| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Medicine (B.E.d.G., S.J.R., C.J.T., J.W.M.L., P.S.), Medical Psychology (S.P.v.d.W.), Chemical Endocrinology (C.G.J.S.), and Pharmacology-Toxicology (P.S.), University Medical Center Nijmegen, Nijmegen 6500HB, The Netherlands
Address all correspondence and requests for reprints to: Paul Smits, M.D., Ph.D., Professor of Pharmacology, Department of Pharmacology-Toxicology 233, University Medical Center Nijmegen, PO Box 9101, 6500 HB Nijmegen, The Netherlands. E-mail: p.smits{at}pharmtox.umcn.nl.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Despite the undisputed role of hypoglycemia per se in the development of hypoglycemia unawareness (4, 5, 6), the underlying mechanism remains incompletely understood. Increased blood-to-brain glucose transport has been suggested as underlying mechanism (7) but may not explain the blunted counterregulatory responses after short duration of hypoglycemia (8). Another proposed mechanism suggests the cortisol component of the metabolic response to hypoglycemia as mediator (9). Antecedent increases in plasma cortisol have been found to reduce responses of adrenaline, noradrenaline, and GH to subsequent hypoglycemia (9, 10, 11, 12), but they do not invariably seem to affect symptom responses. Therefore, additional factors are likely to be involved.
Adrenaline is a key component of the hypoglycemia response profile and closely related to the appearance of hypoglycemic warning symptoms (13). Conversely, the loss of hypoglycemic warning symptoms, which characterizes clinical hypoglycemia unawareness, can result from either a reduced adrenaline response to hypoglycemia or from a reduced sensitivity to catecholamines (14, 15, 16, 17). Because adrenaline still increases at least severalfold in response to hypoglycemia even in patients with severe counterregulatory failure, patients with repeated hypoglycemic events may have chronically elevated adrenaline levels. Animal studies have indicated that prior elevation of adrenaline (18) and repeated stress (which primarily stimulates adrenaline) (19) down-regulate sympathetic responses to novel stress, whereas prior exposure to ß-adrenergic agonists (20, 21) and strenuous exercise (22) reduce ß-adrenergic sensitivity. Based on these data, we hypothesized that the adrenaline response to antecedent hypoglycemia may contribute to the hypoglycemia unawareness syndrome by exerting a down-regulating effect on adrenaline responses to and adrenergic sensitivity during subsequent hypoglycemia. Therefore, the aim of the present study was to determine the effect of prior elevation of plasma adrenaline on hormone and symptom responses to subsequent hypoglycemia.
| Subjects and Methods |
|---|
|
|
|---|
We studied 16 healthy nonsmoking normotensive volunteers (eight males and eight females), aged 22.5 ± 0.5 yr and with a body mass index of 21.6 ± 0.4 kg·m-2. Subjects were studied on two occasions performed at least 3 wk apart. Female subjects were studied at exact 4-wk intervals to ensure that experiments were performed during corresponding periods of the menstrual cycle. The order of the experiments was randomized and performed in a single blind fashion. Studies were approved by the medical ethics committee of the University Medical Center Nijmegen and written informed consent was obtained before participation.
Experimental design
Each participant was admitted to the University Medical Center Nijmegen at 0800 h, having abstained from alcohol for 24 h and from caffeine-containing substances for 48 h. Under local anesthesia (Xylocaine 2%), the brachial artery of the nondominant arm was cannulated (Angiocath 20-gauge, Becton Dickinson, Sandy, UT) for blood sampling and for hemodynamic monitoring (Monitor 378341A, Hewlett-Packard GmbH, Boeblingen, Germany). The antecubital vein of the dominant contralateral arm was cannulated for administration of adrenaline (International Medication Systems Ltd., Slough, UK) or placebo (NaCl 0.9%), and for glucose 20% and insulin (Actrapid, Novo Nordisk, Bagsvaerd, Denmark) infusions.
