Glucagon release and glucose counter-regulation during hypoglycaemia. Modifying effect of the previous glucose level

The importance of a short-term elevation of the ambient glucose level for the release of counter-regulatory hormones and the glucose recovery rate during a subsequent hypoglycaemia was studied in healthy subjects. Hypoglycaemia was induced with insulin infusion after a previous 80 min of euglycaemic (E: 5 mmol/l) or hyperglycaemic (H: 15 mmol/l) glucose clamp. By infusing insulin during the euglycaemic clamp similar levels were reached during both glucose clamps. The same level of hypoglycaemia was reached in both studies (E: 1.5 +/- 0.1, H: 1.5 +/- 0.2 mmol/l) and the insulin levels were also similar both at glucose nadir and during the recovery period. In spite of this, both the mean glucagon levels at nadir at the mean individual maximal increase were significantly lower after the hyperglycaemic clamp (E: 101 +/- 25, H: 54 +/- 7 pg/ml, P less than 0.05). The glucose recovery rate was also significantly impaired following the hyperglycaemic clamp. The results show that a short-term elevation of the ambient glucose level impairs the glucagon release during a subsequent hypoglycaemia. This finding may be of importance for the development of the blunted glucagon release in response to low glucose levels in diabetics.

Fat cell metabolism in different regions in women. Effect of menstrual cycle, pregnancy, and lactation

Adipose tissue lipolysis and lipoprotein lipase (LPL) activity were studied in biopsies from the femoral and abdominal depots in healthy women during early or late menstrual cycle, pregnancy, and the lactation period. When the differences in cell size were taken into account, basal lipolysis was similar in both regions in nonpregnant women. During lactation, however, lipolysis was significantly higher in the femoral region. The lipolytic effect of noradrenaline (10(-6) M) was significantly less in the femoral region in the nonpregnant women and during early pregnancy. However, the lipolytic response was the same in both regions in lactating women. LPL activity was higher in the femoral than in the abdominal region except during lactation when a marked decrease in the LPL activity was seen in the femoral region. The LPL activity in the abdominal region remained unchanged in all patient groups. The results imply that in both nonpregnant and pregnant women lipid assimilation is favored in the femoral depot. During lactation, however, the metabolic pattern changes; the LPL activity decreases and lipid mobilization increases in this depot. These changes are much less pronounced in the abdominal region. Thus, fat cells from different regions show a differential response during pregnancy and lactation. These results suggest that the adipose tissue in different regions may have specialized functions.

Changes in growth hormone binding and metabolic effects of growth hormone in rat adipocytes following hypophysectomy

Growth hormone (GH) binding and the effect of GH and insulin on glucose metabolism in rat adipocytes were studied at various time periods following hypophysectomy. Male rats were hypophysectomized at 33-34 days of age. After 6 h, 20 h or 3, 7 and 14 days adipocytes were prepared from epididymal fat pads by mild collagenase digestion (0.5 mg X ml-1, 60 min, 37 degrees C). Glucose metabolism was studied by determining the production of CO2 from [14C]glucose and the incorporation of [14C]glucose into lipids. GH binding was measured in cell aliquots using [125I]hGH. No difference in GH binding to adipocytes was observed between control rats and rats hypophysectomized or sham-operated 6 h earlier. GH binding was significantly decreased 20 h after hypophysectomy and declined further with time after hypophysectomy. Adipose tissue from normal rats is usually refractory to the insulin-like effect of GH. Adipocytes isolated from normal rats were, however, usually responsive to GH immediately after cell isolation, suggesting that refractoriness to the insulin-like effect of GH was lost during the time required for the preparation of adipocytes. The magnitude of the response to GH in adipocytes progressively declined with time after hypophysectomy. The decreased responsiveness to GH with time after hypophysectomy parallelled the decrease in GH binding. The results suggest that the pituitary, directly or indirectly, is necessary for the maintenance of GH binding sites in adipose tissue and that these binding sites are related to the insulin-like effect of GH.

Bacitracin enhances intracellular accumulation of insulin in rat adipocytes

Bacitracin (1 mg/ml) markedly increased (approx. 75%) the cell-associated specifically bound 125I-labelled insulin without altering the affinity of the binding sites. Bacitracin also exerted a modest inhibitory effect on the degradation of insulin in the incubation medium determined as radioactivity not precipitated by trichloroacetic acid (from 9.6 to 4.8%). The effect on insulin binding was about 5-times as sensitive as the effect on degradation. The increased binding was due to intracellular accumulation of radioactivity which could not be removed by treating the cells with trypsin. This increase was not seen when the internalization process was reduced by ATP-depletion or low temperature. Since the trypsin-sensitive fraction of cell-associated radioactivity was apparently not altered, it is suggested that bacitracin, in addition to its well-known inhibition of extracellular degradation, also inhibits the intracellular degradation of insulin.

