Feedback inhibition of insulin secretion is altered in cirrhosis

Hyperinsulinemia in human cirrhosis is generally considered an expression of reduced hepatic insulin degradation. To determine whether hyperinsulinemia may also depend on an altered feedback inhibition of insulin secretion, we performed euglycemic hyperinsulinemic clamp studies, infusing 40, 372, or 1280 mU/m2 X min biosynthetic human insulin in 30 compensated cirrhotic patients with portal hypertension and impaired glucose tolerance and 25 normal subjects, matched for age, sex, and weight. Mean fasting plasma insulin was significantly higher in cirrhotic patients [26.1 +/- 2.3 vs. 12.4 +/- 0.6 (+/- SE) microU/ml; P less than 0.001], while fasting plasma glucose levels were similar in the 2 groups. The mean plasma C-peptide level was significantly higher in cirrhotic patients, both basally (2.7 +/- 0.1 vs. 1.7 +/- 0.1 ng/ml; P less than 0.001) and during the clamp studies. Suppression of C-peptide at 120 min of the clamp was significantly less in cirrhotic patients (37 +/- 7% vs. 79 +/- 4%, 52 +/- 9% vs. approximately 100%, and 54 +/- 4% vs. approximately 100% during the 40, 372, and 1280 mU/m2 X min insulin infusions, respectively). The fasting C-peptide to insulin molar ratio was significantly lower in cirrhotic patients (5.4 +/- 0.3 vs. 6.4 +/- 0.3; P less than 0.005). The MCR of insulin at the three steady states was not significantly different between the 2 groups, whereas the basal systemic delivery rate of insulin was significantly higher in cirrhotic patients (14.7 +/- 1.7 vs. 6.5 +/- 0.4 mU/m2 X min; P less than 0.001). These results suggest that reduced feedback inhibition of insulin secretion may contribute to the hyperinsulinemia associated with cirrhosis.     

Glucose intolerance and insulin resistance in cirrhosis are normalized after liver transplantation.

Cirrhosis is often associated with insulin resistance and glucose intolerance. We evaluated if these alterations are restored by liver transplantation (LT). Glucose tolerance (oral glucose tolerance test [OGTT]), peripheral insulin sensitivity (euglycemic insulin clamp technique), glucose oxidation (indirect calorimetry), nonoxidative glucose disposal, and insulin secretion (hyperglycemic clamp technique) were measured in 6 patients (Group 1) before and 6 months after LT, in 12 patients (Group 2) who underwent LT 6 to 30 months previously, and in 6 healthy individuals (controls). In Group 1, glucose tolerance and insulin sensitivity (3.24 +/- 0.37 mg/kg/min) were normalized after LT (8.6 +/- 0.77 mg/kg/min; P <.0001; P = not significant vs. controls). The improved insulin-mediated glucose uptake was the result of a normalization of nonoxidative glucose disposal. Fasting insulin and C-peptide decreased from 24.6 +/- 3.3 microU/mL and 4.37 +/- 0.46 ng/dL, respectively, to 12.7 +/- 1.9 microU/mL and 2.46 +/- 0.5 ng/dL (controls: 10.0 +/- 3 microU/mL and 1.45 +/- 0.34 ng/dL). The glucose-induced increase of insulin concentration, which was higher before LT, showed a significant reduction, although the first phase of beta-cell secretion remained significantly higher compared with that of controls. All these findings were also confirmed in Group 2. The present data indicate that LT normalizes glucose tolerance and insulin sensitivity in cirrhotic patients through an improvement of both hepatic glucose clearance and the peripheral glucose disposal. The latter effect may be the result of the correction of chronic hyperinsulinemia. An increased first-phase beta-cell insulin secretion in response to high glucose levels persists, suggesting that a memory of previous insulin resistance is maintained.     

Hepatic and peripheral insulin resistance following streptozotocin-induced insulin deficiency in the dog

