Exercise-related hypoglycemia is common in intensively treated patients with type 1 diabetes. The underlying mechanisms are not clearly defined. In nondiabetic subjects, hypoglycemia blunts counterregulatory responses to subsequent exercise. It is unknown whether this also occurs in type 1 diabetes. Therefore, the goal of this study was to test the hypothesis that prior hypoglycemia could result in acute counterregulatory failure during subsequent exercise in type 1 diabetes. A total of 16 type 1 diabetic patients (8 men and 8 women, HbA1c 7.8 ± 0.3%) were investigated during 90 min of euglycemic cycling exercise, following either two 2-h periods of previous-day hypoglycemia (2.9 mmol/l) or previous-day euglycemia. Patients’ counterregulatory responses (circulating levels of counterregulatory hormones, intermediary metabolites, substrate flux via indirect calorimetry, tracer-determined glucose kinetics, and cardiovascular measurements) were comprehensively assessed during exercise. Identical euglycemia and basal insulin levels were successfully maintained during all exercise studies, regardless of blood glucose levels during the previous day. After resting euglycemia, patients displayed normal counterregulatory responses to exercise. Conversely, when identical exercise was repeated after hypoglycemia, the glucagon response to exercise was abolished, and the epinephrine, norepinephrine, cortisol, endogenous glucose production, and lipolytic responses were reduced by 40–80%. This resulted in a threefold increase in the amount of exogenous glucose needed to maintain euglycemia during exercise. Our results demonstrate that antecedent hypoglycemia, in type 1 diabetes, can produce acute counterregulatory failure during a subsequent episode of prolonged moderate-intensity exercise. The metabolic consequence of the blunted neuroendocrine and autonomic nervous system counterregulatory responses was an acute failure of endogenous glucose production to match the increased glucose requirements during exercise. These data indicate that counterregulatory failure may be a significant in vivo mechanism responsible for exercise-associated hypoglycemia in type 1 diabetes.

It is now clearly established that microvascular complications of type 1 diabetes can be prevented or delayed by maintaining near-normoglycemia in patients with this disease (1). The target of near-normoglycemia, however, is associated with a threefold increased incidence of severe hypoglycemia and coma (2). Further, in patients with type 1 diabetes, hypoglycemia is frequently associated with physical activity, often resulting in limitation of the patient’s daily activities. The reasons for the high incidence of exercise-associated hypoglycemia in patients with type 1 diabetes are not fully elucidated.

When exposed to hypoglycemia, the human body reacts by triggering a series of neuroendocrine, metabolic, and autonomic nervous system (ANS) responses, aimed at restoring euglycemia. In type 1 diabetes, some counterregulatory responses to hypoglycemia (glucagon) are permanently lost shortly after the onset of the disease (3). In the absence of diabetic autonomic neuropathy, sympathoadrenal responses are preserved and can retain the ability to mount an adequate defense against hypoglycemia. Unfortunately, these residual counterregulatory responses to hypoglycemia can also be reduced. After an episode of hypoglycemia, counterregulatory responses to further hypoglycemia are reduced, rendering hypoglycemia itself more likely to occur. Subsequent episodes of hypoglycemia prolong the duration of blunted responses and therefore perpetuate the likelihood of further hypoglycemia. This phenomenon, defined as “hypoglycemia-associated autonomic failure,” has been documented in healthy subjects, insulinoma patients (4), and type 1 diabetic patients (59).

Recently, an additional concept was also introduced stating that hypoglycemia and exercise may reciprocally blunt their respective counterregulatory responses (10,11). In patients with type 1 diabetes, hypoglycemia is often associated with physical activity. However, limited data exist examining the role of deficient counterregulatory responses per se on the increased incidence of exercise-associated hypoglycemia. Typically, glucagon responses, which are permanently lost during hypoglycemia after a few years’ duration of type 1 diabetes, are preserved during exercise. However, Bottini et al. (12) reported that epinephrine responses to exercise can be reduced in patients with classic diabetic autonomic neuropathy. Furthermore, Schneider et al. (13) have reported blunted neuroendocrine responses during exercise in metabolically well-controlled type 1 diabetic patients. Recently, a study from our laboratory in nondiabetic subjects demonstrated a significant reduction in counterregulatory responses to exercise after antecedent hypoglycemia (11). Rattarasarn et al. (14), on the other hand, reported unchanged counterregulatory responses during exercise after hypoglycemia in a group of type 1 diabetic patients. Thus, the effects of prior hypoglycemia on counterregulatory responses during subsequent exercise in patients with type 1 diabetes are unclear.

The present study was therefore designed to test the hypothesis that antecedent hypoglycemia induces acute counterregulatory failure during next-day exercise. A total of 16 patients with type 1 diabetes (8 men and 8 women) were studied. Each participant performed 90-min bouts of euglycemic cycling exercise of moderate intensity (50% Vo2max) after either two 2-h episodes of hypoglycemia (∼2.9 mmol/l) or resting euglycemia (∼5.0 mmol/l).

Subjects.

We studied 16 patients with type 1 diabetes (8 men and 8 women) aged 28 ± 2 years, BMI 22 ± 1 kg/m2, and HbA1c 7.8 ± 0.3% (normal range 4.0–6.5%). Patients had been diagnosed with type 1 diabetes 13 ± 2 years before recruitment and had no evidence of tissue complications of the disease (retinopathy, renal impairment, or hypertension) or of diabetic autonomic neuropathy (normal increase in R-R interval after Valsalva maneuver, average drop of systolic blood pressure 1 min after standing 4 ± 3 mmHg). Each subject had a normal blood count, plasma electrolytes, and liver function. All gave written informed consent. Studies were approved by the Vanderbilt University human subjects institutional review board.

Preliminary exercise testing.

At least 2 weeks before the initial study, patients performed an incremental work test on a stationary cycle ergometer to determine Vo2max and anaerobic threshold (AT). Air flow, O2, and CO2 concentrations in inspired and expired air were measured by a computerized open-circuit indirect calorimetry cart (Medical Graphics Cardio2 cycle) with a mouthpiece and nose clip system. AT was determined by the V-slope method (15). AT determined by gas exchange corresponds to the onset of an increased lactate/pyruvate ratio in blood and indicates the level of exercise above which anaerobic mechanisms supplement aerobic energy production (16). At workloads below the AT, exercise can be continued for a prolonged period, whereas above the AT, fatigue will occur considerably faster (17). The experimental work rate was established by calculating 80% AT. The AT was detected at 59 ± 3% of Vo2max, and 80% AT corresponded to 47 ± 2% of the subjects Vo2max. This workload was chosen because it is close enough to the AT to produce a physically challenging stress (i.e., large experimental signal) but is sustainable for a prolonged period of time. Subjects studied ranged from sedentary to regularly exercising, although not actively participating in competitive sports. Mean Vo2max for the group was 31 ± 2 ml · kg−1 · min−1 (range 21–43 ml · kg−1 · min−1).