Regional differences in adipocyte metabolism and possible consequences in vivo

Recent studies have shown that adipose tissue metabolism varies in different regions. Thus, hormonal responsiveness and sensitivity to both lipolytic and anti-lipolytic agent is increased in abdominal as compared to femoral cells. Abdominal obesity is also associated with greater aberrations in metabolism than peripheral obesity. The increased lipolytic response in abdominal fat cells may lead to higher FFA concentrations, which may attenuate both glucose uptake and insulin clearance by the liver.

Influence of metformin on metabolic effect of insulin in human adipose tissue in vitro

To study the mechanism(s) of action of metformin, fragments of human subcutaneous adipose tissue were incubated with therapeutic blood concentrations of metformin. In the absence of insulin no effect of metformin was seen on either lipolysis or glucose metabolism. When insulin was present, however, metformin stimulated glucose conversion into both triglycerides and CO2. In marked contrast, no effect of metformin was observed on the antilipolytic effect of insulin. In agreement with this selective effect no change in insulin binding was found. In conclusion, metformin seems to exert its effect on glucose metabolism by potentiating the action of insulin at a post-receptor level, possibly on the rate of glucose transport.

Diabetes mellitus in phaeochromocytoma. Fasting blood glucose levels before and after surgery in 60 patients with phaeochromocytoma

Sixty patients undergoing operation for phaeochromocytoma were investigated in the pre- and postoperative states with respect to fasting blood glucose levels. When 6 previously known or suspected diabetics were excluded, preoperative diabetes (fasting blood glucose levels greater than or equal to 7.0 mmol/l) were found in 3 of 13 (23%) with sustained hypertension, in 6 of 12 (50%) with sustained hypertension associated with paroxysms and in 4 of 24 (17%) with paroxysmal hypertension. None of the 5 patients with atypical clinical symptoms had glucose levels greater than or equal to 7.0 mmol/l. In the groups of patients with particularly high urinary excretion of catecholamines and vanilmandelic acid higher blood glucose levels were also found. The postoperative blood glucose levels in the follow-up study were normal and less than 5.8 mmol/l in all cases except in 3 of the 4 still living patients with a previously known diabetes and in 1 patient with a malignant tumour. Thus, manifest diabetes, defined as fasting glucose levels greater than or equal to 7.0 mmol/l, is frequently present in patients with phaeochromocytoma (24% in the present study) and the diabetes is reversed by removal of the tumour.

Counter-regulation of insulin-stimulated glucose transport by catecholamines in the isolated rat adipose cell

The interaction between catecholamines and insulin in regulating glucose transport in isolated rat adipose cells has been evaluated. In the absence of insulin, 1 microM isoproterenol stimulates 3-O-methylglucose transport approximately 2-fold. However, isoproterenol in combination with adenosine deaminase inhibits glucose transport activity approximately 60%. N6-Phenylisopropyladenosine, a nonmetabolizable adenosine analogue, substantially reverses this inhibitory effect and actually stimulates glucose transport activity approximately 2-fold in the absence of isoproterenol. Dibutyryl cAMP inhibits glucose transport activity approximately 75% regardless of adenosine deaminase. While none of these agents significantly influences the basal concentration of plasma membrane glucose transporters, as assessed by specific D-glucose-inhibitable cytochalasin B binding, isoproterenol or dibutyryl cAMP in combination with adenosine deaminase reduces that in the low density microsomes 19 and 58%, respectively. In the presence of insulin, both isoproterenol and adenosine deaminase alone inhibit glucose transport activity approximately 25%. However, only the latter is accompanied by a corresponding decrease in the insulin-stimulated concentration of plasma membrane glucose transporters. Together, isoproterenol and adenosine deaminase inhibit insulin-stimulated glucose transport activity approximately 75%, even in the presence of 5 mM glucose to maintain cellular ATP levels. A similar inhibition is observed with dibutyryl cAMP. However, these agents decrease the insulin-stimulated concentration of plasma membrane glucose transporters only approximately 45%. Nevertheless, all of these inhibitory effects occur through decreases in the transport Vmax. In addition, N6-phenylisopropyladenosine partially reverses the inhibitory effects induced by the presence of adenosine deaminase. These results suggest that catecholamines counter-regulate basal and insulin-stimulated glucose transport in rat adipose cells through a cAMP-mediated mechanism, but only in part by modulating the translocation of glucose transporters.

