The mechanism behind exercise-induced decreases in plasma insulin concentrations was examined in eight healthy young men. In addition, the influence of specific α1- and α2-adrenoceptor blockade on glucose kinetics during exercise was studied. To test the hypothesis that exercise-induced decreases in insulin secretion are mediated via α2-adrenoceptors, all subjects exercised for 60 min on separate occasions under four conditions: with and without α1-receptor blockade (1 mg prazosin) and with and without or α2-receptor blockade (15 mg yohimbine). Glucose kinetics were measured using [3-3H]glucose. During exercise with α2-receptor blockade, the insulin concentration initially increased (first 20 min) then decreased, whereas it continually decreased in the corresponding control experiment. The C-peptide concentration did not change during exercise with α2-receptor blockade but decreased in the control experiment. During exercise with α1-receptor blockade and corresponding control experiments, insulin and C-peptide levels always decreased. With α1-receptor blockade, the glucose concentration increased (first 30 min) and then decreased, whereas it slightly decreased in all other experiments. In addition, with α1-receptor blockade, the glucose rate of appearance (Ra) increased rapidly (because of higher catecholamine concentrations in α1-receptor blockade versus control) and the glucose rate of disappearance (Rd) was higher compared with control. During exercise with α2-receptor blockade, the Ra and Rd were always lower compared with control. Therefore, we conclude that exercise-induced decreases in insulin secretion are mediated via α2-adrenoceptors and that blockade of α1- and α2-adrenoceptors during exercise elicits opposite responses in glucose Ra and Rd.

Insulin is produced and secreted from the β-cells in the pancreas. The secretion of insulin is regulated by a complex interplay among neurohumoral stimuli originating from food entering the gastrointestinal tract, circulating substrates, and hormones.

During exercise, plasma insulin concentrations decrease with a magnitude closely linked to the intensity and duration of the exercise. This decrease is primarily attributable to a decrease in insulin secretion (1). However, the precise mechanism behind the exercise-induced decrease in insulin secretion is not known. Most likely, it comes from increases in sympathoadrenal activity that accompany exercise. Thus, it has been found that the decrease in plasma insulin during exercise can be partly inhibited by prior administration of phentolamine, a nonselective blocker of α-adrenergic receptors (2,3). Later animal studies have shown that the decrease in insulin secretion is mediated via postsynaptic α2-receptors (4,5), the existence of which has now been demonstrated in human β-cells (6). The exercise-induced inhibition of insulin secretion seen in humans may therefore be attributable to the exercise-induced increase in sympathetic activity, which, by the release of norepinephrine in the synaptic cleft, activates postsynaptic α2-receptors.

Results from previous studies using α2 antagonists during physical exercise have been contradictory regarding plasma concentrations of insulin. In humans, plasma insulin concentrations decreased during 30 min of ergometer cycle exercise at 60% of Vo2max after the administration of yohimbine (a selective α2 antagonist); however, in that study, the concentration of C-peptide was not measured (7). In swimming mice, however, yohimbine was shown to inhibit the exercise-induced decrease in plasma insulin (8).

It cannot be a priori excluded that α1-receptors also play a role in the exercise-induced inhibition of insulin secretion in humans, but to our knowledge, no studies on this have been carried out. To elucidate the mechanism behind the exercise-induced inhibition of insulin secretion in humans, we studied insulin secretion and concentrations before, during, and in the recovery phase of a 60-min ergometer bicycle exercise bout with and without selective pharmacological blockade of the α1- and α2-receptors.

In contrast to insulin secretion, hepatic glucose production increases with exercise. Feedback signals from the contracting muscles, mediated both neurally and via the blood stream, adjust the stimulus for hepatic glucose production to maintain euglycemia in the blood (9). A rise in blood glucose directly inhibits glucose production during exercise, whereas a drop in blood glucose, via stimulation of counterregulatory hormones, increases hepatic glucose production (9). Also, central mechanisms, coupled with the degree of motor center activity, stimulate epinephrine release, which in turn enhances hepatic glucose production (9). Sympathetic nerves to the liver do not, however, play a role in the exercise-induced increase in hepatic glucose production, and glucagon in humans is also of minor importance (9). In the present study, the use of selective α-adrenergic antagonists made it possible to evaluate the role of α1- versus α2-receptors in hepatic glucose production and glucose disappearance from plasma; to our knowledge, this has not been done previously.

Subjects.