Insulin resistance and insulin deficiency are both present in many patients with diabetes mellitus. We tested the hypothesis that insulin resistance can evolve from a primary lesion of the beta-cell secretory function. Insulin-mediated glucose uptake (insulin clamp), endogenous glucose production, and glucose-stimulated insulin secretion (hyperglycemic clamp) were measured in awake dogs before and four to six weeks after streptozotocin-induced diabetes mellitus. Streptozotocin (30 mg/kg) resulted in a significant rise in the mean fasting plasma glucose concentration from 104 +/- 2 mg/100 mL to 200 +/- 34 mg/100 mL, (P less than 0.05), and a slight decrease in the mean fasting plasma insulin concentration (from 21 +/- 2 microU/mL to 15 +/- 2 microU/mL). Under conditions of steady-state hyperglycemia (+75 mg/100 mL hyperglycemic clamp, insulin secretion was reduced by 75% in the streptozotocin-treated dogs (P less than 0.025), and the total amount of glucose metabolized decreased from 13.56 +/- 1.04 to 4.74 +/- 0.70 mg/min X kg (P less than 0.001). In the postabsorptive state, endogenous glucose production was slightly, although not significantly, higher in the diabetic dogs (3.05 +/- 0.46 v 2.51 +/- 0.22 mg/min . kg), while the glucose clearance rate was 35% lower (P less than 0.001). When the plasma insulin concentration was increased to approximately 45 microU/mL (insulin clamp) while holding plasma glucose constant at the respective fasting levels (99 +/- 1 and 186 +/- 30 mg/100 mL), endogenous glucose production was completely suppressed in control dogs but suppressed by only 51% (1.46 +/- 0.37 mg/min . kg, P less than 0.025) in diabetic animals.(ABSTRACT TRUNCATED AT 250 WORDS)     

Elevated growth hormone levels and insulin resistance in patients with cirrhosis of the liver

Carbohydrate intolerance is frequently seen in patients with hepatic cirrhosis. To study the role of the counter regulatory hormones, glucagon, cortisol and growth hormone in this disease, these hormones were measured in 11 patients with hepatic cirrhosis and six controls during a 4-hour oral glucose tolerance test (OGTT) and in five normal and cirrhotic subjects during steady-state plasma insulin and glucose concentrations (SSPGI) achieved with the euglycemic clamp technique. Fasting plasma glucose was 103 +/- 4.3 mg/dl in cirrhotics and 88 +/- 3.3 mg/dl in controls (p less than 0.001). Immunoreactive insulin (IRI) was 24.3 microU/ml in cirrhotics and 12.7 +/- 2.2 microU/ml in controls (p less than 0.001); immunoreactive glucagon (IRG) was 263 +/- 30 pg/ml in cirrhotics and 122 +/- 17.5 pg/ml in controls (p less than 0.001); serum growth hormone (GH) was 4.4 +/- 0.9 ng/ml in cirrhotics and 0.5 +/- 0.1 ng/ml in controls (p less than 0.001). During OGTT, the 2-hour glucose concentration was 201 +/- 9.7 mg/dl in cirrhotic subjects and 147 +/- 10.0 mg/dl in controls (p less than 0.001). IRG levels were suppressed by 20% of basal values in patients with cirrhosis, while controls showed 10% suppression after an oral glucose load. At 60 minutes, the serum GH was 14.7 +/- 3.9 ng/ml in cirrhotics and 0.3 +/- 0.1 ng/ml in controls (p less than 0.001). The normal suppressive effect of hyperglycemia on GH secretion in controls was sharply contrasted by a paradoxical elevation of serum GH in the cirrhotic group.(ABSTRACT TRUNCATED AT 250 WORDS)     

Peripheral and hepatic insulin antagonism in hyperthyroidism

Eight hyperthyroid and eight normal subjects underwent 2-h oral glucose tolerance tests (OGTT) and euglycemic clamp studies to assess the presence of peripheral and hepatic insulin antagonism in hyperthyroidism. Although the mean total glucose area during the OGTT was similar in the hyperthyroid patients and normal subjects [16.4 +/- 0.8 (+/- SE) vs. 15.8 +/- 0.7 mmol/L.h], the mean insulin area was significantly elevated in the hyperthyroid group (1413 +/- 136 vs. 1004 +/- 122 pmol/L.h; P less than 0.05). Basal hepatic glucose production was measured during the second hour of a primed [3-3H]glucose infusion. A two-insulin dose euglycemic clamp study with [3-3H]glucose and somatostatin (500 micrograms/h) was carried out during the next 6 h. The insulin infusion rate was 0.05 mU/kg.min during the third, fourth, and fifth hours and 0.60 mU/kg.min during the sixth, seventh, and eighth hours. Hepatic glucose production and glucose utilization were measured during the final 0.5 h of each clamp period. Serum C-peptide concentrations were measured in the initial sample and in the last sample of each clamp period. The mean equilibrium serum insulin concentrations were similar in both groups during the final 0.5 h of the low (90 +/- 8 vs. 79 +/- 6 pmol/L) and high (367 +/- 11 vs. 367 +/- 15 pmol/L) insulin infusion rates. Basal serum C-peptide levels were significantly increased in the hyperthyroid patients (596 +/- 17 vs. 487 +/- 43 pmol/L; P less than 0.05) but were suppressed equally in both groups at the end of both clamp periods. The MCRs of insulin were similar in the hyperthyroid and normal subjects during the low (6.7 +/- 1.1 vs. 5.6 +/- 0.5 mL/kg.min) and high (11.9 +/- 0.4 vs. 12.1 +/- 0.5 mL/kg.mm) insulin infusion rates. Glucose production was significantly increased in the hyperthyroid patients during the basal state (17.6 +/- 0.9 vs. 11.5 +/- 0.5 mumol/kg.min; P less than 0.001) and remained elevated during the final 0.5 h of the low (12.1 +/- 1.1 vs. 5.9 +/- 1.7; P less than 0.01) and high (3.2 +/- 1.2 vs. 0.5 +/- 0.3; P less than 0.05) insulin infusion rates. Peripheral insulin action, assessed by Bergman's sensitivity index, was significantly decreased in the hyperthyroid patients (7.4 +/- 2.2 vs. 15.6 +/- 2.1 L/kg min-1/pmol/L; P less than 0.02). In conclusion, hyperthyroidism is characterized by 1) hyperinsulinemia after oral glucose loading, 2) increased basal hepatic glucose production, 3) impairment of insulin-mediated suppression of hepatic glucose production, and 4) antagonism to insulin-stimulated peripheral glucose utilization.     