Improved but not normalized glucose counter-regulation during glucagon infusion in Type 1 (insulin-dependent) diabetes

Glucose counter-regulation during insulin-induced hypoglycaemia was studied in Type 1 diabetic patients without evidence of autonomic neuropathy and compared with that of a non-diabetic control group. The glucose recovery rate following hypoglycaemia was delayed in the diabetic compared with the control subjects and this was most pronounced for the initial, rapid phase of glucose increase (glucose increase in 15 min, control: 1.1 +/- 0.1 versus 0.4 +/- 0.1 mmol/l; p less than 0.01). The release of glucagon during hypoglycaemia was blunted in the diabetic patients (maximal plasma levels, control: 148 +/- 25 versus 70 +/- 10 pg/ml; p less than 0.01). The adrenaline levels were also lower compared with the control subjects (maximal plasma levels, control: 7.23 +/- 1.21 versus 3.27 +/- 0.87 nmol/l; p less than 0.05). To evaluate the importance of the blunted glucagon response for the delayed glucose compensation, glucagon was infused during the hypoglycaemia. Overall glucose recovery rate was improved but did not return to normal. Consequently impaired glucagon release in the diabetic patients cannot alone explain impaired glucoregulation; the lower adrenaline levels and/or an effect of the previous glucose levels per se on hepatic glucose production are probably also of importance.

Effects of physical training on insulin, connecting peptide (C-peptide), gastric inhibitory polypeptide (GIP) and pancreatic polypeptide (PP) levels in obese subjects

Ten severely obese women were subjected to physical training for three months on ad libitum diet. Under metabolic ward conditions oral glucose tolerance test was performed before and after the training period with the same energy intake quantitatively and qualitatively, and glucose, insulin, connecting (C)-peptide, gastric inhibitory polypeptide (GIP) and pancreatic polypeptide (PP) were determined. In confirmation of previous work, physical training caused no decrease in body fat in these severely obese subjects, and no change in body cell mass or glucose tolerance, while insulin and blood pressure decreased. The control of dietary conditions demonstrated that the latter phenomena were not due to quantitative or qualitative changes in the diet. C-peptide concentrations decreased also, indicating effects of physical training in obesity on insulin production. GIP is believed to be a gastrointestinal factor facilitating insulin secretion (Incretin). Previous work has indicated that gastrointestinal factor(s) are involved in the insulin lowering effect seen after physical training. It is possible that GIP is contributing to this phenomenon.

Reversal of insulin resistance in type I diabetes after treatment with continuous subcutaneous insulin infusion

Insulin responsiveness was studied with the euglycaemic glucose clamp technique in seven patients with type I diabetes and in six control subjects matched for age and weight. The glucose disposal rate was significantly reduced in the diabetic subjects when they were receiving conventional insulin treatment compared with the control group, showing insulin resistance in the diabetics. The diabetic patients were again studied after eight days of intensified metabolic control achieved with continuous subcutaneous insulin infusion. During the infusion a more physiological insulin regimen was used compared with their regular treatment, less of the total insulin dose being given as continuous infusion and more as bolus doses before meals. The insulin resistance in the diabetics was largely reversed after this improved metabolic control. Dose response studies showed an increased glucose disposal rate at all plasma insulin concentrations, including the maximum insulin concentration, indicating a predominant effect of the continuous infusion regimen at the postreceptor level. The improved insulin effect seen with continuous subcutaneous insulin infusion could be due to the improved metabolic control achieved as well as the more physiological regimen.

Effects of physical training on adrenergic sensitivity in obesity

To examine the possibility that the decrease of hyperinsulinemia and blood pressure in obesity associated with physical training is mediated via adaptations in the adrenergic nervous system, a pure beta-adrenergic agonist (isoproterenol) or an alpha-adrenergic antagonist (phentolamine) was infused before and during an oral glucose tolerance test before and after physical training. A number of circulatory, metabolic, and endocrine factors under adrenergic control were followed. Physical training was associated with an augmented beta-agonist response in blood pressure, heart rate, blood glucose, plasma insulin, connecting (C) peptide, and pancreatic polypeptide (PP) but not in plasma glucagon and gastric inhibitory polypeptide. Physical training also resulted in higher values of C-peptide and PP values after alpha-adrenergic blockade. It was concluded that physical training probably is associated with an augmented sensitivity of the beta-adrenergic nervous system. This might also be the case with the alpha-adrenergic system. It was suggested that this in turn might be due to a decreased firing in the adrenergic nervous system leading secondarily to an increased sensitivity in the effector cells. It was hypothesized that such decreased firing could provide a background to explain lower blood pressure and plasma insulin after physical training.

