The "dawn phenomenon"--a common occurrence in both non-insulin-dependent and insulin-dependent diabetes mellitus

The dawn phenomenon is a condition recently described in patients with insulin-dependent diabetes mellitus (IDDM) that is characterized by abrupt increases in fasting levels of plasma glucose or insulin requirements or both between 5 and 9 a.m., in the absence of antecedent hypoglycemia. To determine its potential clinical relevance, we assessed its frequency and reproducibility in 20 patients with IDDM and in 13 patients with non-insulin-dependent diabetes mellitus (NIDDM) during overnight closed-loop (feedback-controlled) intravenous insulin infusion. After 6 a.m., plasma glucose levels increased similarly in NIDDM (89 +/- 2 mg per deciliter, midnight to 6 a.m., vs. 98 +/- 2 mg per deciliter, 6 to 9 a.m.; P less than 0.01). Insulin requirements increased at least 50 per cent for 1 1/2 hours in 77 per cent of patients with NIDDM and in 75 per cent of patients with IDDM. In five patients with IDDM who were studied on four occasions, the phenomenon occurred during 17 of the 20 observation periods, with insulin requirements after 6 a.m. increasing 225 +/- 34 per cent; coefficients of variation in individual patients ranged from 4 to 25 per cent. Thus, the dawn phenomenon occurs commonly in both NIDDM and IDDM, but its potential variability must be taken into consideration when one is attempting to adjust insulin doses.

The adrenergic contribution to glucose counterregulation in type I diabetes mellitus. Dependency on A-cell function and mediation through beta 2-adrenergic receptors

In order to assess the adrenergic contribution to hypoglycemic glucose counterregulation in type I diabetes mellitus and to determine whether the adrenergic contribution is mediated through beta 1- or beta 2-adrenergic receptors, hypoglycemia was induced by an i.v. insulin infusion (30 mU/m2 x min) for 60 min in 11 insulin-dependent diabetic patients (IDDM), 5 with normal plasma glucagon responses and 6 with blunted responses, and also in 7 age-weight-matched nondiabetic subjects. Rates of plasma glucose decrease and postnadir increase, as well as plasma concentrations of free insulin and of counterregulatory hormones, were measured when insulin was infused alone, and when insulin was infused along with propranolol (a beta 1- and beta 2-adrenergic receptor antagonist) or metoprolol (a selective beta 1-antagonist). Postnadir plasma glucose recovery was decreased in IDDM with blunted plasma glucagon responses (21 +/- 0.8 mumol x L-1 x min-1, P less than 0.001), but was normal in patients with normal plasma glucagon responses (30 +/- 0.4 versus 33 +/- 0.5 mumol x L-1 x min-1 in nondiabetic subjects, P = NS). Postnadir plasma glucose recovery was not affected by either propranolol or metoprolol in normal subjects and in IDDM with normal glucagon responses. However, in IDDM with blunted plasma glucagon responses, postnadir plasma glucose recovery was further decreased by propranolol (14 +/- 0.6 mumol x L-1 x min-1, P less than 0.01), but was unaffected by metoprolol (22 +/- 0.9 mumol x L-1 x min-1, P = NS).(ABSTRACT TRUNCATED AT 250 WORDS)

Abnormal glucose counterregulation in insulin-dependent diabetes mellitus. Interaction of anti-insulin antibodies and impaired glucagon and epinephrine secretion

