Cardiac responses to insulin-induced hypoglycemia in nondiabetic and intensively treated type 1 diabetic patients.

Insulin-induced hypoglycemia occurs commonly in intensively treated patients with type 1 diabetes, but the cardiovascular consequences of hypoglycemia in these patients are not known. We studied left ventricular systolic [left ventricular ejection fraction (LVEF)] and diastolic [peak filling rate (PFR)] function by equilibrium radionuclide angiography during insulin infusion (12 pmol. kg(-1). min(-1)) under either hypoglycemic (approximately 2.8 mmol/l) or euglycemic (approximately 5 mmol/l) conditions in intensively treated patients with type 1 diabetes and healthy nondiabetic subjects (n = 9 for each). During hypoglycemic hyperinsulinemia, there were significant increases in LVEF (DeltaLVEF = 11 +/- 2%) and PFR [DeltaPFR = 0.88 +/- 0.18 end diastolic volume (EDV)/s] in diabetic subjects as well as in the nondiabetic group (DeltaLVEF = 13 +/- 2%; DeltaPFR = 0.79 +/- 0.17 EDV/s). The increases in LVEF and PFR were comparable overall but occurred earlier in the nondiabetic group. A blunted increase in plasma catecholamine, cortisol, and glucagon concentrations occurred in response to hypoglycemia in the diabetic subjects. During euglycemic hyperinsulinemia, LVEF also increased in both the diabetic (DeltaLVEF = 7 +/- 1%) and nondiabetic (DeltaLVEF = 4 +/- 2%) groups, but PFR increased only in the diabetic group. In the comparison of the responses to hypoglycemic and euglycemic hyperinsulinemia, only the nondiabetic group had greater augmentation of LVEF, PFR, and cardiac output in the hypoglycemic study (P < 0.05 for each). Thus intensively treated type 1 diabetic patients demonstrate delayed augmentation of ventricular function during moderate insulin-induced hypoglycemia. Although diabetic subjects have a more pronounced cardiac response to hyperinsulinemia per se than nondiabetic subjects, their response to hypoglycemia is blunted.

Oral glucose augments the counterregulatory hormone response during insulin-induced hypoglycemia in humans.

It has been suggested that the counterregulatory hormone (CRH) response to acute hypoglycemia is triggered via glucose sensors situated in either the hypothalamus or the portohepatic area. If the latter were critical during hypoglycemia, one would anticipate that ingestion of glucose, by raising glucose levels in the portal circulation, should attenuate CRH responses previously described in animal studies. To evaluate the effect of raising portal, but not peripheral, glucose levels during insulin-induced hypoglycemia, we performed hypoglycemic clamp studies in five healthy adult males on two occasions. On one occasion, subjects received oral glucose (OG) (25 g) during hypoglycemia; and on one occasion, noncarbohydrate-containing drink of equal volume, while maintaining plasma glucose at 55 +/- 2 mg/dL (3.08 mmol/L). As a result, there were no significant differences in systemic plasma glucose levels between the two hypoglycemic clamp studies, and basal CRH concentrations were also similar. As expected, there was a brisk rise in all CRH during the control (hypoglycemia+noncarbohydrate drink) study. In the experimental study, administration of OG (hypoglycemia+OG), to raise intraportal glucose levels during systemic hypoglycemia, did not attenuate CRH responses. Indeed, OG enhanced the rise in epinephrine, glucagon, and GH. Increases in cortisol and norepinephrine did not differ between the two studies. Therefore, our data suggest that increasing the level of glucose in the portal vein above that in the systemic circulation, during hypoglycemia, enhances (rather than suppresses) CRH responses. Thus, ingestion of glucose may reverse hypoglycemia directly by provision of substrate, as well as indirectly by stimulating counteregulatory mechanisms.

Impaired in vivo stimulation of lipolysis in adipose tissue by selective beta2-adrenergic agonist in obese adolescent girls.