The relationship between insulin sensitivity and skeletal muscle enzyme activities in hepatic cirrhosis

We have examined the hypothesis that insulin insensitivity in hepatic cirrhosis is related to abnormalities of glycogen deposition and skeletal muscle enzyme activities. Otherwise well patients with biopsy-proven hepatic cirrhosis secondary to previous excess alcohol intake were studied. Prior to study, in basal state, patients had identical blood glucose concentrations but raised serum insulin concentrations (cirrhotic: 8.5 +/- 0.8 mU per liter; matched control subjects: 5.7 +/- 0.5 mU per liter, p less than 0.01). Muscle glycogen content, glycogen synthase activity and pyruvate dehydrogenase activity were normal in the basal state. The cirrhotic patients required less glucose to maintain the clamp in response to 0.1 unit per kg per hr insulin (6.7 +/- 0.5 vs. control 8.3 +/- 0.4 mg per kg per min, p less than 0.05) and deposited less glycogen in muscle during the clamp (8.6 +/- 0.5 vs. 12.0 +/- 1.4 mg per gm protein, p less than 0.05). Glycogen deposition correlated with clamp glucose requirement in the cirrhotic patients (r = 0.78, p less than 0.05). The expressed activity of glycogen synthase activity was significantly lower in cirrhotic patients at the end of the clamp (26.5 +/- 1.1% vs. 30.9 +/- 1.6%) and again correlated with clamp glucose requirement (r = 0.82, p less than 0.05). Skeletal muscle pyruvate dehydrogenase activity was not different in patients and control subjects. Insulin insensitivity in hepatic cirrhosis appears to be related to abnormalities of glucose deposition as glycogen in skeletal muscle.     

Resistance to insulin suppression of plasma free fatty acids in liver cirrhosis

Insulin action on carbohydrate metabolism is known to be reduced in liver cirrhosis. However, little is known about the effect of insulin on free fatty acid (FFA) metabolism in these patients. To investigate this aspect we performed a two-step insulin euglycemic clamp in 11 cirrhotic patients and 6 controls. Insulin was infused at 0.25 mU/Kg min from 0 to 100 min and at 1 mU/Kg from 100 to 200 min. The FFA lowering capacity of insulin was studied during the first step; the glucose metabolizing capacity (M) was evaluated during the second step. In the cirrhotic patients, the M value was lower than in controls (3.91 +/- 0.48 vs 7.75 +/- 1.09 mg/kg/min, respectively). During the low insulin infusion, FFA and glycerol plasma levels were decreased in both groups. However, the ability of insulin to suppress plasma FFA and glycerol was lower in cirrhotics than in controls. In fact, at 100 min, FFA were 50% of basal values in cirrhotics and 20% in controls (p less than 0.01), while glycerol plasma levels decreased to 70% of basal values in patients and to 56% in controls. The slope of the linear regression obtained between Ln-FFA concentrations vs time was significantly less in cirrhotic patients than in controls (p less than 0.001). In addition, a positive correlation was found between the M value (r = 0.70; p less than 0.01) and the slope of the Ln-FFA in each patient. These findings suggest that in cirrhotic patients the effects of insulin on both FFA and glucose metabolism are reduced.     

