Psychiatric Annals


The HPA Axis and Psychopathology: Cushing's Syndrome

Monica N Starkman, MD, MS


Cushing's disease, the major type of spontaneous Cushing's syndrome, is the classic endocrine disease characterized by hypersecretion of Cortisol and diminished suppression of Cortisol by dexamethasone.


Cushing's disease, the major type of spontaneous Cushing's syndrome, is the classic endocrine disease characterized by hypersecretion of Cortisol and diminished suppression of Cortisol by dexamethasone.

A substantial percentage of patients with major depressive disorder (MDD) hypersecrete Cortisol, and I many escape early from the normal feedback inhibition by administered steroids such as dexamethasone. Thus, hypercortisolemia and dysregulation of the hypothalamicpituitary-adrenocortical (HPA) axis have become of major interest in psychiatry.1

Endocrine diseases are experiments of nature that can be exploited to study the relationships between neuroactive substances and psychopathology. Patients with pheochromocytoma, for example, provide the opportunity to explore the association of elevated peripheral levels of catecholamines and anxiety.2,3 In a similar fashion, patients with Cushing's syndrome afford an excellent opportunity to study the association of hypercortisolemia and disorders of mood and cognition.

Cushing's disease, the major type of spontaneous Cushing's syndrome, is the classic endocrine disease characterized by hypersecretion of Cortisol and diminished suppression of Cortisol by dexamethasone. Intriguingly, emotional disturbances were already recognized as a feature of the disease in Harvey Cushing's original description.4 This association was later confirmed by retrospective chart reviews.5 More recently, evaluations of patients with Cushing's syndrome prior to treatment have shown that the majority manifest many of the clinical features seen in patients with a primary depressive disorder, such as depressed mood, irritability, decreased concentration, decreased libido, and middle and late insomnia.6-7 In fact, the majority of patients with Cushing's syndrome were found to meet DSM-III criteria for a major depressive disorder (except for the exclusion of an organic mental disorder).7,8

From a theoretical point of view, defining and exploring the clinical and biochemical similarities between patients with spontaneous Cushing's syndrome and patients with primary depression provides a unique opportunity to examine in greater detail the relationship between HPA axis dysregulation and abnormalities in mood and cognition. For the clinician, however, these similarities may lead to a problem in differential diagnosis, particularly early in the course OfCuShUIg1S syndrome when the characteristic physical stigmata are less prominent.

This paper will use the vantage point of Cushing's disease to focus on the HPA axis with emphasis on the psychopathogenic effects of its products, and consider aspects of the differential diagnosis between Cushing's disease and major depressive disorder.


Under normal conditions, the secretion of Cortisol from the adrenal gland is regulated by a system extraordinarily sensitive to changes within the organism and changes in its environment. Hypothalamic neurosecretory cells produce corticotropin-releasing hormone (CRH), and neurotransmitter pathways modulate its release. CRH, in turn, acts on the adrenocorticotrophic hormone (ACTH)-producing cells of the anterior pituitary, and ACTH is secreted into the systemic circulation, acting on the adrenal cortex to elicit the secretion of corticosteroids, including Cortisol.

Regulatory mechanisms are present at various levels along the axis. Cortisol feedback occurs at pituitary, hypothalamic, and suprahypothalamic brain levels. An endogenous timer superimposes a Orcadian pattern on CRH, and thereby on ACTH and Cortisol secretion. The suprachiasmatic nucleus is a major anatomic component of the timer, regulating the periodicity of other physiologic functions such as body temperature, eating, drinking, and activity. In addition, central nervous system centers regulate CRH release in response to environmental and endogenous inputs such as emotional and physical stress, infection, and the metabolic milieu.

Plasma ACTH and Cortisol concentrations fluctuate over the 24-hour period. Highest concentrations are present just prior to and at the time of awakening, while lowest concentrations occur prior to and during the early part of sleep. With a normal sleep cycle, Cortisol secretion is virtually absent between midnight and 4:00 AM. This is followed by a period of increasing HPA activity in the form of seven to nine short secretory episodes until around 9:00 AM.

