Predictive neural networks for learning the time course of blood glucose levels from the complex interaction of counterregulatory hormones

Diabetes mellitus is a widespread disease associated with an impaired hormonal regulation of normal blood glucose levels. Patients with insulin-dependent diabetes mellitus (IDDM) who practice conventional insulin therapy are at risk of developing hypoglycemia (low levels of blood glucose), which can lead to severe dysfunction of the central nervous system. In large retrospective studies, up to approximately 4% of deaths of patients with IDDM have been attributed to hypoglycemia (Cryer, Fisher, & Shamoon, 1994; Tunbridge, 1981; Deckert, Poulson, & Larsen, 1978). Thus, a better understanding of the complex hormonal interaction preventing hypoglycemia is crucial for treatment. Experimental data from a study on insulin-induced hypoglycemia in healthy subjects are used to demonstrate that feedforward neural networks are capable of predicting the time course of blood glucose levels from the complex interaction of glucose counterregulatory (glucose-raising) hormones and insulin. By simulating the deficiency of single hormonal factors in this regulatory network, we found that the predictive impact of glucagon, epinephrine, and growth hormone secretion, but not of cortisol and norepinephrine, were dominant in restoring normal levels of blood glucose following hypoglycemia.     

Physiologic hyperinsulinemia enhances counterregulatory hormone responses to hypoglycemia in IDDM

We evaluated the effect of physiologic hyperinsulinemia (plasma insulin 329 +/- 62 vs 687 +/- 62 pmol/L) on counterregulatory hormone responses in 8 IDDM subjects studied during a 2-hour hypoglycemic clamp study with an equivalent degree of hypoglycemia (plasma glucose 3.1 +/- 0.1 and 3.0 +/- 0.1 mmol/L, respectively). Plasma epinephrine levels were increased by 71% during the last 60 minutes of hypoglycemia in the high insulin study (840 +/- 180 vs 1440 +/- 310 pmol/L, respectively p = 0.006). In addition, plasma cortisol and norepinephrine were also increased in the high insulin study (by 19% and 24% respectively, p < 0.01, for both). Plasma growth hormone and glucagon concentrations were not altered by high dose insulin infusion. In spite of increased epinephrine secretion, the glucose infusion rate required to maintain glucose was 2-fold greater in the high insulin study, and there was greater suppression of lipolysis in that group. We conclude that hyperinsulinemia may enhance counterregulatory hormone secretion in IDDM.     

Prolactin and beta-endorphin responses to hypoglycemia are reduced in well-controlled insulin-dependent diabetes mellitus

Several pituitary hormones, including corticotropin (ACTH), growth hormone (GH), prolactin, and beta-endorphin (but not thyrotropin, follicle-stimulating hormone, or luteinizing hormone), are released in response to hypoglycemia in normal subjects. In patients with insulin-dependent diabetes mellitus (IDDM), the degree of glycemic control is known to alter ACTH and GH responses to hypoglycemia. The current study was performed to examine the effect of glycemic control on prolactin and beta-endorphin responses to hypoglycemia in subjects with IDDM. We performed 3-hour stopped hypoglycemic-hyperinsulinemic clamp studies (12 pmol/kg/min) during which plasma glucose was decreased from 5.0 mmol/L to 2.2 mmol/L in steps of 0.6 mmol/L every 30 minutes in 20 subjects with uncomplicated IDDM (12 males and eight females; age, 26 +/- 2 years; IDDM duration, 10 +/- 1 years; body mass index, 23.6 +/- 0.6 kg/m2) and 10 healthy subjects (five males and five females aged 30 +/- 1 years). The 10 diabetic subjects in good glycemic control (mean hemoglobin A1 [HbA1], 7.5% +/- 0.3%; normal range, 5.4% to 7.4%) were compared with the 10 poorly controlled patients (mean HbA1, 12.6% +/- 0.5%; P < .001 v well-controlled diabetic group). During hypoglycemia, prolactin levels in the well-controlled diabetic group did not change (7 +/- 1 microgram/L at plasma glucose 5.0 mmol/L to 9 +/- 2 micrograms/L at plasma glucose 2.2 mmol/L), whereas prolactin levels increased markedly in the poorly controlled diabetic group (7 +/- 2 micrograms/L to 44 +/- 17 micrograms/L) and healthy volunteers (12 +/- 2 micrograms/L to 60 +/- 19 micrograms/L, P < .05 between IDDM groups). The plasma glucose threshold required for stimulation of prolactin secretion was 2.2 +/- 0.1 mmol/L in well-controlled IDDM, 3.0 +/- 0.4 mmol/L in poorly controlled IDDM, and 2.4 +/- 0.1 mmol/L in healthy subjects (P < .05 between IDDM groups). Responses in males and females were similar. The increase in beta-endorphin levels was also attenuated in well-controlled IDDM patients (4 +/- 1 pmol/L at plasma glucose 5.0 mmol/L to 11 +/- 4 pmol/L at plasma glucose 2.2 mmol/L) versus poorly controlled IDDM patients (5 +/- 1 pmol/L to 26 +/- 7 pmol/L) and healthy subjects (8 +/- 1 pmol/L to 56 +/- 13 pmol/L). The plasma glucose threshold required for stimulation of beta-endorphin release was again lower in well-controlled IDDM versus poorly controlled IDDM patients (2.2 +/- 0.1 v 3.0 +/- 0.3 mmol/L) and healthy subjects (2.5 +/- 0.4 mmol/L, P < .05 between IDDM groups). In conclusion, prolactin and beta-endorphin responses to a standardized hypoglycemic stimulus (plasma glucose, 2.2 mmol/L) are reduced and plasma glucose levels required to stimulate release of prolactin and beta-endorphin are lower in well-controlled IDDM compared with poorly controlled IDDM and healthy subjects. Thus, stress hormones not previously considered to have a primary role in plasma glucose recovery from hypoglycemia are affected by glycemic control, suggesting a more generalized alteration of hypothalamic-pituitary responses to hypoglycemia in IDDM patients with strict glycemic control.     

