Insulin-induced posthypoglycemic hyperglycemia as a cause of "brittle" diabetes. Clinical clues and therapeutic implications

Studies of insulin resistance following hypoglycemia in insulin-dependent diabetes mellitus

Insulin resistance was assessed after a hypoglycemia induced by insulin (1.5 mU X kg-1 X min-1) between 7 and 8 a.m. in 10 well-insulinized patients with insulin-dependent diabetes mellitus (IDDM). Blood glucose levels during a somatostatin (100 micrograms X h-1)-insulin (0.4 mU X kg-1 X min-1)-glucose (4.5 mg X kg-1)-infusion test (SIGIT) performed between 11 a.m. and 3 p.m. served as an indicator of total body insulin resistance. Plasma epinephrine, growth hormone, and cortisol increased in response to hypoglycemia, while blunted responses of glucagon were simultaneously registered. At the start of the subsequent SIGIT, blood glucose and plasma-free insulin concentrations were similar to those obtained in the control study without preceding hypoglycemia, and at this point all counter-regulatory hormones had returned to basal. During the SIGIT close to identical levels of plasma-free insulin and counter-regulatory hormones were registered, despite which a significant hyperglycemia was seen 2 hours after the start of the SIGIT when preceded by hypoglycemia. In a separate study, the SIGIT was shown to have a good reproducibility in IDDM patients. We conclude that hypoglycemia evokes a state of insulin resistance for several hours, as demonstrated by elevated blood glucose levels during a somatostatin-insulin-glucose-infusion test.

The effect of asymptomatic nocturnal hypoglycemia on glycemic control in diabetes mellitus

To assess the effect of asymptomatic nocturnal hypoglycemia on glycemic control in insulin-dependent diabetes mellitus, we studied, on three nights, 10 patients receiving their usual regimens of continuous subcutaneous insulin infusion. During a control night, the patients' mean (+/- SE) plasma glucose level reached a nadir of 4.5 +/- 0.2 mmol per liter at 3 a.m.; the fasting glucose level was 5.9 +/- 0.3 mmol per liter at 7:30 a.m., and a peak glucose level of 8.6 +/- 0.3 mmol per liter was reached at 10 a.m., after breakfast. During nights two and three, supplemental insulin was infused intravenously from 10 p.m. to 2 a.m. to simulate a clinical overdose of insulin. On these nights, either hypoglycemia (2.4 +/- 0.2 mmol per liter) was permitted to occur or a nearly normal glucose level (5.5 mmol per liter) was maintained by infusion of glucose. The subjects were asymptomatic on all three nights. Despite comparable plasma free insulin levels from 4 to 11 a.m., both fasting (7.3 +/- 0.2 mmol per liter) and postbreakfast (12.5 +/- 0.4 mmol per liter) plasma glucose levels were significantly higher after hypoglycemia than when hypoglycemia was prevented (6.2 +/- 0.2 mmol per liter and 8.7 +/- 0.4 mmol per liter, respectively; P less than 0.001 in both cases). Fasting levels of plasma glucose correlated directly with overnight plasma levels of epinephrine (r = 0.78, P less than 0.001), growth hormone (r = 0.57, P less than 0.009), and cortisol (r = 0.52, P less than 0.02) but correlated inversely with the overnight nadir of plasma glucose (r = -0.62, P less than 0.005). We conclude that asymptomatic nocturnal hypoglycemia can cause clinically important deterioration in glycemic control (the Somogyi phenomenon) in patients receiving intensive insulin therapy, and should therefore be considered in the differential diagnosis of unexplained morning hyperglycemia.

Failure of nocturnal hypoglycemia to cause fasting hyperglycemia in patients with insulin-dependent diabetes mellitus

To test the hypothesis that nocturnal hypoglycemia causes fasting hyperglycemia (the Somogyi phenomenon) in patients with insulin-dependent diabetes mellitus, we studied 10 patients, who were on their usual therapeutic regimens, from 10 p.m. through 8 a.m. on three nights. On the first night, only a control procedure was performed (blood sampling only); on the second night, hypoglycemia was prevented (by intravenous glucose infusion, if necessary, to keep plasma glucose levels above 100 mg per deciliter [5.6 mmol per liter]); and on the third night, hypoglycemia was induced (by stepped intravenous insulin infusions between midnight and 4 a.m. to keep plasma glucose levels below 50 mg per deciliter [2.8 mmol per liter]). After nocturnal hypoglycemia was induced (36 +/- 2 mg per deciliter [2.0 +/- 0.1 mmol per liter] [mean +/- SE] from 2 to 4:30 a.m.), 8 a.m. plasma glucose concentrations (113 +/- 18 mg per deciliter [6.3 +/- 1.0 mmol per liter]) were not higher than values obtained after hypoglycemia was prevented (182 +/- 14 mg per deciliter [10.1 +/- 0.8 mmol per liter]) or those obtained after blood sampling only (149 +/- 20 mg per deciliter [8.3 +/- 1.1 mmol per liter]). Indeed, regression analysis of data obtained on the control night indicated that the 8 a.m. plasma glucose concentration was directly related to the nocturnal glucose nadir (r = 0.761, P = 0.011). None of the patients was awakened by hypoglycemia. Scores for symptoms of hypoglycemia, which were determined at 8 a.m., did not differ significantly among the three studies. We conclude that asymptomatic nocturnal hypoglycemia does not appear to cause clinically important fasting hyperglycemia in patients with insulin-dependent diabetes mellitus on their usual therapeutic regimens.