After a 30-min equilibration period following insertion of the cannulae, blood was sampled for baseline measurements. Thereafter, adrenaline (or an equivalent volume of placebo solution) was administered intravenously at an initial rate of 0.04 µg·kg-1·min-1 for 20 min, followed by 0.06 and 0.08 µg·kg-1·min-1 for another 20 min each (total amount, 3.6 µg·kg-1 body weight over 60 min). Target plasma adrenaline levels were 69 nmol·liter-1 to match those obtained in response to hypoglycemia (23, 24). At the end of the final infusion step, blood was sampled and the infusion discontinued.
After a 3-h rest period, a stepped hyperinsulinemic (360 pmol·m-2·min-1) hypoglycemic glucose clamp procedure was performed, as described before (24). Using a variable infusion of glucose 20%, the arterial plasma glucose concentration was sequentially clamped at 5.0, 3.5, and 2.5 mmol·liter-1 at hourly intervals, guided by plasma glucose levels measured in duplicate every 5 min by the glucose oxidation method (Beckman Glucose Analyzer II, Beckman, Fullerton, CA). Arterial blood samples for all analytes other than glucose were drawn before and after adrenaline or placebo infusion, before the start of the clamp and at each glycemic plateau during the clamp, and kept on melting ice until centrifugation. After centrifugation at 4 C, plasma was stored at -20 C until analysis.
At 15-min intervals during the clamp, a semiquantitative symptom questionnaire was administered. Subjects registered six neuroglycopenic symptoms (blurred vision, difficulty speaking, feeling faint, difficulty thinking, confused, and sleepiness), six autonomic symptoms (tingling, sweating, feeling hungry, palpitations, anxiety, trembling), four general symptoms (dry mouth, weakness, nausea, and headache), and two dummy symptoms (yellow vision and pain in the legs) on a scale from 0 (absent) to 6 (severe). The appearance of a sweating response was detected objectively using a sensor applied to the forearm (Evaporimeter EP1, Servomed AB, Stockholm, Sweden) (25).
Two cognitive tests were applied at 20-min intervals during the clamp, the choice reaction time task (CRT), to test vigilance and attention, and the symbol digit test (SDT) from the Dutch version of the Wechsler Adult Intelligence Scale (26), to test psychomotor and cognitive speed. On the CRT, a computer pad is used that contains four buttons with corresponding lights, of which only the light corresponding to the start button is illuminated. Subjects are asked to press the start button until the illumination switches to one of the other three lights, after which they have to press the corresponding button of that light. The reaction timedefined as the sum of the actual reaction time (i.e. the time to depress the start button) and movement time (i.e. the time to press the subsequent button)is recorded. On the SDT, subjects are provided with a sheet of paper containing a total of 100 blank squares in four rows paired with one of 10 randomly selected symbols. Using the digit-symbol key displayed at the top of the sheet, the subjects are asked to manually insert appropriate digits (09) in the squares below the symbols. The score is calculated as the number of correct substitutions within 60 sec. Subjects were provided with sufficient practice on each test before the experiments.
Analytical methods
Plasma insulin and plasma glucagon were measured by RIAs (24). Plasma adrenaline and noradrenaline were analyzed by HPLC using a modification of an earlier described laboratory procedure (27). Plasma cortisol was assayed using the TDx batch analyzer of Abbott Laboratories (Abbott Diagnostics, Hoofddorp, The Netherlands), as described before (24).
