Determination and kinetic analysis of non-insulin mediated glucose uptake in Type 1 (insulin-dependent) diabetes mellitus

Summary In man, total glucose uptake is the sum of insulin mediated glucose uptake and non-insulin mediated glucose uptake. The latter pathway has not been examined in Type 1 (insulin-dependent) diabetes mellitus. In order to assess non-insulin mediated glucose uptake in Type 1 diabetes, we measured steady-state rates of glucose uptake during glucose clamps at 5.27, 9.71 and 12.5 mmol/l using low (0.25 mU· kg^–1·min^–1), intermediate (0.75 mU·kg^–1·min^–1) and high (1.50 mU·kg^–1·min^–1) insulin infusion rates in 10 subjects with Type 1 diabetes. For insulin infusion rates of 0.25, 0.75 and 1.50 mU·kg^–1·min^–1 as plasma glucose rose from 5.27 to 9.71 mmol/l, total glucose uptake increased by 35, 43 and 52 percent respectively (p<0.05 for each insulin infusion rate). For all three insulin infusion rates, there was no significant increase in total glucose uptake as plasma glucose increased from 9.71 to 12.5 mmol/l. At each glycaemic level, glucose uptake correlated significantly with plasma free insulin (r=0.81, p<0.01 at 5.71 mmol/l; r=0.84, p<0.01 at 9.71 mmol/l; r=0.73, p<0.02 at 12.5 mmol/l). Linear regression analysis to a point corresponding to plasma free insulin equalling zero, yielded values for non-insulin mediated glucose uptake (mmol·kg^–1·min^–1) of 0.11,0.14,0.18 at plasma glucose of 5.27, 9.7 and 12.5 mmol/l respectively. Thus, increasing plasma glucose concentrations were associated with increasing rates of non-insulin mediated glucose uptake. For each insulin infusion rate used, the percent of total glucose uptake accounted for by non-insulin mediated glucose uptake remained independent of plasma glucose concentration, but decreased as insulin infusion rate increased. During the insulin infusion at 0.25 mU·kg^–1·min^–1, this percentage ranged from 83.7 to 91.4%. Analysis of glucose uptake data derived for theoretical plasma insulin levels of 0, 40, 80 and 160 U/ml yielded linear Eadie-Hofstee plots (r=– 0.83 to – 0.99), suggesting that insulin increased Vmax but did not alter Km. Hence, in these subjects with Type 1 diabetes, glucose uptake, both insulin mediated and non-insulin mediated can be described by Michaelis-Menten kinetics. Comparison of values obtained for Vmax and Km in the present studies of Type 1 diabetes with those obtained from non-diabetic subjects indicates that non-insulin dependent glucose uptake in Type 1 diabetes is quantitatively similar to that of non-diabetic subjects.     

Contribution of glucose/glucose 6-phosphate cycle activity to insulin resistance in Type 2 (non-insulin-dependent) diabetes mellitus

Summary It has been suggested that increased glucose/glucose 6-phosphate substrate cycling impairs net hepatic glucose uptake in Type 2 (non-insulin-dependent) diabetes mellitus and contributes to hyperglycaemia. To investigate glucose/glucose 6-phosphate cycle activity and insulin action in Type 2 diabetes we studied eight patients and eight healthy control subjects, using the euglycaemic glucose clamp and isotope dilution techniques with purified [2-³H]- and [6-³H] glucose tracers, in the post-absorptive state and eight patients and five healthy control subjects during consecutive insulin infusions at rates of 0.4 and 2.0 mU·kg^–1·min^–1. [2-³H]glucose and [6-³H]glucose radioactivity in plasma samples were determined using selective enzymatic detritiation, allowing calculation of glucose turnover rates for each isotope, the difference being glucose/glucose 6-phosphate cycling. Endogenous glucose production ([6-³H]glucose) was greater in diabetic than control subjects in the post-absorptive state (15.6±1.5 vs 11.3±0.4 mol·kg^–1·min^–1, p<0.05) and during the 0.4 mU insulin infusion (10.1±1.3 vs 5.2±0.3 mol·kg^–1·min^–1, p<0.01) indicating hepatic insulin resistance. Glucose/glucose 6-phosphate cycling was significantly greater in diabetic than in control subjects in the post-absorptive state (2.6±0.4 vs 1.6±0.2 mol·kg^–1·min^–1, p<0.05) but not during the 0.4 mU insulin infusion (2.0±0.4 vs 2.0±0.3 mol·kg^–1·min^–1). During the 2.0 mU insulin infusion endogenous glucose production was suppressed to a similar degree in both groups (2.6±0.5 vs 3.4±0.7 mol · kg^–1·min^–1) but glucose disappearance was lower in the diabetic subjects (30.8±2.0 vs 52.4±4.6 mol·kg^–1·min^–1, p<0.01). During the 2.0 mU insulin infusion glucose/glucose 6-phosphate cycling was greater in the diabetic subjects (3.8±0.7 vs 0.8±0.6 mol·kg^–1·min^–1, p<0.05). In conclusion, both hepatic and peripheral insulin action are impaired in Type 2 diabetes. Increased glucose/glucose 6-phosphate cycling is seen in the post-absorptive state and also during marked hyperinsulinaemia, when insulin resistance is predominantly due to reduced peripheral tissue glucose uptake.     