The study subjects, eight healthy young men (age [means ± SE] 25 ± 0.9 years, height 180 ± 2 cm, weight 73 ± 2 kg), gave their informed consent to participate in this study, which was approved by the Ethical Committee for Copenhagen and Frederiksberg, Denmark. The subjects were recruited through announcements in a student magazine. None of the subjects had any family history of diabetes, and none was taking medication.

Within the 4–30 days before the start of the study, each subject’s Vo2max was measured during exercise on a bicycle ergometer (Ergoline 900 l; Siemens, Denmark), with the subject seated in a semirecumbent position. The Vo2max measured in this position corresponds to 91–93% of the Vo2max measured during upright ergometer bicycling (10).

Experimental protocol.

Two independent crossover studies were performed: bicycling exercise with and without pharmacological α1-receptor blockade and with and without pharmacological α2-receptor blockade. The four experiments were performed during a 3-month period, with at least 7 days between each experimental day. The experiments were carried out in a random fashion, apart from the trial with the α1-receptor blockade, which was always performed before its control study. Each subject participated in all four experimental trials.

On each experiment day, the subjects arrived at the hospital at 9:00 a.m. in the fasting state, having abstained from smoking and physical activity during the foregoing 10 h. The subject was placed on an ergometer bicycle (Ergoline 900 l; Siemens), which could be tilted, enabling the subject to perform bicycling in a position anywhere from an upright to a supine position. In the study, the semisupine position was used to prevent possible hypotensive episodes. A catheter was inserted in a brachial artery for continuous measurement of blood pressure. One intravenous catheter was inserted in the retrograde direction in a dorsal hand vein for collection of arterialized blood samples. This hand was kept in a heating pad (42°C) throughout the experiment. A second intravenous catheter was inserted either contralaterally to or more proximally than the first for the infusion of radioactive-labeled glucose.

Each experiment lasted 4 h (from −120 to 120 min). The subject remained awake and resting in the first 2 h, then performed 1 h of bicycling exercise followed by 1 h of recovery. The subject’s heart rate was continuously monitored by a Polar heart rate monitor (Pulstester PE3000; Polar Instruments, Oulu, Finland).

Infusion of glucose tracer.

[3-3H]glucose (Amersham), pharmaceutically prepared and dissolved in isotonic saline (91.1 ± 1.5 kBq/ml; Isotopapoteket, Copenhagen) was given as a primed (1,138 ± 21 kBq) constant infusion (11.2 ± 0.2 kBq/min) throughout the entire experiment (t = −120 to 120 min).

Exercise and oxygen uptake.

After 2 h of rest, at t = 0 min, the subjects performed bicycle exercise for 60 min in a semisupine position at 60 rpm. Individual workloads were targeted at 60% (α1-receptor blockade study) and 70% (α2-receptor blockade study) of the previously measured Vo2max. During each experiment, Vo2 was measured by a cardiopulmonary test system (Medical Graphics, St. Paul, MN) before and three times during the exercise (15–19, 35–39, and 55–59 min). The first measurement (15–19 min) was used to adjust the workload, which then was not changed during the last 30 min of exercise. If the workload was adjusted during the first 30 min of exercise, exactly the same adjustment was carried out in the corresponding control or blockade experiment. In two of the α1-receptor blockade experiments, the workload had to be diminished because of signs of fatigue (plasma lactate 10 and 15 mmol/l) that would have made the subjects unable to complete the exercise.

Blockade of α-receptors.

In the α1-receptor blockade experiment, the subjects were given selective α1-receptor antagonist tablets orally (1 mg prazosin [Hexapress]; Durascan, Odense, Denmark) 2 h before exercise. In the α2-receptor blockade experiment, the subjects were given selective α2-receptor antagonist tablets orally (15 mg yohimbin [Yohimbin UNP]; United Nordic Pharma, Rungsted, Denmark) 45 min before exercise. The corresponding control experiments were performed without medication.

Analytical techniques.

Blood was sampled at the indicated intervals shown on the figures. All blood samples were kept at −20°C until analysis, except for samples for free fatty acids (FFAs) and catecholamines, which were kept at −80°C. The procedures, stabilization of blood samples, and analysis of all hormones and metabolites have been described previously (11). Blood samples for the determination of glucose and lactate were immediately centrifuged, and the plasma was analyzed by an automatic glucose analyzer (YSI 23AM; Yellow Springs Instruments, Yellow Springs, OH). Concentrations of lactate in plasma were converted to whole-blood values, as previously described (12). From measurements of plasma glucose concentrations and specific activity of the radiolabeled glucose in the last 30 min of the resting period, the individual distribution volume of glucose was calculated (13). The rates of glucose appearance (Ra) and disappearance (Rd) were calculated by using the formulas for non–steady-state conditions (14,15) and the consecutive sliding-fit technique (16). The metabolic clearance rate (MCR) for glucose was calculated as the glucose Rd divided by the corresponding plasma glucose concentration.