To evaluate the roles of counterregulatory hormones and insulin antibodies in the impairment of plasma glucose recovery from hypoglycemia in diabetes mellitus, and to assess the relationship between the glucagon response and duration of the disease, 21 insulin-dependent diabetic patients and 10 nondiabetic subjects were studied. The diabetics consisted of 5 patients with recent onset of diabetes (less than 1 mo); 11 with 2.6 +/- 0.3 (mean +/- SEM) yr duration of diabetes, 5 of whom had insulin antibodies; and 5 patients with long-term diabetes (21 +/- 3 yr), insulin antibodies, and autonomic neuropathy. During insulin-induced hypoglycemia (28 mU/m2 X min for 60 min) in patients with recent-onset diabetes, plasma free insulin, glucose, and counterregulatory hormone concentrations did not differ from those of nondiabetic subjects. In patients with insulin antibodies, the disappearance of insulin after insulin infusion was delayed, and both restitution of normoglycemia and plasma glucagon response were blunted compared with patients without antibodies. When glucagon was infused (80-130 ng/m2 X min) during hypoglycemia in diabetics with impaired glucagon responses in order to simulate normal glucagon responses, plasma glucose recovery was normalized in patients without antibodies but not in those with antibodies. In patients with long-standing diabetes, restitution of normoglycemia was further impaired and this was associated with an absent plasma glucagon response and a diminished plasma epinephrine response. Plasma glucagon responses to hypoglycemia were inversely correlated to the duration of diabetes (r = -0.943; P less than 0.0005). It is concluded that impaired A-cell secretion is the predominant mechanism for the delayed glucose recovery after hypoglycemia in diabetic patients without insulin antibodies and normal epinephrine responses. Slowed disappearance of insulin due to the presence of insulin antibodies further delays the restoration of normoglycemia. Patients with long-standing diabetes and autonomic neuropathy exhibit decreased epinephrine secretion, which leads to an additional retardation of glucose recovery. Since plasma glucagon and epinephrine responses to hypoglycemia were normal at the onset of diabetes but diminished in long-term diabetes, it appears that the impaired glucagon and epinephrine responses to hypoglycemia are acquired defects that develop subsequent to B-cell failure.

Important role of adrenergic mechanisms in acute glucose counterregulation following insulin-induced hypoglycemia in type I diabetes. Evidence for an effect mediated by beta-adrenoreceptors

During hypoglycemia induced by an i.v. insulin infusion for 60 min, rates of plasma glucose (PG) decrease and recovery, PG nadir, and plasma counter-regulatory hormone and free fatty acid responses were studied in eight type I uncomplicated diabetic subjects and eight nondiabetic subjects. Each subject was tested three times at two different rates of insulin infusion (25 and 32 mU/m2/min): (1) during infusion of saline, (2) during infusion of phentolamine + propranolol (combined alpha, beta-blockade), and (3) during infusion of propranolol alone (isolated beta-blockade) for 150 min. At the time of the studies, the diabetic subjects had been made euglycemic by an overnight i.v. insulin infusion. During infusion of insulin (25 mU/m2/min) and saline, the rates of PG decrease and recovery were slower (P less than 0.01) and PG nadir was delayed in the diabetic subjects. Moreover, their plasma glucagon response was blunted while plasma epinephrine, norepinephrine, growth hormone, and cortisol responses were similar in both groups. Infusion of insulin at 32 mU/m2/min caused larger decreases in PG than had been observed when insulin was infused at 25 mU/m2/min. Plasma glucagon responses increased in the nondiabetic subjects (P less than 0.05) but not in the diabetic subjects. However, in the diabetic subjects, plasma epinephrine increased more than in the nondiabetic subjects (P less than 0.05). There was an inverse correlation between the individual plasma epinephrine responses and the plasma glucagon responses in the diabetic subjects (r = -0.72) but not in the nondiabetic subjects. Alpha, beta-adrenergic blockade decreased the plasma glucose nadir and impaired the rate at which normoglycemia was restored in the diabetic subjects (P less than 0.005 vs. saline) but not in the nondiabetic subjects. Plasma catecholamine and growth hormone responses were increased and plasma FFA recovery was suppressed in both groups (P less than 0.05 vs. saline), while the cortisol responses were unaltered. During isolated beta-adrenergic blockade, changes in plasma glucose, counterregulatory hormones and FFA were essentially identical to those observed during combined alpha, beta-adrenergic blockade in both groups except that the augmented plasma norepinephrine responses were no longer apparent.

CONCLUSIONS

although epinephrine is not essential for prompt restoration of normoglycemia in normal man following insulin-induced hypoglycemia, it plays a major role in glucose counterregulation in diabetics who have an impaired glucagon secretion in response to hypoglycemia. These counterregulatory effects of epinephrine are mediated by beta-adrenoreceptors.

