Hormonal and metabolic adaptation to protein-supplemented fasting in obese subjects

Thirty hospitalized, severely obese patients (40 +/- 2 yr, 82 +/- 4 percent weight excess) were submitted to a 13-d protein-supplemented fast (PSF) with 70 g milk proteins/d (1.26 MJ or 300 kcal). The mean weight loss during PSF was 5.4 +/- 0.3 kg corresponding to 422 +/- 39 g/d. Comparison of the urinary nitrogen excretion with daily protein intake revealed that the nitrogen balance was equilibrated during PSF. Blood glucose decreased moderately but significantly during the whole PSF period whereas plasma insulin was only reduced during the first 9 d and tended to rise thereafter. Plasma FFA increased rapidly and remained elevated until the end of the study (+ 60 per cent); serum total cholesterol and plasma triglycerides showed a 26 and a 35 per cent decrease respectively. Basal plasma glucagon was slightly increased. Due to the low sodium intake (42 mmol/d) urinary sodium excretion dropped rapidly. Simultaneously both systolic (-13 mmHg) and diastolic (-7 mmHg) arterial blood pressure decreased significantly. The biological tolerance was good: metabolic acidosis was prevented with sodium bicarbonate, excessive rise in serum uric acid was corrected with allopurinol and a marked decrease in serum potassium was avoided with an appropriate dose of spironolactone. Twenty-six patients could be weighed 6 to 15 months after PSF: 12 showed a further weight reduction (6.6 +/- 1.6 kg) and seven a discrete weight gain (1.0 +/- 0.4 kg). Thus, PSF was well accepted and was profitable in 19 out of our 30 patients. It should be restricted to cases of severe and refractory obesity and performed under careful medical supervision.

Improvement of metabolic control in insulin dependent diabetics treated with the alpha-glucosidase inhibitor acarbose for two months

Acarbose, an alpha-glucosidase inhibitor, delays starch digestion and inhibits intestinal sucrase and maltase activity. Twenty-eight insulin dependent diabetics were given Acarbose (3 x 100 mg daily) over a two month period, preceded and followed by a two month placebo period. Acarbose reduced post-breakfast and post-dinner blood glucose values by 25% (p less than 0.001) and 24% (p less than 0.05) respectively. It also significantly reduced mean daily blood glucose by 18% (p less than 0.05) and mean amplitude of glycaemic excursions from 8.0 +/- 0.6 to 5.5 +/- 0.4 mmol/l (p less than 0.0005). Weight did not change significantly. Daily caloric and carbohydrate intake remained constant throughout the study while insulin requirements decreased slightly but significantly. Out of the 28 patients, 18 had absent while ten had slight residual B cell function as assessed by plasma C-peptide measurements. Treatment with Acarbose did not significantly affect residual B cell function. The beneficial effect of Acarbose on blood glucose control was seen in patients both with and without residual B cell secretion. The major side-effect was flatulence which was never severe enough to interrupt treatment, but led to a 50% reduction of the dose in one patient. It is concluded that Acarbose represents a useful additional means of improving metabolic control in insulin dependent diabetics.

Inhibition by somatostatin of carbamylcholine-induced gastrin and glucagon release from the isolated perfused canine stomach

At an arterial plasma concentration of 61 nmol/l (100 ng/ml) synthetic cyclic somatostatin completely abolished basal glucagon and gastrin release as well as carbamylcholine-induced glucagon and gastrin release from the isolated perfused dog stomach. These observations are compatible with the view that endogenous somatostatin previously reported to be released during vagal stimulation might be involved to explain the lack of gastric-glucagon response in this situation. They do not, however, rule out the alternative proposal that the dog fundic A-cell may simply be a non-innervated cell.

Effect of bicarbonate on the arginine-induced insulin and glucagon secretion in vitro

The isolated perfused rat pancreas was used to investigate the effect of extracellular bicarbonate concentration (25 and 40 mmol/l) on the arginine-induced insulin and glucagon release with or without 5.5 mmol/l glucose in the perfusate. In the absence of glucose, the insulin response to arginine was increased at 40 mmol/l bicarbonate, while no potentiation by glucose of the arginine-induced insulin release, at this bicarbonate concentration, was observed. At 40 mmol/l bicarbonate, glucose inhibition of basal glucagon release was abolished. In contrast, the glucagon response to arginine was not affected by the bicarbonate concentration in the perfusion medium. These results confirm that the bicarbonate concentration in the medium markedly influences the insulin release from the isolated perfused rat pancreas, while the glucagon response to arginine seems to be less sensitive to changes in extracellular bicarbonate concentration.