ß-endorphin, an opiate-like peptide, is also secreted by the pituitary gland. Pro-opiomelanocortin, the precursor of ACTH, contains the entire sequence of both ACTH and ß-lipotropin, from which ß-endorphin (BE) is derived. When an ACTH-releasing stimulus is given, both ACTH and BE are secreted concomitantly.9


In spontaneous Cushing's syndrome, adrenocortical hyperfunction develops in one of two ways: excessive secretion of ACTH, which then stimulates the adrenal gland; or pathology in the adrenal gland, such as a neoplasm, leading directly to the overproduction of Cortisol. Thus, Cushing's syndrome is divided into two types: ACTH-dependent and ACTH-independent.

In Cushing's disease, the most common form of spontaneous Cushing's syndrome, excessive ACTH is produced by the pituitary gland. In addition to excessive secretion of ACTH, the normal circadian rhythm of ACTH release is blunted and the sensitivity of the feedback control system to levels of circulating Cortisol is diminished.

Much more rarely, ACTH-dependent Cushing's syndrome is caused by an ectopic CRH or ACTH-producing tumor such as a thymoma. These nonpituitary neoplasms have the capacity to synthesize and release peptides resembling CRH and ACTH, which are physiologically active. The tumor-produced ACTH under these circumstances leads to excessive stimulation of Cortisol secretion by the adrenal cortex, which appropriately suppresses ACTH release from the normal pituitary gland. Thus, the high ACTH levels in these patients come from the neoplasm and not from the pituitary gland. Oat-cell carcinoma of the lung is the most common of these neoplasms.

Patients with excessive ACTH secretion derived from either the pituitary or an ectopic tumor may also secrete melanocyte stimulating hormone as well as breakdown fragments of ACTH that have melanotropic activity, leading to hyperpigmentation of the skin and mucous membranes.

ACTH-independent Cushing's syndrome occurs when a tumor develops in the adrenal cortex with a capacity to secrete Cortisol in an autonomous fashion. Adrenocortical tumors leading to Cushing's syndrome may be benign adenomas or malignant carcinomas. In comparison to pituitary ACTH-dependent disease, adrenal adenomas and carcinomas are quite rare.


The etiology (or etiologies) of spontaneous pituitary ACTH-dependent Cushing's syndrome, known as Cushing's disease, is still controversial. There are currently several major hypotheses:

1. The primary disorder is in the neoplastic pituitary cells, which comprise the pituitary microadenoma seen in a large percentage of patients with Cushing's disease. The exaggerated ACTH secretory response of many of these patients to stimulation by administered CRH hormone is consistent with this hypothesis.10

2. The primary disorder is in the central nervous system. Abnormality in the limbic system, in particular, may lead to an overproduction of hypothalamic CRH, which drives the pituitary to increased secretion of ACTH and thence, the adrenal to secrete Cortisol.11,12 The disordered circadian periodicity of ACTH, growth hormone, and prolactin secretion and the remission of a small percentage of patients treated with cyproheptadine, a serotonergic blocking agent, are considered consistent with this view.

3. There may be two forms of Cushing's disease, one involving CRH secretion and one not. Studies analyzing episodic Cortisol secretion support this hypothesis.13

4. The idea of a continuum of HPA axis abnormalities has been proposed, so that Cushing's disease and MDD lie on one endocrinologie spectrum, separated by an overlapping "gray zone." 14

5. A psychosomatic hypothesis has been proposed: that Cushing's disease is a pathophysiologic reaction to bereavement that occurs in predisposed individuals with lifelong disturbances in personality structure, homeostatic self-regulation, and neuroendocrine responsiveness.15 Significantly more stressful life events in the year prior to onset of Cushing's syndrome have been observed, suggesting that such stress may be part of a multifactorial model of pathogenesis.16


Glucocorticoids have catabolic and antianabolic effects on protein. There is loss of protein from tissues such as skin, muscles, blood vessels, and bone. Clinically, the skin atrophies and breaks down easily, and wounds heal slowly. Rupture of elastic fibers in the skin causes purple stretch marks, or striae. Muscles atrophy and become weak. Thinning of blood vessel walls and weakening of perivascular supporting tissue result in easy bruising. The protein matrix of bone becomes weak, causing osteoporosis. Bones become brittle and develop pathological fractures. Osteoporosis occurs most frequently in the spine, leading to vertebral collapse with back pain and loss of height.