Impaired insulin action in newly diagnosed type 1 (insulin-dependent) diabetes mellitus

Summary Hepatic and peripheral insulin sensitivity were investigated in five newly diagnosed Type 1 (insulin-dependent) diabetic subjects before and after 1 week of twice daily insulin therapy. Eight weight-matched control subjects were also studied. Hepatic glucose production and glucose utilization were measured basally and during two sequential 2-h insulin (25 and 40 mU· kg^–1· h^–1)/glucose infusion periods. In the untreated hyperglycaemic diabetic patients hepatic glucose production was 16.3±2.6, 8.1±1.1 and 3.6±2.8|mol· kg^–1· min^–1 respectively for each of the three periods (mean±SEM), and fell with treatment to 12.5±1.4, 0.5±0.5 and 0.5±0.5 mol· kg^–1· min^–1. Hepatic glucose production for normal subjects was 13.4±0.7, 2.3±0.8 and <0.1 mol-kg^–1· min^–1. Glucose utilization was 12.7±1.4,18.2±0.7 and 22.1±3.4mol· kg^–1· min^–1 before treatment in the diabetic subjects, and 11.8±1.7, 20.9±3.3 and 30.1±3.6 after treatment. These values compare with those in the euglycaemic control subjects (13.4±0.7, 18.7±1.6 and 36.3±2.7 mol · kg^–1· min^–1). The pre-treatment metabolic clearance rate of glucose in all diabetic studies with insulin levels >30mU/l was 2.6 ±0.4 and rose to 3.9 ±0.5 ml· kg^–1· min^–1 following insulin therapy. This was significantly lower than in the control subjects (6.7±0.8 ml· kg^–1 · min^–1; p<0.005). Basal nonesterified fatty acid levels were high in the untreated, but normal in the treated diabetic subjects, and fell in response to insulin infusion. Basal -hydroxybutyrate levels were high in both diabetic groups, but also fell in response to insulin infusion. Erythrocyte insulin receptor binding was normal in the untreated diabetic subjects, and was not changed by treatment. Therefore, treatment of newly diagnosed Type 1 diabetic subjects with insulin reverses the hepatic insensitivity to insulin. In contrast, treatment only partially improves peripheral glucose disposal. Since erythrocyte insulin receptor binding is normal, it is likely that a post-receptor defect in peripheral glucose metabolism exists in Type 1 diabetic patients despite insulin therapy and good diabetic control for a period of 1 week.     

Contrasting metabolic effects of continuous and pulsatile growth hormone administration in young adults with Type 1 (insulin-dependent) diabetes mellitus

Summary Plasma growth hormone profiles in adolescents with Type 1 (insulin-dependent) diabetes mellitus are characterized by both increases in pulse amplitude and higher baseline concentrations. To determine which of these abnormalities adversely affect metabolic control, we studied six young adults overnight on three occasions. On each night somatostatin (50–100 g·m^2–1·h^–1) and glucagon (1ng· kg^–1·min^–1) were infused continuously and 18mU/kg of growth hormone was given as either: three discrete pulses of 6 mU·kg^–1· h^–1 at 180-min intervals or a 12-h infusion (1.5 mU·kg^–1· h^–1) or buffer solution only on a control night. Euglycaemia was maintained by an insulin-varying clamp. Blood samples were taken every 15 min for glucose and growth hormone and every hour for intermediate metabolites and non-esterified fatty acids. Comparable normoglycaemic conditions were achieved on all three nights. Growth hormone levels achieved (mean±SEM) on study nights were: 32.8±2.2 mU/l (peak level during growth hormone pulses); 9.8± 0.8 mU/l (continuous growth hormone) and 1.1±0.3 mU/l (control level). Pulsatile growth hormone administration led to an increase in insulin requirements (mean±SEM: 0.17±0.03 vs control 0.09±0.01 mU·kg^–1· min^–1, p < 0.05) whereas insulin requirements following continuous growth hormone administration were unchanged. Cross-correlation confirmed an increase in insulin requirements occurring 135 min after a growth hormone pulse (r=0.21, p < 0.001). Growth hormone administration (continuous and pulsatile) led to a significant increase in B-hydroxybutyrate levels compared to the control night: 0.21±0.01 mmol/l (mean±SEM), 0.29±0.01 mmol/l, 0.08±0.01 mmol/l (p< 0.001) during the night with pulsatile growth hormone, continuous growth hormone and control respectively. Mean plasma non-esterified fatty acids were also increased following growth hormone administration: 0.94±0.04 mmol/l (mean±SEM), 1.09±0.07 mmol/l, 0.61±0.05 mmol/l (p<0.003), during the night with pulsatile growth hormone, continuous growth hormone and control respectively. It appears that the pulsatile and baseline growth hormone signals have contrasting metabolic effects in young adults with Type 1 diabetes mellitus.     