Experimental design.

Each patient was further studied during two separate visits, each lasting 2 consecutive days and including two overnight stays in our Clinical Research Center. During day 1 of one of the visits, patients underwent morning and afternoon 2-h hyperinsulinemic-hypoglycemic clamps (anteHypo). During day 1 of the other visit, patients underwent morning and afternoon hyperinsulinemic-euglycemic control experiments (anteEugly). During day 2 of both visits, all patients performed an identical 90-min exercise protocol under euglycemic conditions. The sequence of anteHypo and anteEugly studies was randomized, and at least 6 weeks were allowed to elapse between the two visits.

Patients were asked to avoid hypoglycemia during the 7 days preceding each visit. Patients checked their blood glucose four times per day and twice weekly at night and reported the recorded values to the investigators before admission. Detection of any value <3.9 mmol/l resulted in rescheduling of the study. Patients were also asked to avoid any exercise and consume their usual weight-maintaining diet for 3 days before each study. Intermediate- or long-acting insulin was administered into the arms for 3 days before a study to eliminate exaggerated insulin absorption from a working muscle during cycle exercise. Each subject was admitted to the Vanderbilt Clinical Research Center at 4:00 p.m. on the afternoon before an experiment. Upon admission, patients were asked to discontinue their usual insulin therapy, and two intravenous cannulas were inserted under 1% lidocaine local anesthesia. One cannula was placed in a retrograde fashion into a vein on the back of the left hand. This hand was placed in a heated box (55–60°C) so that arterialized blood could be obtained. The other cannula was placed in the contralateral arm so that insulin and 20% glucose (when needed) could be infused via a variable rate volumetric infusion pump (I-med, San Diego, CA). An insulin infusion was immediately started at a basal rate. Patients then consumed an evening meal and a 7:30 p.m. snack and were requested not to ingest any food after 10:00 p.m. The insulin infusion rate was increased during meal consumption. Throughout the night, blood glucose was measured every 30 min, and the insulin infusion rate constantly adjusted to maintain glycemic levels of 4.4–6.7 mmol/l.

Day 1 procedures.

Day 1 procedures (Fig. 1) started at 8:00 a.m. after a 10-h overnight fast and lasted 480 min, divided into an equilibration period (0–120 min), a morning hyperinsulinemic clamp period (120–240 min), a rest period (240–360 min), and an afternoon hyperinsulinemic clamp period (360–480 min). At 120 min, in all studies, a primed continuous infusion of insulin (9 pmol · kg−1 · min−1) was started (18). In prior-hypoglycemia studies, plasma glucose was allowed to fall over a 30-min period to a target hypoglycemic plateau of ∼2.9 mmol/l. Plasma glucose was measured every 5 min and maintained at the desired level via a variable rate infusion of 20% dextrose (19). In prior-euglycemia studies, plasma glucose was held constant at basal levels by a similar technique (20). At 240 min, the insulin infusion was decreased to the morning basal rate, and euglycemia was restored in prior-hypoglycemia studies and maintained in the prior-euglycemia studies. At 360 min, a second 2-h euglycemic or hypoglycemic clamp identical to that performed in the morning was carried out. At 480 min, the insulin infusion was decreased to the morning basal rate, euglycemia was restored in prior-hypoglycemia studies, and all patients were allowed to consume a standardized meal. Evening and overnight procedures were then identical to those of admission night.

Day 2 procedures.

Day 2 procedures started at 8:00 a.m. after a 10-h overnight fast and lasted 210 min (time 0–210 min), divided into an equilibration period (0–90 min), a basal period (90–120 min), and an exercise period (120–210 min). A primed (18-μCi) infusion (0.18 μCi/min) of [3-3H]glucose was started at 0 min and continued throughout the experiment. Exercise consisted of 90 min continuous pedaling (at 60–70 rpm) on an upright cycle ergometer (Medical Graphics, Yorba Linda, CA) at 80% of the individual’s AT (∼50% Vo2max). Plasma glucose was measured every 5 min and maintained equivalent to baseline levels throughout the study via variable rate infusion of 20% dextrose. In an attempt to reproduce the drop in insulin levels that physiologically occurs with exercise of this intensity, the basal insulin infusion rate was decreased by 40% after the first 30 min of exercise, providing that the resulting reduced rate was at least 6 nmol/h (1 unit/h). In cases in which a 40% reduction of the basal rate would have resulted in an insulin infusion rate of <6 nmol/h, a minimum rate of 6 nmol/h was maintained. Potassium chloride was also infused (5 mmol/h) during exercise. After completion of the exercise protocol, patients consumed a meal and were discharged.

Tracer methodology.

Rates of glucose appearance (Ra), endogenous glucose production (EGP), and glucose utilization were calculated according to the methods of Wall et al. (21). EGP was calculated by determining the total Ra (this comprises both EGP and any exogenous glucose infused to maintain euglycemia) and subtracting from it the amount of exogenous glucose infused. It is now recognized that this approach is not fully quantitative because underestimates of total Ra and rate of glucose disposal can be obtained. This underestimate can be largely overcome by use of high-pressure liquid chromatography (HPLC)-purified tracer and taking measurements under steady-state conditions (i.e., constant specific activity). To minimize changes in specific activity, the tracer infusion rate was gradually doubled during the first 30 min of exercise. During the last 60 min of exercise, proportional additional increases of the tracer infusion rate were made commensurate with the changes of the exogenous glucose infusion rate.

Analytical methods.