Characteristics of insulin resistance in Turner syndrome

The characteristics of insulin resistance, in Turner syndrome are still unclear. For this purpose in 4 patients with Turner syndrome and in 8 control females we performed an euglycaemic hyperinsulinemic glucose clamp at the following insulin infusion rates (50 and 100 mU/Kg x h), each period lasting 120 min. A simultaneous infusion of D-3-H-glucose allowed us to determine in basal conditions and during the clamp hepatic glucose output and glucose disappearance rate (Rd). In basal conditions plasma glucose (4.8 +/- 0.1 vs 4.6 +/- 0.2 mmol/1 p = NS) and plasma glucagon (102 +/- 7.5 vs 112 +/- 11.3 ng/l p = NS) were similar in both groups despite higher plasma insulin (19 +/- 1.8 vs 7 +/- 2.2 mU/l p less than 0.05) and C-peptide (1.0 less than 0.1 vs 0.8 +/- 0.06 pmol/l p less than 0.05) levels in patients with Turner syndrome. In the last 60 min of the lower insulin infusion rate glucose infusion rate (4.1 +/- 0.3 vs 2.9 +/- 0.4 mg/Kg x min p less than 0.05) and glucose disappearance rate (3.89 +/- 0.12 vs 2.63 +/- 0.11 mg/Kg x min p less than 0.01) were significantly reduced in patients with Turner. On the contrary hepatic glucose output was similarly suppressed in both groups of subjects. Doubling the insulin infusion rate, we obtained similar results in patients and controls respectively. So we conclude that in Turner syndrome the insulin resistance state is mainly due to a muscular receptor defect.     

Simultaneous assessment of insulin secretion and insulin sensitivity using a hyperglycemia clamp.

A hyperglycemic clamp is an established method to assess insulin secretion and is generally used only for this purpose. To determine whether it could also be used to assess insulin sensitivity, we compared insulin sensitivity indices (ISI) obtained during euglycemic and hyperglycemic clamp experiments in 22 nonobese volunteers (body mass index, 23.9 +/- 0.6 kg/m2) and in 20 obese individuals (body mass index, 30.8 +/- 1.3 kg/m2) matched for age and gender. The ISI values (micromoles per kg.min/pmol) of the obese group assessed during hyperglycemic (0.088 +/- 0.011) and euglycemic (0.050 +/- 0.005) clamp experiments were both significantly lower than the ISI of the nonobese group assessed in hyperglycemic and euglycemic clamp experiments (0.179 +/- 0.024 and 0.096 +/- 0.009, respectively; both P less than 0.01). Although the ISI values obtained with hyperglycemic clamps were consistently greater than those obtained with euglycemic clamp (0.137 +/- 0.016 vs. 0.075 +/- 0.007; P less than 0.001), they were highly correlated (r = 0.84; P less than 0.0001). Moreover, when these indices were converted to clearance rates, thereby correcting for the mass action effects of glucose on glucose disposal, the values obtained with the hyperglycemic clamp (0.0137 +/- 0.0016 mL/kg.min/pmol) were statistically identical to those obtained with the euglycemic clamp (0.0142 +/- 0.0013 mL/kg.min/pmol), as indicated by a regression equation having an intercept of 0 and a slope (1.03) not different from 1. We, therefore, conclude that the hyperglycemic clamp and the euglycemic clamp yield comparable estimates of insulin sensitivity and that, under appropriate conditions, the hyperglycemic clamp technique may be used to assess both insulin sensitivity and insulin secretion in the same individual in a single experiment.     

Increased insulin sensitivity is associated with reduced insulin and glucagon secretion and increased insulin clearance in man

Insulin secretion is increased in insulin resistance. In this study, we examined whether high insulin sensitivity results in low insulin secretion. Twelve male master athletes [age 25.6 +/- 4.1 (mean +/- SD) yr] and seven male sedentary students (age 25.0 +/- 2.0 yr) underwent a hyperinsulinemic, euglycemic clamp and a glucose-dependent arginine stimulation test. Athletes had high insulin sensitivity [230 +/- 18 vs. 92 +/- 12 (nmol glucose/kg.min)/(pmol insulin/liter), P < 0.001] and low insulin response to arginine (at fasting glucose 135 +/- 22 vs. 394 +/- 60 pmol/liter, P < 0.001), which resulted in unaltered disposition index (32.8 +/- 3.8 vs. 33.5 +/- 3.3 micro mol glucose/kg.min, NS). Also, the C-peptide response to arginine was reduced (at fasting glucose 0.71 +/- 0.09 vs. 0.89 +/- 0.09 nmol/liter, P = 0.034). However, the C-peptide reduction was not as large as the insulin reduction yielding increased disposition index in athletes when calculated from C-peptide data (184 +/- 9 vs. 76 +/- 11 micro mol glucose/kg.min, P < 0.001). This difference is explained by increased insulin clearance among the athletes during the first 5 min after arginine (81.1% +/- 1.8% vs. 53.6% +/- 4.7%, P < 0.001). Also, the glucagon response to arginine was reduced in the athletes (58.8 +/- 6.7 vs. 90.1 +/- 9.9 ng/liter at fasting glucose, P = 0.009). We conclude that high insulin sensitivity results in low islet hormone secretion and increased insulin clearance.     