Deficient counterregulation: a possible risk factor for excessive fetal growth in IDDM pregnancies

OBJECTIVE: The rate of macrosomia in infants born to women with IDDM remains high despite intensive insulin therapy and good glycemic control. We hypothesized that one of the factors contributing to this high rate of macrosomia is deficient counterregulatory hormonal responses to hypoglycemia. RESEARCH DESIGN AND METHODS: Hypoglycemia was induced in 17 women with IDDM and 10 normal control subjects at 24-28 and at 32-34 weeks' gestation, using the hypoglycemic clamp technique. Plasma glucose concentrations were decreased to 3.3 mmol/l and maintained at this level for 1 h. Blood samples were drawn every 15 min for measurement of counterregulatory hormone concentrations. RESULTS: All 17 women with IDDM had diminished epinephrine responses to hypoglycemia, compared with control subjects. Eight of the women with IDDM (nonresponders) had minimal or no responses (< 165 pmol/l above baseline) and nine women (responders) had a moderate response (244-764 pmol/l). Of the eight nonresponders, seven had large infants (birth weight in the upper quartile), while only three of the nine responders had large infants (P < 0.05). CONCLUSIONS: Severely impaired counterregulatory epinephrine responses to hypoglycemia in pregnant women with IDDM may be a factor contributing to excessive fetal growth. We speculate that in these women, recurrent episodes of hypoglycemia may result in frequent bouts of increased caloric intake, with repeated episodes of transient hyperglycemia leading to fetal hyperinsulinism and excessive fetal growth.     

Reduced beta-adrenergic sensitivity in patients with type 1 diabetes and hypoglycemia unawareness

OBJECTIVE: We tested the hypothesis that impaired tissue sensitivity to catecholamines contributes to hypoglycemia unawareness in subjects with type 1 diabetes. RESEARCH DESIGN AND METHODS: A total of 21 subjects with type 1 diabetes underwent a standardized insulin infusion protocol to produce a stepwise decrease in plasma glucose to 45-min plateaus of 4.3, 3.6, 3.0, and 2.3 mmol/l. Glycemic thresholds, maximum responses for adrenergic and neuroglycopenic symptoms, and counterregulatory hormones were determined. Patients were classified as hypoglycemia unaware if the initiation of adrenergic symptoms occurred at a plasma glucose level 2 SD below that of nondiabetic volunteers. beta-Adrenergic sensitivity was measured as the dose of isoproterenol required to produce an increment in heart rate of 25 beats per minute above baseline (I25) in resting subjects. RESULTS: Subjects with type 1 diabetes and hypoglycemia unawareness experienced the onset of adrenergic symptoms at a lower plasma glucose level than did those with awareness (2.5+/-0.1 vs. 3.7+/-0.1 mmol/l, P < 0.001), whereas neuroglycopenic symptoms occurred at similar glucose levels (2.7+/-0.2 vs. 2.8+/- 0.1 mmol/l). The plasma glucose levels for counterregulatory hormone secretion (epinephrine 2.9+/-0.2 vs. 4.1+/-0.2 mmol/l; norepinephrine 2.7+/-0.1 vs. 3.2+/-0.2 mmol/l; cortisol 2.5+/-0.2 vs. 3.3+/-0.2 mmol/l, P < 0.01) were also lower in subjects with unawareness. The maximal epinephrine (1,954+/-486 vs. 5,332+/- 1,059 pmol/l, P < 0.01), norepinephrine (0.73 +/- 0.14 vs. 1.47+/-0.21 nmol/l, P = 0.04), and cortisol (276+/-110 vs. 579+/-83 nmol/l, P < 0.01) responses were reduced in the unaware group. I25 was greater in unaware subjects than in subjects without unawareness (1.5+/-0.3 vs. 0.8+/-0.2 microg), where I25 was not different from that of controls (0.8 +/-0.2 microg). CONCLUSIONS: We conclude that subjects with type 1 diabetes and hypoglycemia unawareness have reduced beta-adrenergic sensitivity, which may contribute to their impaired adrenergic warning symptoms during hypoglycemia.     