The collection and processing of blood samples have been described elsewhere (22). Plasma glucose concentrations were measured in triplicate using the glucose oxidase method with a glucose analyzer (Beckman, Fullerton, CA). Glucagon was measured according to a modification of the method of Aguilar-Parada et al. (23) with an interassay coefficient of variation (CV) of 12%. Free insulin was measured as previously described (24) with an interassay CV of 9%. Catecholamines were determined by HPLC (25) with an interassay CV of 12% for epinephrine and 8% for norepinephrine. We made two modifications to the procedure for catecholamine determination: 1) we used a five-point rather than a one-point standard calibration curve, and 2) we spiked the initial and final samples of plasma with known amounts of epinephrine and norepinephrine so accurate identification of the relevant respective catecholamine peaks could be made. Cortisol was assayed using the Clinical Assays Gamma Coat Radioimmunoassay (RIA) kit with an interassay CV of 6%. Growth hormone was determined by RIA (26) with a CV of 8.6%. Pancreatic polypeptide was measured by RIA using the method of Hagopian et al. (27) with an interassay CV of 8%. Lactate, glycerol, alanine, and β-hydroxybutyrate were measured in deproteinized whole blood using the method of Lloyd et al. (28). Nonesterified fatty acids were measured using the WAKO kit adopted for use on a centrifugal analyzer (29).

On day 2, blood samples for glucose flux were taken every 10 min throughout the basal period and every 15 min during exercise. Blood for hormones and intermediary metabolites were drawn twice during the basal period and every 15 min during the exercise period. Cardiovascular parameters (pulse and systolic and diastolic arterial pressure) were measured every 10 min from −30 min to 90 min. Respiratory quotient, carbohydrate, and lipid oxidation were measured by gas exchange during the basal period and the final 10 min of exercise.

Materials.

HPLC-purified [3-3H]glucose (New England Nuclear, Boston, MA) was used as the glucose tracer (11.5 mCi · mmol−1 · l−1). Human regular insulin was purchased from Eli Lilly (Indianapolis, IN). The insulin infusion solution was prepared with normal saline and contained 3% (vol/vol) of the subjects’ own plasma.

Statistical analysis.

Data are expressed as mean ± SE unless otherwise stated and were analyzed using standard parametric two-way ANOVA with repeated-measures design. This was coupled with the Duncan post hoc test to delineate at which time points statistical significance was reached. A value of P < 0.05 indicated significant difference.

Day 1: Plasma glucose and insulin levels.

Basal plasma glucose levels were comparable in the two experimental groups both in the morning (anteEugly 5.4 ± 0.1 mmol/l, anteHypo 5.3 ± 0.1 mmol/l) and in the afternoon (anteEugly 5.0 ± 0.1 mmol/l, anteHypo 5.0 ± 0.1 mmol/l) (Fig. 2). During the last 30 min of the clamp periods, plasma glucose was 5.1 ± 0.1 mmol/l in the morning and 5.2 ± 0.1 mmol/l in the afternoon during anteEugly and 2.8 ± 0.1 mmol/l in the morning and 2.9 ± 0.1 mmol/l in the afternoon during anteHypo.

Plasma insulin concentrations were similar in the two groups at all times during day 1 procedures (Morning basal levels: anteEugly 78 ± 12 pmol/l, anteHypo 78 ± 12 pmol/l; morning steady-state levels: anteEugly 558 ± 48 pmol/l, anteHypo 552 ± 24 pmol/l. Afternoon basal levels: anteEugly 84 ± 18 pmol/l, anteHypo 72 ± 12 pmol/l; afternoon steady-state levels: anteEugly 582 ± 60 pmol/l, anteHypo 576 ± 30 pmol/l).

Day 2: insulin, glucose, and counterregulatory hormone levels.

Basal pre-exercise plasma glucose levels were comparable in the two experimental groups (anteEugly 5.2 ± 0.1 mmol/l, anteHypo 5.1 ± 0.1 mmol/l). During the last 30 min of exercise, plasma glucose was 5.3 ± 0.2 mmol/l in anteEugly and 5.4 ± 0.1 mmol/l in anteHypo.

Before exercise, basal plasma insulin levels were 75 ± 8 pmol/l in anteEugly and 71 ± 7 pmol/l in anteHypo; during the last 30 min of exercise, insulin was 62 ± 6 pmol/l in anteEugly and 61 ± 6 pmol/l in anteHypo (Fig. 2).

After anteEugly, plasma glucagon increased during exercise from 44 ± 4 to 54 ± 5 ng/l. After anteHypo, basal glucagon levels (39 ± 40 ng/l) remained unchanged during exercise (39 ± 4 ng/l, P < 0.001 vs. anteEugly) (Fig. 3). Plasma cortisol increased during exercise from 359 ± 55 to 635 ± 83 nmol/l in anteEugly. After anteHypo, the increase of cortisol levels was significantly blunted during exercise (386 ± 55 to 469 ± 55 nmol/l, P < 0.001) (Fig. 3).

Exercise increased plasma epinephrine from a basal value of 186 ± 22 to 797 ± 120 pmol/l in anteEugly. After day 1 hypoglycemia, the exercise-induced increase in epinephrine was reduced by 50% (P < 0.02) (224 ± 16 to 529 ± 55 pmol/l) (Fig. 4). Norepinephrine increased during exercise from 1.6 ± 0.2 to 5.3 ± 0.6 nmol/l in anteEugly. After anteHypo, the exercise-induced increase in norepinephrine was 38% smaller than in anteEugly (1.5 ± 0.2 to 4.1 ± 0.6 nmol/l) (Fig. 3).

Basal levels of growth hormone were similar after anteEugly (2.1 ± 1.1 μg/l) and anteHypo (1.2 ± 0.5 μg/l) (Table 1). By the last 30 min of exercise, growth hormone increased to 19 ± 5 μg/l in anteEugly and to 14 ± 3 μg/l in anteHypo (NS). The exercise-induced increase in pancreatic polypeptide was also not different between the two experimental groups (from 12 ± 1 to 30 ± 5 pmol/l in anteEugly and from 12 ± 1 to 25 ± 5 pmol/l in anteHypo), although the incremental area under the curve of this hormone during exercise was 39% greater in the anteEugly group than in the anteHypo group.

Day 2: glucose kinetics and gas exchange measurements.

By the last 30 min of exercise, EGP was significantly greater in the anteEugly group than the anteHypo group (16 ± 3 vs. 6 ± 3 μmol · kg−1 · min−1, P < 0.01) (Fig. 4). Conversely, the infusion rate of exogenous glucose required to maintain euglycemia during the last 30 min of exercise was reduced after anteEugly compared with in anteHypo (9 ± 2 vs. 24 ± 5 μmol · kg−1 · min−1, P < 0.01) (Table 2). Therefore, after day 1 euglycemia, the total Ra was comprised of 62% EGP and 38% exogenous glucose infusion. After day 1 hypoglycemia, EGP accounted for only 23% of the total Ra and exogenous glucose infusion for the remaining 77% (Table 2).

Carbohydrate oxidation, oxidation, and respiratory quotient were similar in both groups at the start and during the last 30 min of exercise.