Fractional hepatic extraction of insulin in man: is it constant?

The present study was designed to compare insulin extraction by the liver following oral glucose administrations of different size, in order to evaluate insulin removal by the liver in relation to the insulin exposure, and to the amount of ingested glucose. Insulin secretion by the pancreas was estimated by the measurement of peripheral C-peptide levels, and insulin extraction by the liver by the analysis of peripheral C-peptide to insulin ratios and relations. Ten healthy subjects (5 males and 5 females), aged 16 to 66 yr, with normal bw, and without family history of diabetes mellitus were investigated by means of the administration, on alternate days, of 50 and 150 g oral glucose loads. After the 150 g oral glucose load plasma glucose levels were significantly higher than after the 50 g oral glucose administration: glucose incremental areas of 1.45 +/- 0.12 vs. 0.55 +/- 0.04 mmol/l X min, respectively (p less than 0.001). Similarly, insulin concentrations were significantly higher following 150 g than after 50 g glucose ingestion: insulin incremental areas of 0.52 +/- 0.09 vs. 0.20 +/- 0.04 nmol/l X min (p less than 0.001). Also C-peptide levels were higher after 150 vs. 50 g oral glucose load: C-peptide incremental areas of 1.85 +/- 0.41 vs. 0.64 +/- 0.13 nmol/l X min (p less than 0.01). C-peptide to insulin molar ratios were similar during the two glucose challenge, and averaged 5.25 +/- 0.42 vs. 5.08 +/- 0.50 after 50 and 150 g oral glucose loads, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)     

Further evidence that insulin metabolism is a major determinant of peripheral insulin response to oral glucose in subjects with mild glucose intolerance

In mild glucose intolerance plasma concentration of C-peptide seems to give an estimate of pancreatic B cell secretion more reliable than plasma insulin itself. In the present study we measured the plasma levels of insulin and C-peptide after oral glucose load in 100 mildly glucose intolerant subjects, focusing our attention on high and low insulin responders. According to an insulin incremental area after oral glucose higher or lower than the mean +/- SD of the mean, 16 subjects were classified as "high insulin responders", and 17 as "low insulin responders". The two groups were similar for sex, age and bw. Mean insulin incremental area was almost 9-fold greater in high insulin responders than in low insulin responders (0.88 +/- 0.03 vs 0.10 +/- 0.01 pmol/ml min, p less than 0.001). Also mean C-peptide incremental area was significantly greater in high insulin responders than in low insulin responders, but the differences between the two groups were smaller. Indeed, mean C-peptide area was approximately 2.5-fold greater in high insulin responders than in low insulin responders (1.58 +/- 0.12 vs 0.66 +/- 0.07 pmol/ml min, p less than 0.001). These results give further support to the concept that in mild glucose intolerance insulin metabolism is a major determinant of peripheral insulin response to oral glucose load.     

Insulin action and insulin secretion in newly diagnosed type 2 diabetic patients

To clarify the insulin action and insulin secretion in newly diagnosed type 2 diabetic subjects, we investigated insulin and C-peptide response to an oral glucose tolerance test (OGTT) in 15 newly diagnosed type 2 diabetic patients and 17 healthy subjects. For insulin action, we found fasting hyperinsulinemia (8.4 +/- 0.8 vs. 6.0 +/- 0.5 microIU/ml, p = 0.014), higher insulin resistance by homeostasis model assessment (HOMA) (4.33 +/- 0.2 vs. 1.34 +/- 0.1 microIU/ml.mmol/l, p < 0.001), and lower insulin sensitivity index (ISI) (51.0 +/- 0.7 vs. 104.0 +/- 0.8, p < 0.001) in newly diagnosed diabetic patients compared to normal subjects. For insulin secretion, the increments of AUCI (area under curve of insulin) and AUCC-P (area under curve of C-peptide) (increment of AUCI: 26.1 +/- 1.4 vs. 82.8 +/- 4.5 microIU/ml.hour, p < 0.001; increment of AUCC-P: 3.9 +/- 0.2 vs. 11.4 +/- 0.6 ng/ml.hour, p < 0.001), insulin secretion by HOMA model (20.7 +/- 1.2 vs. 79.1 +/- 3.8 IU/mol, p < 0.001), and ratio of 30 min increment of fasting insulin to glucose during OGTT (1.14 +/- 0.1 vs. 13.1 +/- 0.5 IU/mol, p < 0.001) were significantly lower in the newly diagnosed diabetic patients than normal subjects. In addition, body mass index (BMI) in our type 2 diabetes is relatively lower (24 +/- 0.65 kg/m2) than those in western countries. These findings revealed poor insulin action and decreased insulin secretion in relatively less obese Taiwanese with newly diagnosed type 2 diabetes.     