Carbohydrate metabolism is affected by abnormally high levels of glucocorticoids. Glucocorticoids stimulate gluconeogenesis and interfere with the action of insulin in the peripheral cells. While some Cushing's syndrome patients compensate by increasing insulin secretion and subsequently normalizing glucose tolerance, others with diminished insulinsecreting capacity (prediabetic Cushing's syndrome or subclinical diabetic Cushing's syndrome) develop abnormal glucose tolerance tests, fasting hyperglycemia, and clinical manifestations of overt diabetes.

Excessive glucocorticoid levels affect the distribution of adipose tissue that accumulates in the central areas of the body. Patients develop truncal obesity, round face (moon facies), supraclavicular fossa fullness, and cervicodorsal hump (Buffalo hump). Truncal obesity occurs together with thinning of the upper and lower extremities as a result of muscular atrophy.

Although glucocorticoids normally have minimal effects on serum electrolytes, in very large concentrations they may cause sodium retention and potassium loss, leading to edema, hypokalemia, and metabolic alkalosis.

Glucocorticoids inhibit the immune response. They impair humoral antibody production and the induction and proliferation of immunocompetent lymphocytes. They also suppress delayed hypersensitivity reactions, so that skin tests for tuberculosis may convert from positive to negative. Clinically, these patients have an increased susceptibility to infectious disease.


The diagnosis of Cushing's syndrome is confirmed by the measurement of abnormally high levels of Cortisol in plasma and urine, impairment of the sensitive feedback control mechanism, and diminution or absence of circadian rhythm of Cortisol secretion. Not all patients demonstrate the full picture of these abnormalities, however. Diminished or absent sensitivity of the feedback control system is demonstrated by lack of suppression of Cortisol following administration of the synthetic steroid dexamethasone. In order to define the subtype of Cushing's syndrome, metyrapone, an inhibitor of adrenal corticosteroid synthesis that lowers Cortisol levels, can be used to test hypothalamic-pituitary responsiveness to the removal of the strong inhibitory feedback exerted by Cortisol. More recently, CRH has been used to demonstrate the exaggerated secretion of ACTH seen in patients with Cushing's disease.


TABLE 1Frequency of Psychiatric Symptoms in 35 Patients with Cushing's Syndrome


Frequency of Psychiatric Symptoms in 35 Patients with Cushing's Syndrome


The frequency of neuropsychiatrie symptoms reported in 35 consecutive patients with Cushing's syndrome (26 of these with Cushing's disease) whom we studied prior to treatment is shown in Table 1.

It should be noted that the majority of patients with Cushing's syndrome do not have very severe neuropsychiatrie disturbances of a psychotic or confusional nature, as patients receiving exogenous corticosteroids sometimes may. Only four of the 35 patients (11%) manifested a thought disorder and/or confusional state. Of these, one of the two with Cushing's disease manifested depression with paranoid ideation, while the other was episodically confused. The other two patients had ectopic ACTH production; one was manic with paranoid ideation, while the other was depressed and confused with paranoid ideation.

This observation can be understood by considering that patients with spontaneous Cushing's syndrome differ from those receiving high-dose exogenous steroids in several respects. The level of circulating Cortisol in the former is not as high as the equivalent amount of steroid administered to the latter. We found the mean Cortisol secretion rate of the 35 patients we studied to be 73 mg per day, an amount equivalent to 20 mg of prednisone.17

The Boston Collaborative Drug Surveillance Program, reporting on administration of high doses of prednisone to hospitalized patients, noted a dose-response relationship between prednisone and acute major psychiatric reaction (psychoses and/or euphoria). This correlation was not present for other adverse reactions such as gastrointestinal side effects.18

The probability of developing an acute psychotic reaction to steroids is highest when dosages of more than 40 mg per day of prednisone or its equivalent are administered.19 It is of interest that those patients with Cushing's syndrome in our study who did have a very severe psychiatric disability with delusional and/or confusional symptoms had a mean Cortisol secretion rate of 157 mg per day, an amount of Cortisol that is equivalent to 40 mg per day of prednisone.17

Patients with Cushing's syndrome also differ from those receiving exogenous steroids in that they are exposed to sustained elevated Cortisol levels for months to years, and are less subject to sudden acute shifts and rapid rates of change of steroid levels that occur during short-term treatment with high-dose corticosteroids.