Day 2: intermediary metabolism.

Blood lactate levels (Fig. 4) were similar at baseline in the two groups. Blood lactate increased by a greater increment during anteEugly compared with anteHypo (1.34 ± 0.14 vs. 0.77 ± 0.20 mmol/l, respectively; P < 0.02). Free fatty acid (FFA) basal levels (Fig. 4) were also similar in both groups, but the exercise-induced increase was significantly less in anteHypo (87 ± 0.58 μmol/l) than in anteEugly (224 ± 68 μmol/l, P < 0.05). Similarly, circulating levels of glycerol (Fig. 4), an indicator of whole-body lipolysis, increased significantly less after anteHypo (66 ± 11 μmol/l) compared with anteEugly (103 ± 14 μmol/l, P = 0.002). No difference in the circulating levels of alanine or of the ketone body β-hydroxybutyrate were measured between the two groups either at baseline or during exercise.

Day 2: cardiovascular parameters.

No significant difference occurred in heart rate and systolic, diastolic, and mean arterial pressure between the two groups at baseline or during exercise. However, unlike heart rate and systolic and mean blood pressure (which increased over baseline during exercise), diastolic blood pressure fell by ∼10 mmHg by the last 30 min of exercise (Table 3).

The results of this study indicate that after two episodes of hypoglycemia of ∼2.9 mmol/l, key neuroendocrine and metabolic counterregulatory responses to next-day prolonged moderate exercise are severely blunted in patients with type 1 diabetes when compared with identical exercise performed after resting euglycemia. It would therefore appear that a state of acute counterregulatory failure induced by antecedent hypoglycemia might be one of the factors responsible for exercise-associated hypoglycemia in patients with type 1 diabetes.

Glycemic levels were carefully controlled at all times during our 2-day studies. During overnight stays at our Clinical Research Center, hypoglycemia was carefully avoided by constant adjustments of exogenous insulin and/or glucose. Uncontrolled hypoglycemia during the first part of the study would have confounded interpretation of blunted responses during subsequent exercise. Additionally, euglycemia was also strictly maintained during day 2 exercise. During exercise, hyperglycemia inhibits neuroendocrine responses, whereas hypoglycemia would have induced counterregulatory responses independent of those induced by exercise per se.

Another important point of our experimental design was controlling the prevailing insulinemia during exercise. In nondiabetic individuals, exercise conditions similar to those used in the present study result in a 40–50% fall in peripheral insulinemia below pre-exercise levels. Trying to reproduce this pattern in patients with type 1 diabetes, however, may result in relative hepatic hypoinsulinemia, particularly in patients who need very low basal insulin infusion rates, and can result in hyperglycemia of 8–9 mmol/l during exercise. Furthermore, a drop in insulin levels during exercise will not reflect real-life conditions for diabetic patients on typical basal/bolus multiple insulin injection regimens who are unable to reduce insulin levels. Therefore, insulin concentrations during exercise were controlled at levels of 60–70 pmol/l, which reflect typical postabsorptive insulinemia found in type 1 diabetic patients (30). Indeed, the inability of type 1 diabetic patients to suppress insulin levels during exercise may per se be a factor responsible for exercise-associated hypoglycemia.

Blunting of the glucagon response during exercise after antecedent hypoglycemia is particularly relevant for patients with type 1 diabetes. In these patients, the glucagon response to hypoglycemia is gradually lost over the first few years after diagnosis. Interestingly, however, secretion of the hormone during exercise is preserved, indicating that the pancreatic α-cell deficit is stimulus specific. After antecedent euglycemia, our patients were able to mount a glucagon response similar to that previously observed in nondiabetic subjects during exercise of similar duration and intensity (11). Two episodes of prior hypoglycemia of 2.9 mmol/l, however, abolished this response, therefore increasing the risk of hypoglycemia. In fact, maintaining euglycemic conditions during exercise required the infusion of an additional 15 μmol · kg−1 · min−1 exogenous glucose after prior hypoglycemia, as compared with prior euglycemia. Had this supplemental exogenous glucose not been administered, hypoglycemia, possibly severe, would have occurred during exercise.

Plasma epinephrine and norepinephrine responses to exercise were reduced after prior hypoglycemia, a finding consistent with reduced sympathetic drive. Additionally, there was a trend for the pancreatic polypeptide response (an index of parasympathetic input to the pancreas) to be reduced during exercise (39% smaller incremental area under the curve) after antecedent hypoglycemia compared with antecedent euglycemia. Taken together, these data indicate that prior hypoglycemia resulted in a widespread reduction of the ANS drive during subsequent exercise in type 1 diabetes. It should also be noted that the pancreatic polypeptide response after antecedent euglycemia (i.e., without the blunting effect of prior hypoglycemia) was only about half that reported during similar experimental conditions in nondiabetic subjects (11). Our data therefore indicate that, in type 1 diabetes, the pancreatic polypeptide response may be an index of reduced ANS activation not only during hypoglycemia, as previously reported (31), but also in response to different forms of stress (exercise).

Lactate, FFA, and glycerol responses were reduced during exercise after antecedent hypoglycemia compared with euglycemia. During exercise, ANS activation and increased catecholamine secretion stimulate lipolysis and increase peripheral release of gluconeogenic precursors, such as lactate and amino acids. The reduced ANS drive and catecholamine levels during exercise, caused by antecedent hypoglycemia, are therefore the most likely explanation for the blunting in lactate and lipolytic responses in our study. It should be noted that insulin levels during exercise were identical in the two experimental conditions, and therefore, decreased lipolysis could not be ascribed to a difference in the antilipolytic effect of this hormone. Greater lactate and glycerol uptake by the liver during exercise could also have reduced the circulating levels of these metabolites. This possibility, however, appears unlikely because greater uptake of these precursors should be reflected by an increase in EGP. Instead, and quite to the contrary, a reduction in EGP occurred during exercise after antecedent hypoglycemia.

Cardiovascular responses to exercise were similar in the two experimental conditions tested in the present study. However, although heart rate and systolic blood pressure increased as expected during exercise, diastolic blood pressure decreased by ∼10 mmHg. Diastolic dysfunction during exercise, not accompanied by systolic dysfunction or altered cardiac output response, has been reported by several authors (3234) in type 1 diabetic patients with no evidence of autonomic neuropathy, leading to the hypothesis that alterations in diastolic filling may represent the earliest functional defect of diabetic cardiomyopathy. Surprisingly, however, an exercise-induced reduction in diastolic blood pressure in type 1 diabetes has not been previously reported. We believe that the experimental design used in previous exercise studies obscured recognition of this finding. The large majority of previous exercise studies in type 1 diabetes used incremental maximal tests (33,35,36) or brief periods of submaximal exercise (37); furthermore, diastolic blood pressure was seldom reported, and glycemic levels during exercise were never controlled (introducing the confounding effect of concomitant hypoglycemia, which may independently lower blood pressure and stimulate the ANS).