Lack of insulin inhibition on insulin secretion in non-diabetic morbidly obese patients.

OBJECTIVE: Insulin inhibition of insulin secretion has been described in normal lean subjects. In this study, we examined whether this phenomenon also occurs in the morbidly obese who often have severe peripheral insulin resistance. SUBJECTS: Twelve obese patients, normotolerant to glucose (8 F/4 M, body mass index (BMI)=54.8+/-2.5 kg/m(2), 39 y) and 16 lean control subjects (10 F/6 M, BMI=22.0+/-0.5 kg/m(2), 31 y). DESIGN AND MEASUREMENTS: An experimental study using various parameters, including an euglycemic hyperinsulinemic clamp (280 pmol/min/m(2) of body surface), an oral glucose tolerance test (OGTT), electrical bioimpedance and indirect calorimetry. RESULTS: The obese subjects were insulin resistant (M=19.8+/-1.6 vs 48.7+/-2.6 micromol/min kg FFM, P<0.0001) and hyperinsulinemic in the fasted state and after glucose ingestion. Fasting plasma C-peptide levels (obese 1425+/-131 pmol/l vs lean 550+/-63 pmol/l; P<0.0001) decreased less during the clamp in the obese groups (-16.9+/-6.9% vs -43.0+/-5.6% relative to fasting values; P=0.007). In the lean group, the C-peptide decrease during the clamp (percentage variation) was related to insulin sensitivity, M/FFM (r=0.56, P=0.03), even after adjustment for the clamp glucose variation. CONCLUSION: We conclude that, in lean subjects, insulin inhibits its own secretion, and this may be related to insulin sensibility. This response is blunted in morbidly obese patients and may have a role in the pathogenesis of fasting hyperinsulinemia in these patients.     

Influence of gliclazide on glucose-stimulated insulin release in man

Although sulfonylureas (SU) are widely used in the management of patients with non-insulin-dependent diabetes mellitus (NIDDM), there is still debate about their mechanism of action on the pancreatic beta cell. It is unclear whether the effect of SU on insulin release is additive to the effect of glucose, or whether SU act by increasing pancreatic beta-cell sensitivity to glucose (a shift in the dose-response curve of glucose-stimulated insulin release without a change in maximum release). To address this issue, we assessed the influence of the SU gliclazide on glucose-stimulated insulin release in eight healthy male volunteers. Sixty-minute hyperglycemic glucose clamps (blood glucose levels: 8 mmol/L, a submaximal stimulus; and 32 mmol/L, a maximally stimulating concentration) were performed with and without prior oral administration of gliclazide (80 mg) 30 minutes before the glucose clamp. Mean plasma C-peptide increment at 5 minutes (first-phase secretion) obtained during the 8-mmol/L hyperglycemic clamp, was higher on the gliclazide study day than on the control day (1.07 +/- 0.10 v 0.88 +/- 0.10 mmol/L, P less than .05), whereas no difference in plasma C-peptide response was observed during the 32-mmol/L hyperglycemic clamp. Mean plasma C-peptide increment obtained at the end 60 minutes; (second-phase secretion) of the 8-mmol/L hyperglycemic clamps was higher on the gliclazide study day than on the control day (1.36 +/- 0.13 v 1.09 +/- 0.09 mmol/L, P less than .02). No difference was observed in plasma C-peptide response at the end of the 32-mmol/L hyperglycemic glucose clamps.(ABSTRACT TRUNCATED AT 250 WORDS)     

Continuous subcutaneous insulin infusion therapy decreases insulin resistance in type 1 diabetes