Patients with Cushing's syndrome manifest a consistent constellation of symptoms that includes impairments in affect (irritability and depressed mood), vegetative functions (decreased libido and middle insomnia), and cognitive functions (decreased concentration and memory).

Mood and Affect

In our study, increased irritability, a very frequent symptom, was the earliest psychiatric symptom to appear. It began simultaneously with the onset of weight gain and prior to the appearance of other physical manifestations of Cushing's syndrome. Patients described themselves as having become overly sensitive, unable to ignore minor irritations, feeling impatient with or pressured by others. Since some patients reported that external noises bothered them excessively, this may reflect a generalized hypersensitivity to stimuli. In addition, an overreactivity and easy development of anger were reported. Patients described feeling that they were often on the verge of an emotional explosion and that the intensity of anger experienced was also increased. Although most could restrain themselves, patients were frightened by their irritability and potential for verbal or physical dyscontrol. They described being on guard for a flare-up of anger and fleeing from confrontations to avoid a feared loss of control.

Depressed mood was reported by 75% of the patients. There was a range in the intensity of depressed mood among patients who reported it. Some patients described short spells of sadness; others experienced feelings of hopelessness and giving up. Six patients had suicidal thoughts, and two of them had made suicide attempts since the onset of Cushing's syndrome. Some patients described hypersensitivity and oversentimentality as determinants leading to crying spells. For some, crying was experienced as their only available behavioral response to anger, frustration, and feeling unable to respond effectively. Patients also experienced spontaneous onset of depressed mood or crying in the absence of any preceding upsetting thought or event.

The time course of the mood disturbances was noteworthy. Most patients reported that their mood disturbances were intermittent rather than sustained. Sometimes they woke up depressed and remained depressed throughout the day or the next day as weD. Alternatively, the onset of depressed mood and/or crying might occur during the day, sometimes suddenly, with a rapid shift.

Although there were intervals when they might not experience pleasure, these patients did not describe persistent anhedonia. That is, most did not experience the unrelenting, unremitting inability to experience pleasure that is characteristic of patients with endogenous major depressive illness. There were intervals when they retained the capacity for pleasure and could still find enjoyment in hobbies and interpersonal relationships. At times, some patients found it difficult to initiate such activities, although once others mobilized them, they would enjoy them.

The duration of each depressive episode was usually one to two days and rarely longer than three days at a time. A frequent weekly total of dysphoria was reported as being three days per week. There was not a regular cycle, however, so patients could not predict when a depressive day would occur.

Social withdrawal was often related to shame because of physical appearance, as well as a decreased sense of focus, alertness, and clarity in the unstructured group setting. Most patients reported an increased desire to have contact with significant family members. However, sporadic withdrawal might occur because of the patient's need to remove himself or herself from a situation of overstimulation that elicited the fear of impending emotional dyscontrol.

Guilt was infrequent and, if present, was not excessive, self-accusatory, or irrational. Instead, it was related primarily to remorse about the uncontrollable angry outbursts and inability to function well at work and in the family.

Hopelessness, if present, was attributed to the existence of a chronic illness with increasing physical and emotional disability that so far had proved undiagnosable and un treatable.

A minority of patients had experienced episodes of elation-hyperactivity early in the course of the disease. As the disease progressed and new physical signs of Cushing's syndrome began to appear, this type of episode became rare or disappeared entirely. The quality of these episodes of elation was described most frequently as a "high." Patients were more ambitious than usual and might attempt to do more than their ability and training made reasonable. Increased motor activity was present, with restlessness and rapidly performed activities. Three patients reported finding, to their embarrassment, that their speech was both loud and rapid. In one of the most severely disturbed patients, who had an ectopic ACTH-secreting thymoma, a typical full-blown manic syndrome with increased pressure of speech, rhyming of words, and paranoid ideation was present.

A substantial percentage of patients in our study reported generalized anxiety. Newonset panic disorder has also been observed in patients with Cushing's syndrome.20 In addition, we have observed that even patients who do not experience psychic anxiety describe episodic symptoms of autonomic activation such as shaking, palpitations, and sweating.

Dialogic Drives

Abnormalities in three areas of basic biological vegetative drives were observed in our study.


