Constant Insulin Infusion Research Articles

Impaired Counter Regulation of Hypoglycemia in a Group of Insulin‐dependent Diabetics with Recurrent Episodes of Severe Hypoglycemia

The counterregulatory response to insulin-induced hypoglycemia was investigated in 22 insulin-dependent diabetics (IDD) with recurrent hypoglycemia and in 6 healthy volunteers. Hypoglycemia was induced by a constant rate infusion of insulin (2.4 U/h) up to four hours. Conventional insulin therapy was changed to an i.v. infusion of regular insulin 24 hours prior to the experiment. The presence of diabetic autonomic neuropathy was evaluated by respiratory sinus arrhythmia and Valsalva maneuver. In healthy subjects, blood glucose was decreased to 2.5 mmol, here reaching steady state level and giving rise to marked glucagon and growth hormone (GH) responses. The majority of IDD (group A) reached a slightly lower steady state glucose level and exhibited similar glucagon and GH responses while the epinephrine response was augmented. Six IDD (group B) showed a continuous decrease in blood glucose to 1.2 +/- 0.1 mmol/l at which level the infusion of insulin was discontinued due to neuroglucopenic symptoms. These subjects had no glucagon and epinephrine responses while their GH and cortisol responses were normal. A comparison of the diabetic groups revealed a longer duration of diabetes and a more impaired autonomic nervous function in group B while glycosylated hemoglobin was similar. It is concluded that most IDD have normal hormonal responses (epinephrine, glucagon, GH, cortisol) and normal counterregulartory capacity to hypoglycemia induced by a prolonged infusion of a moderate dose of insulin. Some patients with long-term diabetes and impaired capacity to counteract hypoglycemia exhibit deficient glucagon and epinephrine responses to hypoglycemia.

Failure to find amyloidosis in dogs treated with long-term intravenous insulin delivered by a totally implantable pump.

We examined tissues of seven non-diabetic mongrel dogs and four diabetic beagle dogs treated with constant insulin infusion via totally implantable pumps for from 210 to 880 days. Kidney and skeletal muscle tissue from all dogs were stained with Congo Red and thioflavin-T and appropriately examined. Kidney tissues from the beagle dogs were examined by electron microscopy. No amyloid deposits were found in any of these tissues. Thus, we cannot confirm an earlier report of amyloid occurring in dogs given long-term intravenous insulin. It is concluded that amyloidosis is not a necessary complication of long-term intravenous insulin infusion in dogs.

Examination of the role of the pituitary-adrenocortical axis, counterregulatory hormones, and insulin clearance in variable nocturnal insulin requirements in insulin-dependent diabetes.

In insulin-dependent diabetics, insulin requirements increase significantly after 0600 h, resulting in prebreakfast hyperglycemia with either conventional insulin therapy or constant insulin infusions with insulin infusion devices. In order to clarify the role of the pituitary-adrenocortical axis and further examine the mechanisms of the phenomenon of nocturnal variability in insulin requirements, we studied five IDDs using a closed-loop insulin infusion device (Biostator, GCIIS). The subjects were given saline (SAL) or dexamethasone (DEX) i.v. from 1800 to 0900 h on successive nights. From 2400-0300 to 0600-0900 h, mean insulin infusion rates required to maintain blood glucose values between 109 and 120 mg/dl increased by 0.21 +/- 0.05 mU/kg/min during the SAL infusion, and 0.16 +/- 0.04 mU/kg/min during the DEX infusion, when plasma cortisols were suppressed to less than or equal to 2 micrograms/dl. Mean free insulin concentrations did not increase and remained constant throughout both study nights in spite of the significantly higher 0600-0900-h insulin infusion rates. Growth hormone, glucagon, epinephrine, and norepinephrine concentrations showed normal nocturnal and early morning patterns during both study nights. We conclude that the nocturnal variability in insulin requirements persists despite suppression of the pituitary-adrenocortical axis, and that increased free insulin clearance or degradation may contribute to the "dawn phenomenon" of rising prebreakfast glucose despite constant insulin infusion.