Studies performed in adults with long-standing obesity suggest a reduced lipolytic sensitivity to catecholamines in subcutaneous abdominal adipose tissue (AT). We used microdialysis to study the in situ lipolytic effects of dobutamine (selective beta1-agonist) and terbutaline (selective beta2-agonist) on glycerol release (lipolytic index) in abdominal subcutaneous AT in 10 obese girls aged 13-17 years, BMI 38 +/- 2.1 kg/m2, and in 7 lean girls aged 11-17 years, BMI 21 +/- 1.1 kg/m2, and compared them with 10 obese women aged 21-39 years, BMI 36 +/- 1.6 kg/m2, and 10 lean women aged 18-42 years, BMI 21 +/- 0.4 kg/m2. Terbutaline at 10(-6) mol/l stimulated glycerol release more efficiently in lean girls than in obese girls (peak response approximately 350 vs. 150% of control, P < 0.01). At the lower concentration of agonist, no significant difference was seen. In women, terbutaline was more effective in lean than in obese women in stimulating glycerol release at both 10(-8) mol/l (peak response lean approximately 175% vs. obese 125% of control) and 10(-6) mol/l (approximately 300 vs. 150% of control, P < 0.05). No significant difference in glycerol release between obese and lean girls or women was detected with selective beta1-stimulation. Our data demonstrate a specific impairment in the capacity of beta2-adrenergic agonists to promote lipolysis in subcutaneous abdominal adipose tissue of obese adolescent girls and women. Thus, decreased mobilization of fat during activation of the adrenergic system might be present early in the development of adolescent obesity.

Efficacy and metabolic effects of metformin and troglitazone in type II diabetes mellitus.

BACKGROUND: Combination therapy is logical for patients with non-insulin-dependent (type 2) diabetes mellitus, because they often have poor responses to single-drug therapy. We studied the efficacy and physiologic effects of metformin and troglitazone alone and in combination in patients with type 2 diabetes. METHODS: We randomly assigned 29 patients to receive either metformin or troglitazone for three months, after which they were given both drugs for another three months. Plasma glucose concentrations during fasting and postprandially and glycosylated hemoglobin values were measured periodically during both treatments. Endogenous glucose production and peripheral glucose disposal were measured at base line and after three and six months. RESULTS: During metformin therapy, fasting and postprandial plasma glucose concentrations decreased by 20 percent (58 mg per deciliter [3.2 mmol per liter], P<0.001) and 25 percent (87 mg per deciliter [4.8 mmol per liter], P<0.001), respectively. The corresponding decreases during troglitazone therapy were 20 percent (54 mg per deciliter [2.9 mmol per liter], P=0.01) and 25 percent (83 mg per deciliter [4.6 mmol per liter], P<0.001). Endogenous glucose production decreased during metformin therapy by a mean of 19 percent (P=0.001), whereas it was unchanged by troglitazone therapy (P=0.04 for the comparison between groups). The mean rate of glucose disposal increased by 54 percent during troglitazone therapy (P=0.006) and 13 percent during metformin therapy (P= 0.03 for the comparison within the group and between groups). In combination, metformin and troglitazone further lowered fasting and postprandial plasma glucose concentrations by 18 percent (41 mg per deciliter [2.3 mmol per liter], P=0.001) and 21 percent (54 mg per deciliter [3.0 mmol per liter], P<0.001), respectively, and the mean glycosylated hemoglobin value decreased 1.2 percentage points. CONCLUSIONS: Metformin and troglitazone have equal and additive beneficial effects on glycemic control in patients with type 2 diabetes. Metformin acts primarily by decreasing endogenous glucose production, and troglitazone by increasing the rate of peripheral glucose disposal.

Counterregulation in peripheral tissues: effect of systemic hypoglycemia on levels of substrates and catecholamines in human skeletal muscle and adipose tissue.

We used microdialysis to distinguish the effects of hyperinsulinemia and hypoglycemia on glucose, gluconeogenic substrate, and catecholamine levels in adipose and muscle extracellular fluid (ECF). Ten lean humans (six males and four females) were studied during baseline and hyperinsulinemic (3 mU x kg-1 x min-1 for 3 h) euglycemia (5.0 mmol/l) and hypoglycemia (2.8 mmol/l). In muscle and adipose, basal ECF glucose was lower (muscle, 3.5 +/- 0.2 mmol/l; adipose tissue, 3.3 +/- 0.2 mmol/l) and lactate was higher (muscle, 2.2 +/- 0.2 mmol/l; adipose, 1.5 +/- 0.3 mmol/l) than respective plasma values (glucose, 4.9 +/- 0.1 mmol/l; lactate, 0.7 +/- 0.1 mmol/l), whereas alanine was higher in muscle ECF (379 +/- 22 micromol/l) than adipose tissue (306 +/- 22 micromol/l) and plasma (273 +/- 33 micromol/l). Plasma catecholamines (unchanged during euglycemia) rose during hypoglycemia with epinephrine, increasing approximately fivefold more than norepinephrine. In contrast, the hypoglycemia-induced increments in muscle dialysate norepinephrine and epinephrine were similar, suggesting local generation of norepinephrine. Compared with euglycemia, hypoglycemia produced a greater increase in lactate and a smaller reduction in alanine in muscle ECF, whereas hypoglycemia caused a greater relative fall in ECF glucose concentrations in muscle (72 +/- 16%) and adipose tissue (69 +/- 9%) than in plasma (42 +/- 3%) (P < 0.05). We conclude that hypoglycemia increases the generation of norepinephrine and gluconeogenic substrates in key target tissues, while increasing the plasma-tissue concentration gradient for glucose. These changes suggest the stimulation of glucose extraction by peripheral tissues, despite systemic counterregulatory hormone release and local sympathetic activation.