TABLE 2Frequency of Abnormal Findings in the Mental Status Examination in 35 Patients with Cushing's Syndrome


Frequency of Abnormal Findings in the Mental Status Examination in 35 Patients with Cushing's Syndrome

Libido. A decrease in libido was very common. In fact, this is one of the earliest manifestations of Cushing's disease, beginning when the patient is experiencing the first onset of weight gain.

Appetite and Eating Behavior. Appetite changes were noted in 54% of patients, In 34%, appetite increased; in 20%, it decreased.

Sleep and Dreams. Difficulty with sleep, particularly middle insomma and late insomnia (early morning awakening), was found in more than 50% of the patients. Difficulty with early insomnia (not falling asleep at bedtime) was not as frequent. One third of the patients reported an alteration in the frequency or quality of their dreams. Some noted an absence of dreams; others reported a marked increase in their frequency and intensity. Their dreams had become bizarre, often frightening, and very vivid. These patients also reported that they had lost the ability to wake themselves out of a nightmare.


Cognitive symptoms are a prominent part of the clinical picture. Difficulty with concentration, inattention, distractibility, and shortened attention span were very frequent symptoms. Difficulty with reasoning ability, comprehension, and processing of new information were also reported often. Some patients reported episodes of rapid scattered thinking while others complained of slow and ponderous thinking. Thought blocking occurred. Patients complained of using incorrect words while speaking, and of misspelling simple words. Perceptual distortions were very rare.

Impairment of memory was one of the most frequent symptoms. Patients reported problems with registration of new information, which could be related in part to impaired concentration. They commonly repeated themselves in ongoing conversations. They easily forgot items such as appointments made, names of people, and location of objects. Difficulty occurred with recall of important dates in their personal or medical histories.

Overall, the psychiatric symptoms that develop in Cushing's syndrome and their quality argue for a pathogenesis over and above a nonspecific response to severe physical illness. Irritability and decreased libido occur early, often before patients are aware that they have any physical problems other than a steady increase in weight. Later in the course, when depressed mood makes its appearance, it is experienced not simply as the demoralization common to patients with medical illness, but as episodic sadness and crying, sometimes occurring in the absence of depressive thought content. The difficulties with memory and concentration seen in these patients appear in the absence of disorientation or overt clouding of consciousness, and the EEGs in almost all cases are not characteristic of delirium.7,21


The frequency of abnormal findings in the mental status examination is shown in Table 2. Notable are the difficulties with serial seven subtractions, interpretation of proverbs, and recall of presidents seen in close to 50% of the patients. Difficulty with recall of three cities was seen in more than 30% of the patients. Nevertheless, the great majority of patients had no disorientation or overt clouding of consciousness. Of 35 patients in the study, the electroencephalographs (EEGs) of only two were characteristic of delirium (both of whom had been classified prior to knowledge of EEG findings as being very severely psycbiatrically affected.)

Neuropsychological studies reveal a variability in severity of cognitive dysfunction, with one third demonstrating moderate to severe deficits in a wide variety of subtests. In addition, there is a hierarchy of a differential vulnerability of neuropsychologic functions, so that some are affected more frequently and severely than others, although in severely affected subjects there is a decline in almost all tests of verbal, nonverbal, sensory, and motor functions.22

Because of their hypercortisolemia and their cognitive deficits, patients with Cushing's syndrome provide the opportunity to investigate the intriguing relationship between glucocorticoid excess and hippocampal damage.

The hippocampal formation (HF), a part of the limbic system, plays an important role in new learning and intermediate memory. Interestingly, the hippocampal formation contains the highest concentration of corticosteroid binding sites in the entire brain.23-24 Because studies in animals indicate that prolonged stress and elevation of corticosteroids are toxic to a subpopulation of hippocampal cells,25,26 we hypothesized that damage to the hippocampus may occur in Cushing's syndrome patients, with subsequent deleterious effects on memory functions. We used magnetic resonance imaging (MRI) to investigate relationships between HF volume, memory dysfunction, and Cortisol levels in patients with Cushing's syndrome.27

The 12 Cushing's syndrome patients we studied exhibited variability in hippocampal formation volume, as well as variability in severity of learning and memory dysfunction as measured by neuropsychologic testing. There were strong and significant associations between reduced HF volume and lower scores on verbal learning and memory tests. These tests, components of the Wechsler Memory Scale with Russell modification, included paired associative learning and verbal recall with a 30-minute delay. This relationship between hippocampal formation volume and cognitive functioning was memory-specific, in that there were no significant correlations with full-scale IQ, a general measure of intellectual ability, or with psychomotor performance on the Trails test. In addition, there was a significant association between higher mean plasma Cortisol level and smaller hippocampal formation volume (r= - .73,p<.05).