Calculations and statistical analyses
The glycemic threshold for detection of sweat production was defined for each individual as the plasma glucose level at which dew-point electrode readings showed at least doubling of baseline values. Serial data were compared between groups by repeated measures ANOVA and differences in means were tested using paired Students t test. Correlations were examined by Pearsons correlation analysis. For calculations and statistical analyses, the SPSS personal computer software package (SPSS, Chicago, IL) was used. P < 0.05 was considered statistically significant. Results in tables and figures are expressed as means ± SEM, unless otherwise indicated.
| Results |
|---|
|
|
|---|
After adrenaline infusion, arterial plasma adrenaline levels increased from 0.10 ± 0.02 to 8.11 ± 0.42 nmol·liter-1 (P < 0.001), values comparable to the usual adrenaline response to hypoglycemia (24). Adrenaline transiently increased plasma glucose (P < 0.001) (Fig. 1
), heart rate, and systolic blood pressure (sBP), and decreased diastolic blood pressure (dBP) and mean arterial pressure (MAP) (all P < 0.01) (Table 1
). After cessation of adrenaline, glucose levels normalized at levels slightly, but significantly, lower than after placebo. Placebo infusion had no cardiovascular or metabolic effects, except for a slight fall in heart rate (P = 0.032).
|
|
Plasma insulin levels increased from 65 ± 8 pmol·liter-1 at baseline to stable levels of 521 ± 24 pmol·liter-1 during the clamp in the placebo control study. Corresponding values after prior adrenaline infusion were 59 ± 6 pmol·liter-1 at baseline and 524 ± 16 pmol·liter-1 during the clamp. Mean (±SD) glucose levels at target glycemic plateaus were similar in both study-arms, and measured 4.9 ± 0.1 [coefficient of variation (CV), 4.1%], 3.4 ± 0.1 (CV, 4.3%), and 2.6 ± 0.1 (CV, 5.0%) mmol·liter-1 (Fig. 1
).
Responses to hypoglycemia
In response to hypoglycemia, all of the following parameters increased: plasma levels of glucagon, adrenaline, noradrenaline, and cortisol (all P < 0.001) (Table 2
), autonomic and neuroglycopenic symptom scores (both P < 0.001) (Fig. 2
), and heart rates (from 64 ± 2 to 68 ± 2 b·min-1, P < 0.001). Hypoglycemia caused decrements in sBP, dBP, and MAP (all P < 0.001) (Fig. 3
), and in performance on cognitive function tests, as reflected by an increase in reaction time on the CRT (from 466 ± 13 to 535 ± 15 msec, P < 0.001) and a fall in number of correct substitutions on the SDT (P = 0.001) (Fig. 4
), respectively.
|
|
|
|
The glucose infusion rate (GIR) to maintain hypoglycemic nadir was significantly higher after prior adrenaline than after prior placebo (10.1 ± 1.3 vs. 7.3 ± 1.3 µmol·kg-1·min-1, P = 0.017). Prior administration of adrenaline did not affect the release of glucagon, adrenaline, noradrenaline or cortisol in response to hypoglycemia (Table 2
). Responses of autonomic and neuroglycopenic symptom scores to hypoglycemia were similar between prior placebo and prior adrenaline studies (Fig. 2
), and sweating responses were elicited at comparable glycemic thresholds [2.8 ± 0.1 mmol·liter-1 (prior placebo) vs. 2.7 ± 0.1 mmol·liter-1 (prior adrenaline), P = 0.31]. Prior adrenaline administration reduced the falls in sBP (-7 ± 3 vs. -1 ± 3 mm Hg, ANOVA P = 0.010), dBP (final values, -15 ± 2 vs. -10 ± 1 mm Hg, ANOVA P = 0.005), and MAP (-13 ± 2 vs. -8 ± 2 mm Hg, ANOVA P = 0.006) in response to hypoglycemia (Fig. 3
). There was a significant correlation between the hypoglycemia-induced fall in dBP and GIR at hypoglycemic nadir (R = 0.39, P = 0.027). Prior adrenaline had no effect on the heart rate response to hypoglycemia. After prior adrenaline infusion, hypoglycemia caused a similar increase in reaction time on the CRT (from 475 ± 15 to 550 ± 29 msec, P = 0.007), but did no longer interfere with the number of correct substitutions on the SDT (ANOVA, P = 0.34) (Fig. 4
).