Transient insulin resistance following infusion of adrenaline in Type 1 (insulin-dependent) diabetes mellitus

Summary Insulin resistance was assessed after an intravenous infusion of adrenaline (50 ng·kg^–1·min^–1) or saline (control study) given between 08.00 and 08.30 hours in nine patients with Type 1 (insulin-dependent) diabetes mellitus. The blood glucose level during a somatostatin (100g/h)-insulin (0.4mU·kg^–1·min^–1)-glucose (4.5 mg·kg^–1·-min^–1)-infusion-test performed between 1030 and 14.30 hours served as an indicator of the total body insulin resistance. Blood glucose was maintained around 7 mmol/l between 08.00 and 10.30 hours by a constant infusion of regular insulin (0.57 mU·kg^–1· min^–1) and a variable infusion of a 20% glucose solution. The infusion of adrenaline raised plasma adrenaline to 2.7±0.3 nmol/l (mean±SEM) at the end of the infusion; thereafter it returned to its basal level within 30 min. The plasma levels of free insulin, glucagon, cortisol and growth hormone were similar in the adrenaline and the control studies from 08.00 to 14.30 hours. In comparison with the control study the infusion of adrenaline decreased the need for intravenous glucose significantly over the initial 2 h. Furthermore, during the somatostatin-insulin-glucose infusion test the blood glucose rose significantly (p<0.05) over the initial 2h; thereafter no significant differences between the two studies were seen. It is concluded that a short term infusion of adrenaline, resembling the adrenergic hormone response to hypoglycaemia, induces a diabetogenic effect which subsides within 6 h after omission of the adrenaline infusion.     

Decreased hepatic glucagon responses in Type 1 (insulin-dependent) diabetes mellitus

Summary The effect of glucagon infusion on hepatic glucose production during euglycaemia was evaluated in seven Type 1 (insulin-dependent) diabetic patients and in ten control subjects. In the diabetic subjects normoglycaemia was maintained during the night preceding the study by a variable intravenous insulin and glucose infusion. During the study endogenous insulin secretion was suppressed by somatostatin (450 g/h) and replaced by insulin infusion (0.15 mU·kg^–1·min^–1). ³H-glucose was infused for isotopic determination of glucose turnover. Plasma glucose was clamped at 5 mmol/1 for 2 h 30 min and glucagon (1.5 ng· kg^–1·min^–1) was then infused for the following 3 h. Hepatic glucose production and glucose utilisation were measured during the first, second and third hour of the glucagon infusion. Basal hepatic glucose production (just prior to glucagon infusion) was similar in diabetic (1.2±0.3 mg·kg^–1·min^–1) and control (1.6±0.1 mg·kg^–1·min^–1) subjects. In diabetic patients hepatic glucose production rose slowly to 2.1±0.5 mg·kg^–1·min^–1 during the first hours of glucagon infusion and stabilized at this level (2.4±0.5 mg·kg^–1·min^–1) in the third hour. In control subjects hepatic glucose production increased sharply to higher levels than in the diabetic subjects (3.4±0.3 mg·kg^–1·min^–1) during the first and second hour of glucagon infusion (p<0.05) and then gradually fell (2.9±0.4 mg·kg^–1·min^–1) during the third hour. In conclusion, when stimulated with glucagon at a physiologic plasma concentration diabetic patients had 1) an overall reduced hepatic glucose production response and 2) an abnormal sluggish response pattern. These abnormalities may imply inappropriate counter-regulation following a hypoglycaemic episode.     

Characterization of the insulin-antagonistic effect of growth hormone in man

Summary The insulin-antagonistic effect of growth hormone was characterized by infusing the hormone at three different infusion rates (6, 12 or 24 mU·kg^–1·min^–1) for one h in 11 healthy subjects. The insulin effect was measured with the euglycaemic clamp technique combined with D-(3-³H)-glucose infusion to evaluate glucose production and utilization. A control study with NaCl (154 mmol·l^–1) infusion was also performed. The insulin levels during the clamps were similar in all studies (36±0.2 mU·l^–1). Peak growth hormone levels were reached at 60 min (growth hormone 6mU·kg^–1·h^–1: 31±5; growth hormone 12 mU·kg^–1·h^–1: 52±4 and growth hormone 24 mU·kg^–1·h^–1: 102±8mU·l^–1). The insulin-antagonistic effect of growth hormone started after 2 h, was maximal after 4–5 h (39% inhibition of glucose infusion rate between control and growth hormone 24 mU·kg^–1·h^–1) and lasted for 6–7 h after peak levels. The resistance was due to a less pronounced insulin effect both to inhibit glucose production and to stimulate glucose utilization. Growth hormone infusion of 12 mU·kg^–1·h^–1 induced a similar insulin-antagonistic effect as the higher infusion rate whereas 6 mU·kg^–1·h^–1 induced a smaller response with a duration of 1 h between 3–4 h after peak levels of growth hormone. The present study demonstrates that growth hormone levels similar to those frequently seen in Type 1 (insulin-dependent) diabetic patients during poor metabolic control or hypoglycaemia, have pronounced insulin-antagonistic effects. The effects starts after about 2–3 h, is maximal after 4–5 h and lasts for about 6–7 h. Both duration and inhibitory effect of growth hormone are related to the plasma levels, where a maximal effect is seen at about 50 mU·l^–1 or higher.     