Failure of indomethacin to affect arginine-induced C-peptide and glucagon release in insulin-treated diabetics. Major role of residual B cell function in conditioning the magnitude of the blood glucose rise after intravenous arginine

Fourteen insulin-treated diabetics were submitted to an arginine infusion test performed with either 11.7 or 5.85 mg kg-1 min-1 arginine monohydrochloride infused during 40 min with or without previous oral administration of a low (75 + 50 mg) or a high (75 mg + 3 mg/kg) dose of indomethacin. Blood glucose, plasma non-esterified fatty acids, insulin, C-peptide and glucagon were determined at regular intervals before, during and after the arginine infusion. These parameters were totally unaffected by the two doses of indomethacin both in the basal state and during the arginine infusions at the two loads tested. Eight subjects had a basal C-peptide level above 0.07 pmol/ml and a mean (+/- SEM) maximal rise of 0.21 +/- 0.04 pmol/ml during the arginine infusion, whereas the remaining six patients had virtually zero values throughout the tests. The arginine-induced plasma glucagon rise was similar for the two rates of arginine infusion; the sum of the increments in plasma glucagon averaged 877 +/- 120 and 647 +/- 92 pg/ml (p greater than 0.1) for the high and low rates of arginine infusion, respectively. The magnitude of the blood glucose rise appeared independent of the amount of arginine infused. Confirming previous reports, we found that the blood glucose rise after arginine was three to four times higher in subjects without C-peptide than in subjects with C-peptide. The mean glucagon response did not differ significantly between subjects with or without C-peptide. Thus, residual B cell function determines the magnitude of the blood glucose rise but not the glucagon response after intravenous arginine.

Effect of indomethacin on the metabolic and hormonal response to a standardized breakfast in normal subjects

We have investigated the influence of a single oral administration of indomethacin on blood glucose, plasma free fatty acids (FFA), alpha-amino-nitrogen, insulin and glucagon concentrations in young healthy subjects. Two groups of 6 subjects were studied, the first received a standardized 500 kcal mixed meal without any previous drug administration (controls) whereas the second group received 50 mg indomethacin 2 h before ingesting an identical meal. Plasma indomethacin concentration reached its maximum (2.36 +/- 0.36 micro g/ml) 15 min after administration and declined to 0.45 +/- 0.04 micro g/ml after 2 h. Indomethacin ingestion was followed by a significant increase in blood glucose and plasma FFA reaching their maximum value at 45 min and returning to basal levels at 120 min. No simultaneous changes in plasma alpha-amino-nitrogen, insulin or glucagon levels were detected during this period. The meal was followed by a rise in blood glucose and plasma insulin as well as by a decrease in plasma FFA concentration. No significant differences were detected between the controls and the subjects receiving indomethacin. In controls, the meal was followed by a rise in plasma alpha-amino-nitrogen and a modest although significant increase in glucagon levels. In indomethacin-treated subjects, the increment of alpha-amino-nitrogen was less marked and the increase in plasma glucagon was not observed. Thus, indomethacin by itself can exert several metabolic effects; however, it does not deteriorate the blood glucose or insulin profile after a regular meal. The present work is the first to demonstrate that an inhibitor of prostaglandin synthesis inhibits the plasma glucagon rise occurring after a physiological stimulus such as a normal meal. On the basis of previous in vitro experiments, we suggest that this effect results from an inhibition of glucagon secretion by the PG synthesis inhibitor.

