protocol, insulin (0.1 U/kg/h) was infused 2 h before ingestion of 100 g fat and continued for 2 h thereafter. In protocol II, saline was infused 2 h before the fat load and the insulin infusion started at the time of fat ingestion. During both insulin infusion studies, glucose levels were clamped at the fasting level by means of the Biostator and plasma levels of insulin, C-peptide, and GIP were estimated by radioimmunoassay. The response of GIP to oral fat was inhibited by 63% if insulin infusion was started at the time of fat ingestion, whereas no inhibition was seen if a 2-h hyperinsulinemic period preceded the fat load. The plasma insulin levels were comparable at the end of each experiment, ranging from 110 to 130 microU/ml. Plasma C-peptide levels decreased during the insulin infusion and increased after fat ingestion. These findings were not the result of inhibition of gastric emptying by insulin because they could be confirmed in four volunteers with intraduodenal infusion of fat. The present data show that insulin does inhibit fat-induced GIP secretion in normal man, but preceding hyperinsulinemic glucose clamping masks this insulin effect, probably by decreasing the sensitivity of the GIP cells to insulin.

The plasma levels of glucose, immunoreactive insulin (IRI), and immunoreactive GIP (IR-GIP) of eight healthy normal-weight subjects were compared following administration of oral glucose load of 10, 30, 60, 90, and 120 g, each given in a volume of 300 and 600 mL of water. By increasing the glucose load from 30 to 60 g, the integrated glucose response was more than doubled, irrespective of the volume. If more than 60 g glucose was given, the venous blood glucose levels did not significantly increase further. The IRI concentrations peaked between 30 and 40 minutes, irrespective of the size and volume of the glucose load. The peak values were significantly higher if 30, 60, and 90 g glucose was given in 600 mL than in 300 mL water. The integrated IRI output increased gradually if the glucose concentration of the 300 mL load was increased, whereas a maximal IRI response occurred already with 60 g glucose if dissolved in 600 mL water. At identical amounts of glucose (60 g) the integrated IRI response was doubled by increasing the ingested volume of the drink from 300 to 600 mL. Also the peak and integrated IR-GIP response increased in a dose-dependent manner by increasing the size of the glucose load. Larger amounts of glucose mainly prolong the GIP response. Significantly greater amounts of IR-GIP were released with 60 and 90 g glucose when given in 600 mL instead of 300 mL water. The parallel increment of IR-GIP and IRI is compatible with an important role of GIP as an insulinotropic gut factor.(ABSTRACT TRUNCATED AT 250 WORDS)

This study has assessed the effect of oral or intraduodenal HCl, administered alone or in combination with glucose, on gastric inhibitory polypeptide (GIP) and insulin secretion. Eight young males were given the following three oral tests: 30 g glucose, 150 cc 0.1 N HCl and the glucose-acid combination. Another group of eight controls received intraduodenal infusions of 20 g glucose, 75 cc 0.1 N HCl and their combination. All tests were performed in a random fashion at weekly intervals. When given alone, HCl did not influence glucose or insulin levels. However, HCl did produce an increase in GIP. The GIP response to acid was less than that to glucose and was delayed. The peak insulin and GIP responses with the glucose and glucose/acid combinations were similar with both the oral and intraduodenal routes. There was, however, a potentiation of both the GIP and insulin responses when intraduodenal acid was given with glucose. This effect on GIP and insulin was not evident with the oral glucose/acid load. It is concluded that HCl by itself is capable of stimulating GIP secretion. Since there was only a potentiation of insulin and GIP secretion when large doses of HCl were given together with glucose via the intraduodenal route, the physiological relevance of acid-induced GIP secretion remains to be resolved.

Cyclosporine was fed to male Wistar rats in a dose of 5, 10, or 50 mg/kg b.wt. for 7 days, and the effect on insulin secretion from the isolated perfused pancreas was investigated. Dose-dependently plasma insulin and pancreatic insulin content decreased while whole-blood CsA levels increased. An increase in blood glucose was only observed after feeding 50 mg/kg b.wt. CsA resulting in whole-blood CsA levels of 7735 ng/ml. Glucose (20 mM)-stimulated total insulin secretion (ng/50 min) was not affected during feeding 5 mg/kg b.wt. CsA, but was significantly reduced after feeding 10 or 50 mg/kg b.wt. CsA. The biphasic insulin secretion was reduced after 5 mg/kg b.wt. during the initial peak (0-10 min) but not during the second peak (10-50 min), whereas after 10 or 50 mg/kg b.wt. CsA both peaks were markedly reduced. The arginine (20 mM) and the arginine (20 mM)-plus-glucose (20 mM) stimulated insulin secretion was less affected after feeding 10 mg/kg b.wt. CsA than after stimulation with glucose (20 mM) alone. The addition of CsA to the perfusate did not influence glucose-stimulated insulin release from normal rat pancreas. Our results demonstrate a toxic effect of CsA on the pancreatic beta cell that is dose dependent and possibly influences both insulin secretion and biosynthesis.