These findings require caution in interpretation because the sample size was small. However, they are consonant with the association between hippocampal atrophy defined by MRI and the severity of memory dysfunction observed in patients with Alzheimer's disease.28 They are also consonant with studies in primates demonstrating cortisol-induced neuropathologic markers of hippocampal damage such as dendritic atrophy and soma shrinkage.26 Significant correlations between elevated Cortisol levels (induced by glucose tolerance test) and severity of hippocampal atrophy have been reported in patients with Alzheimer's disease as well.29

In fact, the pathophysiologic role of cortisol in the cognitive dysfunction seen in these patients likely extends even further beyond its specific effects on the hippocampus. Earlier studies using computed tomography (CT) revealed ventricular enlargement and cortical atrophy in patients with hypercortisolism due to Cushing's disease,30 which, interestingly, were reversible after treatment.


Cushing's syndrome is treated in a variety of ways. In patients with pituitary ACTHdependent disease, a transsphenoidal resection is usually attempted to remove a defined pituitary microadenoma. If there is evidence of pituitary hyperfunction but such a microadenoma tumor is not found or completely removed, cobalt irradiation of the pituitary gland is performed, although several years may pass before an effect on ACTH is seen. In the meantime, Cortisol can be controlled by chemical agents capable of blocking (metyrapone) or blocking and destroying (mitotane) the cortisol-secreting adrenocortical cells. Adrenal adenomas and carcinomas are treated by surgical removal, followed by chemotherapy in patients with carcinoma.

In nonmalignant conditions, improvement in hypertension can occur promptly, while remission of other physical signs and symptoms takes place over a period of six to 12 months. Similarly, while some neuropsychiatrie symptoms such as irritability improve rapidly, there may be a delay of months to years from the correction of elevated Cortisol levels to remission of psychiatric symptoms in patients with pituitary ACTH-dependent Cushing's syndrome31 or adrenal adenomas.8

In the treated patients we studied, improvements in depressed mood were manifested by a decrease in the frequency of days when the patient felt depressed. In addition, each episode lasted a shorter period of time, perhaps only a few hours instead of one to two days. The patients also described a change in quality of the depressive mood, so that they no longer experienced being in a total depression, or felt the depression as deep or allencompassing. They no longer felt depressed without some external precipitating reason. They noted that now the mood change came on gradually and disappeared gradually rather than appearing suddenly as before. Crying was less frequent, less easily elicited by environmental upsets, occurred only with some identifiable external precipitant, and was of shorter duration.31 These improvements occurred without the addition of any psychotropic medications such as antidepressants, neuroleptics, or minor tranquilizers.

The time course of improvement in depressed mood compared to improvement in other neurovegetative symptoms is of interest. In Cushing's disease patients who manifested depressed mood at initial evaluation and who were subsequently studied during the first 12 months after treatment was initiated, we have observed that improvement in symptoms other than depressed mood occurred prior to improvement in depressed mood. Depressed mood was less likely than irritability and sleep, for example, to be among the first cluster of symptoms to improve. Interestingly, this lag is similar to that seen in patients with major depressive episodes treated with antidepressants, in whom improvements in sleep and psychomotor activity often occur prior to improvement in depressed mood.


Using a modified Hamilton Rating Scale for Depression, a standard measure of the multiple components of the depressive syndrome, we found that individuals with massively elevated urinary-free Cortisol (1000 µg per day and above) had the highest depression severity score. For our sample of patients as a whole, the Hamilton depression scores at the pretreatment first visit showed a trend toward correlation with urinary free Cortisol (UFC) levels (r= +.39; p = .08). Scatter plots showed that at the higher levels of 24-hour UFC, higher scores on the Hamilton Rating Scale for Depression occurred consistently; at the lower levels of 24-hour UFC elevation, Hamilton scores could be either high or low.31