Computer-assisted instruction for diabetics. An original project developed at the University of Liège, Belgium

An experiment conducted at the University of Liège, Belgium in the computer-assisted teaching of juvenile-type insulin-dependent diabetics is reported. The course was designed to individualize teaching in order to give a minimum of basic information to a maximum of patients. The original computer-assisted instruction system DOCEO II is described; methods and programming techniques are summarized. The course was evaluated on fifty patients who attended the first four lessons. The improvement in knowledge was impressive: scores corrected for guessing were 9.59 +/- 4.43 out of 20 before the course and 17.04 +/- 2.26 out of 20 after the course (paired t test: p less than 0.001). The system was enthusiastically accepted by the patients and is now routinely used in our Institution. Long-term studies are in progress to determine whether increased knowledge leads to better self-care, improved control and, ultimately, reduced incidence of diabetic complications.

Insulin, prostaglandin E2 and glucagon release by human tissue incubated in vitro. Influence of indomethacin

In view of reports that prostaglandins influence insulin and glucagon secretion, we have studied PGE2, insulin and glucagon release from fragments (15-20 mg) of human insulinoma tissue incubated in vitro in the absence or presence of indomethacin (100 mumol/liter) an inhibitor of prostaglandin synthesis. Acid-ethanol extraction of this tissue showed the following hormonal contents : insulin : 7.17 U and glucagon 84.4 ng per g of tissue (wet weight). In the absence of indomethacin, the mean release of PGE2, insulin and glucagon into the incubation medium was 3.65 +/- 1.3 pmol, 10.5 +/- 1.2 mU and 708.4 +/- 141.8 pg in two hours (mean of 5 vials containing 2 fragments of 15-20 mg of tissue). PGE2 release was significantly inhibited in the presence of indomethacin (0.89 +/- 0.23 pmol). This effect was associated with a significantly higher insulin (16.8 +/- 1.9 mU/2 hours) and lower glucagon (176 +/- 19.7 pg/2 hours) release. These results support the view that insular tissue possesses a prostaglandin synthesis system which positively modulates glucagon secretion whereas it negatively influences insulin release.

Neurotransmitters and glucagon release from the isolated, perfused canine stomach

The isolated, perfused, canine stomach was used to investigate the effect of three neurotransmitters--norepinephrine, acetylcholine (or its analogue carbamylcholine), and VIP (vasoactive intestinal peptide)--on gastric glucagon release. Norepinephrine at the two concentrations tested (3.10-8 and 7.10-7 M) did not influence gastric glucagon release. In contrast, acetylcholine or carbamylcholine (5.10-6 M) as well as VIP (46-60 ng/ml) unequivocally stimulated gastric glucagon release, an effect apparently independent of the changes in blood flow. These results are in sharp contrast with the previously reported lack of effect of an electric stimulation of the vagus nerves on the release of glucagon from the dog stomach. An absence of innervation of the canine gastric A-cell would probably best explain this situation.

Pancreatic B-cell response to a test-meal in lean and obese diabetic patients: relation to metabolic control

We have measured fasting C-peptide reactivity (CPR) as well as CPR responses to a test meal in 83 diabetic patients and 41 non diabetic controls. In comparison to controls, basal CPR was decreased in lean insulin-treated diabetics with stable or brittle diabetes and in obese patients with brittle diabetes. Lean and obese maturity-onset diabetics had increased CPR levels and so had obese insulin-treated patients. Nevertheless, the CPR response to the test meal was clearly inadequate in all diabetics. In control patients, there was a positive correlation between fasting blood glucose and CPR levels. On the contrary, lean diabetics demonstrated a negative correlation between these parameters. Hemoglobin A1 levels were negatively correlated to fasting CPR levels in lean diabetics, indicating the importance of residual B-cell function for diabetes control. These correlations were obscured in obese diabetics. In our patients, circulating insulin antibodies had apparently no deleterious effect on metabolic control.