Nocturnal hypoglycaemia in patients receiving conventional treatment with insulin

The prevalence of nocturnal biochemical hypoglycaemia--that is, blood glucose concentrations below 3 mmol/l (55 mg/100 ml)--was evaluated in a random sample of 58 insulin dependent diabetics receiving twice daily insulin. Seventeen patients had at least one blood glucose value below 3 mmol/l (55 mg/100 ml) and five a value below 2 mmol/l (36 mg/100 ml) during the night. Both bedtime (2300) and fasting morning (0700) blood glucose concentrations were significantly lower in the group with nocturnal hypoglycaemia compared with the group without (p less than 0.00001). If the bedtime blood glucose concentration was below 6 mmol/l (108 mg/100 ml) the risk of nocturnal hypoglycaemia was 80% (95% confidence limits 51-96%). If the bedtime blood glucose concentration was above 6 mmol/l the likelihood of hypoglycaemia not occurring during the night was 88% (74-96%). The mean glycosylated haemoglobin A1c (HbA1c) concentration in the group with nocturnal biochemical hypoglycaemia (8.2 (range 5.0-12.4)%) was significantly lower than that in the group without (9.4(7.0-14.2)%) (p less than 0.02). The prevalence of nocturnal hypoglycaemia in the patients receiving twice daily insulin (29%) was compared with that in 15 patients receiving thrice daily insulin (47%) and was not found to be significantly different. The likelihood of this risk being greater with thrice daily insulin was, however, 88%. No patient with nocturnal biochemical hypoglycaemia woke up during the night with symptomatic hypoglycaemia. Nocturnal biochemical hypoglycaemia is common during twice daily treatment with insulin, and low values of HbA1c might be associated with a higher risk of such hypoglycaemia. The blood glucose concentration at bedtime is a significant predictor of nocturnal biochemical hypoglycaemia, and HbA1c values might be of help in identifying patients at risk.

Glucose counterregulation, hypoglycemia, and intensive insulin therapy in diabetes mellitus

The prevention or correction of hypoglycemia is the result of both dissipation of insulin and activation of counterregulatory systems. In the models studied to date, glucagon and epinephrine have been shown to be the key counterregulatory factors; the potential roles of other hormones, neural factors, or substrate mechanisms in other models and during more gradual recovery from hypoglycemia remain to be defined. Deficient glucagon responses to decrements in plasma glucose, which are common in patients with IDDM and occur in some patients with NIDDM, result in altered counterregulation. But counterregulation is generally adequate, because epinephrine compensates for it. Defective glucose counterregulation due to combined deficiencies of glucagon and epinephrine secretory responses occurs in many patients, typically those with longstanding diabetes, and must be added to the list of factors known to increase the risk of hypoglycemia, at least during intensive therapy. From the material reviewed, it should be apparent that much has been learned about glucose counterregulation. It should be equally clear that much remains to be learned. Among the many possibilities, we consider four worthy of emphasis. First of all, we need to examine the physiology and pathophysiology of glucose counterregulation in additional models (e.g., during exercise) and over longer periods. Secondly, we need to determine whether central nervous system adaptation to antecedent glycemia occurs and, if so, identify its mechanisms. Thirdly, we need to develop better methods of insulin delivery or learn to correct or compensate for defective counterregulatory systems, if we are to achieve euglycemia safely in diabetic patients with defective glucose counterregulation. Finally, we need to know whether effective control of diabetes mellitus prevents development of defective glucose counterregulation.