Pharmacokinetics of continuous subcutaneous insulin infusion

One of the reasons for the variability of blood glucose regulation in Type 1 (insulin-dependent) diabetic patients is the huge variation in subcutaneous absorption of intermediate-acting insulin. We have investigated the variation in insulin absorption during continuous subcutaneous insulin infusion in eight such patients. The content of insulin in the subcutaneous tissue was measured using 125I-labelled insulin. The concentration of free serum insulin and blood glucose was followed from 1 h before and from 7 h after breakfast on two consecutive days. The amount of insulin absorbed during 24 h differed in all cases by less than 3% from the daily insulin dose given by the pumps. Mean insulin absorption rates and mean free insulin concentration showed peak values 30-90 min after meal bolus injections; this was sufficient to maintain near-normal blood glucose. Mean free serum insulin correlated strongly with disappearance of insulin from the subcutaneous tissue (r = 0.98). From the insulin absorption rates and free insulin concentrations during basal constant insulin infusion, the half-time of serum insulin was calculated as 6 min. Compared with the known large variability in the absorption of intermediate-acting insulin, continuous subcutaneous insulin infusion offers a precise and reproducible way of insulin administration resulting in post-prandial serum insulin peaks sufficient to maintain near-normal blood glucose levels. The half-time of serum insulin during subcutaneous infusion corresponds to values for intravenous infusion given in the literature, indicating that local degradation of insulin in subcutaneous tissue is of minor importance.

Effect of prolonged insulin treatment on blunted plasma catecholamine and glucagon increase during insulin hypoglycemia in streptozotocin diabetic rats

The role of prolonged insulin treatment of diabetic rats on the lack of glucagon and catecholamine increases in response to insulin-hypoglycemia was investigated. Streptozotocin diabetic rats were maintained in a normoglycemic state for a long period by intraperitoneal constant infusion of insulin with the Alzet osmotic minipumps. Hyperglucagonemia observed in diabetic rats was normalized with insulin treatment. However, the blunted response of glucagon during insulin-hypoglycemia was not altered with insulin treatment. On the other hand, in diabetic rats the diminished catecholamine response to insulin-induced hypoglycemia was rendered normal after insulin treatment. The data indicate that the lack of glucagon and catecholamine responses in diabetic rats in response to insulin-hypoglycemia can vary independently. The persistent decreased glucagon response in normoglycemic diabetics could be the cause of their impaired recovery from hypoglycemia.

Absence of exercise-induced hypoglycaemia in type i (insulin-dependent) diabetic patients during maintenance of normoglycaemia by short-term, open-loop insulin infusion.

To assess the risk and possible mechanisms of hypoglycaemia during moderate exercise in Type I (insulin-dependent) diabetic patients receiving constant insulin infusion, five insulin-dependent male diabetic patients were exercised 18 h after their last meal and 30 h after their last injection of intermediate acting insulin. Intravenous insulin was initially delivered via a closed-loop infusion system programmed to lower mean blood glucose from 11.3 +/- 1.8 to 4.8 +/- 0.4 mmol/l over approximately 3.5 h. Blood glucose was then maintained at this level for 4 h. At this time, the closed-loop infusion was discontinued and replaced by an open-loop system. The average amount of insulin infused per min during the 4 h normoglycaemic closed-loop period was calculated and this amount was infused at a constant rate during both a 30 min period of exercise on a bicycle ergometer (approximately 65% maximum oxygen uptake) and a 30 min rest period which followed. Five nondiabetic males served as control subjects. Despite significantly higher free insulin concentrations (p less than 0.05) and identical pre-exercise blood glucose concentrations, blood glucose rose during exercise only in the diabetic group (0.5 +/- 0.2 mmol/l; p less than 0.01). Changes in the serum concentrations of lactate, glycerol, glucagon, cortisol, non-esterified fatty acids and growth hormone were similar in the two groups and did not account for the increment of blood glucose in the diabetic patients. Beta-hydroxybutyrate concentrations were, however, higher in the diabetic patients at the onset of exercise (p less than 0.01) and decreased significantly more than the control subjects during exercise. We conclude that exercise under these conditions in diabetic patients is not attended by hypoglycaemia.

A regular meal and insulin infusion regimen: its use in the treatment of acute-onset ketotic diabetes and in stabilization of poorly controlled established diabetic subjects.