Peripheral and hepatic insulin sensitivity in subjects with impaired glucose tolerance

Summary Recent evidence suggests that the post-prandial hyperglycaemia in impaired glucose tolerance is primarily due to impaired suppression of basal hepatic glucose output. This in turn appears to be secondary to decreased first phase insulin secretion, although decreased hepatic insulin sensitivity, which is a feature of non-insulin-dependent diabetes mellitus, might also play a role. Eight mildly overweight subjects with impaired glucose tolerance and eight closely matched control subjects with normal glucose tolerance underwent an intravenous glucose tolerance test to assess first phase insulin secretion. Insulin sensitivity was examined by a 150-min hyperinsulinaemic-euglycaemic clamp. Somatostatin was infused from 150 min to suppress endogenous insulin secretion, and glucagon and insulin were replaced by constant infusion. Glucose with added dideuterated glucose (labelled infusion technique) was infused to maintain euglycaemia. First phase insulin secretion ( 0–10 min insulin area ÷ 0–10 min glucose area) was significantly decreased in the subjects with impaired glucose tolerance (median [range]: 1.2 [0.2–19.4] vs 9.1 [2.6–14.5] mU·mmol^–1; p<0.01). During the clamp, circulating insulin (93±8 [mean±SEM] and 81±10 mU·l^–1) and glucagon (54±4 and 44±6 ng·l^–1) levels were comparable. Total glucose disposal was decreased in subjects with impaired glucose tolerance (2.78±0.27 vs 4.47±0.53 mg·kg^–1·min^–1; p<0.02), and was primarily due to decreased non-oxidative glucose disposal. However, hepatic glucose output rates were comparable during the clamp (0.38±0.10 and 0.30±0.18 mg·kg^–1·min^–1). Therefore, the main defects in subjects with impaired glucose tolerance are decreased first phase insulin secretion and peripheral non-oxidative glucose disposal, but hepatic glucose output shows normal responsiveness to insulin.Abbreviations FPIS First phase insulin secretion - PG plasma glucose - NIDDM non-insulin-dependent diabetes mellitus - IGT impaired glucose tolerance - HGO hepatic glucose output - IVGTT intravenous glucose tolerance test - OGTT oral glucose tolerance test     

Acetyl-salicylic acid impairs insulin-mediated glucose utilization and reduces insulin clearance in healthy and non-insulin-dependent diabetic man

Summary The effect of acetyl-salicylic acid (ASA, 3 g per day for 3 days) on glucose utilization and insulin secretion was studied in healthy volunteers and Type 2 diabetic patients using the hyperglycaemic and euglycaemic insulin clamp technique. When in healthy subjects arterial plasma glucose was acutely raised and maintained at +7 mmol/l above fasting level, the plasma insulin response was enhanced by ASA (70±7 vs. 52±7mU/l), whereas the plasma C-peptide response was identical. Despite higher insulin concentrations, glucose utilization was not significantly altered (control, 61±7; ASA, 65±6mol·kg^–1·min^–1) indicating impairment of tissue sensitivity to insulin by ASA. Inhibition of prostaglandin synthesis was not likely to be involved in the effect of ASA, since insulin response and glucose utilization were unchanged following treatment with indomethacin. In the euglycaemic insulin (1 mU·kg^–1·min^–1) clamp studies, glucose utilization was unaltered by ASA despite higher insulin concentrations achieved during constant insulin infusion (103±4vs. 89±4mU/l). In Type 2 diabetic patients, fasting hyperglycaemia (10.6 ±1.1 mmol/l) and hepatic glucose production (15±2 mol·kg^–1·min^–1) fell upon ASA treatment (8.6±0.7 mmol/l; 13±1 mol·kg^–1· min^–1). During the hyperglycaemic clamp study, the plasma response of insulin, but not of C-peptide, was enhanced by ASA, whereas tissue sensitivity to insulin was reduced by 30 percent. It is concluded that in healthy and Type 2 diabetic man, ASA impairs tissue sensitivity to the action of insulin. This effect is counterbalanced by an augmented plasma insulin response to glucose, which results from a reduced insulin clearance rate. In Type 2 diabetic patients, the reduction in hepatic glucose production may be responsible for the amelioration of hyperglycaemia following ASA treatment.     

Differential effects of insulin therapy on hepatic and peripheral insulin sensitivity in Type 2 (non-insulin-dependent) diabetes

Summary Hepatic glucose production and metabolic clearance rate of glucose were measured using (3-³H) glucose at steady state, basally and during two sequential 2 h insulin (25 and 40mU · kg^–1 · h^–1)/glucose(2 and 3mg · kg^–1 · min^–1) infusion periods. Eight diabetic subjects were studied before and after 1 week of twice daily insulin therapy; six control subjects matched for age, weight and degree of obesity were also studied. In the diabetic patients, pre-treatment hepatic glucose production was 20.0 ± 2.2, 9.9 ± 2.9, and 1.4 ± 0.8 mol · kg^–1 · min^–1 respectively (± SEM) for each of the three periods, and fell significantly with treatment to 12.8 ± 1.7,4.0 ± 1.5 and 1.9 ± 1.0 mol · kg^–1 · min^–1. Hepatic glucose production in normal subjects was 13.2 ± 0.6, 2.2 ± 0.8 and < 1 mol · kg^–1 · min^–1. The pre-treatment metabolic clearance rate in all diabetic studies with insulin levels 30 mU/l was 1.10 ± 0.14 ml · kg^–1 · min^–1 and remained virtually unchanged following insulin therapy; this was significantly lower than in the control subjects (6.83 ± 1.02, p < 0.001). Basal non-esterified fatty acid levels were higher (p < 0.02) in the pre-treated diabetic patients compared to post-treated diabetic patients and control subjects. Non-esterified fatty acids in each group fell to similar levels during the insulin infusions, but the rate of fall was slower in the pre-treated diabetic patients. Insulin receptor binding to erythrocytes was normal in the diabetic subjects and unchanged by treatment. Therefore, following insulin treatment of uncontrolled Type 2 (non-insulin-dependent) diabetes, the initially increased basal hepatic glucose production, and decreased hepatic sensitivity, return towards normal. However, the glucose clearance remains low, despite good diabetic control, and appears to be a major factor in the continuing glucose intolerance. As insulin receptor binding is normal, the defect of glucose clearance in Type 2 diabetes appears compatible with a post-receptor defect of glucose metabolism.     