One-way analysis of variance (ANOVA) for repeated measures was used to detect significant changes with time. Two-way ANOVA for repeated measures was used to detect significant differences between blockade and corresponding control studies. Single differences were determined using the paired Student’s t test. The level of significance was set at P < 0.05 in two-tailed testing. All data are reported as means ± SE.

Vo2 and workload.

Vo2max was 3.6 ± 0.1 l/min. During exercise, the absolute workload was 151 ± 7 (α1) and 188 ± 8 (α2) W (P < 0.05). Oxygen uptake rates were 2.3 ± 0.1 and 2.0 ± 0.1 l/min for α1-receptor blockade and control, respectively (P < 0.05), and 2.4 ± 0.1 and 2.5 ± 0.1 l/min for α2-receptor blockade and control, respectively (P > 0.05). Thus, the relative workload was 63 ± 1 and 55 ± 2% of Vo2max for α1-receptor blockade and control, respectively (P < 0.05), and 68 ± 2 and 69 ± 1% of Vo2max for α2-receptor blockade and control, respectively (P > 0.05).

Hemodynamic parameters.

In the α1-receptor blockade study, heart rates were higher (P < 0.05) during rest, exercise, and recovery compared with control (Fig. 1). In contrast, the α2-receptor blockade did not influence the heart rate in any phase of the experiments (Fig. 1). Mean arterial blood pressure was not different in the blockade experiments compared with the corresponding control experiments in any of the phases (rest, exercise, or recovery) (Fig. 1).

Glucose kinetics.

In the α1-receptor blockade and control experiments, plasma glucose concentrations were similar during rest and recovery (P > 0.05) (Fig. 2). In the blockade experiment, plasma glucose concentrations increased during the first 30 min of exercise then decreased, whereas there was a continuous slight decrease in the control experiment. During exercise, plasma glucose concentrations were significantly higher in the blockade compared with the control experiment (P < 0.01) (Fig. 2).

In the α2-receptor blockade and control experiments, plasma glucose concentrations decreased (P < 0.05) in response to exercise (Fig. 2). During exercise, there was a tendency for the plasma glucose concentration to be lower during the blockade than during the control experiment (P = 0.0505) (Fig. 2).

The Ra and Rd were always similar between blockade and control experiments during rest and recovery. However, during exercise with α1-receptor blockade, the Ra tended to be higher compared with the control experiment (P = 0.052), and at one time point (t = 52.5 min), a significant difference was seen (P < 0.01; paired t test) (Fig. 2). The Rd continued to increase throughout the exercise and generally tended (P = 0.08) to be higher compared with during the control experiment (Fig. 2). In the α1 control experiments, the Ra and Rd leveled off after 20 min of exercise. A significant difference in the Rd between α1-receptor blockade and control was seen at the end of exercise (P < 0.02, paired t tests) (Fig. 2). In the α2-receptor blockade experiments, the Ra and Rd were always lower during exercise compared with the control experiments (P < 0.002) (Fig. 2).

Insulin, C-peptide, and glucagon concentrations.

Plasma insulin concentrations were similar at rest in all experiments (Fig. 3). However, resting values during α1-receptor blockade tended to be higher (P < 0.1) compared with all other experiments. In the α1-receptor blockade and corresponding control experiments, plasma insulin concentrations decreased during exercise (P < 0.05), but the values were significantly higher during exercise and recovery in the blockade compared with the control experiment (P < 0.05). However, the relative decrease in plasma insulin concentrations in response to exercise was not different during blockade compared with control (P > 0.05) (Fig. 3). During exercise with α2-receptor blockade, an increase in plasma insulin concentrations was seen in the first 20 min of the exercise, followed by a decrease. In the corresponding control experiment, a decrease was seen throughout the exercise period (P < 0.05) (Fig. 3). During exercise, plasma insulin concentrations were significantly higher with α2-receptor blockade compared with control (P < 0.05), whereas no difference was seen in the recovery period (Fig. 3).

At rest, plasma C-peptide concentrations were similar in all four experiments (Fig. 4). During exercise, plasma C-peptide decreased with and without α1-receptor blockade (P < 0.05). The decrease in C-peptide concentrations from resting values was less with α1-receptor blockade compared with control values (P < 0.05). During exercise with α2-receptor blockade, plasma C-peptide concentrations were significantly higher compared with control values (P < 0.05), and the decrease of C-peptide seen in the control experiment was not seen in the α2-receptor blockade experiment (P < 0.05) (Fig. 4).