Interestingly, despite significant reductions in catecholamine responses after day 1 hypoglycemia, there were similar cardiovascular responses during exercise in both series of studies. Cardiovascular responses during stress can depend on the balance between the sympathetic and parasympathetic drive. An off-setting blunting effect on the two limbs of the ANS in the anteHypo group may have therefore resulted in a new equilibrium with unchanged cardiovascular responses. However, this hypothesis remains speculative because, in the present study, only indirect measurements of sympathetic and parasympathetic impulses were obtained.

Current clinical practice stresses the importance of physical activity in diabetes management. The beneficial effects of exercise (including weight control, improved insulin sensitivity, and protection from cardiovascular disease) are particularly important for type 1 diabetic patients, already at high risk for long-term cardiovascular complications. Consequently, growing numbers of patients participate in physical activities such as football, tennis matches, outdoor hiking, or bike rides, which are all forms of physical activity with duration and intensity similar to our exercise model. Importantly, the blunting effect of antecedent hypoglycemia on counterregulatory responses to exercise, as indicated by inadequate supply of glucose and lipids to the working muscle, was apparent after just 30 min of exercise. This indicates that if patients with type 1 diabetes are exposed to prior hypoglycemia and later engage in a form of physical activity comparable to that performed in our study, they may expect hypoglycemia to begin early and persist or worsen through the completion of exercise. In this context, our observations may be relevant for day-to-day management of type 1 diabetes. The concept that antecedent episodes of hypoglycemia may affect counterregulatory responses during next-day exercise may help provide the conceptual basis for prevention of undesired exercise-associated hypoglycemia.

The results from the present study are consistent with our previous observations in a group of 16 healthy subjects who underwent a comparable experimental protocol (11). Similar to the findings reported here, antecedent hypoglycemia resulted in a widespread blunting of neuroendocrine and metabolic counterregulatory responses during exercise, including glucagon, catecholamines, cortisol, EGP, and lipolysis. Changes in other parameters were also qualitatively in agreement with the present study. An earlier study by Rattarasarn et al. (14), on the other hand, reported no blunted response to exercise after hypoglycemia. In that study, type 1 diabetic patients exercised after antecedent afternoon euglycemia or 2 h of hypoglycemia. On both occasions, however, patients had been hypoglycemic for 2 h during the morning of the day preceding the exercise bout. This result may have generated a strong enough blunting stimulus to render superfluous the presence or absence of afternoon hypoglycemia. Other differences, such as the shorter duration of exercise (only 60 min), prevailing hyperglycemia, and smaller sample size (n = 8), may also have affected the results.

The underlying mechanisms causing the high incidence of exercise-associated hypoglycemia in type 1 diabetes are still incompletely understood. One mechanism points to the role played by acute increases in insulin sensitivity (and also in relatively elevated insulin levels) that occur during exercise (38). The hyperinsulinemia that typically occurs during clinical practice in patients with type 1 diabetes, however, is modest and very unlikely to cause marked hypoglycemia if counterregulatory responses are intact (19). Furthermore, if hypoglycemia occurs during exercise, neuroendocrine responses are in fact increased (39). Catecholamine responses to hypoglycemic exercise, on the other hand, are reduced in well-controlled type 1 diabetic patients (i.e., exposed to repeated antecedent hypoglycemia) (13). Furthermore, patients with classic diabetic autonomic neuropathy have blunted epinephrine responses during euglycemic exercise (12). These reports suggest that an alteration in counterregulatory responses, similar to that observed during repeated hypoglycemia, may be an important cause of the increased incidence of hypoglycemia associated with physical activity. The mechanisms responsible for acute counterregulatory failure during repeated hypoglycemia, however, are also speculative and could include prior hypercortisolemia (40), alterations in cerebral glucose extraction (41), and elevations of circulating levels of lactate (42) and ketone bodies (42) during the subsequent stress. Whether one or more of these mechanisms are also responsible for blunted responses during exercise is unclear.

In summary, this study has demonstrated that after two 2-h episodes of hypoglycemia of ∼2.9 mmol/l, a widespread blunting of counterregulatory responses occurs in patients with type 1 diabetes during next-day moderate exercise. Among neuroendocrine counterregulatory responses, glucagon, epinephrine, and cortisol were the most severely blunted, and this reduction was paralleled by proportional blunting of EGP and lipolysis.

We conclude that in patients with type 1 diabetes, antecedent hypoglycemia induces acute counterregulatory failure not only during subsequent hypoglycemia, but also during subsequent, moderate exercise. This acute state of counterregulatory impairment may be one of the causes of exercise-associated hypoglycemia in patients with type 1 diabetes.

FIG. 1.

Schematic diagram of experimental protocols.

FIG. 1.

Schematic diagram of experimental protocols.

FIG. 2.

Plasma glucose levels from arterialized venous blood during day 1 (two 120-min clamps with either hypoglycemia of ∼2.9 mmol/l [anteHypo] or euglycemia [anteEugly]) and day 2 (90 min exercise at ∼50% Vo2max) and plasma insulin levels from day 2. n = 16 patients (8 men/8 women) with type 1 diabetes. Data are group means (SD).

FIG. 2.

Plasma glucose levels from arterialized venous blood during day 1 (two 120-min clamps with either hypoglycemia of ∼2.9 mmol/l [anteHypo] or euglycemia [anteEugly]) and day 2 (90 min exercise at ∼50% Vo2max) and plasma insulin levels from day 2. n = 16 patients (8 men/8 women) with type 1 diabetes. Data are group means (SD).

FIG. 3.

Plasma incremental catecholamine, glucagon, and cortisol levels from arterialized venous blood at baseline and during 90 min of exercise at ∼50% Vo2max in 16 patients (8 men and 8 women) with type 1 diabetes. On the previous day, patients had undergone two 120-min clamps with either hypoglycemia of ∼2.9 mmol/l (anteHypo group) or euglycemia (anteEugly group). Data are group means (SD). Epinephrine, glucagon, and cortisol: P < 0.05, anteHypo vs. anteEugly.

FIG. 3.