A simple regimen, consisting of a constant intravenous insulin infusion at either a basal, nocturnal rate or at a daytime rate matched by seven small, isocaloric meals taken every 2 h, has been applied to two clinical situations requiring optimal blood glucose control. In eight poorly controlled established diabetic subjects, quantitative estimates of daily insulin requirements were possible, with consequent improved control upon reinstitution of twice-daily subcutaneous insulin. In five acute-onset, ketotic diabetic subjects first treated by intravenous saline and low-dose intramuscular insulin, the regimen was used to achieve and maintain basal and postprandial normoglycemia. Ketonuria was abolished quickly in these patients, and falling insulin requirements and large doses of insulin were handled easily. In both clinical situations, subsequent subcutaneous insulin doses required little adjustment. The regimen is cheap, convenient to use, and widely applicable.

In vivo insulin sensitivity and insulin binding to erythrocytes in children: insulin resistance in obese children.

This aim of this study was to determine whether RBC insulin receptor assay represents a clinically useful way of assessing insulin sensitivity in obese children. Steady state plasma glucose (SSPG) was established by a constant infusion of glucose (6 mg/kg/min), insulin (0.8 mU/kg/min) and somatostatin (125 micrograms/m2/h), following the loading dose of somatostatin (125 micrograms/m2). Insulin binding to RBCs was measured by a modified method of Gambhir and was compared with SSPG. Of 21 children with various relative body weight, 8 hyperinsulinemic obese children had a decreased insulin binding to RBCs due to decreased receptor concentrations. The insulin binding was inversely correlated with the fasting serum insulin level and with the insulin area under the O-GTT insulin response curve. In 11 children with various relative body weight, a highly significant inverse relationship was found between SSPG and insulin binding. SSPG was also correlated with the fasting serum insulin level. It was concluded that RBC insulin receptor may quantitatively reflect insulin resistance in obese children, and may be a useful tool for clinical evaluation of tissue insulin sensitivity in children.

In vivo insulin sensitivity and insulin binding to erythrocytes in children

The aim of this study was to determine whether RBC insulin receptor assay represents an useful means for assessment of the clinical sensitivity of insulin. Steady State Plasma Glucose(SSPG) was established by constant infusion of somatostatin (125μg loading followed by 125μg/m2/hr), glucose(6mg/kg/min) and insulin(0.8 mU/kg/min). Insulin binding to RBC receptors was measured and compared with SSPG. Insulin binding to RBC was assayed by the modified method of Gambhir. Plasma insulin area was measured by GTT. A decreased specific insulin binding(SB) to RBC, due to reduced receptor concentrations was observed in hyperinsulinemic obese children. In 21 children with various body weights, SB was inversely related to fasting plasma insulin level (r=-0.651, p<0.01) and also with plasma insulin area (r=-0.547, p<0.01). A highly significant inversed correlation was noted between SSPG and SB(r=-0.807, p<0.005). SSPG was, also, correlated with the fasting plasma insulin level(r=0.669, p<0.05). It was concluded that RBC insulin receptor is an useful tool for clinical evaluation of tissue insulin sensitivity.

Effect of Insulin Therapy on Insulin Resistance in Type II Diabetic Subjects: Evidence for Heterogeneity