Plasma concentrations of glucagon always increased during and immediately after exercise (P < 0.05), but no difference between the experiments was seen during rest, exercise, or recovery (Fig. 5).

Catecholamines.

Plasma epinephrine concentrations were similar at rest in all four experiments; in response to exercise, epinephrine always increased (P < 0.05) (Fig. 6). During exercise with α1-receptor blockade, epinephrine concentrations were higher compared with control values (P < 0.05), whereas no general differences were seen between α2-receptor blockade and control values (Fig. 6).

Plasma norepinephrine concentrations were always higher during α1-receptor blockade compared with control values, at rest and during exercise and recovery (P < 0.02) (Fig. 6). In the α2 experiments, norepinephrine concentrations were similar at rest and during recovery, but during exercise tended to be higher with α2-receptor blockade compared with control values (P < 0.1) (Fig. 6).

Metabolites.

The lactate concentrations in blood were similar at rest in all four experiments (Fig. 7). During α2-receptor blockade and corresponding control studies, there was no difference in lactate concentrations at any time. During exercise and recovery with α1-receptor blockade, lactate concentrations were higher compared with (P < 0.05) control values (Fig. 7).

Plasma concentrations of d-β-hydroxybutyrate (representing ketone bodies) did not change in response to exercise during α1-receptor blockade compared with control (0.15 ± 0.05 vs. 0.10 ± 0.04 mmol/l and 0.18 ± 0.05 vs. 0.15 ± 0.06 mmol/l for rest and exercise, respectively) or during α2-receptor blockade compared with control (0.14 ± 0.03 vs. 0.15 ± 0.07 mmol/l and 0.17 ± 0.04 vs. 0.13 ± 0.04 mmol/l for rest and exercise, respectively). However, during recovery, the concentration was always increased (P < 0.05), and was similar in all four experiments (α1-receptor blockade, 0.41 ± 0.11; α1 control, 0.61 ± 0.19; α2-receptor blockade, 0.63 ± 0.09; α2 control, 0.53 ± 0.15).

Plasma concentrations of FFAs (Fig. 8) were similar during rest in all experiments. No changes were seen with exercise, except for a slight increase during α2-receptor blockade (P < 0.05). FFA concentrations always peaked in the early recovery phase, then quickly decreased toward resting values (Fig. 8).

Plasma concentrations of glycerol were similar in both α1- and α2-receptor blockade and corresponding control studies in all three phases of the experiment. Glycerol always peaked at the end of exercise (Fig. 8).

The main conclusion of the present study is that the exercise-induced decrease in plasma insulin concentrations is mediated via α2-receptors. When yohimbine, a specific blocker of α2-receptors, was given, plasma insulin concentrations during exercise were significantly higher compared with control values (Fig. 3). In fact, with α2-receptor blockade, plasma insulin concentrations increased during the first part of the exercise, then decreased. An increase in plasma insulin concentrations during exercise is remarkable and can only be attributed to blockade of the α2-receptors. The fact that plasma insulin concentrations decreased slightly in the latter part of the exercise may have been because of a fading effect of yohimbine, but other studies using the same dosage, time frame, and route of administration of yohimbine have shown an effect on lipolysis (7) and plasma norepinephrine (17) for >105 min. However, it is noteworthy that α2-receptor blockade did not have an effect on plasma insulin concentrations at rest and especially not during recovery. Thus, the rebound increase in plasma insulin concentrations seen immediately after exercise was not diminished by α2-receptor blockade, which, taken together with the fact that plasma insulin concentrations did decrease slightly in the latter part of the exercise with α2-receptor blockade, may indicate that the blockade was not complete throughout the whole exercise period. In support of the view that α2-receptor blockade abolished the normal exercise-induced decrease in plasma insulin secretion was the finding that C-peptide concentrations changed only minimally (Fig. 4). However, because of a considerably lower metabolic clearance rate than insulin, peripheral C-peptide concentrations may not accurately reflect insulin secretion rates when these change rapidly (18).