Plasma incremental catecholamine, glucagon, and cortisol levels from arterialized venous blood at baseline and during 90 min of exercise at ∼50% Vo2max in 16 patients (8 men and 8 women) with type 1 diabetes. On the previous day, patients had undergone two 120-min clamps with either hypoglycemia of ∼2.9 mmol/l (anteHypo group) or euglycemia (anteEugly group). Data are group means (SD). Epinephrine, glucagon, and cortisol: P < 0.05, anteHypo vs. anteEugly.

FIG. 4.

Incremental EGP, blood glycerol, lactate, and plasma FFA levels during 90 min of exercise at ∼50% Vo2max in 16 patients (8 men and 8 women) with type 1 diabetes. On the previous day, patients had undergone two 120-min clamps with either hypoglycemia of ∼2.9 mmol/l (anteHypo group) or euglycemia (anteEugly group). Data are group means (SD). EGP, glycerol, lactate, and FFA: P < 0.05, anteHypo vs. anteEugly.

FIG. 4.

Incremental EGP, blood glycerol, lactate, and plasma FFA levels during 90 min of exercise at ∼50% Vo2max in 16 patients (8 men and 8 women) with type 1 diabetes. On the previous day, patients had undergone two 120-min clamps with either hypoglycemia of ∼2.9 mmol/l (anteHypo group) or euglycemia (anteEugly group). Data are group means (SD). EGP, glycerol, lactate, and FFA: P < 0.05, anteHypo vs. anteEugly.

TABLE 1

Blood lactate, alanine, β-hydroxybutyrate, and plasma FFA, growth hormone, and pancreatic polypeptide during 90 min of exercise at 50% Vo2max in 16 patients with type 1 diabetes (8 men and 8 women) after two episodes of antecedent euglycemia or hypoglycemia of 2.9 ± 0.1 mmol/l

BaselineExercise (min)
306090
Blood lactate (mmol/l)     
 anteEugly 0.8 ± 0.1 3.1 ± 0.2 2.4 ± 0.2 2.1 ± 0.2 
 anteHypo 0.8 ± 0.1 2.2 ± 0.2* 1.7 ± 0.2* 1.5 ± 0.2* 
Blood alanine (μmol/l)     
 anteEugly 287 ± 19 389 ± 13 391 ± 17 376 ± 17 
 anteHypo 280 ± 15 362 ± 17 364 ± 21 354 ± 18 
Blood β-hydroxybutyrate (μmol/l)     
 anteEugly 77 ± 17 52 ± 6 66 ± 10 90 ± 14 
 anteHypo 91 ± 17 67 ± 14 63 ± 12 80 ± 13 
Plasma FFA (μmol/l)     
 anteEugly 313 ± 33 327 ± 35 401 ± 47 531 ± 61 
 anteHypo 364 ± 47 348 ± 46 362 ± 42 449 ± 50* 
Plasma growth hormone (μg/l)     
 anteEugly 2 ± 1 12 ± 4 20 ± 6 18 ± 4 
 anteHypo 1 ± 1 11 ± 3 16 ± 4 13 ± 3 
Plasma pancreatic polypeptide (pmol/l)     
 anteEugly 12 ± 1 18 ± 2 25 ± 3 34 ± 7 
 anteHypo 12 ± 1 15 ± 2 21 ± 4 26 ± 5 
BaselineExercise (min)
306090
Blood lactate (mmol/l)     
 anteEugly 0.8 ± 0.1 3.1 ± 0.2 2.4 ± 0.2 2.1 ± 0.2 
 anteHypo 0.8 ± 0.1 2.2 ± 0.2* 1.7 ± 0.2* 1.5 ± 0.2* 
Blood alanine (μmol/l)     
 anteEugly 287 ± 19 389 ± 13 391 ± 17 376 ± 17 
 anteHypo 280 ± 15 362 ± 17 364 ± 21 354 ± 18 
Blood β-hydroxybutyrate (μmol/l)     
 anteEugly 77 ± 17 52 ± 6 66 ± 10 90 ± 14 
 anteHypo 91 ± 17 67 ± 14 63 ± 12 80 ± 13 
Plasma FFA (μmol/l)     
 anteEugly 313 ± 33 327 ± 35 401 ± 47 531 ± 61 
 anteHypo 364 ± 47 348 ± 46 362 ± 42 449 ± 50* 
Plasma growth hormone (μg/l)     
 anteEugly 2 ± 1 12 ± 4 20 ± 6 18 ± 4 
 anteHypo 1 ± 1 11 ± 3 16 ± 4 13 ± 3 
Plasma pancreatic polypeptide (pmol/l)     
 anteEugly 12 ± 1 18 ± 2 25 ± 3 34 ± 7 
 anteHypo 12 ± 1 15 ± 2 21 ± 4 26 ± 5 

Data are means ± SE.

*

P < 0.05 vs. anteEugly.

TABLE 2

Whole-body glucose kinetics during 90 min of exercise at 50% Vo2max in 16 patients with type 1 diabetes (8 men and 8 women) after two episodes of antecedent euglycemia or hypoglycemia of 2.9 ± 0.1 mmol/l

BaselineDuration of exercise (min)
30607590
Glucose specific activity (dpm/μmol)      
 anteEugly 386 ± 21 294 ± 20 278 ± 22 258 ± 20 248 ± 20 
 anteHypo 387 ± 21 280 ± 18 268 ± 15 267 ± 17 256 ± 16 
Glucose Ra (μmol · kg−1 · min−1     
 anteEugly 11 ± 1 20 ± 2 30 ± 2 34 ± 2 36 ± 3 
 anteHypo 11 ± 1 21 ± 2 33 ± 3 36 ± 4 38 ± 3 
Glucose Rd (μmol · kg−1 · min−1     
 anteEugly 11 ± 1 21 ± 2 32 ± 3 34 ± 2 35 ± 3 
 anteHypo 11 ± 1 21 ± 2 33 ± 3 37 ± 4 36 ± 3 
Glucose infusion/rate (μmol · kg−1 · min−1     
 anteEugly 1 ± 1 6 ± 4 6 ± 2 8 ± 2 9 ± 2 
 anteHypo 2 ± 1 9 ± 3 18 ± 4* 23 ± 5* 26 ± 6* 
BaselineDuration of exercise (min)
30607590
Glucose specific activity (dpm/μmol)      
 anteEugly 386 ± 21 294 ± 20 278 ± 22 258 ± 20 248 ± 20 
 anteHypo 387 ± 21 280 ± 18 268 ± 15 267 ± 17 256 ± 16 
Glucose Ra (μmol · kg−1 · min−1     
 anteEugly 11 ± 1 20 ± 2 30 ± 2 34 ± 2 36 ± 3 
 anteHypo 11 ± 1 21 ± 2 33 ± 3 36 ± 4 38 ± 3 
Glucose Rd (μmol · kg−1 · min−1     
 anteEugly 11 ± 1 21 ± 2 32 ± 3 34 ± 2 35 ± 3 
 anteHypo 11 ± 1 21 ± 2 33 ± 3 37 ± 4 36 ± 3 
Glucose infusion/rate (μmol · kg−1 · min−1     
 anteEugly 1 ± 1 6 ± 4 6 ± 2 8 ± 2 9 ± 2 
 anteHypo 2 ± 1 9 ± 3 18 ± 4* 23 ± 5* 26 ± 6* 

Data are means ± SE.