Insulin resistance in Type 1 (insulin-dependent) diabetes following hypoglycaemia — evidence for the importance of β-adrenergic stimulation

Summary The insulin effect, evaluated with the euglycaemic clamp technique, was studied before and after hypoglycaemia in 7 patients with Type 1 (insulin-dependent) diabetes. Following an initial 2 h clamp (clamp I) hypoglycaemia was induced and 2 h later a second clamp (clamp II), identical to the former, was performed. Each subject was studied twice; during infusion with saline (placebo) or propranolol. Glucose production and disposal were studied with the 3(³H)glucose technique. During placebo infusion, hypoglycaemia elicited an insulin resistance leading to approx. 50% reduction in the steady state glucose infusion rate during clamp II as compared to clamp I (clamp I 2.58±0.32, clamp II 1.26±0.08 mg·kg^–1·min^–1, p<0.02). The insulin resistance was prevented by infusing propranolol (clamp I 2.29±0.29, clamp II 2.85±0.56 mg·kg^–1·min^–1). The posthypoglycaemic insulin resistance was due to a less pronounced insulin effect on both glucose production (clamp I 0.29±0.21, clamp II 0.86±0.19 mg·kg^–1·min^–1, p<0.05) and glucose utilisation (clamp I 2.84±0.26, clamp II 2.13±0.23 mg·kg^–1·min^–1, p<0.05). The insulin resistance on both glucose production and utilisation was prevented by propranolol. Thus, the present study demonstrates that hypoglycaemia elicits a prolonged insulin resistance which is due to a less pronounced effect of insulin to both inhibit splanchnic glucose production and to stimulate peripheral glucose utilisation. The insulin resistance is due to -adrenergic stimulation and can be prevented by propranolol.     

Artificial induction of intravascular lipolysis by lipid-heparin infusion leads to insulin resistance in man

Summary Although extensive evidence indicates that free fatty acids can decrease glucose utilization in vitro, it is still controversial how an increase in lipolysis affects glucose metabolism in man. To test the hypothesis that an increase in lipolysis is related to insulin resistance, we examined the effect of lipid-heparin infusion on glucose metabolism in ten normal subjects by the euglycaemic glucose clamp technique and isotopic determination of glucose turnover. In the control euglycaemic clamp studies with insulin infusion at 0.2 and 1.0 mU·kg^–1·min^–1, endogenous glucose production was suppressed from the basal rate of 2.0±0.3 mg· kg^–1min^–1 to 1.1±0.7 mg·kg^–1·min^–1 and -0.4±0.7mg· kg^–1min^–1 respectively. Glucose utilization increased from the basal rate of 2.0±0.3 mg·kg^–1min^–1 to 2.3±0.5mg· kg^–1min^–1 and 5.9±1.8 mg·kg^–1min^–1 respectively. When the euglycaemic clamp studies were coupled with lipid-heparin infusion at comparable low and high rates of insulin infusion, endogenous glucose production increased (1.8± 0.7 mg·kg^–1·min^–1, p<0.001, and 0.3±0.6 mg·kg^–1· min^–1, p<0.05, respectively), and glucose utilization decreased (2.1±0.3 mg·kg^–1·min^–1, not significant, and 3.2±0.7 mg·kg^–1·min^–1, p<0.001 respectively). These data suggest that the artificial induction of intravascular lipolysis by lipid-heparin infusion leads to a state of insulin resistance in man.     

Increased glucose carbon recycling in severely insulin deficient Type 1 (insulin-dependent) diabetic subjects

Summary Six Type 1 (insulin-dependent) diabetic subjects were studied in order to determine the contribution of recycling of glucose carbon to the overproduction of glucose which is characteristic of the fasting hyperglycaemia produced by insulin withdrawal. The subjects were studied on two occasions, once after an overnight insulin infusion and once following 24 h of insulin withdrawal. The difference in turnover rates of 1-¹⁴C-glucose and 3-³H-glucose was used as a measure of glucose recycling. Insulin withdrawal caused a marked metabolic derangement with a rise in non-esterified fatty acids from 0.69±0.23 to 1.11±0.21 mmol/l (mean±SEM, p<0.05), total ketones from 0.27±0.06 to 2.06±0.51 mmol/l (p<0.01), cortisol from 341±43 to 479±31 nmol/l (p<0.05) and growth hormone from 1.1±0.3 to 19+5-mu/l (p<0.05). Glucose turnover rose from 13.8±2.3 mol·kg^–1·min^–1 at a glucose of 6.9±0.7 mmol/l in the insulin infused study to 25.8±4.4 mol·kg^–1·min^–1 (p<0.05) at a glucose of 16.4±0.7 mmol/l in the insulin withdrawn study. Recycling also rose from 3.0±0.4 mol· kg^–1·min^–1 to 9.4±2.2 mol·kg^–1·min^–1 (p<0.05) when insulin withdrawn, accounting for 23±3% and 36±3% of glucose turnover, respectively. We conclude that in the severely insulin deficient Type 1 diabetic subject recycling of glucose carbon is a major contributor to the excess glucose production.     