In contrast to the experiments with α2-receptor blockade, selective α1-receptor blockade with prazosin did not abolish the normal exercise-induced decrease in plasma insulin or C-peptide concentrations (Figs. 3 and 4). Although a significant difference in plasma insulin and C-peptide concentrations existed during exercise between the α1-receptor blockade and the corresponding control experiments, this was to some extent attributable to differences in concentrations at the onset of exercise (Figs. 3 and 4). The reason for the increased (α1-receptor blockade versus control) insulin and C-peptide concentrations in the resting pre-exercise period and, more importantly, during exercise can be explained by the fact that α1-receptor blockade inhibits vasoconstriction in the splanchnic vascular bed (see the discussion of workloads below). Thus, splanchnic blood flow was most likely higher and its drainage into the peripheral circulation markedly greater during α1-receptor blockade compared with the corresponding control study. This means that delivery of insulin and C-peptide to the sampling site (arterialized venous blood) was greater during α1-receptor blockade versus control, causing the higher plasma concentrations (Figs. 3 and 4). Because of the relatively short half-life of insulin (∼5 min), we found only a minor impact of the higher insulin concentrations on the exercise-induced decrease. In contrast, a less pronounced (compared with control) decrease of C-peptide concentrations during exercise would be expected in a situation with increased delivery (as during the exercise period with α1-receptor blockade) and a relatively long half-life (∼30 min). In the recovery period, rebound increases were more pronounced during the α1-receptor blockade compared with the corresponding control experiment (Figs. 4 and 5). This was most likely attributable to the fact that plasma norepinephrine concentrations were considerably higher during exercise with α1-receptor blockade compared with during control experiments, then quickly declined with the termination of exercise (Fig. 7); therefore, the relief of inhibition had a greater impact on insulin secretion in the α1-receptor blockade study compared with in the corresponding control study.

The present results are in accordance with previous results obtained in swimming mice (8) but are in contrast to one study in humans (7), in which it was found that yohimbine (0.2 mg/kg p.o., given 45 min before exercise at 60% of maximal aerobic capacity) had no effect on the exercise-induced decrease in plasma insulin concentrations. However, the focus of that study (7) was not primarily to study the effect on plasma insulin secretion (for example, plasma C-peptide concentrations were not measured). It is difficult to compare the present with the past (7) study, as exercise-related details (e.g., maximal aerobic capacity, heart rate, and oxygen consumption) were not reported (7). Therefore, we are unable to offer an explanation for the divergence in the effect of α2-receptor blockade on insulin secretion during exercise. The data on plasma insulin concentrations in the present study are in line with those from another study (19), in which a peripherally selective α2-receptor antagonist (MK-467) was infused before and during exercise. However, in that study (19), an exercise-induced decrease in insulin during placebo treatment could not be demonstrated, probably because of an inadequate exercise protocol. Finally, the importance of α2-receptors in the exercise-induced inhibition of insulin secretion is supported by the study from Natali et al. (20), in which selective α2-receptor blockade with deriglidole prevented epinephrine-induced inhibition of insulin secretion.

In the current study, we used different absolute workloads in the α1 and α2 experiments. When choosing the workloads, we aimed at near maximal workloads that could be maintained for 1 h, given that the inhibition of insulin secretion during exercise is directly related to the intensity of exercise. However, during exercise with α1-receptor blockade, shunting of blood from the gastrointestinal compartment to the working muscles is inhibited. Normally, the exercise-induced reduction of splanchnic blood volume and flow may be as much as 40–50% of resting values (21), but some of this reduction does not occur with α1-receptor blockade. The effect of α1-receptor blockade on splanchnic blood flow was not measured in the present study, but the exaggerated heart rate response to exercise did indicate that the effect was not trivial (Fig. 1). The consequences of a lesser reduction in splanchnic blood volume and flow are that a relative lack of blood—and thereby oxygen—supply to the working muscles becomes apparent, nonoxidative glycolysis increases with subsequent considerable lactate release, and the subjects become fatigued. Therefore, the workload had to be diminished during α1-receptor blockade and the corresponding control experiment; comparisons between the responses during α1- and α2-receptor blockade (and corresponding control experiments) should take this into account. The described effect of exercise with α1-receptor blockade on the normal shunting of blood was also reflected by the higher oxygen uptake rates, and thus relative intensity, during α1-receptor blockade compared with the control study.