As for the role of ACTH, Cushing's syndrome patients with low or normal ACTH levels predominantly reported milder rather than pronounced depressed mood. Conversely, patients with a pronounced depressed mood rarely had low or normal ACTH levels. Patients with moderate to high ACTH levels were not necessarily severely depressed and were equally likely to show mild or pronounced depressed mood. Thus, in the milieu of elevated Cortisol levels prior to treatment, low or normal ACTH levels were associated with mild rather than pronounced depressed mood, while elevated ACTH levels were equally likely to be associated with mild or pronounced depressed mood.7

Since patients with adrenal adenoma comprise a major proportion of the Cushing's syndrome patients with low ACTH levels, it is informative to examine them as a separate group. In our study, these patients manifested no or mild depressed mood rather than pronounced depressed mood, even though their mean Cortisol levels were not significantly different from those of the Cushing's disease subgroup.7 Similarly, in a review of 78 cases of Cushing's syndrome reported in the literature, although % of patients with pituitary ACTH elevation had a problem with depression, only 1A of patients with Cushing's syndrome secondary to an adrenal tumor had depression, a highly statistically significant difference.32 In another series, of the four Cushing's syndrome patients who were mentally unaffected, three had adrenal tumors. In yet another study, of five patients with adrenal tumors, four were only mildly depressed.33 These studies, taken together, suggest that elevated Cortisol levels in the absence of elevated ACTH levels in patients with adrenal adenomas are more frequently associated with less severely depressed mood.

Several studies have investigated the relationship of improvement in depression with the decrease in Cortisol induced by treatment. Significant improvement in the Hamilton Rating Scale for Depression scores occur after treatment produces decreases in Cortisol levels.31'34 In fact, normalizing Cortisol levels can improve depressed mood and the depressive syndrome despite the continuation of elevated ACTH levels. In our study, 23 patients with Cushing's disease were treated with mitotane, an adrenal gland inhibitor that reduces cortisol levels while ACTH levels do not change or may even rise above pretreatment levels. Of our Cushing's disease patients with depressed mood prior to treatment who normalized UFC but continued to have elevated ACTH levels, 2/.3 showed improvement in depressed mood scores despite the continuing elevation of ACTH.31 Other groups, using treatments (bilateral adrenalectomy or metyrapone) that lower Cortisol but not ACTH also observed improvements in depression subsequent to Cortisol normalizationA33

In summary, the data from these studies indicate that patients with elevation of ACTH alone (as in mitotane-treated patients), or elevation of Cortisol alone (as in untreated patients with adrenal adenomas) are less likely to have severe depressed mood. In contrast, patients with elevation of both Cortisol and ACTH are more at risk of developing severe depressed mood. One hypothesis we have suggested is that abnormally elevated Cortisol levels may destabilize important psychoneurophysiologic systems, making the central nervous system more vulnerable to the effects of elevated ACTH and/or other neuroactive substances.35


The studies reviewed above lend support to the hypothesis that the neuropsychiatrie changes seen in Cushing's syndrome are related, at least in part, to the effect of elevated levels of Cortisol.

Further support for the role of Cortisol in eliciting neuropsychiatrie abnormalities comes from research with patients with primary MDD. Studies in patients with primary psychiatric disorders also indicate that increased Cortisol levels are associated with behavioral abnormalities. Psychiatric inpatients with high 8:00 AM values of plasma Cortisol, regardless of whether they had primary endogenous depression or some other psychiatric diagnosis, were more symptomatic on admission, and, on discharge, had more symptoms of sleep disturbance and decreased ability to think than patients with normal values of plasma Cortisol.36

Several investigators have reported extremely high urinary free Cortisol levels in subgroups of depressed patients who subsequently attempted suicide.37,38 Cerebrospinal fluid (CSF) Cortisol levels are elevated in depressed patients as compared to normal subjects, and there is a correlation of CSF Cortisol with the severity of depression.39