Erythrocyte concentration of glycolytic phosphorylated intermediates and adenosine nucleotides in subjects with diabetes mellitus

We determined erythrocyte concentration of phosphorylated glycolytic intermediates and adenosine nucleotides in 9 non-ketotic diabetics under different conditions of glycemic control. We showed a significant increase of G6P, F6P and FDP levels in the 9 diabetics with poor glycemic control, whereas other phosphorylated glycolytic intermediates were in the normal range. After achieving fair glycemic control G6P, F6P and FDP fell to normal value. G6P, F6P and FDP were well correlated with simultaneously determined plasma glucose. Another 5 diabetics with poor glycemic control were connected to Biostator in order to determine the time relationship between blood glucose and erythrocyte levels of G6P, F6P and FDP. We demonstrated a prompt decrease to the normal values of the phosphorylated glycolytic intermediates where blood glucose fell to normal 60-90 minutes after automatic insulin delivery. Finally, we determined RBC glycolytic intermediates in 5 subjects with normal OGTT and in 5 subjects with normal fasting plasma glucose but abnormal OGTT. In this group we observed a prompt and persistent increase of G6P, F6P and FDP levels over the three hours hyperglycemia phase which has been induced by glucose load. No significant increase was found in subjects with normal OGTT. Our results demonstrate that erythrocyte glycolytic metabolism in abnormal in diabetics, depending strictly on the degree of glycemic control.

Circulating catecholamine and glucagon responses to insulin-induced blood glucose decrement in a patient with pheochromocytoma

In a gastrectomized woman with an adrenal pheochromocytoma we observed hypertensive crisis in association with postprandial hypoglycemic episodes. To assess whether hypoglycemia could be responsible for the hypertensive crises, we measured circulating catecholamines and glucagon during an insulin-induced blood glucose decrement carried out by an artificial endocrine pancreas. When the blood glucose level reached 36 mg/dl, a severe hypertensive crisis occurred. At this time, circulating catecholamines increased 2-fold (norepinephrine, from 2200 to 3568 pg/ml; epinephrine, from 950 to 1750 pg/ml), while no changes in glucagon were observed. Our observation suggests that in patients with pheochromocytoma, hypoglycemia may trigger a marked release of catecholamines independent of glucagon secretion. This response probably is mediated by activation of the sympathetic nervous system. Our results also suggest that the pancreatic A-cell response to blood glucose decrement is totally suppressed in patients with pheochromocytoma by the chronically high levels of circulating catecholamines. Thus, hypoglycemia may be added to the list of other well known factors which may provoke hypertensive emergencies in patients with pheochromocytoma.

Lack of glucagon response in glucose counter-regulation in type 1 (insulin-dependent) diabetics: absence of recovery after prolonged optimal insulin therapy

Mild hypoglycaemia was induced using an artificial pancreas in five normal subjects (from 5.00 +/- 0.15 to 2.83 +/- 0.15 mmol/l) by infusing 28 mU/m2 per min soluble insulin for 60 min. Six Type 1 (insulin-dependent) diabetic patients were stabilized for 14h using an artificial pancreas. They were then rendered hypoglycaemic (from 4.94 +/- 0.09 to 2.89 +/- 0.11 mmol/l) by infusing 28 mU/m2 per min plus 16 +/- 3.8 mU/min insulin for 60 min. Before the study, the diabetic patients were in optimal blood glucose control (mean blood glucose 6.72 +/- 0.11 mmol/l over the previous 14-20 days; HbA1 8.3 +/- 0.1%). During the insulin infusion test, blood glucose decrement was slower in the diabetic patients than in the control subjects. The blood glucose nadir was delayed in the diabetics until 75 min compared with 55 min in the control subjects. Blood glucose recovery rate in the diabetic subjects was severely impaired. In Type 1 diabetes, the counter-regulatory hormonal response to insulin induced hypoglycaemia is similar to that of non-diabetics, except for that of glucagon, the blunted response of which is not reversed by prolonged optimisation of blood glucose control. This impaired response of the A cell does not seem to be a consequence of insulin deficiency.