Plasma glucose kinetics during exercise were quite influenced by α1-receptor blockade (Fig. 2). Normally, during exercise, a slight decrease of plasma glucose concentrations is seen, but with α1-receptor blockade, plasma glucose concentrations increased markedly during the first 30 min of exercise, then clearly decreased toward the pre-exercise level. Calculations of glucose Ra and Rd revealed that the immediate increase was attributable to an enhanced Ra (compared with control) and not to changes in the Rd (Fig. 2). The rapid increase in the Ra, with the subsequent increase in glucose concentrations (Fig. 2), can be attributed to the initial effect of the increased catecholamine concentrations (Fig. 6). The enhanced glucose Ra seen during exercise with α1-receptor blockade was most likely attributable to hepatic gluconeogenesis, and lactate may have served as substrate (Fig. 7), but the tracer used in the present study did not allow a definite conclusion about this. The decrease in plasma glucose concentrations during the latter part of exercise with α1-receptor blockade (Fig. 2) was because of continuing increases in the glucose Rd, with only a tendency to steady state at the end of the exercise period (Fig. 2). From the data, it must be concluded that during α1-receptor blockade, the disappearance of glucose from plasma during exercise is enhanced (most likely representing increased skeletal muscle glucose uptake). The mechanism is unclear, but it may be attributable to inhibition of the sympathoadrenal vasoconstrictor component in the skeletal muscle, thus allowing increased glucose supply to the working muscles. In addition, during exercise with α1-receptor blockade, the relative workload was higher compared with the corresponding control study, which may have resulted in an amplification of motor unit recruitment, especially considering the fact that the subjects were not highly trained athletes. Furthermore, the differences in the glucose Rd between exercise with and without α1-receptor blockade seemed to be the consequence of differences in the prevailing plasma glucose concentrations, given that there was no difference in the MCR (Fig. 2).

During exercise with and without α2-receptor blockade, plasma glucose concentrations decreased to a similar extent with exercise. There was, however, a tendency (P = 0.0505) to lower plasma glucose concentrations during α2-receptor blockade versus during the control study (Fig. 2), which could be expected, considering the higher plasma insulin concentrations (Fig. 3). During exercise, the glucose Ra was reduced by the α2-receptor blockade (Fig. 2). This may have been a direct effect on hepatic α2-receptors, and/or may have been attributable to the fact that plasma insulin concentrations were significantly higher with than without α2-receptor blockade (Fig. 3). In the resting nonexercise situation, there was no effect of α2-receptor blockade on glucose Ra, as has been also shown by others (20). Surprisingly, the glucose Rd was significantly lower with α2-receptor blockade compared with the corresponding control study (Fig. 2). A decreased peripheral glucose uptake could be attributable to a decrease in glucose delivery (i.e., blood flow). Thus, α2-receptor blockade should result in a pronounced vasoconstriction because of blockade of presynaptic α2-receptors. However, the similar blood lactate concentrations (Fig. 7) during exercise with and without α2-receptor blockade speak against such a mechanism. Hence, we speculated that α2-receptor blockade partially inhibits skeletal muscle glucose uptake during exercise via an unknown mechanism. This was an unexpected finding, in particular because of the well-known additive effect of exercise and insulin on skeletal muscle glucose uptake.

Our use of labeled glucose, with the subsequent calculations and approximations to estimate glucose Ra and Rd during a non–steady-state condition, such as exercise, can be criticized (22). However, in a recent study, this method was compared with the direct arterial-hepatic vein balance technique, and it was concluded that during exercise in humans, determination of hepatic glucose production can be performed equally well with the two techniques (23).

In human fat cells, the lipolytic activity is subject to regulation via activation of both β- and α2-adrenoceptors, with an increase and decrease in lipolytic activity with β- and α2-receptor stimulation, respectively. Thus, treatment with an α2-antagonist would be expected to increase lipolytic activity, resulting in increases in plasma FFAs and glycerol concentrations. Such an effect of α2-antagonism has also been shown in the past in resting humans (7). Furthermore, during combined exercise and α2-receptor blockade, an additive effect on FFAs and glycerol was demonstrated (7). This additive effect was probably attributable to the combined effect of exercise-induced increase in sympathoadrenal activity stimulating fat cell β-receptors and blockade of α2-receptors. However, in that study (7), plasma insulin concentrations were unaffected by α2-receptor blockade; that is, they decreased with exercise. In the present study, we found that the exercise-induced decreased insulin secretion was abolished by α2-receptor blockade, and therefore the antilipolytic effect of insulin was preserved. Thus, in the present study, only a marginal increase in FFA concentrations during exercise was seen with α2-receptor blockade, and no difference in the glycerol concentration with or without α2-receptor blockade was seen (Fig. 8).

In summary, we have now shown that the exercise-induced decrease in insulin secretion is mediated via α2-adrenoceptors. Blockade of α1-adrenoceptors did not affect the normal decrease in plasma insulin concentrations during exercise. Glucose kinetics during exercise were differently influenced by α1- and α2-adrenoceptor blockade. With α1-receptor blockade, plasma glucose concentrations initially increased, then decreased as the exercise continued because of the enhanced glucose Rd. In contrast, during a 60-min exercise bout with α2-receptor blockade, plasma glucose concentrations continuously decreased, and the glucose Ra and Rd were diminished. A decreased glucose Ra during exercise with α2-adrenoceptor blockade can be attributed to the accompanying elevation in plasma insulin concentrations; the reason for a diminished glucose Rd, however, remains obscured.