A novel presentation of Addison disease: hypoglycemia unawareness in an adolescent with insulin-dependent diabetes mellitus

A 16-year-old boy with insulin-dependent diabetes mellitus (IDDM) and a history of marginal glycemic control had severe hypoglycemia unawareness and a marked decrease in insulin requirement. His counterregulatory hormone response at the time of hypoglycemia suggested adrenocortical and adrenomedullary dysfunction. Further testing confirmed Addison disease. The patient's hypoglycemia unawareness was reversed by glucocorticoid replacement, although the plasma epinephrine response to hypoglycemia remained undetectable.     

Inhibitory effect of pregnancy on counterregulatory hormone responses to hypoglycemia in awake rat

Intensive insulin treatment during diabetic pregnancy is complicated by maternal hypoglycemia. To investigate whether pregnancy may contribute as an independent hypoglycemia risk factor, awake pregnant rats that were near term underwent stepped insulin hypoglycemic (3.4 and 2.3 mM) clamp studies in the fasted and nonfasted states. In the fasted state, the glucagon response to hypoglycemia was completely suppressed in the pregnant rats (P < 0.01). Epinephrine, but not norepinephrine, was also diminished by approximately 70-75% at both hypoglycemic steps, and more exogenous glucose was needed to maintain hypoglycemia during pregnancy. To avoid the potential confounding effect of increased ketone levels (beta-hydroxybutyrate was approximately 170% higher in the pregnant rats), experiments were repeated in the nonfasting state when ketosis was eliminated in both groups. The nonfasted pregnant rats continued to show near complete suppression of the glucagon response, even at glucose levels of 2.3 mM. In contrast, a brisk response occurred in nonpregnant controls when glucose fell to 3.4 mM. Although epinephrine levels in the pregnant rats were also markedly suppressed during the milder hypoglycemic stimulus, they approached values seen in nonpregnant controls when glucose was lowered further to 2.3 mM. We concluded that in the rat, pregnancy markedly suppresses glucagon responses to hypoglycemia. The release of epinephrine, but not norepinephrine, is also blunted, especially during mild hypoglycemia. These findings suggest that pregnancy may impair glucose counterregulation by inhibiting glucagon and epinephrine release during hypoglycemia.     

Mothers of adult children with schizophrenia: an ethnographic study

The purpose of this review was to summarize some of the recent advances that have been made in the understanding of the physiology involved in the counter-regulatory response to prolonged hypoglycemia. It is hoped that this review may stimulate thought and increase awareness that other factors, in addition to the glycemic level, can affect the counterregulatory response. However, if the goal of tight metabolic control is to be achieved in subjects with diabetes, the spectre of severe hypoglycemia must be removed. Thus further work is needed to understand the physiologic mechanisms controlling hypoglycemic counterregulation in normal subjects and the syndromes of abnormal counterregulatory responses present in subjects with IDDM.     

Counterregulation by epinephrine and glucagon during insulin-induced hypoglycemia in the conscious dog

We assessed the combined role of epinephrine and glucagon in regulating gluconeogenic precursor metabolism during insulin-induced hypoglycemia in the overnight-fasted, adrenalectomized, conscious dog. In paired studies (n = 5), insulin was infused intraportally at 5 mU.kg-1.min-1 for 3 h. Epinephrine was infused at a basal rate (B-EPI) or variable rate to simulate the normal epinephrine response to hypoglycemia (H-EPI), whereas in both groups the hypoglycemia-induced rise in cortisol was simulated by cortisol infusion. Plasma glucose fell to approximately 42 mg/dl in both groups. Glucagon failed to rise in B-EPI, but increased normally in H-EPI. Hepatic glucose release fell in B-EPI but increased in H-EPI. In B-EPI, the normal rise in lactate levels and net hepatic lactate uptake was prevented. Alanine and glycerol metabolism were similar in both groups. Since glucagon plays little role in regulating gluconeogenic precursor metabolism during 3 h of insulin-induced hypoglycemia, epinephrine must be responsible for increasing lactate release from muscle, but is minimally involved in the lipolytic response. In conclusion, a normal rise in epinephrine appears to be required to elicit an increase in glucagon during insulin-induced hypoglycemia in the dog. During insulin-induced hypoglycemia, epinephrine plays a major role in maintaining an elevated rate of glucose production, probably via muscle lactate release and hepatic lactate uptake.     