Insulin resistance has been demonstrated previously in both glucose-intolerant and untreated type II diabetics. Although the former group had hyperinsulinemia in the fasting state and after an oral glucose load, the hyperglycemic type II subjects were relatively insulin deficient after carbohydrate ingestion. In an attempt to define the role that insulin deficiency might have played in the pathophysiology of insulin resistance in this latter group, we measured the steadystate plasma glucose level (SSPG) during a constant infusion of glucose (6 mg/kg/min) and insulin (80 mU/min) in 15 nonobese insulin-deficient type II diabetics before and after insulin replacement therapy. Hepatic glucose output (HGO) and plasma clearance of glucose (PCG) during these studies were also determined using an infusion of 3-3H-glucose. Before insulin treatment, 13 of 15 subjects had SSPG levels above the range reported in normal subjects. Two type II patients appeared to be normally sensitive to insulin. Although all subjects were more sensitive to insulin after 1–8 wk of insulin therapy, they could be clearly divided into two groups. Thus, five patients were now normally sensitive to insulin (SSPG:75 ± 5.3 mg/dl) while 10 patients were still significantly resistant (SSPG: 227±11.3 mg/dl). HGO during the infusion studies was high in the resistant group pretreatment and did not change posttreatment (3.54 ± 0.32 vs. 3.14 ± 0.78 mg/kg/min). HGO was lower before insulin therapy and decreased by &amp;gt; 50% after therapy in the sensitive group (1.93 ± 0.07 vs. 0.71 ± 0.32 mg/kg/min). PCG determinations revealed a similar pattern of response. Retrospective analysis of the two groups revealed no significant differences in their baseline characteristics. Fasting plasma glucose and hemoglobin A1c levels also did not differ between the two groups before or after insulin therapy. These results indicate that heterogeneity exists among subjects with type II diabetes mellitus. One-third of the subjects studied appear to have only insulin deficiency as the basis of their diabetes, while two-thirds seem to have an underlying resistance to insulin that is not corrected by insulin therapy.

1131 EVALUATION OF THERAPY IN DIABETES MELLITUS: USE OF FREE INSULIN

Diabetic therapy is traditionally evaluated using measurements of blood glucose regulation. The ability to measure free (unbound) insulin enables a more direct study of the relationship between insulin dose and the concentration of free insulin and metabolic parameters in blood.Eight insulin-dependent, C-peptide negative diabetics, 7 to 22 years, were given their usual dose of NPH and regular insulin, and blood samples were serially obtained for levels of free and total insulin, glucose, FFA and ketone bodies. Twenty-four hours later, a standard four hour low dose constant infusion of crystalline insulin was given. Data were analyzed using individual and integrated concentrations of hormones and substrates.In all patients, NPH insulin led to minimal increases in blood insulin (< 15 μU/ml). Low insulin concentrations were associated with elevated parameters of metabolic control. Insulin infusion led to insulin levels between 35 and 50 μU/ml at steady state, and with normalization of glucose, FFA and ketone bodies. A significant reduction in serum [K+] occurred during insulin infusion.These data suggest that: 1) long-acting insulin produces minimal increases in serum insulin; 2) measurement of free insulin can be used to distinguish inadequate insulin dose or aberrant absorption from insulin resistance; 3) conventional NPH therapy cannot lead to adequate metabolic control; 4) hypokalemia may be a marker for states of excess insulin administration.

Chronic maternal hyperinsulinemia and hypoglycemia. A model for experimental intrauterine growth retardation.

Chronic hypoglycemia was produced in pregnant rats during the last third of gestation through the constant infusion of insulin by a subcutaneously implanted, osmotically driven minipump. The offspring, delivered by cesarean section on day 20 of gestation, demonstrated intrauterine growth retardation compared to control fetuses. This effect was associated with fetal hypoglycemia and hypoinsulinemia. Fetal total and branch-chain amino acid concentrations were not affected. Increases in maternal to fetal glucose ratio and fetal insulin to glucose ratio in the experimental group were interpreted to indicate that limitation in the availability of glucose, a major fetal nutrient, was the basis for these effects.

Epinephrine-induced hypoaminoacidemia in normal and diabetic human subjects: effect of beta blockade.

To evaluate the effect of epinephrine on the circulating amino acids, we infused epinephrine into normal human subjects and juvenile-onset diabetic patients given a constant basal infusion of insulin. Epinephrine infusion produced an identical 350--400 pg/ml rise in plasma epinephrine in both groups. In normal subjects, epinephrine caused a progressive 26% reduction in total circulating amino acids, despite unchanged levels of plasma insulin. This effect was most pronounced for the branched amino acids, which fell by 40% (P < 0.001). Plasma alanine was the only amino acid which failed to decline. Similarly, infusion of epinephrine in the insulin-infused diabetics produced a 23% fall in total amino acids, a 37% decline in branched chain amino acids, but no change in plasma alanine. Saline infusion in the insulin-infused diabetics had no effect on plasma amino acid concentrations. In addition, when epinephrine was infused into two insulin-withdrawn diabetics, a comparable hypoaminoacidemic response was observed. The infusion of propranolol in both normal and diabetic subjects totally prevented the epinephrine-induced fall in plasma amino acids. It is concluded that (1) increments in epinephrine similar to those observed in stress cause a decline in circulating amino acids (except alanine) which is greatest for the branched chain amino acids; (2) this hypoaminoacidemic effect occurs in the absence of a rise in plasma insulin and diabetic subjects, as well; and (3) epinephrine-induced changes in amino acid regulation are prevented by beta-adrenergic blockade. Our findings suggest that, in contrast with glucose and fat metabolism, epinephrine and insulin may have similar, rather than antagonistic, effects on plasma amino acid metabolism.