FIG. 1.

Heart rate and blood pressure before, during, and after 60-min ergometer bicycle exercise in eight men with α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study.

FIG. 1.

Heart rate and blood pressure before, during, and after 60-min ergometer bicycle exercise in eight men with α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study.

Close modal
FIG. 2.

Plasma glucose, Ra, Rd, and MCR before, during, and after 60-min ergometer bicycle exercise in eight men with α1-receptor blockade (prazosin; 1 mg at t = −60 min;) and without (control), and with α2-receptor blockade (yohimbine; 15 mg at t = −45 min) and without (control). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study; †significant difference (P < 0.05) at single time points.

FIG. 2.

Plasma glucose, Ra, Rd, and MCR before, during, and after 60-min ergometer bicycle exercise in eight men with α1-receptor blockade (prazosin; 1 mg at t = −60 min;) and without (control), and with α2-receptor blockade (yohimbine; 15 mg at t = −45 min) and without (control). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study; †significant difference (P < 0.05) at single time points.

Close modal
FIG. 3.

Plasma insulin concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with (A) α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and (B) with α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study; #significantly increased (P < 0.05) compared with values at rest and with preceding value.

FIG. 3.

Plasma insulin concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with (A) α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and (B) with α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study; #significantly increased (P < 0.05) compared with values at rest and with preceding value.

Close modal
FIG. 4.

Plasma C-peptide concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with (A) α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with (B) α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study.

FIG. 4.

Plasma C-peptide concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with (A) α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with (B) α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study.

Close modal
FIG. 5.

Plasma glucagon concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with (A) α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with (B) α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE.

FIG. 5.

Plasma glucagon concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with (A) α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with (B) α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE.

Close modal
FIG. 6.

Plasma epinephrine and norepinephrine concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study; †significant difference at single time points (P < 0.05).

FIG. 6.

Plasma epinephrine and norepinephrine concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study; †significant difference at single time points (P < 0.05).

Close modal
FIG. 7.

Blood lactate concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with (A) α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with (B) α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study.

FIG. 7.

Blood lactate concentrations before, during, and after 60-min ergometer bicycle exercise in eight men with (A) α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with (B) α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant difference (P < 0.05) between blockade and corresponding control study.

Close modal
FIG. 8.

Plasma FFAs and glycerol concentration before, during, and after 60-min ergometer bicycle exercise in eight men with α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant increase (P < 0.05) with time in blockade study.

FIG. 8.

Plasma FFAs and glycerol concentration before, during, and after 60-min ergometer bicycle exercise in eight men with α1-receptor blockade (prazosin; 1 mg at t = −60 min; ▾) and without (control; ○), and with α2-receptor blockade (yohimbine; 15 mg at t = −45 min; ▪) and without (control; ○). Data are means ± SE. *Significant increase (P < 0.05) with time in blockade study.

Close modal

This study was supported by grant from the Danish National Research Foundation (no. 504-14), the Novo-Nordic Foundation, the Danish Sports Research Council, the P. Carl Petersen Foundation, the Fondation of 1870, the Direktør Jacob Madsen og hustru Olga Madsens Fondation, and the Direktør Ib Henriksens Fondation.

We thank Regitze Kraunsøe and Jeppe Bach for excellent technical assistance.