| Discussion |
|---|
|
|
|---|
At hypoglycemic nadir, GIR is inversely related to the glucose-stimulating capacity of metabolic counterregulation as a whole; i.e. the higher the GIR, the lower the ability of metabolic counterregulation to raise plasma glucose. In the present study, counterregulatory hormone responses to hypoglycemia were not different according to pretreatment infusions, so that the higher GIR after adrenaline pretreatment can only be explained by reduced responsiveness to counterregulatory hormones to stimulate glucose. To what extent this reduced responsiveness is caused by reduced endogenous glucose production or by reduced suppression of insulin-induced glucose uptake cannot be determined from our data. In studies investigating the effect of antecedent hypoglycemia or cortisol on responses to subsequent hypoglycemia, it has been reported that increased exogenous glucose requirements during subsequent hypoglycemia are primarily the result of reduced glucose production (9). It could then be hypothesized that the reduction in glucose production was the result of depletion of glycogen stores by prior adrenaline. However, this is unlikely because systemic adrenaline has a very modest effect on (hepatic) glycogenolysis (28) and because any fall in glycogen content would be corrected promptly by the sufficient supply of glucose (and insulin) during the initial phase of the clamp.
Based on adrenalines role in acute glucose counterregulation and because adrenaline exerts its glucose-stimulating effects on liver and muscle primarily through the ß-adrenergic receptor (29), it seems plausible that the reduced responsiveness after prior adrenaline specifically involves the ß-adrenergic receptor (30). In line with this reasoning is the finding that blood pressure responses to hypoglycemia were attenuated after prior adrenaline, and are previous findings of reduced ß-adrenergic sensitivity after short- (20) or long-term (21) exposure to ß-adrenergic agonists. The fact that we were unable to detect an effect of prior adrenaline on autonomic symptom responses may be due to insensitivity of the checklists, or because other factors besides ß-adrenergic stimulation are involved in the development of hypoglycemic symptoms (e.g. cerebral glucopenia) (7). Reduced adrenergic sensitivity has been suggested to play a role in hypoglycemia unawareness (1), to explain observations of unawareness despite normal plasma catecholamine responses to hypoglycemia (31). Recently, ß-adrenergic sensitivity was found to be reduced in type 1 diabetic patients with hypoglycemia unawareness when compared with patients with normal hypoglycemic awareness and controls (14). After avoidance of hypoglycemias, this reduced ß-adrenergic sensitivity improved in parallel with the increase in autonomic warning symptoms (17), whereas ß-adrenergic sensitivity appeared to decrease after an antecedent hypoglycemic episode (32, 33). The present data suggest that an inhibiting effect of hypoglycemia on ß-adrenergic sensitivity (15, 16) is mediated by adrenaline. The extent to which this phenomenon can be extrapolated to diabetic patients with hypoglycemia unawareness requires further study.
Surprisingly, prior adrenaline administration was associated with absence of hypoglycemia-induced deterioration of performance on the SDT cognitive function test. Because prior exposure to stressful stimuli may reduce anxiety caused by certain cognitive tests (34), a lower level of anxiety after adrenaline pretreatment may have led to a better performance. Alternatively, adrenaline-induced desensitization of
1-adrenergic receptors in the brain, stimulation of which has been associated to impaired cognitive function in rats (35), might also have inhibited the hypoglycemia-induced fall in cognitive function after adrenaline. However, neither habituation to stress nor desensitization of
1-adrenergic receptors provide an explanation for the lack of prior adrenaline to affect performance on the CRT cognitive function test. It cannot be excluded that the results obtained on the SDT during hypoglycemia in the prior adrenaline study were to some extent flawed by the lower peak performance obtained during normoglycemia. Moreover, potential limitations apply to the use of cognitive function tests during studies of acute hypoglycemia. Both CRT and SDT are appropriate tests for this purpose as they are susceptible to hypoglycemia, require little time, provide measures with clinical relevance, and do not lead to sustained arousal (36). However, the SDTbut not the CRTrequires fine motor coordination of the (dominant) writing hand. It cannot be excluded that hypoglycemia-induced autonomic symptomse.g. tremor, sweating, agitationto some extent interfered with the performance on the SDT, independent of cortical malfunction (37). This would, again, suggest some sort of reduction in end-organ sensitivity caused by prior adrenaline.