Insulin resistance,hypertension and microalbuminuria in patients with Type 2 (non-insulin-dependent) diabetes mellitus

Summary We examined the impact of hypertension and microalbuminuria on insulin sensitivity in patients with Type 2 (non-insulin-dependent) diabetes mellitus using the euglycaemic insulin clamp technique in 52 Type 2 diabetic patients and in 19 healthy control subjects. Twenty-five diabetic patients had hypertension and 19 had microalbuminuria. Hypertension per se was associated with a 27% reduction in the rate of total glucose metabolism and a 40% reduction in the rate of non-oxidative glucose metabolism compared with normotensive Type 2 diabetic patients (both p<0.001). Glucose metabolism was also impaired in normotensive microalbuminuric patients compared with normotensive normoalbuminuric patients (29.4±2.2 vs 40.5±2.8 mol · kg lean body mass^–1 · min^–1; p=0.012), primarily due to a reduction in non-oxidative glucose metabolism (12.7±2.9 vs 21.1±2.6 mol · kg lean body mass^–1 ·min^–1; p=0.06). In a factorial ANOVA design, however, only hypertension (p=0.008) and the combination of hypertension and microalbuminuria (p=0.030) were significantly associated with the rate of glucose metabolism. The highest triglyceride and lowest HDL cholesterol concentrations were observed in Type 2 diabetic patients with both hypertension and microalbuminuria. Of note, glucose metabolism was indistinguishable from that in control subjects in Type 2 diabetic patients without hypertension and microalbuminuria (40.5±2.8 vs 44.4±2.8 mol · kg lean body mass^–1 · min^–1). We conclude that insulin resistance in Type 2 diabetes is predominantly associated with either hypertension or microalbuminuria or with both.     

Insulin resistance in Type 2 (non-insulin-dependent) diabetic patients with hypertriglyceridaemia

Summary Hypertriglyceridaemia, which is frequently seen in Type 2 (non-insulin-dependent) diabetes mellitus, is associated with insulin resistance. The connection between hypertriglyceridaemia and insulin resistance is not clear, but could be due to substrate competition between glucose and lipids. To address this question we measured glucose and lipid metabolism in 39 Type 2 diabetic patients with hypertriglyceridaemia, i. e. mean fasting serum triglyceride level equal to or above 2 mmol/l (age 59±1 years, BMI 27.4±0.5 kg/m², HbA_1c8.0±0.2%, serum triglycerides 3.2±0.2 mmol/l) and 41 Type 2 diabetic patients with normotriglyceridaemia, i. e. mean fasting serum triglyceride level below 2 mmol/l (age 58±1 years, BMI 27.0±0.7 kg/m², HbA_1c7.8±0.2 %, serum triglycerides 1.4±0.1 mmol/l). Insulin sensitivity was assessed using a 340 pmol·(m²)^–1· min^–1 euglycaemic insulin clamp. Substrate oxidation rates were measured with indirect calorimetry and hepatic glucose production was estimated using a primed (25 Ci)-constant (0.25 Ci/min) infusion of [3-³H]-glucose. Suppression of lipid oxidation by insulin was impaired in patients with hypertriglyceridaemia vs patients with normal triglyceride levels (3.5±0.2 vs 3.0±0.2mol·kg^–1· min^–1; p<0.05). Stimulation of glucose disposal by insulin was reduced in hypertriglyceridaemic vs normotriglyceridaemic patients (27.0±1.3 vs 31.9±1.6 mol·kg^–1·min^–1; p<0.05) primarily due to impaired glucose storage (9.8±1.0 vs 14.6±1.4mol·kg^–1·min^–1; p<0.01). In contrast, insulinstimulated glucose oxidation was similar in patients with hypertriglyceridaemia and in patients with normal triglyceride concentrations (16.9±0.8 vs 17.2±0.7mol·kg^–1·min^–1). Hepatic glucose production in the basal state and during the clamp did not differ between the two groups. We conclude therefore that oxidative substrate competition between glucose and lipids does not explain insulin resistance associated with hypertriglyceridaemia in Type 2 diabetes. The question remains whether the reduced nonoxidative glucose disposal observed in the patients with hypertriglyceridaemia is genetically determined or a consequence of increased lipid oxidation.     

Erythrocyte sodium-lithium countertransport activity and total body insulin-mediated glucose disposal in normoalbuminuric normotensive Type 1 (insulin-dependent) diabetic patients