The influence of continuous sc insulin infusion therapy for 6 weeks on sensitivity to insulin (euglycemic clamp technique) and hepatic glucose production (3-[3H]glucose technique) was measured in 10 type 1 diabetic patients whose mean duration of diabetes was 8 yr. Mean diurnal blood glucose fell from 8.5 +/- 0.8 (SEM) mmol/liter to 6.0 +/- 0.6 mmol/liter (P less than 0.05) and glycosylated hemoglobin from 10.5 +/- 0.4% to 8.7 +/- 0.3%. Insulin requirements declined by 23% from 47 +/- 4 U/day prepump to 36 +/- 2 U/day after 6 weeks of pump therapy (P less than 0.01). During the insulin clamp, plasma insulin was maintained at approximately 90 mU/liter and plasma glucose at approximately 5.0 mmol/liter in all studies. The rate of glucose metabolism in diabetic patients during conventional therapy (4.65 +/- 0.41 mg/kg X min) was 35% lower than in normal subjects (7.20 +/- 0.42 mg/kg X min, n = 14, P less than 0.001). After 6 weeks of pump therapy, total glucose uptake increased by 27% to 5.90 +/- 0.60 mg/kg X min, P less than 0.05 vs. prepump). This was still 18% lower than in the normal subjects (P less than 0.05). Basal hepatic glucose production in the diabetic patients during conventional therapy (3.07 +/- 0.14 mg/kg X min) was 70% higher than in the normal subjects (1.79 +/- 0.07 mg/kg X min, n = 7, P less than 0.001). After 6 weeks of pump therapy, hepatic glucose production fell to 2.48 +/- 0.19 mg/kg X min (P less than 0.05), which was still 40% higher than in the normal subjects (P less than 0.01). Basal hepatic glucose production was directly related to the fasting plasma glucose level (r = 0.67, P less than 0.001) and inversely proportional to fasting insulin concentration (r = -0.48, P less than 0.05) in the diabetic patients. Specific tracer insulin binding to erythrocytes in the diabetic patients (19.4 +/- 1.5%) was comparable to that in the normal subjects (19.6 +/- 1.2%) and remained unchanged during pump therapy. Thus the improved metabolic control resulting from pump therapy is associated with enhancement in sensitivity to insulin, and reduction in basal hepatic glucose production.     

Effects of physiologic and supraphysiologic hyperglycemia on early and late-phase insulin secretion in chronically dialyzed uremic patients

To test secretory capacity of the beta-cell to a glucose stimulus in uremic patients on chronic dialysis, three hyperglycemic clamps (plasma glucose increments: 1, 4.5 and 11 mmol/l) were performed in 8 uremic and 8 healthy subjects. Early-phase insulin and C-peptide responses (delta I and delta C) during the initial 6 min were consistently exaggerated at all three steps in uremic patients compared with controls (delta I. 16 +/- 4 vs 4 +/- 2, 41 +/- 11 vs 15 +/- 4 and 60 +/- 12 vs 24 +/- 5 mU/l; delta C. 0.39 +/- 0.13 vs 0.07 +/- 0.02, 0.40 +/- 0.13 vs 0.16 +/- 0.02 and 0.73 +/- 0.15 vs 0.29 +/- 0.04 nmol/l, p less than 0.05 in all cases). Similarly, late-phase insulin secretion defined as the insulin increment between 90 and 120 min after initiation of the glucose challenge was enhanced in uremic patients at the two highest glycemic steps (44 +/- 10 vs 16 +/- 2 and 123 +/- 29 vs 44 +/- 5 mU/l, both p less than 0.01). The raised late-phase insulin response allowed comparable glucose disposal in the two groups (uremic patients: 9.2 +/- 1.0 and 15.5 +/- 1.6 mg.kg-1.min-1. Controls: 9.0 +/- 1.3 and 19.9 +/- 2.4 mg.kg-1.min-1). The slopes of potentiation, i.e. the slopes of the regression lines expressing the relationship between changes in insulin increments and changes in glucose, were markedly steeper in uremic patients (0.45 +/- 0.09 and 0.66 +/- 0.20, early and late-phase respectively) than in controls (0.20 +/- 0.06 and 0.25 +/- 0.03).(ABSTRACT TRUNCATED AT 250 WORDS)     

Insulin and glucagon secretion are suppressed equally during both hyper- and euglycemia by moderate hyperinsulinemia in patients with diabetes mellitus