Effects of acetylsalicylic acid on blood glucose, plasma FFA, glycerol, 3-hydroxybutyrate, alanine, C-peptide, glucagon and growth hormone responses to arginine in insulin-dependent diabetics

This study aimed at evaluating the effect of acetylsalicylic acid (ASA) on blood glucose, plasma FFA, glycerol, 3-hydroxybutyrate, alanine, C-peptide, glucagon and growth hormone responses to arginine in subjects with insulin-dependent diabetes. For this purpose, seven insulin-requiring diabetics were submitted to a standard arginine tolerance test before and after a three day treatment with ASA (50 mg/kg/daily, plus 1 g before the second test). ASA treatment resulted in no significant changes in either basal or arginine-stimulated blood glucose, but it significantly decreased the basal concentrations of plasma FFA (p less than 0.05), 3-hydroxybutyrate (p less than 0.05) and glycerol (p less than 0.05). In addition, the fall in plasma FFA concentrations during arginine infusion was significantly less after ASA than levels observed without ASA (--262 +/- 100 microEq/l vs --35 +/- 57 microEq/l, p less than 0.02). No significant changes in either basal or arginine-stimulated glucagon concentrations were observed after ASA; by contrast, the growth hormone peak was significantly reduced after ASA (11.3 +/- 4.2 ng/ml vs 5.1 +/- 1.1 ng/ml, p less than 0.05). These metabolic effects exerted by ASA in insulin-dependent diabetes seem not to be related to alterations in endogenously secreted insulin since C-peptide circulating levels were similar during the pre- and post-treatment arginine tests.

Effect of phosphate on the arginine-induced insulin release by the isolated perfused rat pancreas

The isolated perfused rat pancreas was used to investigate the effect of extracellular phosphate on the arginine-induced insulin release. In the absence of any metabolic substrate, the insulin response to arginine was monophasic. In the absence of phosphate in the medium, the insulin release as unaffected until the 15th minute of the stimulation period, but was significantly augmented from that time onward. In the presence of oleic acid in the perfusate, the insulin response to arginine was also monophasic but occurred earlier than in controls. In this conditin, phosphate omission resulted in an increase of the insulin response to arginine from the 3rd minute of the stimulatory period onward. In the presence of glucose 5.5 mM in the medium the insulin response to argnine was biphasic and was not affected by extracellular phosphate omission.

Plasma C-peptide response to arginine in insulin-dependent diabetic subjects

Plasma C-peptide concentrations have been determined in the basal state and in response to intravenous arginine in 10 insulin-dependent diabetics. Five patients had fasting C-peptide levels above 0.08 pmol/ml and responded to the arginine infusion with a rise in C-peptide levels of more than 0.2 pmol/ml (responsive diabetics). The remaining 5 patients had fasting C-peptide below 0.03 pmol/ml and showed no C-peptide response to arginine (nonresponsive diabetics). Fasting blood glucose and the rise in blood glucose in response to arginine were higher in non-responsive than in responsive diabetics. The magnitude of blood glucose rise in response to arginine was inversely correlated with increments in plasma C-peptide. In addition, the fasting levels of FFA and 3-hydroxybutyrate were significantly lower in C-peptide responsive than in nonresponsive patients. These data give further support to the concept that measurements of fasting plasma C-peptide permit to distinguish secretors from nonsecretors, and demonstrate that residual beta-cell function is associated with a lesser degree of aminoacid-induced hyperglycemia.

Effect of oleic acid on arginine-induced glucagon secretion by the isolated perfused rat pancreas

The isolated perfused rat pancreas was used to investigate the effect of oleic acid on glucagon secretion in response to 10 mmol/l arginine. In the absence of oleic acid and at 2.5 mmol/l calcium, arginine induced a biphasic glucagon secretion. At lower extracellular calcium concentration (1.0 mmol/l), the second phase of glucagon release was reduced, the first phase being unchanged. In the presence of 1,500 mumol/l oleic acid, the glucagon response to arginine was also biphasic, but second phase release was markedly inhibited, the first phase glucagon release being unchanged. Such an effect was not obtained when oleic acid concentration in the medium was 750 mumol/l. These results demonstrate that high concentrations of oleic acid inhibit glucagon secretion in response to arginine from the isolated perfused rat pancreas and support the concept that circulating free fatty acid levels are involved in the control of glucagon secretion.