Summary Insulin resistance in Type 1 (insulin-dependent) diabetes mellitus may be associated with raised erythrocyte sodium-lithium countertransport activity in patients with hypertension, or nephropathy, or both. However, in these circumstances it is difficult to separate the impact of hypertension, hyperlipidaemia and nephropathy on erythrocyte sodium-lithium countertransport from that of insulin resistance. We have therefore examined the relationship between insulin-mediated glucose disposal and erythrocyte sodiumlithium countertransport in 41 normotensive (mean blood pressure 120/74 mm Hg), normoalbuminuric (mean albumin excretion 6.2 g/min), normolipidaemic (mean serum cholesterol 4.3 mmol/l and mean serum triglycerides 1.0 mmol/l) Type 1 diabetic patients. Erythrocyte sodium-lithium countertransport was on average 0.31 mmol Li · h^–1 · l erythrocytes ^–1 (range 0.07–0.69). Nine patients had values above 0.40 mmol Li · h^–1 erythrocytes^–1 (0.51±0.10 mmol Li · h^–1 · l erythrocytes^–1). The patients with high erythrocyte sodium-lithium countertransport were matched for age, sex, BMI, HbA₁ and duration of diabetes, with nine patients with normal erythrocyte sodium-lithium countertransport. Insulin-mediated glucose disposal was evaluated during the last hour of a euglycaemic clamp (insulin 0.015 U · kg^–1 · h^–1; blood glucose clamped at 7.0 mmol/l). The free insulin levels were comparable between the patients with high and normal erythrocyte sodium-lithium countertransport (37.2±14.7 mU/l and 34.7±17.2 mU/l respectively). Insulin-mediated glucose disposal was on average 3.1±1.5 (range 0.8–6.8) mg · kg^–1 · min^–1. Erythrocyte sodium-lithium countertransport did not correlate with insulin-mediated glucose disposal in all 41 cases (r _s=–0.14), but when the matched groups were compared, patients with raised erythrocyte sodium-lithium countertransport had lower insulin-mediated glucose disposal rates compared to those with normal erythrocyte sodium-lithium countertransport (2.7±1.1 vs 3.9±1.3 mg · kg^–1 · min^–1; p=0.044). In these 18 patients a significant inverse relationship was found between erythrocyte sodium-lithium countertransport and insulin-mediated glucose disposal (r _s=–0.62; p=0.003). Raised erythrocyte sodium-lithium countertransport appears to be associated with insulin insensitivity in Type 1 diabetes, even in the absence of hyperlipidaemia, hypertension and nephropathy.     

Comparison of the effects of human recombinant insulin-like growth factor I and insulin on plasma amino acid concentrations and leucine kinetics in humans

Summary We examined the effects of recombinant human insulin-like growth factor I (IGF-I) and insulin on the plasma amino acid (AA) profile and leucine kinetics in eight normal subjects. IGF-I was infused at 52 pmol·kg^–1·min^–1, in combination with prime-continuous [1-¹⁴C] leucine infusion, to obtain steady-state plasma concentrations of total (54±3 nmol/l) and free (7.3±1 nmol/l) IGF-I (study 1). In response to IGF-I, plasma AA levels declined by 37±3% (1975±198 to 1368±120 mol/l) and total branched chain amino acids (BCAA) declined by 34±3% (390±21 to 256±13 mol/l). This hypoaminoacidaemic effect was associated with a decline in endogenous leucine flux of 17±2% (1.88±0.05 to 1.57±0.04 mol·kg^–1·min^–1) and leucine oxidation of 17±1% (0.31±0.02 vs 0.26±0.02 mol·kg^–1·min^–1) (both p<0.01 vs basal). The same subjects underwent a second study (study 2) in which insulin was infused at 6.22 pmol·kg^–1·min^–1 to obtain a steady-state plasma insulin concentration of 530±25 pmol/l while maintaining euglycaemia. The infusion rate was designed to match the declines in plasma BCAA levels and leucine turnover observed during IGF-I infusion. The rates of glucose infusion necessary to maintain euglycaemia during IGF-I and insulin infusion were 4.9±1.0 and 7.8±0.6 mg·kg^–1 ·min^–1, respectively. During insulin infusion total BCAA declined by 39% from 369±23 to 226±20 mol/l, leucine flux declined by 16±2% from 1.90±0.05 to 1.61±0.03 mol·kg^–1·min^–1, and leucine oxidation declined by 19±2% from 0.32±0.02 to 0.26±0.02 mol·kg^–1·min^–1. On a molar basis IGF-I was 7.3% as potent as insulin in inhibiting proteolysis. These results demonstrate that in humans: (i) the hypoaminoacidaemic response to IGF-I can be entirely ascribed to the inhibition of proteolysis; (ii) qualitatively, the effects of IGF-I and insulin on plasma AA profile and protein metabolism are similar; (iii) quantitatively, IGF-I is 14-fold less potent than insulin in suppressing protein degradation.Abbreviations IGF-I Insulin-like growth factor I - AA aminoacid - BCAA branched chain amino acids - KIC alphaketoisocaproate - ELF endogenous leucine flux - NOLD non-oxidative leucine disposal     

The effects of different plasma insulin concentrations on lipolytic and ketogenic responses to epinephrine in normal and Type 1 (insulin-dependent) diabetic humans