*

P < 0.05 vs. anteEugly. Rd, rate of disappearance.

TABLE 3

Heart rate and systolic, diastolic, and mean arterial blood pressures during 90 min of exercise at 50% Vo2max in 16 patients with type 1 diabetes (8 men and 8 women) after two episodes of antecedent euglycemia or hypoglycemia of 2.9 ± 0.1 mmol/l

BaselineExercise (min)
306090
Heart rate (bpm)     
 anteEugly 84 ± 3 143 ± 2* 144 ± 2* 149 ± 2* 
 anteHypo 87 ± 3 141 ± 2* 143 ± 2* 143 ± 2* 
Arterial blood pressure (mmHg)     
 Systolic     
  anteEugly 116 ± 3 151 ± 5* 148 ± 4* 147 ± 4* 
  anteHypo 118 ± 3 146 ± 4* 144 ± 2* 147 ± 4* 
 Diastolic     
  anteEugly 78 ± 2 71 ± 2 67 ± 2* 68 ± 2* 
  anteHypo 78 ± 2 70 ± 2 69 ± 2* 69 ± 2* 
 Mean     
  anteEugly 91 ± 2 98 ± 2 94 ± 2 94 ± 2 
  anteHypo 91 ± 2 95 ± 2 94 ± 2 95 ± 2 
BaselineExercise (min)
306090
Heart rate (bpm)     
 anteEugly 84 ± 3 143 ± 2* 144 ± 2* 149 ± 2* 
 anteHypo 87 ± 3 141 ± 2* 143 ± 2* 143 ± 2* 
Arterial blood pressure (mmHg)     
 Systolic     
  anteEugly 116 ± 3 151 ± 5* 148 ± 4* 147 ± 4* 
  anteHypo 118 ± 3 146 ± 4* 144 ± 2* 147 ± 4* 
 Diastolic     
  anteEugly 78 ± 2 71 ± 2 67 ± 2* 68 ± 2* 
  anteHypo 78 ± 2 70 ± 2 69 ± 2* 69 ± 2* 
 Mean     
  anteEugly 91 ± 2 98 ± 2 94 ± 2 94 ± 2 
  anteHypo 91 ± 2 95 ± 2 94 ± 2 95 ± 2 

Data are means ± SE.

*

P < 0.05 vs. corresponding baseline value.

This work was supported by a grant from the Juvenile Diabetes Foundation International (JDFI), by National Institutes of Health Grant R01 DK45369, by Diabetes Research and Training Center Grant 5P60-AM20593, by Clinical Research Center Grant M01-RR00095, and by a VA/JDFI Diabetes Research Center grant. P.G. was supported by a JDFI research fellowship grant.

We thank Eric Allen, Angelina Penalosa, and Wanda Snead for expert technical assistance. We also appreciate the skill and help of the nurses of the Vanderbilt General Clinical Research Center in the performance of the studies included in this report.