Co-existence of severe insulin resistance and hyperinsulinaemia in pre-adolescent obese children.

To determine the time course of changes in insulin action and secretion that occur early during the development of obesity, we studied children before the onset of puberty. The reason for choosing the prepubertal stage of development is that it is metabolically characterized by both a high sensitivity to insulin and low glucose stimulated insulin responses. Fifteen obese preadolescents (8 male/7 female, age 10 +/- 0.4 years, body mass index (BMI) 31 +/- 1.2 kg/m2 Tanner Stage I) with a duration of obesity of less than 5 years and 10 non-obese preadolescents (6 male/4 female, age 10 +/- 0.4 years, BMI 18 +/- 0.9 kg/m2) matched for gender were studied. In a cross-sectional analysis, we compared responses in obese preadolescents, with those in obese adolescents and obese adults with a longer duration of obesity. The euglycaemic hyperinsulinaemic clamp with 1-13C-glucose (Hot Ginf) and indirect calorimetry were used to quantitate insulin action and the hyperglycaemic clamp used to assess beta-cell function. Insulin-stimulated glucose uptake measured at two physiological levels of hyperinsulinaemia (approximately 180 and 480 pmol) was reduced by 20 and 45% in all three groups of obese compared to non-obese subjects (p < 0.01). Defects in oxidative and non-oxidative glucose metabolism were observed in all three groups of obese subjects at the higher insulin infusion rate. The ability of insulin to inhibit lipid oxidation was impaired in all three obese groups at both levels of hyperinsulinaemia. Increases in basal and glucose-stimulated insulin levels during the hyperglycaemic clamp mirrored the reductions in glucose uptake during the insulin clamp in all obese groups. These results indicate that insulin resistance and hyperinsulinaemia co-exist in preadolescent children with moderate to severe obesity.

Interstitial fluid concentrations of glycerol, glucose, and amino acids in human quadricep muscle and adipose tissue. Evidence for significant lipolysis in skeletal muscle.

To determine the relationship between circulating metabolic fuels and their local concentrations in peripheral tissues we measured glycerol, glucose, and amino acids by microdialysis in muscle and adipose interstitium of 10 fasted, nonobese human subjects during (a) baseline, (b) euglycemic hyperinsulinemia (3 mU/kg per min for 3 h) and, (c) local norepinephrine reuptake blockade (NOR). At baseline, interstitial glycerol was strikingly higher (P < 0.0001) in muscle (3710 microM) and adipose tissue (2760 microM) compared with plasma (87 microM), whereas interstitial glucose (muscle 3.3, fat 3.6 mM) was lower (P < 0.01) than plasma levels (4.8 mM). Taurine, glutamine, and alanine levels were higher in muscle than in adipose or plasma (P < 0.05). Euglycemic hyperinsulinemia did not affect interstitial glucose, but induced a fall in plasma glycerol and amino acids paralleled by similar changes in the interstitium of both tissues. Local NOR provoked a fivefold increase in glycerol (P < 0.001) and twofold increase in norepinephrine (P < 0.01) in both muscle and adipose tissues. To conclude, interstitial substrate levels in human skeletal muscle and adipose tissue differ substantially from those in the circulation and this disparity is most pronounced for glycerol which is raised in muscle as well as adipose tissue. In muscle, insulin suppressed and NOR increased interstitial glycerol concentrations. Our data suggest unexpectedly high rates of intramuscular lipolysis in humans that may play an important role in fuel metabolism.