Insulin resistance in fat cells from insulin-treated type I diabetic individuals

Fat biopsies from the lower abdominal wall were obtained from 13 insulin-treated type I diabetic subjects and from 12 age-, weight-, and sex-matched control subjects. Insulin binding and the antilipolytic effect of insulin were studied. Insulin binding was significantly reduced in the diabetic subjects (34% reduction at tracer binding, P less than 0.05) due to a decreased number of binding sites. In agreement with this, the dose-response curve for the antilipolytic effect of insulin was shifted to the right in the diabetic subjects. Furthermore, the maximal antilipolytic effect of insulin was also reduced (64%, P less than 0.05). Thus, fat cells from conventionally treated type I diabetic individuals are resistant to insulin. This resistance is due to a combination of a decreased number of insulin binding sites and an unspecified intracellular (postreceptor) defect involving the antilipolytic effect of insulin. These findings are in accord with recent in vivo studies showing that type I diabetic patients are also resistant to the stimulating effect of insulin on glucose disposal.

Impact of obesity on metabolism in men and women. Importance of regional adipose tissue distribution

The distribution of adipose tissue thickness, fat cell weight (FCW), and number (FCN) were studied in four regions in randomly selected middle-aged men and women and in 930 obese individuals. Both the obese and the randomly selected men were found to have the largest adipose tissue thickness in the abdominal region. Women, however, showed a relative preponderance for the gluteal and femoral regions. FCW increased with expanding body fat up to a maximal size of approximately 0.7-0.8 micrograms/cell in each region. After this increase in FCW, a more rapid increase in FCN was found. For the same degree of relative overweight, men had higher triglyceride, fasting glucose, and insulin levels; higher sums of glucose and insulin levels during an oral glucose tolerance test; and higher blood pressure. Furthermore, elevated fasting glucose levels (greater than 7.4 mM) occurred twice as often in the males. These differences between males and females persisted even after body fat matching. A male risk profile was seen in women characterized by abdominal obesity (high waist/hip circumference ratio) as compared to women with the typical peripheral obesity. Stepwise multiple regression analyses in both women and men showed the obesity complications to be associated in a first step to waist/hip circumference or body fat and in a second to abdominal fat cell size. It may thus be concluded that: (a) In both obese and nonobese subjects, regional differences exist between the sexes with regard to adipose tissue distribution. (b) Moderate expansion of body fat is mainly due to FCW enlargement, which is subsequently followed by increased FCN. (c) Men and women with a male abdominal type of obesity are more susceptible to the effect of excess body fat on lipid and carbohydrate metabolism.

Insulin binding and responsiveness in fat cells from patients with reduced glucose tolerance and type II diabetes

Adipose tissue was obtained from 66 individuals including 21 patients with type II diabetes of different severity (16 SU-treated and 5 diet-treated only) as well as 9 obese subjects with reduced glucose tolerance. Adipocyte insulin binding, antilipolytic effect of insulin, and glucose incorporation into triglycerides were measured in the diabetic and the obese subjects and the data compared with that of normal controls of similar age and relative weight. Insulin binding per cell was normal in the diabetic patients and was significantly increased at low insulin concentrations in the obese patients with reduced glucose tolerance, suggesting increased affinity. Furthermore, insulin binding correlated negatively with age but, when age was corrected for, did not correlate significantly with fasting insulin or glucose levels, relative body weight, or fat cell size. Insulin sensitivity, measured as the antilipolytic effect of insulin, was similar in all patient groups. Patients with the most severe type II diabetes (SU-treated group) demonstrated, in contrast to the less severely diabetic patients, a marked reduction in both basal and insulin-stimulated glucose incorporation into triglycerides showing the presence of a pronounced postreceptor defect. The insulin effect on glucose incorporation correlated negatively with the fasting glucose levels, suggesting that the postreceptor defect seen in the adipocyte reflects perturbations in other organs, like muscle or liver, of greater importance for glucose homeostasis.