Morphologic and cytochemical characteristics of blood cells from Hawaiian green turtles

Previous studies have shown that hypoglycemia may reduce counterregulatory responses to subsequent hypoglycemia in healthy subjects and in patients with diabetes. The effect of hypoglycemia on the hormonal response to a nonhypoglycemic stimulus is uncertain. To test the hypothesis that the cortisol response to corticotropin (ACTH) infusion is independent of antecedent hypoglycemia, 10 healthy subjects received a standard ACTH infusion (0.25 mg Cosyntropin [Organon, West Orange, NJ] intravenously over 240 minutes) at 8:00 AM on day 1 and day 3 and a hypoglycemic insulin clamp study (1 mU/kg/min) at 8:00 AM on day 2. During the hypoglycemic clamp, plasma glucose decreased from 5.0 mmol/L to 2.8 mmol/L for two periods of 120 minutes (mean glucose, 2.9 +/- 0.03 and 2.8 +/- 0.02 mmol/L, respectively) separated by a 60-minute interval of euglycemia (mean glucose, 4.7 +/- 0.01 mmol/L). Seven subjects also had paired control studies in random order during which a 330-minute euglycemic clamp (mean glucose, 5.0 +/- 0.11 mmol/L) instead of a hypoglycemic clamp was performed on day 2. Basal ACTH (4.6 +/- 0.7 v 2.6 +/- 0.4 pmol/L, P < .02) and basal cortisol (435 +/- 46 v 317 +/- 40 nmol/L, P < .02) both decreased from day 1 to day 3 following intervening hypoglycemia. In contrast, with intervening euglycemia, neither basal ACTH (5.9 +/- 1.5 v 4.5 +/- 1.0 pmol/L) nor basal cortisol (340 +/- 38 v 318 +/- 60 nmol/L) were reduced significantly on day 3 compared with day 1. Following interval hypoglycemia, the area under the curve (AUC) for the cortisol response to successive ACTH infusions was increased (4,734 +/- 428 nmol/L over 240 minutes [day 3] v 3,526 +/- 434 nmol/L over 240 minutes [day 1], P < .01). The maximum incremental cortisol response was also significantly increased (805 +/- 63 nmol/L (day 3) v 583 +/- 58 nmol/L (day 1), P < .05). In contrast, the AUC for the cortisol response to successive ACTH infusions with interval euglycemia (3,402 +/- 345 nmol/L over 240 minutes [day 3] v 3,709 +/- 391 nmol/L over 240 minutes [day 1] and the incremental cortisol response (702 +/- 62 nmol/L [day 3] v 592 +/- 85 nmol/L [day 1] were unchanged. Following exposure to intermittent hypoglycemia in healthy humans, fasting morning ACTH and cortisol levels are reduced and the incremental cortisol response to an infusion of ACTH is enhanced. The enhanced cortisol response to exogenous ACTH infusion after intervening hypoglycemia (but not intervening euglycemia) may reflect priming of the adrenal gland by endogenous ACTH produced during the hypoglycemia. These data suggest that adrenal function testing by exogenous ACTH administration is not impaired by prior exposure to hypoglycemia. Moreover, the reduced cortisol response to recurrent hypoglycemia in patients with well-controlled diabetes is not likely the result of impaired adrenal responsiveness.     

Impact of nocturnal hypoglycemia on hypoglycemic cognitive dysfunction in type 1 diabetes