Analysis of short-term changes in reversibly and irreversibly glycosylated haemoglobin AI: relevance to diabetes mellitus

We have determined the stable (irreversibly glycosylated) fraction of haemoglobin AI (HbAI) on Bio-Rex 70 after incubation of red blood cells in 0.9 % saline solution for 6 h at 37 degrees C. The total (reversibly + irreversibly glycosylated fractions) HbAI was determined before each incubation. Labile (reversibly glycosylated) HbAI represented the difference between total and stable HbAI fractions. Total and stable HbAI fractions were determined during insulin- or meal-induced blood glucose fluctuations in 24 insulin-dependent diabetics and in seven subjects with impaired glucose tolerance. In the diabetics, the maximal fluctuation of total HbAI was 1.47% over 2-12 h, while the simultaneous plasma glucose variation was 21.5 mmol/l. The stable HbAI fraction did not change significantly. In diabetics the differences between the maximal and minimal values of plasma glucose and total HbAI were significantly correlated. Plasma glucose correlated with simultaneously determined total and labile HbAI fractions, but not with stable HbAI. In subjects with impaired glucose tolerance, similar changes in total but not in stable HbAI were observed during an oral glucose tolerance test. We conclude that, although rapid changes in chromatographically determined HbAI are relatively small, the determination of stable HbAI should be performed to circumvent this problem and to ensure a more accurate index of blood glucose control.

The importance of determining irreversibly glycosylated hemoglobin in diabetics

The concentration of glycohemoglobins (HbA1(a+b+c), HbA1,) was measured before and after incubation of normal and diabetic erythrocytes for 6 h at 37°C in saline. This procedure removes as much as 80-90% of the labile glucose-HbA0 adduct (labile HbA1,), thus allowing accurate estimation of irreversibly glycosylated hemoglobin (stable HbA1,). The concentration of HbA1 measured before such an incubation is total HbA1, (stable + labile). We determined the concentration of total, stable, and labile HbA1, in the same blood samples used to measure fasting plasma glucose (FPG) every day, for 4 consecutive days, in two groups of hospitalized insulin-treated diabetics. Group A subjects (N = 7) were type I, C-peptide negative, unstable diabetics, while group B subjects (N = 15) were type II, C-peptide positive, stable diabetics. Individual day-to-day variations of total HbA1, were wide in group A (Δ = 1.58 ± 0.14%), and slight in group B (Δ = 0.12 ± 0.01%; P < 0.001), paralleling similar plasma glucose fluctuations. Day-to-day variations of stable HbA1, were virtually absent not only in group B subjects with stable glycemic values (Δ = 0.08 ± 0.01%), but also in those of group A with marked glycemic instability (Δ = 0.07 ± 0.01%; P = NS). Day-to-day variations of labile HbA1, were marked in group A (Δ = 1.31 ± 0.14%), but negligible in group B (Δ = 0.15 ± 0.03%; P < 0.001). On admission, FPG correlated with labile HbA1, in group A (r = 0.89) and B (r = 0.71). FPG correlated with stable HbA1, in group B (r = 0.73) but not in group A subjects and with total HbA1, more closely in group B (r = 0.73) than in A (r = 0.61). A very close correlation was found between the concentration of total and labile HbA1 in subjects of group B (r = 0.82). In group B, fasting, post-breakfast, and mean daily plasma glucose values, determined every 3-6 days during the 2 mo before admission, significantly correlated both with total and stable HbA1 determined on admission, while in group A they did not. In group A, the correlation was significant when stable instead of total HbA1 was considered. We conclude that the significant fluctuations of total HbA1 reduce its value as an index of long-term control in unstable diabetics. On the other hand, a single determination of stable HbA1, totally independent of simultaneous blood glucose values, closely reflects blood glucose control over the previous 2 mo. We propose routine estimation of stable HbA1, which is simple and straightforward, to carry out follow-up studies of unstable diabetics.

[Early diagnosis of pheochromocytoma in pregnancy]

The authors report a case of pheochromocytoma revealed by a de novo appeared hypertension in a young female patient at her 3rd trimester of pregnancy. An early diagnosis and an appropriate medical management allowed a successful maternal outcome. Since maternal and fetal prognosis of pheochromocytoma associated with pregnancy is strictly related to an antepartum diagnosis, the importance of suspecting pheochromocytoma as underlying cause of de novo appearing hypertension (or hypertension of unknown origin) during pregnancy is emphasized. Biochemical tests (plasma and/or urinary catecholamine or metabolite determination) are recommended as the most suitable approach to the screening of pheochromocytoma in pregnancy.