Clofibrate and diabetes control in patients treated with oral hypoglycaemic agents

1. Twenty-two maturity-onset type diabetics treated with oral hypoglycaemic agents entered a single-blind crossover study using placebo (periods A and C, 2 months each) and clofibrate (2 g/day; period B; 2 months). 2. In thirteen patients, under reasonably good control, clofibrate did not reduce fasting or post-prandial blood glucose, nor 24 h glycosuria; no improvement was noted in the M-value, an index of diabetes control. 3. In contrast, in nine patients, with poor diabetes control, clofibrate reduced 24 h glycosuria and significantly improved the M-value. 4. In all patients, clofibrate therapy was associated with a significant 19-23% reduction in plasma fibrinogen. 5. It is suggested that addition of clofibrate may be useful in maturity-onset diabetics not adequately controlled by diet combined with oral hypoglycaemic agents.

Effect of oleic acid on insulin secretion by the isolated perfused rat pancreas

The isolated perfused rat pancreas was utilized to investigate the effect of oleic acid on insulin secretion. In the absence of glucose, a continuous infusion of oleic acid (1500 micromol/l) induced a biphasic insulin release. This effect was reduced at low extracellular calcium concentration. In the presence of oleic acid 1500 micromol/l, the insulin response to 10 mmol/l arginine occurred earlier, the total amount of insulin released in response to the amino acid being unchanged. Such an effect was not obtained when oleic acid in the medium was 750 micromol/l, but it was observed in the presence of oleic acid 1500 micromol/l when the concentration of albumin in the perfusate was increased from 2 g/100 ml to 4 g/100 ml. The insulin response to a continuous infusion of glucose (4.4 mmol/l and 16.7 mmol/l) was potentiated by the presence of oleic acid 1500 micromol/l in the perfusate. No modification of the biphasic pattern of insulin response to glucose 16.7 mmol/l was observed. These results demonstrate that high concentrations of oleic acid stimulate insulin release from the isolated perfused rat pancreas and modulate the insulin response to arginine or glucose.

[Hemoglobin A1c and diabetes control (author's transl)]

Hemoglobin A1c (HbA1c) is a minor component of human hemoglobin resulting from a non enzymatic linkage of glucose with the NH2-terminal amino acid of the beta chain of hemoglobin. Under normal conditions, HbA1c represent about 5% of total hemoglobin. The HbA1c blood concentration increases in direct proportion of the duration and degree of hyperglycemia. Available procedures for measuring HbA1c include column chromatography, high pressure liquid chromatography, a colorimetric procedure based on the formation of 5-hydroxymethylfurfural and isoelectrofocusing. In a group of 138 patients, we have confirmed that HbA1c provides a useful means of evaluating the degree of diabetic control: the highest values have been recorded in cases of poor control, the lowest in cases of excellent control. In the latter case, the HbA1c values recorded were not statistically different from those obtained in a control group of 92 non-diabetic subjects. The interest of evaluating this parameter in diabetes is briefly analyzed.

Pituitary-testicular axis in diabetic men with and without sexual impotence

In nine impotent diabetics (ID), eight diabetics with normal sexual function (ND), eight impotent non diabetics (IN) and seven normal patients (NN), we found normal basal FSH and LH levels. Plasma testosterone (T) values were significantly lower in ID and IN in comparison with NN and ND. After LHRH injection, plasma T increased in ID up to normal values. In non diabetic patients (NN and IN), LHRH induced a slight but significant rise in blood glucose. Post LHRH gonadotropins response was in the normal range for all groups and the amplitude of the response was highly correlated with basal levels except for LH in ND. Mean blood glucose during the test and LH response to LHRH were not inversely correlated in the patients studied. We conclude that impotence in diabetic patients is a neurological complication and that low testosterone levels are probably secondary to a decreased coital frequency. ID and IN probably share a common pattern of inadequate hypothalamic feedback control in the presence of low testosterone levels.