To test the hypothesis that glycemic thresholds for cognitive dysfunction during hypoglycemia, like those for autonomic and symptomatic responses, shift to lower plasma glucose concentrations after recent antecedent hypoglycemia in patients with type 1 diabetes mellitus (T1DM), 15 patients were studied on two occasions. Cognitive functions were assessed during morning hyperinsulinemic stepped hypoglycemic clamps (85, 75, 65, 55, and 45 mg/dl steps) after, in random sequence, nocturnal (2330-0300) hypoglycemia (48 +/- 2 mg/dl) on one occasion and nocturnal euglycemia (109 +/- 1 mg/dl) on the other. Compared with nondiabetic control subjects (n = 12), patients with T1DM had absent glucagon (P = 0.0009) and reduced epinephrine (P = 0.0010), norepinephrine (P = 0.0001), and neurogenic symptom (P = 0.0480) responses to hypoglycemia; the epinephrine (P = 0.0460) and neurogenic symptom (P = 0.0480) responses were reduced further after nocturnal hypoglycemia. After nocturnal hypoglycemia, in contrast to nocturnal euglycemia, there was less deterioration of cognitive function overall (P = 0.0065) during hypoglycemia based on analysis of the sum of standardized scores (z-scores). There was relative preservation of measures of pattern recognition and memory (the delayed non-match to sample task, P = 0.0371) and of attention (the Stroop arrow-word task, P = 0.0395), but not of measures of information processing (the paced serial addition task) or declarative memory (the delayed paragraph recall task), after nocturnal hypoglycemia. Thus, glycemic thresholds for hypoglycemic cognitive dysfunction, like those for autonomic and symptomatic responses to hypoglycemia, shift to lower plasma glucose concentrations after recent antecedent hypoglycemia in patients with T1DM.     

Optimization of glycemic control by insulin therapy decreases the proportion of small dense LDL particles in diabetic patients

Small dense LDL particles (B phenotype) are considered to be more atherogenic than large buoyant LDL particles. The influence of glycemic control on LDL particle size and density is still under debate. The aim of this study was to determine LDL subfraction phenotype in both IDDM and NIDDM patients in poor glycemic control compared with that of respective matched control groups. In addition, we evaluated the effect of a 3-month period of optimized glycemic control on this parameter. Thirty-seven IDDM patients and 33 NIDDM patients, together with two respective age-, sex-, and BMI-matched control groups were studied. Non-A phenotype prevalence in IDDM patients before (19%) and after blood glucose optimization (11%) was similar to that of their control group (12%). However, NIDDM patients displayed a higher proportion of the non-A phenotype (51%) than did the control group (28%), but it became closer (30%, P < 0.05) after glycemic control improved. All subjects with non-A phenotype that changed to A phenotype showed triglyceride levels below 1.63 mmol/l and a greater decrease in HbA1c than did subjects whose phenotype did not change (4.9 +/- 1.5 vs. 3.1 +/- 1.4%, P < 0.05). A higher proportion of small dense LDL was observed in NIDDM women than in nondiabetic women (LDL5 10.0 +/- 4.8 vs. 6.3 +/- 1.5%, LDL6 6.1 +/- 2.2 vs. 4.2 +/- 0.8%, P < 0.05) during both stages of glycemic control, but no differences were observed between NIDDM and nondiabetic men. In conclusion, these findings provide new evidence for the relevance of near-normal glycemic control in the prevention of macrovascular disease and could contribute to an explanation of the loss of protection for cardiovascular disease in diabetic women.     

Preservation of physiological responses to hypoglycemia 2 days after antecedent hypoglycemia in patients with IDDM

OBJECTIVE: To assess the effects of short-term antecedent hypoglycemia on responses to further hypoglycemia 2 days later in patients with IDDM. RESEARCH DESIGN AND METHODS: We studied eight type I diabetic patients without hypoglycemia unawareness or autonomic neuropathy during two periods at least 4 weeks apart. On day 1, 2 h of either clamped hyperinsulinemic (60 mU.m-2.min-1) hypoglycemia at 2.8 mmol/l or euglycemia at 5.0 mmol/l were induced. Hyperinsulinemic hypoglycemia was induced 2 days later with 40 min glucose steps of 5.0, 4.0, 3.5, 3.0, and 2.5 mmol/l. Catecholamine levels and symptomatic and physiological responses were measured every 10-20 min. RESULTS: When compared with the responses measured following euglycemia, the responses of norepinephrine 2 days after hypoglycemia were reduced (peak, 1.4 +/- 0.4 [mean +/- SE] vs.1.0 +/- 0.3 nmol/l [P < 0.05]; threshold, 3.4 +/- 0.1 vs. 2.9 +/- 0.1 mmol/l glucose [P < 0.01]). The responses of epinephrine (peak, 4.0 +/- 1.4 vs. 3.5 +/- 0.8 nmol/l [P = 0.84]; threshold, 3.8 +/- 0.1 vs. 3.6 +/- 0.1 mmol/l glucose [P = 0.38]), water loss (peak, 194 +/- 34 vs. 179 +/- 47 g-1.m-2.h-1 [P = 0.73]; threshold, 2.9 +/- 0.2 vs. 2.9 +/- 0.2 mmol/l glucose [P = 0.90]), tremor (peak, 0.28 +/- 0.05 vs. 0.37 +/- 0.06 root mean square volts (RMS V) [P = 0.19]; threshold, 3.2 +/- 0.2 vs. 3.1 +/- 0.2 mmol/l glucose [P = 0.70]), total symptom scores (peak, 10.6 +/- 2.1 vs. 10.8 +/- 1.9 [P = 0.95]; threshold, 3.3 +/- 0.2 vs. 3.6 +/0 0.1 mmol/l glucose [P = 0.15]), and cognitive function (four-choice reaction time: threshold, 2.9 +/- 0.2 vs. 3.0 +/- 0.2 mmol/l glucose [P = 0.69]) were unaffected. CONCLUSIONS: The effect on hypoglycemic physiological responses of 2 h of experimental hypoglycemia lasts for 1-2 days in these patients with IDDM . The pathophysiological effect of antecedent hypoglycemia may be of shorter duration in IDDM patients, compared with nondiabetic subjects.     

Brief twice-weekly episodes of hypoglycemia reduce detection of clinical hypoglycemia in type 1 diabetes mellitus

We tested the hypothesis that as few as two weekly brief episodes of superimposed hypoglycemia (i.e., doubling the average frequency of symptomatic hypoglycemia) would reduce physiological and behavioral defenses against developing hypoglycemia and reduce detection of clinical hypoglycemia in patients with type 1 diabetes mellitus (T1DM). Compared with nondiabetic controls, six patients with well-controlled T1DM (HbA1c, 7.5 +/- 0.7% [mean +/- SD]) exhibited absent glucagon responses and reduced epinephrine (P = 0.0027), norepinephrine (P = 0.0007), pancreatic polypeptide (P = 0.0030), and neurogenic symptom (P = 0.0451) responses to hypoglycemia as expected. In these patients, 2 h of induced hypoglycemia (50 mg/dl, 2.8 mmol/l) twice weekly for 1 month, compared in a random-sequence crossover design with an otherwise identical 2 h of induced hyperglycemia (150 mg/dl, 8.3 mmol/l) twice weekly for 1 month, further reduced the epinephrine (P = 0.0001) and pancreatic polypeptide (P = 0.0030) responses, tended to further reduce the norepinephrine and neurogenic symptom responses to hypoglycemia, and reduced cognitive dysfunction during hypoglycemia (P = 0.0271), all assessed in the investigational setting. In the clinical setting, induced hypoglycemia did not alter overall glycemic control, but did reduce the total number of symptomatic hypoglycemic episodes detected by the patients from 49 to 30 per month and lowered the mean +/- SE self-monitored blood glucose level during symptomatic hypoglycemia from 51 +/- 2 mg/dl (2.8 +/- 0.1 mmol/l) to 46 +/- 3 mg/dl (2.6 +/- 0.2 mmol/l) (P < 0.01). It also reduced the proportion of low regularly scheduled self-monitored values that were symptomatic by approximately 33%. Thus as little as doubling the frequency of symptomatic hypoglycemia further reduced both the key epinephrine response and clinical awareness of developing hypoglycemia, changes reasonably expected to increase the risk of severe iatrogenic hypoglycemia in T1DM.     

Decreased epinephrine responses to hypoglycemia during sleep

BACKGROUND: In patients with type I diabetes mellitus, hypoglycemia occurs commonly during sleep and is frequently asymptomatic. This raises the question of whether sleep is associated with reduced counterregulatory-hormone responses to hypoglycemia. METHODS: We studied the counterregulatory-hormone responses to insulin-induced hypoglycemia in eight adolescent patients with type I diabetes and six age-matched normal subjects when they were awake during the day, asleep at night, and awake at night. In each study, the plasma glucose concentration was stabilized for 60 minutes at approximately 100 mg per deciliter (5.6 mmol per liter) and then reduced to 50 mg per deciliter (2.8 mmol per liter) and maintained at that concentration for 40 minutes. Plasma free insulin, epinephrine, norepinephrine, cortisol, and growth hormone were measured frequently during each study. Sleep was monitored by polysomnography. RESULTS: The plasma glucose and free insulin concentrations were similar in both groups during all studies. During the studies when the subjects were asleep, no one was awakened during the hypoglycemic phase, but during the final 30 minutes of the studies when the subjects were awake both the patients with diabetes and the normal subjects had symptoms of hypoglycemia. In the patients with diabetes, plasma epinephrine responses to hypoglycemia were blunted when they were asleep (mean [+/-SE] peak plasma epinephrine concentration, 70+/-14 pg per milliliter [382+/-76 pmol per liter]; P=0.3 for the comparison with base line), as compared with when they were awake during the day or night (238+/-39 pg per milliliter [1299+/-213 pmol per liter] P=0.004 for the comparison with base line, and 296+/-60 pg per milliliter [1616+/-327 pmol per liter], P=0.004, respectively). The patients' plasma norepinephrine responses were also reduced during sleep, whereas their plasma cortisol concentrations did not increase and their plasma growth hormone concentrations increased slightly. The patterns of counterregulatory-hormone responses in the normal subjects were similar. CONCLUSIONS: Sleep impairs counterregulatory-hormone responses to hypoglycemia in patients with diabetes and normal subjects.     