1.
Diabetes Control and Complications Trial Research Group: The effect of intensive treatment of diabetes on the development and progression of long term complications in insulin-dependent diabetes mellitus.
N Engl J Med
329
:
977
–986,
1993
2.
Diabetes Control and Complications Trial Research Group: Hypoglycemia in the Diabetes Control and Complication Trial.
Diabetes
46
:
271
–286,
1997
3.
Gerich JE, Langlois M, Noacco C, Karam J, Forsham PH: Lack of a glucagon response to hypoglycemia in diabetes: evidence for an intrinsic pancreatic alpha-cell defect.
Science
182
:
171
–173,
1973
4.
Mitrakou A, Fanelli C, Veneman T, Perriello G, Calderone S, Platanisiotis D, Rambotti A, Raptis S, Brunetti P, Cryer P: Reversibility of unawareness of hypoglycemia in patients with insulinomas.
N Engl J Med
329
:
834
–839,
1993
5.
Cryer PE: Iatrogenic hypoglycemia as a cause of hypoglycemia-associated autonomic failure in IDDM.
Diabetes
41
:
255
–260,
1992
6.
Davis MR, Mellman MJ, Shamoon H: Further defects in counterregulatory responses induced by recurrent hypoglycemia in IDDM.
Diabetes
41
:
1335
–1340,
1992
7.
Davis SN, Shavers C, Mosqueda-Garcia R, Costa F: Effects of differing antecedent hypoglycemia on subsequent counterregulation in normal humans.
Diabetes
46
:
1328
–1335,
1997
8.
Widom B, Simonson D: Intermittent hypoglycemia impairs glucose counterregulation.
Diabetes
41
:
1597
–1602,
1992
9.
Heller SR, Cryer PE: Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after one episode of hypoglycemia in nondiabetic humans.
Diabetes
40
:
223
–226,
1991
10.
Galassetti P, Mann S, Tate D, Neill RA, Costa F, Wasserman DH, Davis SN: Effect of antecedent prolonged exercise on subsequent counterregulatory responses to hypoglycemia.
Am J Physiol
280
:
E908
–E917,
2001
11.
Davis SN, Galassetti P, Wasserman DH, Tate D: Effects of antecedent hypoglycemia on subsequent counterregulatory responses to exercise.
Diabetes
49
:
73
–81,
2000
12.
Bottini P, Boschetti E, Pampanelli S, Ciofetta M, Del Sindaco P, Scionti L, Brunetti P, Bolli GB: Contribution of autonomic neuropathy to reduce plasma adrenaline responses to hypoglycemia in IDDM.
Diabetes
46
:
814
–823,
1997
13.
Schneider S, Vitug A, Ananthakrishnan R, Khachadurian A: Impaired adrenergic response to prolonged exercise in type I diabetes.
Metabolism
40
:
1219
–1225,
1991
14.
Rattarasarn C, Dagogo-Jack S, Zachwieja J, Cryer PE: Hypoglycemia-induced autonomic failure in IDDM is specific for stimulus of hypoglycemia and is not attributable to prior autonomic activation.
Diabetes
43
:
809
–818,
1994
15.
Beaver WL, Wasserman K, Whipp BJ: A new method for detecting anaerobic threshold by gas exchange.
J Appl Physiol
60
:
2020
–2027,
1986
16.
Wasserman K, Beaver WL, Davis JA, Pu JZ, Whipp BJ: Lactate, pyruvate and lactate-to-pyruvate ratio during exercise and recovery.
J Appl Physiol
59
:
935
–940,
1985
17.
Wasserman K: The anaerobic threshold measurement to evaluate exercise performance.
Am Rev Resp Dis
129
:
535
–540,
1984
18.
Sherwin RS, Kramer K, Robin J, Andres R: A model of the kinetics of insulin in man.
J Clin Invest
53
:
1481
–1492,
1978
19.
Amiel SA, Tamborlane WV, Simonson DC, Sherwin RS: Defective glucose counterregulation after strict control of insulin-dependent diabetes mellitus.
N Engl J Med
316
:
1376
–1383,
1987
20.
DeFronzo RA, Tobin K, Andres R: Glucose clamp technique: a method for quantifying insulin secretion and resistance.
Am J Physiol
237
:
E216
–E223,
1979
21.
Wall JS, Steele R, DeBodo RD, Altszuler N: Effect of insulin on utilization and production of circulating glucose.
Am J Physiol
189
:
43
–50,
1957
22.
Cherrington AD, Lacy WW, Chiasson JL: Effect of glucagon on glucose production during insulin deficiency in the dog.
J Clin Invest
62
:
664
–667,
1978
23.
Aguilar-Parada E, Eisentraut AM, Unger RH: Pancreatic glucagon secretion in normal and diabetic subjects.
Am J Med Sci
257
:
415
–419,
1969
24.
Wide L, Porath J: Radioimmunoassay of proteins with the use of sephadex-coupled antibodies.
Biochim Biophys Acta
130
:
257
–260,
1966
25.
Causon R, Caruthers M, Rodnight R: Assay of plasma catecholamines by liquid chromatography with electrical detection.
Annal Biochem
116
:
223
–226,
1982
26.
Hunter W, Greenwood F: Preparation of [131I]-labeled human growth hormone of high specific activity.
Nature
194
:
495
–496,
1962
27.
Hagopian W, Lever E, Cen D, Emmounoud D, Polonsky K, Pugh W, Moosa A, Jaspan JB: Predominance of renal and absence of hepatic metabolism of pancreatic polypeptide in the dog.
Am J Physiol
245
:
171
–177,
1983
28.
Lloyd B, Burrin J, Smythe P, Alberti KGMM: Enzymatic fluorometric continuous-flow assays for blood glucose lactate, pyruvate, alanine, glycerol, and 3-hydroxybutyrate.
Clin Chem
24
:
1724
–1729,
1978
29.
Ho RJ: Radiochemical assay of long chain fatty acid using 63NI as tracer.
Anal Biochem
26
:
105
–113,
1970
30.
Rabasa-Lhoret R, Bourque J, Ducros F, Chiasson JL: Guidelines for premeal insulin dose reduction for postprandial exercise of different intensities and durations in type 1 diabetic subjects treated intensively with a basal-bolus insulin regimen (ultralente-lispro).
Diabetes Care
24
:
625
–630,
2001
31.
White NH, Gingerich RL, Levandoski LA, Cryer PE, Santiago JV: Plasma pancreatic polypeptide response to insulin-induced hypoglycemia as a marker for defective glucose counterregulation in insulin-dependent diabetes mellitus.
Diabetes
34
:
870
–875,
1985
32.
Jermendy G, Bachman B: Left ventricular diastolic dysfunction in diabetic patients.
Q J Med
87
:
519
–520,
1994
33.
Nugent A-M, Steele IC, Al-Modaris F, Vallely S, Moore A, Campbell NPS, Bell PM, Buchanan KD, Trimble ER, Nicholls DP: Exercise responses in patients with IDDM.
Diabetes Care
20
:
1814
–1821,
1997
34.
Ruddy TD, Shumak SL, Liu PP, Barnie A, Seawright SJ: The relationship of cardiac diastolic dysfunction to concurrent hormonal and metabolic status in type 1 diabetes mellitus.
J Clin Endocrinol Metab
66
:
113
–118,
1988
35.
Storstein L, Jervell J: Response to bicycle exercise testing in long-standing juvenile diabetes.
Acta Med Scand
205
:
227
–230,
1979
36.
Vered Z, Battler A, Segal P, Liberman D, Yerushalmi Y, Berezin M, Neufeld HN: Exercise-induced left ventricular dysfunction in young men with asymptomatic diabetes mellitus (diabetic cardiomyopathy).
Am J Cardiol
54
:
633
–637,
1984
37.
Hilsted J, Galbo H, Christensen NJ, Parving H-H, Benn J: Hemodynamic changes during graded exercise in patients with diabetic autonomic neuropathy.
Diabetologia
22
:
318
–323,
1982
38.
Sonnenberg GE, Kemmer FW, Berger M: Exercise in type I (insulin-dependent) diabetic patients treated with continuous subcutaneous insulin infusion: prevention of exercise-induced hypoglycaemia.
Diabetologia
33
:
696
–703,
1990
39.
Sotsky MJ, Shilo S, Shamoon H: Regulation of counterregulatory hormone secretion in man during exercise and hypoglycemia.
J Clin Endocrinol Metab
68
:
9
–16,
1989
40.
Davis SN, Shavers C, Costa F, Mosqueda-Garcia R: Role of cortisol in the pathogenesis of deficient counterregulation after antecedent hypoglycemia in normal man.
J Clin Invest
98
:
680
–691,
1996
41.
Boyle P, Kempers S, O’Connor A, Nagy R: Brain glucose uptake and unawareness of hypoglycemia in patients with insulin-dependent diabetes mellitus.
N Engl J Med
333
:
1720
–1731,
1995
42.
Veneman T, Mitrakou A, Mokan M, Cryer PE, Gerich JE: Effects of hyperketonemia or hyperlacticacidemia on symptoms, cognitive dysfunction, and counterregulatory hormone responses during hypoglycemia in normal humans.
Diabetes
43
:
1311
–1317,
1994