Possible role of endogenous prostaglandins in glucagon secretion by isolated guinea-pig islets

Previous studies have demonstrated that prostaglandins stimulate glucagon secretion in vitro and in vivo. The present work was aimed at investigating the influence of two inhibitors of prostaglandin synthesis, isopropyl-2 nicotinoyl-3 indole (L8027) and indomethacin, on basal and arginine- or noradrenaline-stimulated glucagon release from isolated guinea-pig islets incubated in the absence of glucose. L8027 (10(-4) and 10(-5) mol/l) did not alter basal glucagon release, blocked almost completely the glucagon response to arginine (10(-2) mol/l), had no effect on the glucagon release induced by noradrenaline (10(-4) mol/l), but reduced the stimulatory effect of a lower concentration of noradrenaline (5.10(-7) mol/l). The kinetic study of this inhibitory effect demonstrated that (1) it necessitates preincubation of the islets with L8027 for 30 minutes before the addition of arginine, (2) after a short preincubation period (30 minutes) in the presence of L8027, removal of the inhibitor at the time of arginine stimulation resulted in enhanced glucagon response, (3) on the contrary, after a prolonged incubation period (75 min) with arginine and L8027, the inhibitory effect remained transiently detectable after removal of L8027. Indomethacin similarly blocked arginine- and noradrenaline-induced glucagon secretion. These results suggest that an intra-insular synthesis of prostaglandins is involved in the A cell response to arginine and noradrenaline.

Glucose and insulin in the regulation of glucagon release from the isolated perfused dog stomach

The respective roles of glucose and insulin in the regulation of glucagon release from the canine stomach were investigated using an isolated blood-perfused preparation. At normal blood glucose and plasma insulin levels, the stomach released small amounts of glucagon. Such basal gastric glucagon release was not modified by hyperglycemia. In contrast, gastric glucagon release was increased by hypoglycemia or 2-deoxy-D-glucose-induced cytoglycopenia. Antibody neutralization of basal circulating concentrations of insulin (10 +/- 1 microU/ml) doubled the stimulation induced by hypoglycemia alone. It is concluded that: 1) suppression of gastric glucagon release is observed with very low concentrations of insulin; 2) basal gastric glucagon release is not further suppressed by hyperglycemia; and 3) that hypoglycemia and cytoglycopenia stimulate gastric glucagon secretion.

Influence of elevated plasma free fatty acids on the glucagon response to hypoglycemia in normal and in pregnant women

We investigated the influence of an insulin-induced hypoglycemia on plasma glucagon in nonpregnant healthy young women and in women during the last month of gestation. Both groups were tested either in the basal state or during a period where free fatty acid plasma levels were increased by infusion of a lipid emulsion supplemented with heparin. Regular insulin was injected intravenously at the dose of 0.1 U/kg body wt in controls and 0.3 U/kg in pregnant women in order to obtain a similar lowering of blood glucose in all groups. In controls, the increase in plasma glucagon was maximum 30 and 45 min after insulin injection and averaged 130 pg/ml; the infusion of triglycerides and heparin which raised plasma FFA to about 1300 muEq/liter decreased basal plasma glucagon levels and reduced, by about 70%, the glucagon response to hypoglycemia. During the last month of pregnancy, the glucagon response to insulin-induced hypoglycemia was reduced by 60% (mean maximal increase 52 pg/ml); furthermore, raising plasma FFA to about 1500 muEq/liter completely abolished the glucagon rise induced by the insulin hypoglycemia. These results support the view that the glucagon release from A-cells can be modulated by the level of circulating plasma FFA.

Effect of glucose on plasma glucagon and free fatty acids during prolonged exercise

The effects of glucose ingestion on the changes in blood glucose, FFA, insulin and glucagon levels induced by a prolonged exercise at about 50% of maximal oxygen uptake were investigated. Healthy volunteers were submitted to the following procedures: 1. a control test at rest consisting of the ingestion of 100 g glucose, 2. an exercise test without, or 3. with ingestion of 100 g of glucose. Exercise without glucose induced a progressive decrease in blood glucose and plasma insulin; plasma glucagon rose significantly from the 60th min onward (+45 pg/ml), the maximal increase being recorded during the 4th h of exercise (+135 pg/ml); plasma FFA rose significantly from the 60th min onward and reached their maximal values during the 4th h of exercise (2177 +/- 144 muEq/l, m +/- SE). Exercise with glucose ingestion blunted almost completely the normal insulin response to glucose. Under these conditions, exercise did not increase plasma glucagon before the 210th min; similarly, the exercise-induced increase in plasma FFA was markedly delayed and reduced by about 60%. It is suggested that glucose availability reduces exercise-induced glucagon secretion and, possibly consequently, FFA mobilization.