Lack of association between the Gly40Ser polymorphism in the glucagon receptor gene and NIDDM in Finland

A heterozygous polymorphism changing GGT40 (Gly) to AGT40 (Ser) (Gly40Ser) in the glucagon receptor gene was reported to be associated with non-insulin-dependent diabetes mellitus (NIDDM). A possible involvement of this polymorphism in impaired glucose tolerance was also suggested in a French population. To replicate this finding we screened 311 unrelated NIDDM patients, 101 unrelated individuals with impaired glucose tolerance and 306 control subjects for the presence of the Gly40Ser polymorphism by use of polymerase chain reaction-restriction fragment length polymorphism in a Finnish population. None of the NIDDM or impaired glucose tolerant patients had this polymorphism. Instead, four of the control subjects (1.3%) were heterozygous carriers of the polymorphism (NS). The age, body mass index, 2-h blood glucose level, 2-h insulin level, and incremental insulin are of the four subjects with this polymorphism were similar to those of the control subjects homozygous for the wild type. Taken together, the data do not support the suggested involvement of the Gly40Ser polymorphism in impaired glucose tolerance and the hypothesis of an association between NIDDM and the glucagon receptor gene in this population.     

Different change in lipoprotein(a) levels from lipid levels of other lipoproteins with improved glycemic control in patients with NIDDM

OBJECTIVE: To evaluate change both in lipoprotein(a) [Lp(a)] and lipid levels in other lipoproteins in non-insulin-dependent diabetes mellitus (NIDDM) after short-term improvement of glycemic control. RESEARCH DESIGN AND METHODS: We compared Lp(a) levels in 210 NIDDM patients with those in 46 control subjects and evaluated the relationship between glycemic control and Lp(a) levels in diabetic patients. In addition, changes in Lp(a) levels and lipid levels were assessed after the improvement of glycemic control in 54 poorly controlled NIDDM patients. RESULTS: In NIDDM, Lp(a) levels in all patients, 62 patients with HbA1c < 6.0%, and 75 patients with HbA1c between 6.0 and 8.0%, were significantly higher than those in control subjects (19.1 [1.7-106.6], 19.2 [6.0-106.6], and 20.3 [2.7-75.3] vs. 15.4 [2.0-61.7] mg/dl, median [range], P < 0.05). Lp(a) levels in 73 patients with HbA1c of > or = 8.0% (18.7 [1.7-58.8] mg/dl) were not significantly different from those in control subjects. After glycemic control, lipid levels in plasma and in other lipoproteins fell significantly, but Lp(a) did not change (from 18.3 [1.7-58.8] to 18.4 [6.6-95.3] mg/dl). Changes in lipid levels, including Lp(a), did not correlate with those in fasting plasma glucose or HbA1c. CONCLUSIONS: These results suggest that elevated Lp(a) levels do not reflect poor glycemic control and that Lp(a) levels are independent of lipid levels in other lipoproteins after improved glycemic control in NIDDM.     

Pulmonary stenosis: defect-specific diagnostic accuracy of heart murmurs in children

The role of the adrenals in the polycystic ovary syndrome (PCOS) is debated. Both single steroid-converting enzyme abnormalities and increased adrenal activity have received support. The conventional Synacthen test using pharmacological doses of ACTH results in unphysiological levels of ACTH. Therefore, we used insulin-induced hypoglycemia (0.15 IU/kg BW) to asses the responses of ACTH, cortisol, pregnenolone, 17-hydroxypregnenolone, dehydroepiandrosterone, progesterone, 17-hydroxyprogesterone, and androstenedione in 18 women with PCOS and in 17 normal women of similar age and body mass index. The blood glucose concentration at 30 min was 2 mmol/L or less in all women, i.e. well below the threshold of the hormonal counterregulatory response. The women with PCOS showed a lower ACTH response, expressed as the maximum increment above basal [mean (95% confidence interval): PCOS, 11.1 (6.9-15.3); controls, 19.9 (13.8-26) pmol/L; P < 0.05], but a quantitatively comparable [PCOS, 207.2 (148.5-266.5); controls, 167.1 (100.6-233.2) nmol/L; P = NS] and more prompt cortisol response than the controls (by chi2 test, P < 0.05), resulting in a higher molar ratio between the maximum increments of cortisol and ACTH [PCOS, 13.9 (8.7-19); controls, 8.8 (5.7-12); P < 0.05]. The women with PCOS did, however, show a more rapid decline in cortisol levels than the controls (P < 0.05 at 120 and 180 min). The responses of the androgens and intermediate adrenal steroids were similar in women with PCOS and controls. The findings suggest an adaptation to increased adrenal reactivity to endogenous ACTH in women with PCOS. Exposure to hypoglycemia as a model of stress was not followed by hypersecretion of adrenal androgens and revealed no signs of steroid enzyme disturbances in women with PCOS.     