Hyper- and euglycemic clamp studies were performed in patients with noninsulin-dependent diabetes mellitus to examine the effects of exogenous insulin administration on insulin and glucagon secretion. Plasma glucose was kept at the fasting level [mean, 10.0 +/- 0.2 (+/- SE) mmol/L; hyperglycemic clamp], and graded doses of insulin (1, 3, and 10 mU/kg.min, each for 50 min) were infused. The plasma C-peptide level gradually decreased from 523 +/- 66 to 291 +/- 43 pmol/L (n = 13; P less than 0.005) by the end of the hyperglycemic clamp study. After 90 min of equilibration with euglycemia (5.4 +/- 0.1 mmol/L; euglycemic clamp), the same insulin infusion protocol caused a similar decrease in the plasma C-peptide level. With the same glucose clamp protocol, physiological hyperinsulinemia for 150 min (676 +/- 40 pmol/L), obtained by the infusion of 2 mU/kg.min insulin, caused suppression of the plasma C-peptide level from 536 +/- 119 to 273 +/- 65 pmol/L during hyperglycemia and from 268 +/- 41 to 151 +/- 23 pmol/L during euglycemia (n = 9; P less than 0.005 in each clamp). Plasma glucagon was suppressed to a similar degree in both glycemic states. These results demonstrate that 1) insulin secretion in non-insulin-dependent diabetes mellitus is suppressed by high physiological doses of exogenous insulin in both the hyper- and euglycemic states, the degree of inhibition being independent of the plasma glucose level; and 2) glucagon secretion is also inhibited by such doses of exogenous insulin.     

Influence of rosiglitazone treatment on beta-cell function in type 2 diabetes: evidence of an increased ability of glucose to entrain high-frequency insulin pulsatility

Thiazolidinediones have well-established insulin-sensitizing effects. Their impact on insulin secretion is less clarified. Consequently, we sought to determine potential effects of a thiazolidinedione (rosiglitazone) on the beta-cell function. Twenty type 2 diabetic individuals were randomized to receive rosiglitazone (rosi) 4 mg twice daily or placebo (pla) for 13 wk. Before treatment and at the end of the treatment period, the patients underwent an iv glucose tolerance test (0.3 g/kg), a hyperglycemic (15 mmol/liter) clamp with arginine (5 g) stimulation, assessment of baseline high-frequency insulin pulsatility, and glucose-entrained insulin pulsatility (6 mg/kg.min every 10 min), and a hyperinsulinemic euglycemic clamp. Fasting plasma glucose was reduced (pla, 8.2 +/- 2.1 vs. 8.8 +/- 2.6 mmol/liter; rosi, 8.6 +/- 7.1 vs. 7.1 +/- 1.2 mmol/liter; P < 0.01), and insulin sensitivity was increased by rosiglitazone treatment (M value: pla, 5.3 +/- 1.8 vs. 5.4 +/- 1.6 mg/kg.min; rosi, 5.9 +/- 2.2 vs. 7.4 +/- 1.3 mg/kg.min; P = 0.05). First-phase insulin secretion and insulin secretory capacity were unaffected. Glucose-entrained insulin secretion was increased as assessed by spectral power analysis (P = 0.05). In conclusion, rosiglitazone treatment for 3 months in type 2 diabetic patients exerts no action on insulin secretion per se. Improved glucose-entrained high-frequency insulin pulsatility suggests an increased ability of the beta-cell to sense and respond to glucose changes within the physiological range.     

The relationship between first-phase insulin secretion and glucose metabolism.

A possible pathogenetic link between absence of first-phase insulin secretion and development of impaired glucose metabolism has been suggested by the results of several cross-sectional studies. First-phase insulin secretion measured during a +7 mmol/l hyperglycemic glucose clamp correlated with total glucose disposal during the clamp (r = 0.65, p < 0.001, N = 59). To examine whether restoration of first-phase insulin secretion improves peripheral glucose uptake in subjects with impaired glucose utilization, seven insulin-resistant subjects (age 54 (38-62) years: BMI 29.3 (21.7-35.8); fasting plasma glucose 5.5 (4.8-7.2) mmol/l; fasting insulin 57 (37-105) pmol/l with impaired first-phase (148 (29-587) vs controls 485 (326-1086) pmol/l x 10 min; p < 0.05) and normal second-phase (1604 (777-4480) vs controls (1799 (763-2771) pmol/l x 110 min) insulin secretion were restudied. The impaired first-phase insulin secretion was restored by an iv insulin bolus at the start of the hyperglycemic clamp. Substrate oxidation rates and hepatic glucose production were determined by indirect calorimetry and [3-3H]glucose infusion. Total glucose uptake was impaired in the insulin-resistant subjects with impaired first-phase insulin secretion compared to controls (18.8 (13.2-22.2) vs 34.8 (24.3-62.1) mumol.kg-1 x min-1; p < 0.01). Restoration of first-phase insulin secretion (1467 (746-2440) pmol/l x 10 min) did not affect glucose uptake (20.2 (9.9-23.8) mumol.kg-1.min-1), with no difference in oxidative and non-oxidative glucose metabolism between the experiments. Second-phase insulin secretion was similar during both experiments.(ABSTRACT TRUNCATED AT 250 WORDS)