Influence of ambient glucose and insulin concentrations on adipocyte insulin binding

To elucidate factors of importance for insulin binding, fat cells from humans and rats were incubated under various experimental conditions for different periods of time. Human adipocytes incubated for 24 hours in the absence of insulin showed no significant difference in insulin binding compared with cells from freshly excised tissue. After 48 hours, however, an increased rate of binding (average 54%; P less than 0.05) was obtained. The addition of insulin (2000 microU/ml) to the culture medium resulted in a decrease in insulin binding (average 33%; P less than 0.05) compared with cells maintained in the absence of insulin. There was no apparent difference in receptor affinity, indicating that the altered binding was due to a change in receptor number. In the absence of insulin, elevating the glucose concentration of the medium from 0.8 mM to 22.4 mM did not significantly influence insulin binding. Rat adipocytes showed similar but more rapid changes. Thus, incubation for 24 hours without insulin caused an increase in insulin binding (average 37%; P less than 0.05). This up-regulation was seen even in a high glucose concentration (28 mM) but was completely prevented by the presence of insulin in the medium. Furthermore, when rat adipocytes were incubated with insulin in the presence of a high glucose concentration (28 mM) there was a significant further decrease in insulin binding compared with that of parallel incubations performed in 5.6 mM glucose. Thus, even in the absence of TRIS buffer, insulin-dependent regulation of the number of binding sites is shown for both human and rat adipocyte tissue in vitro. Although this perturbation could be directly due to hormone-receptor interaction at the membrane level, the finding of rat adipocytes that the ambient glucose concentration can modulate this effect suggests the importance of post-receptor events.

beta-Adrenergic dependent downregulation of insulin binding in rat adipocytes

In order to study factors regulating insulin binding to rat adipocytes short- and long-term incubations were performed in the presence or absence of noradrenaline (NA) at a concentration of 3 microM. Culture with NA for 24 hours resulted in a significant decrease (30%) of insulin binding due to a reduced number of binding sites. This reduction was dose-dependent and completely prevented by the addition of timolol, a beta-adrenergic blocking agent, but not by the alpha-blocking agent phentolamine. Addition of 1.0 mM db cAMP to the culture medium resulted in a similar reduction in insulin binding. Also short-term incubations with NA resulted in a downregulation of insulin binding; reproducible reductions seen already after 20 min and about 30% reduction achieved after 2 hours' preincubation. Thus, beta-adrenergic stimulation and subsequent cAMP elevation results in a rapid reduction in the number of insulin binding sites.

Effects of anti-hypertensive therapy on serum lipoproteins. Treatment with metoprolol, propranolol and hydrochlorothiazide

The effects of metoprolol, propranolol and hydrochlorothiazide on lipoprotein metabolism were studied in three different but comparable groups of middle-aged men with previously untreated hypertension (n=10, n=10, and n=11, respectively). All three treatments were associated with an increase in serum triglyceride and VLDL-cholesterol levels. Propranolol treatment was associated with consistent and significant decreases in HDL-cholesterol, apoA-I and A-II levels, whereas these changes during the other treatments were neither significant nor consistent. An increase in adipose tissue LPL-activity and a decrease in serum free fatty acids were seen in the propranolol treatment group. Significant changes were not observed in glucose tolerance or catecholamine excretion. The blood pressure reduction was similar in the three groups. The design of the present study was in some important respects different from that of others. This may help to explain why we found a difference between the two beta-blockers in our study.

Effect of a modified guar gum preparation on glucose and lipid levels in diabetics and healthy volunteers

Six healthy volunteers and 17 diabetics (6 insulin-dependent and 11 diet- and tablet-treated) were treated with a special processed, palatable guar gum (10 g b.i.d. immediately before meals) for periods of one or three weeks or, in some cases, up to 13 weeks. A standardized test meal was given to study the effect of the fiber on postprandial glucose levels. Ten g guar was stirred in water and taken immediately before the test meal. The postprandial blood glucose levels were similar in the healthy volunteers but significantly lower in the diabetics following treatment with guar for one and three weeks, respectively. Furthermore, the fasting blood glucose levels were significantly lower in the diabetics after three, but not one, weeks of treatment. The lower postprandial glucose levels were coupled with attenuated and delayed insulin levels in accordance with an effect of guar gum on the rate of carbohydrate absorption. The cholesterol levels were on average reduced with 14% in the diabetics following three weeks' treatment with guar. The higher the initial cholesterol level, the greater the reduction in cholesterol; 26% reduction was achieved in four patients with initial levels above 7 mM. The alpha-lipoprotein cholesterol levels were not significantly changed, thus an increase in the alpha-lipoprotein cholesterol/total serum cholesterol ratio was obtained. Neither plasma triglycerides nor body weights altered during treatment. The reported side-effects were as expected and were usually mild and transient (e.g. increased flatulence). The data show that guar gum also reduces postprandial glucose levels on a long-term basis and may improve the diabetic control. Additionally, treatment with this fiber leads to a concentration-dependent decrease in cholesterol levels.