Effect of continuous subcutaneous insulin infusion with lispro on hepatic responsiveness to glucagon in type 1 diabetes

OBJECTIVE: People with type 1 diabetes frequently develop a blunted counterregulatory hormone response to hypoglycemia coupled with a decreased hepatic response to glucagon, and consequently, they have an increased risk of severe hypoglycemia. We have evaluated the effect of insulin lispro (Humalog) versus regular human insulin (Humulin R) on the hepatic glucose production (HGP) response to glucagon in type 1 diabetic patients on intensive insulin therapy with continuous subcutaneous insulin infusion (CSII). RESEARCH DESIGN AND METHODS: Ten subjects on CSII were treated for 3 months with lispro and 3 months with regular insulin in a double-blind randomized crossover study After 3 months of treatment with each insulin, hepatic sensitivity to glucagon was measured in each subject. The test consisted of a 4-h simultaneous infusion of somatostatin (450 microg/h) to suppress endogenous glucagon, regular insulin (0.15 mU x kg(-1) x min(-1)), glucose at a variable rate to maintain plasma glucose near 5 mmol/l, and D-[6,6-2H2]glucose to measure HGP During the last 2 h, glucagon was infused at 1.5 ng x kg(-1) x min(-1). Eight nondiabetic people served as control subjects. RESULTS: During the glucagon infusion period, free plasma insulin levels in the diabetic subjects were 71.7+/-1.6 vs. 74.8+/-0.5 pmol/l after lispro and regular insulin treatment, with plasma glucagon levels of 88.3+/-1.8 and 83.7+/-1.5 ng/l for insulin:glucagon ratios of 2.8 and 3.0. respectively (NS). However, plasma glucose increased to 9.2+/-1.1 mmo/l after lispro insulin compared with 7.1+/-0.9 mmol/l after regular insulin (P < 0.01), and the rise in HGP was 5.7 +/-2.8 micromol x kg(-1) x min(-1) after lispro insulin versus 3.1+/-2.9 micromol x kg(-1) x min(-1) after regular insulin treatment (P=0.02). In the control subjects, HGP increased by 10.7+/-4.2 micromol x kg(-1) x min(-1) under glucagon infusion. CONCLUSIONS: Insulin lispro treatment by CSII was associated with a heightened response in HGP to glucagon compared with regular human insulin. This suggests that insulin lispro increases the sensitivity of the liver to glucagon and could potentially decrease the risk of severe hypoglycemia.     

Enhanced neuroendocrine response to insulin tolerance test performed under increased ambient temperature

The hypothesis that an increase in ambient temperature modulates neuroendocrine response in clinically used provocative pituitary function tests was verified. Healthy male volunteers were subjected to insulin tolerance tests in two randomized trials. In the first trial hypoglycemia was induced by a bolus injection of insulin (0.1 U per kg of BW, i.v.) at room temperature. In the second trial, the subjects were exposed to increased ambient temperature for 45 min before insulin injection and for 45 min thereafter. The environmental temperature was selected to increase body temperature less than 1C. Under conditions of increased temperature basal hormone levels as measured in antecubital venous blood samples failed to be modified and the hypoglycemia was less severe. Nevertheless, the responses of most (beta-endorphin, ACTH, prolactin, catecholamines), but not all (growth hormone, cortisol), hormones to hypoglycemia were exaggerated. The remarkable increase in ACTH and beta-endorphin release was not accompanied by concomitant increase of plasma cortisol response. The sympathetic-adrenomedullary system was significantly activated, which was manifested particularly by enhanced norepinephrine release. Growth hormone response to hypoglycemia was not modified, while that of prolactin was enhanced. Thus during evaluation of neuroendocrine function under clinical conditions, changes in ambient and body temperature should not be underestimated.