Counterregulatory hormones during moderate, insulin-induced, blood glucose decrements in man

To verify whether a significant increase in levels of counterregulatory hormones occurs in the course of mild blood glucose decrements, we infused regular insulin iv over 65 min in two groups of healthy volunteers (group A, n= 7; group B, n = 6) at a constant rate (group A, 0.05 U/kg; group B, 0.025 U/kg). All subjects were connected to an artificial endocrine pancreas (Biostator) for continuous blood glucose (BG) monitoring. Plasma insulin, glucagon, and GH were determined by specific RIAs. Plasma norepinephrine, epinephrine, and cortisol were measured by sensitive fluorimetric methods. A moderate fall in BG occurred from 91 +/- 1.5 mg/dl (mean +/- SEM) to a nadir of 56 +/- 4.5 mg/ml at 45 min in group A and from 81 +/- 2.5 to a nadir of 64 +/- 4.9 mg/dl at 45 min in group B. In both groups A and B, the increases in plasma glucagon and catecholamine levels, which remained strictly in the physiological range, appeared concomitant and were significant at 45 min (P less than or equal to 0.05 vs. basal), while the increases in plasma cortisol and GH concentrations were clearly delayed. The increments for all counterregulatory hormones (expressed as the area to minutes ratio) except GH, were significantly greater in group A than in group B ( P less than or equal to 0.01). There was a significant correlation between these increases, including that of GH and the BG decrease, calculated in all subjects investigated. These results suggest that the mechanisms involved for the release of counterregulatory hormones such as glucagon, catecholamines, cortisol, and GH are very sensitive to a moderate decrease in BG concentration and that there is a close relationship between this hormonal response and the degree of the BG decrements obtained.

HbA1 in subjects with abnormal glucose tolerance but normal fasting plasma glucose

HbA1(a+b+c)(HbA1) was determined chromatographically in 107 subjects with normal fasting plasma glucose (FPG) and 112 patients with overt diabetes. Subjects with normal FPG were divided into two groups based on their response to two oral glucose tolerance tests (OGTTs), at an interval of 2 mo. In 40 subjects with normal OGTT (group I), HbA1 ranged from 5.2% to 7.2%, while in 67 subjects with abnormal OGTT (group II), it ranged from 6.3% to 9.6%. HbA1 levels were significantly higher in group II than in group I (7.7 +/- 0.09% versus 6.4 +/- 0.08%, mean +/- SEM, P less than 0.0005), but 14 subjects of group II had HbA1 levels less than 7.2%. No correlation was found between HbA1 and FPG, OGTT peak, and curve area in either group. However, the correlation became significant in all 107 subjects with normal FPG (groups I + II). In patients with overt diabetes, HbA1 ranged from 6.3% to 18% (11.9 +/- 0.22%) and correlated with FPG (r = 0.78, P less than 0.0005). The traditional OGTT seems more sensitive than the HbA1 measurement in detecting subjects with reduced carbohydrate tolerance. HbA1 level, on the other hand, is known to be more specific indicator of structural abnormalities following long-term hyperglycemia. Thus HbA1 determination might be a helpful test along with OGTT to improve both selection and follow-up subjects with true borderline diabetes.

Modification of glycosylated haemoglobin concentration during artificial endocrine pancreas treatment of diabetics. Evidence for a short-term effect on HbA 1 (a+b+c) levels

In order to verify whether or not insulin-induced blood glucose control can acutely lower glycosylated haemoglobin levels, HbAI (a+b+c) (HbAI) was measured in 11 diabetics before, during and after 3 days of treatment with an "artificial endocrine pancreas" (Biostator). Initially 5 patients were in fair glycaemic control (group A), while the other 6 showed poor control (group B). HbAI levels decreased significantly after 3 days in both groups A (from 9.6 +/- 0.2% to 8.5 +/- 0.3%, mean +/- SEM, p less than 0.05) and B (from 13.7 +/- 0.2% to 12.6 +/- 0.3%, p less than 0.05). A further HbAI decrease was observed until day 60 following Biostator treatment, during which period glycaemic control improved, as assessed by fasting and post-lunch plasma glucose values and daily glycosuria determined every 10 days. These results suggest that increased HbAI levels may be reversed early by strict blood glucose control during a 3 day period. It is concluded that HbAI levels not only reflect long-term glycaemic control, but also recent acute variations in mean blood glucose values.