Effect of caffeine on the recognition of and responses to hypoglycemia in humans.

OBJECTIVE: To determine whether two effects of acute caffeine ingestion--decrease in cerebral blood flow and increase in brain glucose use--alter the recognition of and physiologic responses to hypoglycemia. DESIGN: On two occasions, a hyperinsulinemic glucose clamp technique (2 mU/kg body weight per minute) was used to maintain plasma glucose at 5 mmol/L for 90 minutes, followed by 60 minutes at 3.8 mmol/L, and then 2.8 mmol/L. After 30 minutes at 5 mmol/L, participants consumed, using a randomized, double-blind design, caffeine-free cola with or without caffeine (400 mg) added. SETTING: Yale Clinical Research Center. PARTICIPANTS: Eight healthy, nonobese volunteers (5 men; age range, 20 to 33 years). MEASUREMENTS: Middle cerebral artery velocity (V MCA), counter-regulatory hormone levels, hypoglycemic symptoms, and cognitive function (P300 evoked potentials). RESULTS: Caffeine caused an immediate and sustained 23% decrease in VMCA from 64 to 49 cm/s (point estimate of difference, +15 cm/s [95% CI, 10 to 21 cm/s], P < 0.001). At a glucose level of 3.8 mmol/L, only the participants given caffeine had warning symptoms and "felt hypoglycemic." Moreover, the level of epinephrine was 118% ([CI of point difference, 76% to 158%] [CI, P < 0.001]) higher after caffeine consumption compared with placebo. Similarly, levels of norepinephrine (41% [CI, 26% to 60%], P < 0.002), cortisol (65% [CI, 26% to 78%], P < 0.008), and growth hormone (60% [CI, 16% to 143%], P < 0.05) were higher after caffeine consumption compared with placebo. At 2.8 mmol/L, epinephrine (40% [point estimate of the percentage difference], P < 0.05), norepinephrine (27%, P < 0.05), and cortisol (24%, P < 0.05) levels were higher, participants were more aware (P < 0.02) of hypoglycemia, and P300 latency was prolonged in the group that consumed caffeine (7.2%, P < 0.05). CONCLUSIONS: Acute ingestion of caffeine is associated with sympathoadrenal activation and awareness of hypoglycemia at a glucose level not usually considered hypoglycemic. Our data suggest that individuals who ingest moderate amounts of caffeine may develop hypoglycemic symptoms if plasma glucose levels fall into the "low-normal" range, as might occur in the late postprandial period after ingestion of a large carbohydrate load.

Effect of insulin-like growth factor-1 on the responses to and recognition of hypoglycemia in humans. A comparison with insulin.

Recombinant human insulin-like growth factor-1 (rhIGF-1) lowers blood glucose in humans but its effect on counterregulatory responses has not been established. We therefore compared infusions of rhIGF-1 (0.7 micrograms/kg per min) and insulin (0.8 mU/kg.min) for 120 min in 10 healthy volunteers (glucose allowed to fall freely). With both, glucose fell rapidly because of stimulation of glucose uptake and suppression of hepatic glucose production. Despite similar plasma glucose nadirs (2.6 +/- 0.1 vs. 2.7 +/- 0.1 mM), the glucagon response was absent (P < 0.005), growth hormone release was attenuated (P < 0.03), and norepinephrine levels were increased (P < 0.05) by rhIGF-1 compared with insulin. Absent glucagon responses were associated with a blunting of the rebound increase in glucose production (P < 0.05 vs. insulin). After stopping the infusions, glucose recovery was delayed with rhIGF-1 (P < 0.001 vs. insulin). To further evaluate the effects of rhIGF-1 during a standard hypoglycemic stimulus, eight additional healthy subjects received rhIGF-1 or insulin while glucose was clamped at 2.8 mM. Again the rise in glucagon during insulin-induced hypoglycemia was totally abolished by rhIGF-1. Growth hormone responses were delayed, whereas increases in norepinephrine, heart rate, and symptomatic awareness of hypoglycemia were greater with rhIGF-1 compared with insulin (P < 0.05). It was concluded that rhIGF-1 suppression of glucagon release during hypoglycemia impairs glucose recovery. Paradoxically, awareness of hypoglycemia is enhanced with rhIGF-1 in part due to stimulation of the sympathetic activity.