Importance of glucose control for the recovery from hypoglycemia in insulin-dependent diabetics

To evaluate whether the delayed glucose compensation after hypoglycemia in insulin-dependent diabetics was associated with their elevated blood glucose levels, five diabetic patients were studied before and after a period of intensified metabolic control. The glucose recovery rate was found to be improved after better diabetic control. This influence seems to be better reflected by the mean diurnal level rather than the glucose level immediately before hypoglycemia. The improvement occurred despite the same or lower levels of the important glucocompensatory hormones. These results show the importance of antecedent metabolic control for glucose compensation after hypoglycemia.

Beta-adrenoceptor blockade and recovery from hypoglycaemia in diabetic subjects: normalization after lactate and glycerol infusions

1. Previous studied have shown that non-selective beta-adrenoceptor blockade attenuates the blood glucose recovery rate after hypoglycaemia in type I diabetes. Apart from possible effects on hepatic glycogenolysis propranolol also inhibits the release of the important gluconeogenic substrates lactate and glycerol. 2. To determine whether the effect of non-selective beta-adrenoceptor blockade on glucose recovery could be associated with diminished availability of gluconeogenic substrates, lactate and glycerol were infused during hypoglycaemia in four insulin-dependent diabetic patients. Comparisons were made of the blood glucose recovery on placebo, propranolol and propranolol combined with the infusion. 3. The blood glucose recovery rate after hypoglycaemia was less on propranolol than with placebo but was significantly improved and not different from placebo when propranolol treatment was combined with lactate and glycerol infusions. Thus, at least for type I diabetic patients, in whom gluconeogenesis is proportionally greater than in healthy subjects, non-selective beta-adrenoceptor blockade attenuates the glucose recovery rate from hypoglycaemia mainly by reducing the availability of gluconeogenic substrates.

Hypoglycemic symptoms in insulin-dependent diabetics. A prospective study of the influence of beta-blockade

Hypoglycemic attack rate, duration and symptomatology were studied in five insulin-dependent diabetics with borderline hypertension all of whom were prone to hypoglycemia. They were treated in a double-blind, cross-over fashion with the cardioselective beta 1-blocking agent metoprolol and placebo. The treatment period on each drug lasted at least three months. No treatment-associated differences in attack rate and duration of hypoglycemic attacks were recorded. Slight, but no severe masking of hypoglycemic symptoms was recorded in one patient on metoprolol. It is concluded that cardioselective beta 1-blocking agents can be used by insulin-dependent diabetics. However, until further direct experience has been gained caution should be exercised in treating patients with obvious clinical signs of autonomic neuropathy with these drugs.

Effect of beta-adrenoceptor blockade on exercise performance and metabolism

1. Carbohydrate and lipid metabolism and the capacity to perform prolonged submaximal physical exercise were studied in six young healthy subjects treated in a randomized double-blind fashion for 2 days with either placebo, the non-selective beta-adrenoceptor antagonist propranolol (80 mg b.i.d.) or the cardioselective agent metoprolol (100 mg b.i.d.). On day 3, 1 h after the last dose, the subjects exercised for 30 min periods followed up 10 min rest up to the point of exhaustion. 2. The capacity to perform exercise was decreased with both beta-adrenoceptor antagonists. However, at an equal degree of beta 1-adrenoceptor blockade, all subjects could exercise for a longer period of time on the cardioselective agent as compared with the non-selective drug. 3. Blood glucose levels decreased during exercise irrespective of the type of treatment, but the attenuation occurred most rapidly on propranolol. At exhaustion the average non-esterified fatty acid levels had increased 256% on placebo, 148% on metoprolol and 65% on propranolol. A significant positive correlation was found between changes in non-esterified fatty acid levels during exercise and total working time. It is concluded that beta-adrenoceptor blockade diminishes the capacity for prolonged sub-maximal exercise at least in part by reducing the availability of substrates to the working muscles.