Summary This study was performed to verify: (1) the ability of different insulin concentrations to restrict the lipolytic and ketogenic responses to exogenous epinephrine administration; (2) whether the ability of insulin to suppress the lipolytic and ketogenic responses during epinephrine administration is impaired in Type 1 (insulin-dependent) diabetic patients. Each subject was infused on separate occasions with insulin at rates of 0.2, 0.4, and 0.8 mU·kg^–1·min^–1 while normoglycaemic. To avoid indirect adrenergic effects on endocrine pancreas secretions, the so-called islet clamp technique was used. Rates of appearance of palmitic acid, acetoacetate, and 3-hydroxybutyrate were simultaneously measured with an infusion of ¹³C-labelled homologous tracers. After a baseline observation period epinephrine was exogenously administered at a rate of 16 ng·kg^–1·min^–1. At low insulin levels (20 U/ml) the lipolytic response of comparable magnitude was detected in normal and Type 1 diabetic patients. However, the ketogenic response was significantly higher in Type 1 diabetic patients. During epinephrine administration, similar plasma glucose increments were observed in the two groups (from 4.74±0.53 to 7.16±0.77 mmol/l (p<0.05) in Type 1 diabetic patients and from 4.94±0.20 to 7.11±0.38 mmol/l (p<0.05) in normal subjects, respectively). At intermediate insulin levels (35 U/ml) no significant differences were found between Type 1 diabetic patients and normal subjects, whereas plasma glucose levels rose from 4.98±0.30 to 6.27±0.66 mmol/l (p<0.05) in Type 1 diabetic patients, and from 5.05±0.13 to 6.61±0.22 mmol/l (p<0.05) in normal subjects. At high insulin levels (70 U/ml) the lipolytic response was detectable only in Type 1 diabetic patients; the ketogenic response was reduced in both groups. During the third clamp, a significant rise in plasma glucose concentration during epinephrine infusion was observed in both groups. In conclusion this study shows that at low insulin levels Type 1 diabetic patients show an increased ketogenic response to epinephrine, despite their normal nonesterified fatty acid response. Instead, high insulin levels are able to restrict the ketogenic response to epinephrine in both normal and Type 1 diabetic subjects, although there is a still detectable lipolytic response in the latter. In the presence of plasma free insulin levels that completely restrict the ketogenic response in the same group, there is still a distinct glycaemic response. Plasma insulin levels in Type 1 diabetic patients are a major determinant of the metabolic response to epinephrine.     

Importance of glucagon in the control of futile cycling as studied in alloxan-diabetic dogs

Summary In order to determine the role of glucagon in futile or substrate cycling in diabetes, we measured tracer determined glucose kinetics during a combined infusion of 2-³H-glucose (total glucose production) and 6-³H-glucose (glucose production) in six alloxan-diabetic dogs. The animals received either a 420 min infusion of (1) somatostatin alone (0.3 g·kg^–1· min^–1), (2) somatostatin with insulin replacement (100 U·kg^–1min^–1) or (3) glucagon (6 ng·kg^–1· min^–1) together with somatostatin and transient insulin replacement. When somatostatin was given alone, plasma glucagon (p<0.004) and insulin (p<0.0001) were suppressed. Glucose production and disappearance and plasma glucose concentrations fell (p<0.0001), but the metabolic clearance of glucose did not change significantly. In the basal state, futile cycling comprised 29±4%, 33±4% and 33±3% of total glucose production in the three goups of studies, which is high compared to normal dogs. The absolute rate of futile cycling fell slightly but significantly from 10.0±1.7 to 8.3±1.7 mol·kg·^–1min^–1 (p<0.0008). When insulin replacement was given during somatostatin infusion to correct for the small somatostatin-induced insulin suppression, there were similar changes in plasma glucagon, glucose concentrations and glucose kinetics as seen during the infusion of somatostatin alone. Futile cycling decreased to a slightly greater extent from 12.8±2.8 to 9.5±1.7mol·kg^–1·min.^–1 (p<0.02). When glucagon was infused together with somatostatin and insulin replacement, plasma glucagon (p<0.0002) increased and plasma glucose levels rose (p<0.001) due to a transient increase in glucose production. Metabolic clearance of glucose did not change significantly. There was a marked increase in futile cycling from 12.2±1.7 to 21.7±1.7mol· kg^–1·min^–1 (p<0.0001) in response to exogenous glucagon excess. There was a slight (p<0.01) drop in free fatty acid levels with somatostatin. Free fatty acid levels nearly doubled (p<0.025) with the infusion of glucagon together with somatostatin. In conclusion, (a) futile cycling was increased in alloxan-diabetic dogs; (b) glucagon suppression can suppress futile cycling only if total insulin deficiency is prevented; and (3) hyperglucagonaemia increases futile cycling, and this effect is more pronounced during insulin deficiency.     

Reduced glycogen synthase activity in skeletal muscle from obese patients with and without Type 2 (non-insulin-dependent) diabetes mellitus

Summary In order to evaluate the importance of a defect in insulin mediated non-oxidative glucose metabolism and glycogen synthase activity in skeletal muscles in obese subjects with and without Type 2 (non-insulin-dependent) diabetes mellitus we studied: 10 lean and 10 obese control subjects and 12 obese diabetic patients using the euglycaemic hyperinsulinaemic clamp technique (basal, 20 mU · (m²)^–1 · min^–1, 80mU·(m²)^–1·min^–1) in combination with indirect calorimetry. Muscle biopsies were taken from m. vastus lateralis at each insulin level. We found that non-oxidative glucose metabolism could be stimulated by insulin in all three groups (p<0.01). The values obtained at the highest insulin levels (around 140 U/ml) were lower in both obese groups compared to the lean control subjects (118±21, 185±31, 249±14 mg·(m²)^–1·min^–1 (p< 0.01)). Insulin stimulation of the glycogen synthase activity at a glucose-6-phosphate concentration of 0.1 mmol/l was absent in both obese groups, while activities increased significantly in the lean control subjects (19.6±4.2% to 45.6±6.8%, p< 0.01). Glycogen synthase activities at the highest insulin concentrations only differed significantly between lean control subjects and obese diabetic patients (45±7% and 31±5%, p< 0.05). We conclude that insulin resistance in peripheral tissues in obese subjects with and without Type 2 diabetes may be partly explained by a reduced insulin mediated non-oxidative glucose metabolism and that this abnormality might be due to an absent insulin stimulation of glycogen synthase in skeletal muscles. This enzyme defect is correlated to obesity itself.