1.
Galbo H:
Hormonal and Metabolic Adaptation to Exercise.
New York, Thieme-Stratton,
1983
2.
Brisson GR, Malaisse-Lagae F, Malaisse WJ: Effect of phentolamine upon insulin secretion during exercise.
Diabetologia
7
:
223
–226,
1971
3.
Galbo H, Christensen NJ, Holst JJ: Catecholamines and pancreatic hormones during autonomic blockade in exercising man.
Acta Physiol Scand
101
:
428
–437,
1977
4.
Nakaki T, Nakadate T, Kato R: α2-Adrenoceptors modulating insulin release from isolated pancreatic islets.
Nauyun Schmiedebergs Arch Pharmacol
313
:
151
–153,
1980
5.
Morgan NG, Montague W: Studies on the mechanism of inhibition of glucose-stimulated insulin secretion by noradrenaline in rat islets of Langerhans.
Biochem J
226
:
571
–576,
1985
6.
Lacey RJ, Chan SL, Cable HC, James RF, Perrett CW, Scarpello JH, Morgan NG: Expression of alpha 2- and beta-adrenoceptor subtypes in human islets of Langerhans.
J Endocrinol
148
:
531
–543,
1996
7.
Galitzky J, Taouis M, Berlan M, Rivière D, Garrigues M, Lafontan M: α2-Antagonist compounds and lipid mobilization: evidence for a lipid mobilizing effect of oral yohimbine in healthy male volunteers.
Eur J Clin Invest
18
:
587
–594,
1988
8.
Karlsson S, Ahrén B: Insulin and glucagon secretion in swimming mice: effects of autonomic-receptor antagonism.
Metabolism
39
:
724
–732,
1990
9.
Kjær M: Hepatic fuel metabolism during exercise. In
Exercise Metabolism
. Hargreaves M, Ed. Champaign, IL, Human Kinetics,
1995
, p.
73
–97
10.
Kjaer M, Secher NH, Bach FW, Galbo H: Role of motor center activity for hormonal changes and substrate mobilization in humans.
Am J Physiol
253
:
R687
–R695,
1987
11.
Dela F, Mikines KJ, Linstow vM, Secher NH, Galbo H: Effect of training on insulin mediated glucose uptake in human skeletal muscle.
Am J Physiol
263
:
E1134
–E1143,
1992
12.
Wandrup J, Tvede K, Grinsted J, Jordening H: “Stat” measurements of L-lactate in whole blood and cerebrospinal fluid assessed.
Clin Chem
35
:
1740
–1743,
1989
13.
Hetenyi GJ, Norwich KH: Validity of the rates of production and utilization of metabolites as determined by tracer methods in intact animals.
Fed Proc
33
:
1841
–1848,
1974
14.
Radziuk J, Norwich KH, Vranic M: Experimental validation of measurements of glucose turnover in nonsteady state.
Am J Physiol
234
:
E84
–E93,
1978
15.
Steele R, Wall JS, De Bodo RC, Altszuler N: Measurement of size and turnover rate of body glucose pool by the isotope dilution method.
Am J Physiol
187
:
15
–24,
1956
16.
Cherrington AD, Vranic M: Effect of arginine on glucose turnover and plasma free fatty acids in normal dogs.
Diabetes
22
:
537
–543,
1973
17.
Galitzky J, Rivière D, Tran MA, Montastruc JL, Berlan M: Pharmacodynamic effects of chronic yohimbine treatment in healthy volunteers.
Eur J Clin Pharmacol
39
:
447
–451,
1990
18.
Polonsky KS, Frank B, Pugh W, Addis A, Karrison T, Meier P, Tager H, Rubenstein AH: The limitations to and valid use of C-peptide as a marker of the secretion of insulin.
Diabetes
35
:
379
–386,
1986
19.
Sciberras DG, Reed JW, Elliott C, Blain PG, Goldberg MR: The effects of a peripherally selective alpha 2-adrenoceptor antagonist, MK-467, on the metabolic and cardiovascular response to exercise in healthy man.
Br J Clin Pharmacol
37
:
39
–44,
1994
20.
Natali A, Gastaldelli A, Galvan AQ, Sironi AM, Ciociaro D, Sanna G, Rosenzweig P, Ferrannini E: Effects of acute alpha 2-blockade on insulin action and secretion in humans.
Am J Physiol
274
:
E57
–E64,
1998
21.
Rowell LB:
Human Circulation Regulation During Physical Stress.
N York/Oxford, Oxford University Press,
1986
22.
Cobelli C, Mari A, Ferrannini E: Non-steady state: error analysis of Steele’s model and developments for glucose kinetics.
Am J Physiol
252
:
E679
−E689,
1987
23.
Bergeron R, Kjaer M, Simonsen L, Bulow J, Galbo H: Glucose production during exercise in humans: a-hv balance and isotopic-tracer measurements compared.
J Appl Physiol
87
:
111
–115,
1999

Address correspondence and reprint requests to Flemming Dela, MD, Associate Professor, Department of Medical Physiology, the Panum Institute, Bldg. 12.4.7., University of Copenhagen, Blegdamsvej 3, 2200 Copenhagen N, Denmark. E-mail: [email protected].

Received for publication 14 November 2000 and accepted in revised form 24 April 2001.

ANOVA, analysis of variance; FFA, free fatty acid; MCR, metabolic clearance rate; Ra, rate of glucose appearance; Rd, rate of glucose disappearance.