Interaction of fat-stimulated gastric inhibitory polypeptide on pancreatic alpha and beta cell function.

Gastric inhibitory polypeptide (GIP) is considered to be the principal mediator of the enteroinsular axis. A glucose-insulin clamp technique was used to study the effects of differing blood glucose levels on the insulinotropic and glucagonotropic actions of fat-stimulated GIP in seven healthy subjects, as well as the effect of physiologic hyperinsulinemia on GIP secretion. Blood glucose levels were clamped for 4 h at 43+/-2 mg/dl (hypoglycemic clamp), 88+/-1 mg/dl (euglycemic clamp), and 141+/-2 mg/dl (hyperglycemic clamp) in the presence of a constant insulin infusion (100 m U/kg per h). Under hypoglycemic clamp conditions there was no increase in C-peptide nor glucagon after Lipomul ingestion, despite an increase of GIP of 51.7+/-8.7 ng/ml per 120 min. Under euglycemic clamp conditions, small and inconsistent increases in C-peptide and glucagon were observed after fat ingestion and a concomitant increase of GIP of 35.2+/-9.4 ng/ml per 120 min. Under hyperglycemic clamp conditions after fat ingestion and a GIP increase of 24.0+/-5.7 ng/ml per 120 min, C-peptide increased from 6.4+/-5 ng/ml to 11.0+/-1.1 ng/ml (P < 0.01) but glucagon did not change. These findings confirm that in healthy man GIP exerts its insulinotropic properties only under hyperglycemic conditions and indicate that GIP is not glucagonotropic. Under euglycemic clamp conditions (plasma glucose, 89+/-1 mg/dl) and physiologic hyperinsulinemia (serum immunoreactive insulin, 137+/-3 muU/ml) GIP responses to fat ingestion (39.7+/-9.8 ng/ml per 120 min) were not different from the GIP responses to fat ingestion in the absence of hyperinsulinemia (39.7+/-11.1 ng/ml per 120 min). Therefore, insulin under normoglycemic conditions does not exert an inhibitory effect on fat-stimulated GIP secretion. The higher GIP response to oral fat in the hypoglycemic clamp, and the lower GIP response in the hyperglycemic clamp compared to the response in the euglycemic clamp suggests an effect of glycemia itself on GIP secretion in the presence of hyperinsulinemia.

Attenuated response to insulin in very low birthweight infants.

Four hyperglycemia very low birthweight (VLBW) infants were given a constant infusion of insulin with parenteral alimentation in order to improve glucose tolerance. After an initial 45% decrease in serum glucose, intolerance recurred within hours to days despite delivery of as much as 16 units/kg/hour or regular insulin. Resulting serum insulin levels were as high as 72,000 microU/ml. It is suggested that insulin "resistance", instead of insulin deficiency, may be responsible for glucose intolerance in some infants.

Attenuated Response to Insulin in Very Low Birthweight Infants

Summary: Four hyperglycemia very low birthweight (VLBW) infants were given a constant infusion of insulin with parenteral alimentation in order to improve glucose tolerance. After an initial 45% decrease in serum glucose, intolerance recurred within hours to days despite delivery of as much as 16 units/kg/hour of regular insulin. Resulting serum insulin levels were as high as 72,000 μU/ml. It is suggested that insulin “resistance”, instead of insulin deficiency, may be responsible for glucose intolerance in some infants. Speculation: Abnormal hepatic regulatory systems, immature enzyme systems, or abnormal insulin receptors may play a part in the glucose intolerance of the premature infant. Further studies, both in human infants and in appropriate animal models will have to be done to determine the etiology (ies) of the glucose intolerance.