Instability of fasting blood glucose values in noninsulin-dependent diabetic patients with long-term insulin treatment

We studied serum free C-peptide immunoreactivity (CPR) and the coefficient of variation (CV) of fasting blood glucose values (FBG) in 26 insulin-treated patients with non-insulin-dependent diabetes mellitus (NIDDM) in relation to the duration of insulin treatment. Serum free CPR responses during 100 g oral glucose tolerance test (OGTT) were significantly lower in patients with insulin treatment for five years or more than in those with insulin treatment for less than five years although their previous immunoreactive insulin (IRI) responses during OGTT before insulin treatment showed no significant difference. CV of FBG was found to be significantly higher at the time of this study (20.6 +/- 7.8%, mean +/- SD) than at the second year of insulin treatment (15.3 +/- 7.7%, P less than 0.05) in the patients with insulin treatment for five or more years but did not show any significant difference in patients with insulin treatment for less than five years at the corresponding times. Thus we measured CV of the FBG in NIDDM patients at various intervals during the long-term insulin or oral hypoglycemic agent treatment in another study. In 20 patients with insulin treatment, CV of FBG was found to be significantly different among the various intervals during insulin treatment (P less than 0.0025). It was significantly higher at the eight year (22.2 +/- 8.6%) and 12th year (21.9 +/- 9.1%) than at the second year (14.9 +/- 6.1%) and fifth year (15.0 +/- 6.7%) of insulin treatment (P less than 0.025, P less than 0.025; P less than 0.05, P less than 0.01, respectively).(ABSTRACT TRUNCATED AT 250 WORDS)

Glucose counterregulation and waning of insulin in the Somogyi phenomenon (posthypoglycemic hyperglycemia)

To determine the roles of glucose counterregulation and the waning of insulin action in the development of posthypoglycemic hyperglycemia (the Somogyi phenomenon), we studied changes in plasma glucose and glucose turnover in five patients with insulin-dependent diabetes mellitus (IDDM) after subcutaneous injection of insulin under conditions in which hypoglycemic glucose counterregulation and the waning of insulin action were allowed to occur or were prevented. In control experiments, in which both glucose counterregulation and insulin waning were allowed to occur, plasma glucose levels decreased from 94 +/- 3 to 47 +/- 7 mg per deciliter and then increased to 289 +/- 20 mg per deciliter at 12 hours because of a marked increase in glucose production. When the waning of insulin action was prevented by insulin infusion, glucose production increased less (P less than 0.01), but marked rebound hyperglycemia still occurred (188 +/- 26 mg per deciliter at 12 hours). When both insulin waning and glucose counterregulation were prevented by infusion of both glucose and insulin, glucose production did not increase, and rebound hyperglycemia did not occur. We conclude that hypoglycemia can cause rebound hyperglycemia in the absence of insulin waning in patients with IDDM, and that this results primarily from an excessive increase in glucose production due to activation of glucose counterregulatory systems.

Relation between the urinary cortisol:creatinine ratio and hypoglycaemia

The relationships between (i) urinary free cortisol and urinary creatinine concentrations and (ii) the urinary cortisol/creatinine ratio (UCCR) and various glycaemic levels were studied in three groups--normal, insulin-stressed and insulin-treated diabetic subjects. In non-hypoglycaemic subjects, there was a significant positive linear correlationship between urinary free cortisol and urinary creatinine excretion, but in the presence of hypoglycaemia, this relationship was lost. The highest mean urinary cortisol/creatinine ratio (UCCR) was found in subjects after an insulin tolerance test (ITT). The mean post-ITT UCCR was significantly greater than the mean for the pre-ITT samples. There was a significant negative correlation between capillary blood glucose levels at 03.00 and the UCCR of the overnight urine samples of insulin-treated diabetic subjects. We conclude that there is a definite increase in the UCCR after hypoglycaemia in subjects with adequate adrenocortical response to hypoglycaemia and that determination of the UCCR could be helpful in the detection of nocturnal hypoglycaemia.

The role of "diabetogenic" hormones on carbohydrate and lipid metabolism following oral glucose loading in insulin dependent diabetics: effects of acute hormone administration