[Adrenergic activity and glycometabolic compensation in patients with diabetes mellitus]

In an assessment of the degree of adrenergic activity in the course of diabetes mellitus, plasma levels and urinary excretion of norepinephrine and epinephrine were determined in 20 normal subjects and 47 diabetics: 11 in good control (group I), 23 in poor control (group II), 13 with frank ketoacidosis (group III). The study was repeated in groups II and III once good glycometabolic control had been achieved. Slightly above normal catecholamine levels were noted in group I, while there was a marked increase in group II. Group III shaved an enormous increase by comparison with the other two groups. After medical treatment values in group III fell to within the group I range. The conclusion is drawn that a close relationship exists between adrenergic acitivity and the degree of control of diabetes. The sympathetic nervous system, therefore, interferes in the course of diabetes with blood sugar control via numerous, complex mechanisms.

Red cell 2,3-DPG levels in diabetic vasculopathy

RBC 2,3-DPG resulted statistically increased versus normal subjects in 22 insulin-dependent diabetics affected by arterial vasculopathy and in 19 vasculopathic non-diabetic patients. RBC 2,3-DPG, on the other hand, did not differ from the normal mean in 16 insulin-dependent diabetics without vascular involvement. The increase of RBC 2,3-DPG represents principally a consequence of the vasculopathy.

Effect of metabolic control on urinary excretion and plasma levels of catecholamines in diabetics

Urinary excretion and plasma levels of catecholamines were determined in 20 normal and 39 diabetic subjects to evaluate the sympathetic activity. Diabetic patients were divided into 4 groups according to the metabolic control. Sympathetic activity showed no differences between normal and subjects with chemical diabetes (group I, n = 5). In insulin-treated diabetics in good metabolic control (group II, n = 11) only urinary excretion of free norepinephrine was significantly higher than normals (p less than .05). In insulin-treated diabetics in poor metabolic control (group III, n = 16) urinary excretion and plasma levels of norepinephrine showed a marked increase over groups I and II (p less than .001). In insulin-treated diabetics with ketosis (group IV, n = 7) urinary excretion and plasma levels of both norepinephrine and epinephrine showed the highest values (p less than .001 and less than .1). Finally, in groups III and IV, after achieving improved metabolic control, a significant decrease of urinary excretion and plasma levels of catecholamines was observed. The results confirm that there is an increased rate of catecholamine release in poorly controlled diabeties and suggest a close correlation between sympathetic activity and metabolic derangement in diabetes.

Urinary excretion and plasma levels of norepinephrine and epinephrine during diabetic ketoacidosis

Sympathetic activity was determined in 13 ketoacidotic diabetics by evaluation of plasma and urinary catecholamines, before and in the course of medical management. Patients were divided into two groups. Group A (severe ketoacidosis, n = 5) and Group B (moderate ketoacidosis, = 8), depending on plasma glucose, pH and plasma bicarbonate levels. The results showed an enhanced sympathetic activity in all patients before treatment, with significant decrease during therapy. In Group A plasma catecholamines were higher than in Group B, both before and in the course of therapy. A significant correlation was found between basal plasma catecholamines and initial plasma glucose, plasma bicarbonate, hours of therapy and insulin dosage required to obtain plasma glucose levels below 150 mg/100 ml .These results, suggesting a close correlation between glycometabolic control and adrenergic activity, emphasize the role of the sympathetic nervous system as a powerful contrainsular factor in the pathogenesis and metabolic derangement of diabetic ketoacidosis.

[Behavior of plasma androgens in cases of adrenal pathology]

Following chromatographic separation, the RI method has been used to examine the behaviour of plasmatic androgens in a group of normal subjects, in two patients with Addison's disease and two patients with hypercorticoadrenalism, due respectively to corticoadrenal adenoma and bilateral adenomatous hyperplasia. A significant reduction in adrenal androgens was observed in patients suffering from Addison's disease. The most interesting information, however, was the paradoxical reduction in DHEA and delta 4-A in patients with corticoadrenal hyperfunction.