Glucose, insulin and somatostatin infusion for the determination of insulin resistance in liver cirrhosis.

Twelve patients with liver cirrhosis and ten normal subjects were studied. Using a constant intravneous infusion of glucose, insulin and somatostatin over 2 1/2 hours we determined the stteady state plasma glucose level (SSPG) in order to measure insulin resistance. The results demonstrated that the cirrhotic patients were insulin resistant compared to normals and that plasma glucagon does not account for the insulin resistance in these patients.

Hormonal responses during treatment of acute diabetic ketoacidosis with constant insulin infusions.

Changes in glucagon, growth hormone (GH), cortisol, renin and aldosterone accompanying the metabolic disturbances and dehydration of severe diabetic ketoacidosis were studied over a 24 h period in eight patients treated with a constant intravenous insulin infusion. Mean steady state plasma-free insulin levels achieved were 28.6--49 mu/1 in patients receiving 2 u/h but a satisfactory rate of fall of glucose was not always obtained until the infusion dose was increased to 4 u/h or more. The total insulin dose administered was positively correlated with the level of plasma glucagon and cortisol on admission. During insulin infusion, both glucagon and cortisol fell but the rate of fall was not related to dose or plasma level of free insulin achieved. In six of eight patients studied increments in plasma GH above admission levels were observed during insulin treatment. Admission values of both plasma renin activity and plasma aldosterone were raised. The renin levels were highest in newly diagnosed diabetics, and two patients with long-established diabetes showed only small increments despite profound dehydration. Plasma renin activity, but not plasma aldosterone correlated with the fluid and sodium retention over the initial 24 h treatment period, but not with potassium requirements. The urinary excretion rates of the small molecular weight proteins GH and insulin, were considerably elevated over the treatment and convalescent periods.

HYPERGLYCEMIA INHIBITS GLUCOSE PRODUCTION IN MAN INDEPENDENT OF CHANGES IN GLUCOREGULATORY HORMONES

To evaluate the influence of hyperglycemia on hepatic glucose output in the absence of a rise in insulin, glucose was infused for 2 hours into six juvenile-onset diabetics receiving a constant infusion of insulin at a rate of 0.05-0.15 microM kg-1min-1. Prior to the infusion of glucose, insulin administration resulted in stable levels of plasma glucose (76 +/- 8 mg/dl) and glucose output (1.9 +/- 0.1 mg kg-1min-1). The addition of glucose produced a 2-3 fold rise in plasma glucose and a prompt fall in glucose output to 0.2-0.4 mg kg-1min-1, despite the unchanged rate of insulin infusion and the absence of a reduction in plasma glucagon or catecholamines. A similar decline in glucose output was observed when exogenous glucagon (1 ng kg-1-min-1) was added to the glucose infusion. We conclude that in the presence of basal insulin levels hyperglycemia inhibits glucose output independent of a rise in insulin or a fall in anti-insulin hormones.

Validation of I.V. small-dose insulin infusion therapy in diabetic ketoacidosis of depancreatized dogs.

A validation of small-dose insulin infusion therapy was studied by the constant i.v. infusion of various doses of insulin into ketoacidotic depancreatized dogs. Constant insulin infusion of 5 x B, 10 X B, 30 X B and 50 X B (B = 225 microunit/kg/min) was performed for 3 h by mechanical pump. The following results were obtained: (1) plasma concentrations of immunoreactive insulin (IRI) increased proportionally to the dose of infused insulin, but the higher IRI did not result in a greater fall in plasma glucose concentration, correspondingly; the mean rate of fall in plasma glucose concentration of 5 x B was not significantly lower than that of 50 x B; beta-hydroxybutyrate and arterial pH improvements were observed in each group during the 3-h insulin infusion. These data suggested that for the improvement of diabetic ketoacidosis, the insulin infusion rate of more than 30 x B, which raised the plasma IRI levels above the physiological range, was not essential; (2) the necessity of potassium supplementation during the small-dose insulin infusion was suggested if the pre-treatment level of serum potassium was low. These results confirmed that in the absence of infection or severe acidosis small-dose insulin infusion therapy is as effective as the conventional large-dose insulin therapy.