The role of gastrointestinal and pancreatic hormones in regulating liver growth was evaluated by measuring their effect on DNA synthesis in the normal and regenerating liver of rats in vivo and in maintenance cultures of adult rat hepatocytes in vitro. After partial liver resection DNA synthesis reached peak levels after 24 hours while serum concentrations of immunoreactive insulin in portal and peripheral blood at this time were still suppressed. Increase of endogenous insulin levels by intravenous glucose infusion or portal infusion of insulin, glucagon or both together with glucose did not change DNA synthesis in normal or regenerating rat liver. After acute carbon tetrachloride poisoning of rats, survival rate and degree of liver necrosis was not changed by intraperitoneal infusion of glucagon and insulin with glucose. In vitro, insulin, glucagon and somatostatin synergistically stimulated the specific thymidine uptake in seven-day-old maintenance cultures of rat hepatocytes. The hormones did not cause cell multiplication but enhanced cell survival, probably by improving the uptake and utilization of nutrients. Gastrin G-17, secretin and cholecystokinin (contaminated with gastric inhibitory polypeptide) had no effect. It is concluded that the results do not support the contention that liver regeneration is regulated by the known pancreatic hormones. However, a trophic effect of pancreatic hormones on liver cells in vitro could be demonstrated. Gastrointestinal hormones had no such effect.

To investigate the role of endogenous insulin on the secretion of immunoreactive gastric inhibitory polypeptide (IR-GIP) the response of IR-GIP and immunoreactive insulin (IRI) to an oral fat load (100 g triglyceride) alone and during an intravenous glucose infusion (0.7 g/kg/h) was examined in normal weight and obese subjects. In normal weight subjects the fat induced integrated rise of IR-GIP was 112.7 +/- 9.4 ng/ml/120 min. When glucose and fat were given together this IR-GIP response was lowered to 46.2 +/- 2.9 ng/ml/120 min while the serum IRI response to i.v. glucose and the glucose tolerance were enhanced by fat ingestion. In obese subjects with normal glucose tolerance the GIP suppressing effect of i.v. glucose infusion was less marked than in controls. The integrated IR-GIP response to fat ingestion was 225.6 +/- 20.3 mg/ml/120 min and to fat plus glucose 152.6 +/- 14.8 ng/ml/120 min. In obese subjects with glucose intolerance i.v. glucose completely failed to lower the exaggerated secretion of IR-GIP following oral fat. Thus, a graded abnormality of the GIP response to glucose induced insulin release occurs in obesity with normal and pathological glucose tolerance. After reducing the ideal body weight of six obese subjects with glucose intolerance by hypocaloric diet for 3 weeks the exaggerated rise of IR-GIP after oral fat was reversed and the lowering effect of i.v. glucose on the IR-GIP response re-established.

The hepatotrophic effect of portal serum was demonstrated in primary monolayer cultures of adult rat hepatocytes: DNA synthesis in cultures incubated for 4 days with pooled portal serum of fed rats was significantly higher than in cultures with peripheral serum from the same animals. This difference was not observed when portal blood was treated with charcoal or obtained from fasting rats. Portal serum from diabetic rats stimulated DNA synthesis less than portal control serum. The trophic effect of portal serum from normal rats was enhanced after tolbutamide injection. The lost tropic effect of charcoal-treated portal serum was completely restored when insulin was added to reobtain a normal portal immunoreactive insulin concentration. Thus, insulin seems to be the main factor responsible for the hepatotrophic action of portal blood in vitro.

Integrated insulin secretion rates calculated from peripheral venous C-peptide measurements by two-compartment kinetic analysis were measured in six young normal subjects after increasing oral glucose loads of 25, 50, and 100 g and respective isoglycemic glucose infusions. The differences in B-cell secretory responses between oral and iv glucose challenges were attributed to factors other than glycemia itself (incretin effect). Both insulin and C-peptide concentrations as well as calculated integrated insulin secretion rates increased with increasing oral glucose loads. Due to the similarity in the glucose profiles after all oral loads, almost identical amounts of iv glucose (approximately 20 g) were infused in all "isoglycemic" infusion experiments, with resulting similar hormone profiles and insulin secretion rates. The percent contribution of incretin factors to total immunoreactive insulin responses after 25, 50, and 100 g glucose (85.6%, 74.9%, and 93.0%; response to oral load, 100%) was significantly higher than their contribution to integrated C-peptide responses (27.6-62.9%) or calculated integrated insulin secretion rates (19.2-61.0%). These findings indicate that the degree of incretin stimulation of insulin secretion depends on the amount of glucose ingested. A discrepancy between the estimates of the incretin effect derived from peripheral venous insulin responses, on the one hand, and C-peptide responses or calculated insulin secretion rates, on the other hand, exists. Inasmuch as peripheral insulin values reflect both insulin secretion and hepatic insulin removal, this discrepancy suggests that elimination kinetics of insulin differ between oral and iv glucose administration. This difference can be related to a significantly reduced fractional hepatic insulin extraction after oral (46.9-54.6%) compared to iv (63.4-76.5%) glucose administration when calculated by a three-compartment kinetic model. This reduction in fractional hepatic insulin extraction could be caused by gastrointestinal factors (hormones or nerves) stimulated in the course of glucose ingestion.