Our study design differed in some aspects from those of other studies investigating the pathogenesis of hypoglycemia-induced counterregulatory failure. Firstly, in most studies on this subject, an intervention on d 1 is followed by measurement of responses to hyperinsulinemic hypoglycemia on d 2 (9, 10, 11, 12). It could be argued that the interval between the intervention and the subsequent clamp study was too short in the present study because we measured counterregulatory responses to subsequent hypoglycemia on the same day. Conceptually and in clinical practice, however, an interval of 24 h is not required for interaction between hypoglycemic episodes. For instance, a single episode of asymptomatic nocturnal hypoglycemia was found to impair both hormonal and symptom responses to hypoglycemia the following morning (38). Recently, Davis and Tate (39) demonstrated that moderate antecedent morning hypoglycemia blunted all neuroendocrine responses to hyperinsulinemic hypoglycemia 2 h later, and to similar extent as the blunting effect of two previous-day hypoglycemic episodes. As such, it is unlikely that the interval time affected our results. Secondly, in many studies the intervention to be investigated is performed twice to ensure a stronger stimulus, whereas in our study we scheduled one intervention. Nevertheless, it has previously been shown that one recent hypoglycemic episode is sufficient to induce counterregulatory impairments (38). Lastly, it could be claimed that 1 h of elevated adrenaline was too short a stimulus. However, the suppressive effect of prior hypoglycemia on responses to subsequent episodes seems to be relatively independent of the initial episodes duration, as even episodes of less than 30 min duration have been reported to cause counterregulatory impairments (40). In addition, plasma adrenaline levels after adrenaline infusion were at least twice as high as those attained by studies using antecedent (moderate) hypoglycemia as stimulus, so that on balance the adrenaline stimuli were at least similar. Despite these considerations, it cannot be totally excluded that prolonged, repeated or higher doses of adrenaline would have produced more pronounced effects.
In summary, we found that prior elevation of plasma adrenaline did not affect hormonal responses to or symptomatic awareness of subsequent hypoglycemia but was associated with higher exogenous glucose requirements and attenuated blood pressure responses during hypoglycemia. The higher exogenous glucose requirements and the attenuated cardiovascular responses after prior adrenaline administration are compatible with reduced responsiveness to counterregulatory hormone action, most likely explained by reduced ß-adrenergic sensitivity. Because hypoglycemia may suppress ß-adrenergic sensitivity (32), our data raise the intriguing possibility that this suppression of ß-adrenergic sensitivity is mediated by the adrenaline response to hypoglycemia, which then would contribute to compromised counterregulatory function. Despite this potential contribution, however, adrenaline is probably not the missing link to explain the suppressive effect of antecedent hypoglycemia on responses to a subsequent episode.
| Acknowledgments |
|---|
| Footnotes |
|---|
Received March 10, 2003.
Accepted August 11, 2003.
| References |
|---|
|
|
|---|
-1-adrenoceptor mediation in the prefrontal cortex. Biol Psychiatry 46:12661274[CrossRef][Medline]
This article has been cited by other articles:
![]() |
P. E. Cryer Mechanisms of Hypoglycemia-Associated Autonomic Failure and Its Component Syndromes in Diabetes Diabetes, December 1, 2005; 54(12): 3592 - 3601. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Francisco, C. Hernandez, R. Galard, and R. Simo Usefulness of Homeostasis Model Assessment for Identifying Subjects at Risk for Hypoglycemia Failure during the Insulin Hypoglycemia Test J. Clin. Endocrinol. Metab., July 1, 2004; 89(7): 3408 - 3412. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |