NPY attenuates positive cortical DC-potential shift upon food intake in man

Previous studies have shown that the regulation of hunger and satiety is accompanied by coordinate changes in cortical excitability. Starved subjects show a transient negative shift in the scalp-recorded cortical direct current (DC-)potential in the beginning of eating, indicating increased cortical excitability. With increasing satiety, the DC negativity becomes soon replaced by a reward related positive potential shift. Neuropeptide Y (NPY) is known from animal studies to increase food intake and induce weight gain, which might result from increasing hunger drive or reducing satiety. Here we investigated whether NPY affects the cortical sequelae of hunger and satiety regulation as reflected by cortical DC-potentials in man. DC-potentials were recorded over frontal (Fz, F3, F4), central (Cz, C3, C4) and parietal (Pz, P3, P4) electrode positions in 14 subjects who had abstained from eating for 15 h and who were intranasally administered 50 nmol of NPY and placebo 20 min prior to recordings. After a 3-min baseline epoch, subjects consumed 400 ml of liquid food within 5 min. Recordings ended 7 min after food consumption. In the placebo condition during food intake, with some delay a positive DC-potential shift developed which was most pronounced over frontal and central areas and reached maximum values 0-3 min after food consumption. NPY reduced this satiation associated positive shift (p<0.05) over all areas except P3 and Pz. Data suggest that NPY exerts its orexigenic influence by attenuating mechanisms of satiation.

The neuroendocrine control of glucose allocation

Here we propose that glucose metabolism can be understood on the basis of three concept-derived axioms: (I) A hierarchy exists among the glucose-utilizing organs with the brain served first, followed by muscle and fat. (II) Tissue-specific glucose transporters allocate glucose among organs in order to maintain brain glucose concentrations. (III) Exogenous carbohydrate supply compensates for glucose alterations that can temporarily occur in muscle and fat. Derived from the control theory, the simplest solution of allocating supply to 2 organs, e.g. brain and muscle, is a "fishbone"-structured model. We reviewed the literature, searching for neuroendocrine and metabolic mechanisms that can fulfill control functions in such a model: The tissue-specific glucose transporters are differentially regulated. GLUT 1, carrying glucose across the blood-brain-barrier, is independent of insulin. Instead, this trans-endothelial glucose transporter is rather dependent on potent regulators of blood vessel function like vascular endothelial growth factor - a pituitary counterregulatory hormone. GLUT 4, carrying glucose across the membranes of muscle and fat cells, depends on insulin. Thereby, insulin allocates glucose to muscle and fat. The hypothalamus-pituitary-adrenal (HPA) axis, the sympathetic nervous system (SNS), and vascular endothelial growth factor allocate glucose to the brain. Multiple "sensors" (some of which have only recently been identified as ATP sensitive potassium channels) measure glucose or glucose equivalents at various sites of the body: the ventromedial hypothalamus, the lateral hypothalamus, portal vein, pancreatic beta cell, renal tubule, muscle and adipose tissue. Feedback pathways both from the brain and from muscle and fat are involved in regulating glucose allocation and exogenous glucose supply. The main feedback signal from the brain is found to be glucose, that from muscle and fat appears to be leptin. In fact, the literature search revealed two or more biological mechanisms for the function of each component in the model, finding glucose regulation highly redundant. This review focuses on "brain glucose" control. The concept of glucose allocation presented here challenges the common opinion of "blood glucose" being the main parameter controlled. According to the latter opinion, hyperglycemia in the metabolic syndrome is due to a putative defect located within the closed loop including the beta cell, muscle and fat cells. That traditional view leaves some peculiarities of e.g. the metabolic syndrome unexplained. The concept of glucose allocation, however, would predict that weight gain - with abundance of glucose in muscle and fat - increases feedback to the brain (via hyperleptinemia) which in turn results in HPA-axis and SNS overdrive, impaired insulin secretion, and insulin resistance. HPA-axis overdrive would account for metabolic abnormalities such as central adiposity, hyperglycemia, dyslipidemia, and hypertension, that are well known clinical aspects the metabolic syndrome. This novel viewpoint of "brain glucose" control may shed new light on the pathogenesis of the metabolic syndrome and type 2 diabetes.

EEG synchronization upon reward in man

OBJECTIVES

It was tested whether reward in humans is associated with EEG synchronization similar to that seen in animals.

METHODS

In two experiments (I and II) the EEG was recorded from frontal, central, and parietal positions before, during, and after drinking or oral stimulation. In Experiment I, subjects (n=11) who had either been thirsty for 16h or had quenched thirst before recordings, drank 400ml of water. In Experiment II, thirsty subjects (n=11) either drank 400ml of water or sucked on a soother. The recording epochs included a 3min baseline, an interval of about 5min during which subjects drank or sucked on the soother, and a 7min post-drinking interval.

RESULTS

During the drinking epoch, beta band-power (12-30Hz) was enhanced in both conditions of Experiment I and II, respectively. In Experiment I, after drinking, lower alpha power (8-10Hz) was higher when subjects were thirsty than when they were not. Lower alpha was also enhanced in the post-drinking interval of both conditions of Experiment II, and after sucking, lower alpha synchronization was in addition accompanied by increased theta activity (4-8Hz).

CONCLUSIONS

Increased beta activity during drinking and sucking in thirsty subjects presumably reflects non-specific activation related to the motivational strength of sensorimotor regulation during consumatory behavior. The thirst dependent lower alpha synchronization after drinking, generated not only by water consumption but also by surrogate oral stimuli, can be considered a reflection of the drive reducing and rewarding qualities of oral stimulation and consumatory behavior.

[Jeremy, age 6 and one-half months, consulting for hypotrophy]

This is the medical history of a 6 month old baby who suffers from retarded weight gain. The difficulties encountered in establishing the diagnosis are outlined.

The effect of experimentally induced insulin resistance on the leptin response to hyperinsulinaemia

OBJECTIVE

Insulin is thought to be an important regulator of leptin secretion. However, increasing evidence suggests that insulin-mediated glucose uptake rather than insulin per se regulates circulating leptin concentration. Here, we hypothesised that a reduction of insulin sensitivity, ie insulin resistance, will diminish the stimulatory effect of insulin on leptin secretion as a consequence of decreased insulin-mediated glucose uptake.

DESIGN

Changes in serum leptin concentration during 30 hyperinsulinaemic-hypoglycaemic clamps were studied after induction of different levels of insulin resistance in normal-weight men. In 15 subjects insulin sensitivity was reduced by exposing them to a 2.5 h antecedent hypoglycaemia (3.1 mmol/l) induced by a high rate of insulin infusion (15.0 mU/min/kg) on the day before the proper experiment ('ante-hypo' condition). In the other 15 subjects no antecedent hypoglycaemia was induced ('control' condition). The proper experiment on both conditions was a 6 h stepwise hypoglycaemic clamp induced by a constant rate of insulin infusion (1.5 mU/min/kg).

SUBJECTS

Experiments were carried out in 30 lean healthy subjects (age, mean +/- s.e.m., 26 +/- 1 y; body mass index, 23.1 +/- 0.6 kg/m2).

RESULTS

As expected, glucose demand during the clamp was lower in the ante-hypo condition than in the control condition (gram of glucose infused per kilogram body weight, 1.52 +/- 0.16 vs 2.01 +/- 0.17 g/kg; P < 0.05). During the clamp, leptin levels increased by 25.4 +/- 4.3% in the control condition (P < 0.05), but not in the ante-hypo condition (+4.8 +/- 4.5%; P > 0.25). Thus, serum leptin response to the clamp significantly differed between the two conditions (P < 0.01). Across both conditions, the increase of leptin levels during the clamp was correlated with the amount of glucose infused (r = 0.37; P < 0.05).

CONCLUSION

Considering that insulin concentrations were identical during both clamp conditions, the data indicate that experimentally-induced insulin resistance diminishes the stimulatory effect of insulin on leptin secretion.

Drinking related direct current positive potential shift in the human EEG depends on thirst

Scalp recorded direct current (DC)-potential shifts were examined in 11 human subjects who had either thirsted for 16 h or had quenched thirst before recordings. The recording epoch included a 3-min baseline, an interval of about 5 min during which subjects drank 400 ml of water, and a 7-min post-drinking interval. Consistent with previous data, when thirsty, subjects displayed a widespread negative DC-potential shift during drinking which was replaced by a positive DC shift at the transition to the post-drinking interval. The positivity after drinking lasted for about 2 min and averaged 146 microV at frontal recording sites. Quenching thirst before recordings reduced the positive DC-potential shift upon drinking, whereas changes in preceding drinking related DC negativity appeared to be secondary. The post-drinking positive DC-potential shift depending on the subject's motivational state can be considered an indicator of reward associated with quenching thirst, pointing to a lowered frontocortical excitability during reward.

Hypoglycemia, but not insulin, acutely decreases LH and T secretion in men

Hypoandrogenemia is frequently associated with hyperinsulinemia in men with the metabolic syndrome. We questioned whether insulin or changes in blood glucose levels influence pituitary gonadotropin secretion or testicular steroidogenesis in healthy men. Also, the relationship between hypoglycemia-induced activation of the hypothalamus-pituitary-adrenal axis and altered steroidogenesis was examined. Euglycemic and hypoglycemic clamp experiments were performed in 30 healthy men over a period of 6 h. Half of the men were infused with insulin at a rate of 1.5 mU/min.kg; the other half were infused at a rate of 15.0 mU/min.kg. Plasma glucose was held constant during a euglycemic clamp session and was decreased stepwise in a hypoglycemic clamp session. LH and total/free T concentrations decreased under hypoglycemic conditions regardless of the rate of insulin infusion. With euglycemic conditions, LH and T levels remained unchanged. Dehydroepiandrosterone concentrations increased during hypoglycemia, but not during the euglycemic conditions. The FSH concentration was not affected by insulin or glycemic clamps. Hypoglycemia acutely suppresses T secretion, and this effect is apparently mediated by pituitary LH. Insulin is ineffective. As counterregulation to hypoglycemia begins at normoglycemic ranges in poorly controlled type 2 diabetes and probably also in patients with long-term perturbed glucose regulation in the metabolic syndrome, control of glucose-responsive neurons in the brain may contribute to hypoandrogenemia. Apart from down-regulation of hypothalamic release of GnRH, concurrent activation of the pituitary-adrenal axis (i.e. increased release of dehydroepiandrosterone) may add to the suppressive effect of hypoglycemia on gonadal steroidogenesis.

Improving influence of insulin on cognitive functions in humans

Insulin receptors have been identified in limbic brain structures, but their functional relevance is still unclear. In order to characterize some of their effects, we evaluated auditory evoked brain potentials (AEP) in a vigilance task, behavioral measures of memory (recall of words) and selective attention (Stroop test) during infusion of insulin. The hormone was infused at two different rates (1.5 mU/kg x min, "low insulin", and 15 mU/kg x min, "high insulin"), inducing respectively serum levels of 543 +/- 34 and 24,029 +/- 1,595 pmol/l. This experimental design allowed to compare cognitive parameters under two conditions presenting markedly different insulin levels, but with minimal incidence on blood glucose concentrations since these were kept constant by glucose infusion. A "no insulin treatment" group was not included in order to avoid leaving patients infused with glucose without insulin treatment. Measures were taken during a baseline phase preceding insulin infusion and every 90 min during the 360 min of insulin infusion. Compared with "low insulin", "high insulin" induced a slow negative potential shift in the AEP over the frontal cortex (average amplitude, high insulin: 0.27 +/- 0.48 microV; low insulin: 1.87 +/- 0.48 microV, p < 0.005), which was paralleled by enhanced memory performance (words recalled, high insulin: 22.04 +/- 0.93; low insulin: 19.29 +/- 0.92, p < 0.05). Also, during "high insulin" subjects displayed enhanced performance on the Stroop test (p < 0.05) and expressed less difficulty in thinking than during "low insulin" (p < 0.03). Results indicate an improving effect of insulin on cognitive function, and may provide a frame for further investigations of neurobehavioral effects of insulin in patients with lowered or enhanced brain insulin, i.e., patients with Alzheimer's disease or diabetes mellitus.

Losartan attenuates symptomatic and hormonal responses to hypoglycemia in humans

OBJECTIVE

Reduced awareness of hypoglycemic symptoms and compromised hormonal counterregulation increase the risk of severe hypoglycemia in people with diabetes mellitus. Up to the present, angiotensin 1 receptor blockers, which play an important role in controlling diabetic complications, have not been known to increase the risk of hypoglycemia. Nevertheless, we observed 3 cases of diabetic patients complaining of reduced awareness of hypoglycemic symptoms while they were under treatment with losartan in our outpatients clinic. We therefore investigated the effects of losartan on symptomatic and hormonal responses to hypoglycemia in humans.

RESEARCH DESIGN AND METHODS

We carried out a randomized, double-blind, crossover study including 16 healthy men. The subjects received losartan 50 mg/d versus placebo. Treatment periods lasted for 7 days and were followed by a stepwise hypoglycemic clamp session (4.5 to 3.8 to 3.1 to 2.4 mmol/L) with measurement of counterregulatory hormones (epinephrine, norepinephrine, adrenocorticotropin, cortisol, glucagon), symptoms, and hemodynamic parameters (blood pressure, heart rate).

RESULTS

Losartan attenuated the hypoglycemia-induced rise in plasma epinephrine (6480 +/- 490 pmol/L versus placebo 8970 +/- 790 pmol/L; P <.001) and the rise in plasma adrenocorticotropin (21 +/- 2 pmol/L versus 26 +/- 3 pmol/L; P <.01). Losartan also reduced symptom scores during hypoglycemia (P <.05).

CONCLUSION

We conclude that short-term treatment with losartan slightly attenuates symptomatic and hormonal responses to hypoglycemia. At present, for patients who are unaware of hypoglycemia and who require antihypertensive or nephroprotective treatment, we would recommend caution concerning treatment with losartan.

Changes in immune cell counts and interleukin (IL)-1beta production in humans after a somnogenically active growth hormone-releasing hormone (GHRH) administration

Growth hormone-releasing hormone (GHRH) has been shown to enhance slow-wave sleep (SWS) and non-rapid eye movement sleep in animals and humans. In animals the somnogenic effect of interleukin (IL)-1beta appears to be mediated by GHRH. Neuroimmunological interactions in sleep are most frequently studied in humans by sleep deprivation or by cytokine administration. The present study, in contrast, investigates in humans the effect of enhanced sleep through GHRH administration on selected immune parameters. Results reveal that a single intravenous bolus of 50 microg GHRH which enhanced SWS stage 4 in the first half of the night suppressed circulating suppressor T cell (CD3+/CD8+) numbers, with a similar tendency for B cells (CD19+) and suppressed mitogen-stimulated IL-1beta production. When the same amount of GHRH was administered distributed across five repetitive boluses of 10 microg GHRH within 1 h, neither corresponding sleep nor immune parameters were changed significantly compared to placebo. These data suggest that GHRH can modulate immune functions through brain mechanisms which are also involved in the regulation of sleep.

Metformin does not adversely affect hormonal and symptomatic responses to recurrent hypoglycemia

Body weight gain and severe hypoglycemia are the major adverse effects of insulin therapy in type 2 diabetic patients. Metformin has been shown to prevent insulin therapy-induced body weight gain when used in combination with insulin. However, the effects of metformin on hormonal and symptomatic responses to hypoglycemia mediating hypoglycemia awareness have not been assessed to date. Fifteen young healthy men were treated with 850 mg metformin and placebo twice daily for a 16-d period in a double blind, cross-over design. On the last 2 d of the treatment period, the subjects underwent three hypoglycemic clamp experiments, with the first and the last performed with identical patterns of plasma glucose decrease. Differences between the effects of metformin and placebo (effect of metformin) as well as between first and last hypoglycemic clamps (effect of antecedent hypoglycemia) were assessed. Antecedent hypoglycemia significantly reduced epinephrine, ACTH, cortisol, glucagon, GH, and symptomatic responses to hypoglycemia (P < 0.05 for all variables). There was no detectable effect of metformin on epinephrine, norepinephrine, ACTH, cortisol, glucagon, or autonomic symptomatic response to hypoglycemia (P > 0.05 for all comparisons), except that metformin slightly increased the response of GH to hypoglycemia (P = 0.039). The latter finding may be due to an IGF-I-reducing effect of metformin, as after 14 d of metformin treatment baseline levels of IGF-I were significantly lower than in the placebo condition (236.9 +/- 13.9 vs. 263.2 +/- 14.4 microg/liter; P = 0.015). The data indicate that metformin does not adversely affect hormonal and symptomatic responses to hypoglycemia. This finding appears to be relevant with regard to the safety of the combination of metformin with insulin therapy.

Time course of intranasally administered cholecystokinin-8 on central nervous effects

The gut and brain peptide cholecystokinin (CCK) exerts a number of central nervous effects. Among them are effects on attention and stimulus processing as revealed by modulations of event-related potentials (ERPs). In the present study the time course of central nervous effects after an intranasal administration of CCK-8 was investigated by means of ERPs. ERPs were recorded in an oddball paradigm 15, 30, 60, 90, 120, and 240 min after administration. Following the double-blind intranasal administration of CCK-8 and placebo, the late positive complex (LPC) of the ERP was significantly increased following CCK-8 compared to placebo. This effect was more pronounced in women than in men. The enhancement of the LPC by intranasal CCK-8 was not restricted to a specific recording time but reached its maximum 120 min after administration in men and women. Moreover, results tentatively indicate that 30 min after administration of CCK-8 the LPC increased only in women but not in men. The early effect of intranasal CCK-8 on LPC in women is unlikely to be caused by changes in plasma CCK-8 levels and suggests a direct nose-brain pathway.

The Effect of Food Deprivation on ERP During Identification of Tachistoscopically Presented Food-Related Words

Abstract The present study served to investigate the effects of food deprivation on the identification of subliminally presented food-related words by means of event-related potentials (ERPs). ERPs were recorded in 16 hungry and 16 satiated subjects during repeated tachistoscopic presentation of food-related words (food names) and food-unrelated words (neutral words, sexual words) as controls. ERPs were recorded during each presentation of a word prior to identification and during the first presentation after identification and exhibited N1, P2, and slow-wave components. The number of tachistoscopic presentations until identification was not affected by hunger and satiety. However, ERPs were differentially affected by hunger and satiety: the P2 to food-related words was larger in hungry subjects compared to satiated subjects in all presentations. Additionally, the P2 was also larger to sexual words in hungry subjects in all presentations except the one preceding the identification response. The slow wave was not affected by hunger but increased with progressing stimulus identification. Following the identification of the words, all ERP components markedly declined in amplitude. The results indicate that hunger affects the processing of food and sexual stimuli during identification at an early ERP component (P2) even if the stimuli are not fully identified. In contrast, the later slow wave is sensitive to progressing stimulus identification, irrespective of hunger and stimulus meaning.

Translocation (16;17)(q22;p13) is a recurrent anomaly of aneurysmal bone cysts

Recently, Panoutsakopoulos et al. (1999) reported 2 cases of aneurysmal bone cysts with a recurrent (16;17)(q22;p13) translocation. We present here two additional cases harboring the same translocation as well as additional chromosomal changes.

Hyperinsulinemia causes activation of the hypothalamus-pituitary-adrenal axis in humans

OBJECTIVE

Hyperactivity of the hypothalamus-pituitary-adrenal (HPA) axis is frequently found in hyperinsulinemic subjects, such as patients with diabetes or abdominal obesity. Here, the question has been posed as to whether hyperinsulinemia increases HPA secretory activity.

METHODS

We performed paired-euglycemic and stepwise hypoglycemic (76-66-56-46 mg/dl)-clamp experiments in two groups (each of 15 healthy men) at different insulin infusions rates, ie, 1.5 mU/min/kg (low-insulin condition) and 15.0 mU/min/kg (high-insulin condition).

RESULTS

During the euglycemic clamp, the high rate insulin infusion increased plasma ACTH levels, whereas plasma ACTH levels remained essentially unchanged during the low-insulin condition (condition by time interaction, P=0.008). Likewise, serum cortisol levels were higher during the high- vs low-insulin condition (condition by time interaction, P=0.004). During the hypoglycemic clamp, plasma ACTH levels did not differ between the low- vs high-insulin condition, while serum cortisol levels were higher during the high- vs low-insulin condition at the beginning of the clamp (plasma glucose approximately 76 mg/dl; P=0.032).

CONCLUSION

Data indicate that hyperinsulinemia acutely increases HPA secretory activity in healthy men. This finding appears to be relevant to the pathogenesis of many clinical abnormalities associated which diabetes and abdominal adiposity, often referred to as the metabolic syndrome.

Hypoxia regulates avian cardiac Arnt and HIF-1alpha mRNA expression

The aryl hydrocarbon receptor nuclear translocator (Arnt) and hypoxia-inducible factor (HIF)-1alpha mediate cellular responses to hypoxia. We investigated the ability of hypoxia to regulate Arnt and HIF-1alpha mRNA in the heart in vivo. We cloned avian Arnt, developed an in vivo model of chronic cardiac hypoxia, and measured expression of cardiac Arnt and HIF-1alpha mRNA by quantitative RT-PCR. Chronic hypoxic exposure (24 h to 15% O(2)) of day 9 chick embryos resulted in a 30-fold increase in covalent binding of (3)H-misonidazole, a hypoxic tissue marker, to cardiac tissue, and a 2-fold induction of cardiac inducible nitric oxide synthase mRNA, compared to normoxic controls. In this same model, cardiac Arnt mRNA expression decreased by 35%, while HIF-1alpha mRNA expression increased 400%. These data suggest that regulation of Arnt and HIF-1alpha mRNA expression may contribute to the physiological responses of the heart during prolonged hypoxia.

Increase in systolic blood pressure and catecholamine level during hyperinsulinemia in a placebo-controlled euglycemic clamp in healthy subjects

Hyperinsulinemia has been shown to induce vasodilation and activation of the sympathetic nervous system. Whether these effects result in changes in blood pressure (BP) is discussed controversially. We measured BP and catecholamine levels in 50 healthy subjects during a 40-min baseline phase and during a 100-min euglycemic clamp phase. In a double-blind, between-subject comparison, 30 subjects were infused with 1.5mU insulin/kg x min, 20subjects were infused with saline solution. Insulin levels increased during insulin infusion from (mean+/-SE) 23.7 0.6 pmol/l to 406.2+/-3.0 pmol/l, but remained unchanged during placebo infusion. Blood glucose levels were identical during both conditions. Systolic BP increased from 116.6+/-1.5 mmHg to 119.8+/-1.8 mmHg during insulin infusion and decreased from 116.6+/-2.3 mmHg to 114.0+/-2.4 mmHg during placebo infusion (p<0.001, for the difference between the effects of insulin vs. placebo). Heart rate was higher during insulin infusion as compared to placebo infusion (63.8+/-1.9 vs. 60.9+/-2.4 beats/min, p<0.05). Norepinephrine levels increased from 1.25+/-0.09 to 1.58+/-0.12 nmol/l during insulin infusion and remained unchanged during placebo infusion (1.24+/-0.09 vs. 1.29+/-0.11 pmol/l; p<0.001). Epinephrine levels were also higher during insulin as compared to placebo infusion (249.8+/-17.4 vs. 212.8+/-21.1 pmol/l, p<0.001). The changes did not depend on whether the subject experienced his first or second clamp. Data demonstrate reproducable increasing effects of hyperinsulinemia within the normal physiological range on catecholamine release and systolic BP in healthy humans.

Rhythms of pituitary-adrenal activity during sleep in patients with Cushing's disease

Previous studies have indicated a dependence of nocturnal pituitary-adrenal secretory activity on central nervous sleep processes in healthy humans: Under normal physiological conditions the release of ACTH/cortisol is inhibited during early sleep and becomes entrained to periods of NonREM sleep during late sleep. Here, we compared nocturnal dynamics in plasma concentrations of ACTH/cortisol in 7 patients with Cushing's disease with those of 7 healthy controls matched in age and sex with the patients. The patients in part were repeatedly tested. The total of 13 nights is composed of 7 nights of hyperpulsatile secretion pattern (5 patients) and 6 nights from hypopulsatile secretion pattern (4 patients). After an adaptation night polysomnographic sleep recordings were obtained and blood was sampled every 15 min between 23.00 and 7.00 h. Controls displayed the typical minimum in ACTH/cortisol concentrations during the early part of the night and maximum concentrations during the late part of the night, whereas ACTH/cortisol levels of Cushing patients indicated a relatively constant elevated pituitary-adrenal activity throughout the night, lacking any circadian variation. Autocorrelation functions revealed the presence of cortisol secretory rhythms with a similar period length in healthy controls (155.6+/-17.4 min) and patients with a hyperpulsatile pattern (142.4+/-6.6 min). In patients displaying hypopulsatility, no significant rhythmicity was observed. However, regardless of the type of secretory pulsatility, adrenal secretory activity started predominantly during periods of NonREM sleep (p<0.01) in healthy controls as well as in patients with Cushing's disease. This data indicates that the normal nocturnal circadian oscillation of pituitary-adrenal activity is absent in Cushing patients, whereas a link between pituitary-adrenal activity and ultradian rhythms of sleep appears to be preserved.