To evaluate the relative role of "diabetogenic" hormones as insulin antagonists in severe derangements of diabetic control, glucagon, cortisol, growth hormone and adrenaline were administered by continuous intravenous infusion, separately and in combination, to ketosis-prone insulin-dependent diabetics (n = 11). The amount of insulin required for the assimilation of a 50 g glucose load during the various hormone infusions was determined by means of an automated glucose-controlled insulin infusion system and used as an index of insulin effectiveness. Raising plasma hormone concentrations acutely into the range seen in severe diabetic states (glucagon 517 +/- 70 pg/ml; cortisol 32 +/- 3 micrograms/dl; growth hormone 14 +/- 3 ng/ml) did not alter significantly blood glucose profile and insulin requirement (control 11.3 +/- 1.1 U; glucagon 11.6 +/- 2.0 U; cortisol 11.1 +/- 0.4 U; growth hormone 12.9 +/- 1.4 U), except for adrenaline (plasma level 550 +/- 192 pg/ml), which caused a marked rise in blood glucose levels and a threefold increase in insulin demand (31.1 +/- 3.7 U). Combined infusion of all hormones did not potentiate significantly the latter effect (38.3 +/- 4.7 U). The effectiveness of metabolic control by insulin was assessed by a marked decrease in plasma nonesterified free fatty acids and ketone bodies upon its administration after glucose ingestion in all groups studied. It is concluded that from the hormones investigated within this study adrenaline exerts the strongest diabetogenic action during its short term administration followed by that of growth hormone. Whereas it may well be that over-insulinization of the patients by the glucose controlled insulin infusion system has overcome and disguised the smaller diabetogenic effects of cortisol and glucagon.

Profiles of metabolic control in diabetic children-frequency of asymptomatic nocturnal hypoglycemia

Twenty-four hr glucose and hormonal monitoring was conducted in 34 randomly selected children with insulin dependent diabetes. Asymptomatic nocturnal hypoglycemia was present in 18% (6/34). The nocturnal plasma glucose decline of 20-25 mg/dl/hr reached a mean nadir of 50 mg/dl. The mean rebound hyperglycemia of 300 mg/dl over the subsequent 6.4 hrs. was significantly greater than any glucose excursion in diabetic children with daytime, symptomatic hypoglycemia (n = 5) or in those with non-hypoglycemic profiles (n = 23). Coincident with the nocturnal decline, but preceding the glucose nadir, was a marked release of growth hormone which was significantly greater (p less than .05) than that observed in the other diabetic groups. This release of growth hormone, and the nocturnal hypoglycemia, were reflected in the ratio of awake/sleep mean concentrations of glucose and growth hormone. These data support the speculation that growth hormone release contribute to the hyperglycemic rebound observed. Mean 24 hr growth hormone concentrations varied considerably from patient to patient such that a generalization for growth hormone concentrations in insulin dependent diabetes cannot be made. Asymptomatic nocturnal hypoglycemia is a frequent complication of the therapy of insulin dependent diabetes. Subsequent hyperglycemic rebound (the "Somogyi Effect") is associated with exuberant counterregulatory release of growth hormone. The precise pathophysiological role of this growth hormone release is unclear.

In search of the Somogyi effect

Insulin-treated diabetic patients may show a rapid swing to hyperglycaemia after episodes of hypoglycaemia. This rebound hyperglycaemia, or Somogyi effect, is thought to be caused by the unopposed actions of hormonal antagonists to insulin secreted in response to hypoglycaemia. To test this theory a study was made of 15 patients who had 17 episodes of asymptomatic untreated hypoglycaemia (blood-glucose less than 2 mmol/l) between 11 P.M. and 3 A.M. After nocturnal hypoglycaemia, mean fasting blood-glucose concentrations at 7 A.M. ranged from 0.7-17 mmol/l and were over 7 mmol/l in 6 patients. These 6 patients with apparent rebound hyperglycaemia did not have higher levels of growth hormone, cortisol, or glucagon than those who had little or no recovery of blood-glucose. There was a close inverse correlation (r = -0.996, p < 0.001) between blood-glucose and free insulin, suggesting that hyperglycaemia, when present, was due to relative insulin deficiency in the latter part of the night. Early changes in blood-glucose after untreated hypoglycaemia seem to be primarily due to changes in free insulin rather than a response to antagonist hormones.

The Somogyi phenomenon. A short review

The Somogyi phenomenon or effect is a paradoxical situation of insulin-induced post-hypoglycemic hyperglycemia. The historical aspects of this phenomenon and the subsequent hypotheses and controversy are reviewed. The clinical situation is explained, with regard to its recognition, management and importance as an etiological factor in "brittle" diabetes. Hormone immunoassay techniques at present show human growth hormone (HGH) to be the major consequence of insulin-induced hypoglycemia leading to post-hypoglycemia glucose intolerance, but further studies will probably show glucagon to have a major role.

Managing insulin-dependent diabetic patients

About 95% of insulin-dependent diabetics can be managed satisfactorily with one or a combination of the follwoing insulin preparations: single-peak U-100 beef-pork NPH, Lente, and Regular. Complications of insulin therapy are commonly attributable to poor regulation of insulin dosage, irregular or excessive food intake, or unusual physical activity. One form of hyperglycemia is induced by insulin. Generalized allergic reactions to insulin sometimes require desensitization. Insulin-resistant patients are treated with a glucocorticoid.