Possible mechanisms for the effect of Cortisol on behavior include a direct effect on the cells of the central nervous system, on neurotransmitter synthesis or function, on intracellular electrolytes, or by increasing the sensitivity of the brain to other neuroactive substances. Cortisol has effects on a wide variety of enzymes important in neurochemistry, and one mechanism of its action may be to induce alterations in mood-regulating neurotransmitter systems within the brain. Hydrocortisone increases liver tryptophan pyrolase with a subsequent decrease in brain tryptophan and serotonin.40 In tissue culture, hydrocortisone and dexamethasone stimulate tyrosine hydroxylase, the rate-limiting enzyme in catecholamine synthesis, and inhibit choline acetyltransferase, markedly inhibiting the synthesis of acetylcholine.41 In rat brain, corticosteroids alter the sensitivity to norepinephrine of the norepinephrine receptor-coupled adenylate cyclase system.42 Corticosteroids also interact with the GABA receptor complex, altering neuronal excitability.35

ACTH has potent behavioral effects. ACTH 4-10 increases the state of arousal in limbic midbrain structures.43 Microinjection of ACTH 1-24 into the periaqueductal gray matter of rats produces a syndrome of fearful hyperreactivity in which the animals shriek shrilly and jump repetitively upon exposure to mild auditory and visual stimuli normally neutral to the rat.44 Interestingly, microinjection of ß-endorphin (BE) elicits the opposite effect: sedated immobility. Since normal levels of ACTH and BE apparently do not cross the blood-brain barrier, how the elevated ACTH and ß-endorphin produced by the pituitary of patients with Cushing's disease directly affects the brain is not clear. One hypothesis we suggest is that the elevated levels of pituitary ACTH (and BE) produced by the pituitary are diffused by retrograde flow back up the portal vein to the hypothalamus, and regulate the brain's production of its own ACTH and BE, which occurs solely in the hypothalamus.

It is important to note that this paper has focused primarily on a single adrenal steroid, Cortisol, and a single pituitary peptide, ACTH. This reflects the current state of knowledge. Many other adrenal glucocorticoids as well as adrenal sex steroids are also elevated in patients with Cushing's disease, as are the pituitary peptides lipotropin and its product, endorphin. These substances likely contribute to the psychopathology of Cushing's disease - and possibly primary psychiatric disorders such as MDD - and remain to be studied.


Differentiation of Cushing's syndrome patients from those in the general population with similar physical signs and symptoms (obesity, hirsutism, menstrual irregularities, hypertension, diabetes mellitus) is difficult in the early stages of the disease. Manifestations of protein catabolism (muscle weakness and atrophy, thinning of skin and easy bruising) have been found to be the clinical features of greatest discriminatory value in distinguishing these patients from those with simple obesity. Since neuropsychiatrie symptoms such as irritability, decreased libido, sleep disturbance, and depressed mood appear consistently in the course of Cushing's syndrome, these symptoms too should raise the levels of suspicion for this diagnosis.

Deciding whether a patient with hypercortisolemia has primary depression or early Cushing's syndrome may also be difficult. Some patients who eventually have a full-blown picture of Cushing's syndrome begin with only intermittent elevations of Cortisol and symptoms of a major affective disorder.45 Conversely, patients have been reported in whom the physical presentation and hypercortisolemia are suggestive of Cushing's disease, yet the clinical course and response to psychotropic medication favor primary psychiatric illness.14 In fact, there is some evidence for somatic effects of glucocorticoid excess in patients with MDD. Patients who were abnormal responders on the dexamethasone suppression test showed hematologic and blood chemistry findings consistent with the effects of hypercortisolemia.46

Although similar in many respects to the major depressive disorders, the depression seen in Cushing's syndrome does have certain distinguishing clinical characteristics that may aid in differential diagnosis. Summarizing these once more, irritability is a prominent and consistent feature, as are symptoms of autonomic activation such as shaking, palpitations, and sweating. Depressed affect is often intermittent, with episodes of one to three days' duration, recurring very frequently at irregular intervals. Patients usually feel their best, not their worst, in the morning. Psychomotor retardation, although present in many patients, is usually not so pronounced as to be clinically obvious and is usually apparent only in retrospect after improvement with treatment. The majority of these patients are not withdrawn, apathetic, monosyllabic, unspontaneous, or hopeless. Their guilt, when present, is not irrational or self-accusatory, and is primarily related to their realistic inability to function effectively. Significant cognitive impairment, including disorder of memory, is a consistent and prominent clinical feature of patients with Cushing's syndrome.

Are there any procedures currently available that may aid in the differential diagnosis? Sleep EEG studies indicate that there are many similarities between Cushing's disease patients and patients with MDD, which is theoretically interesting but is not diagnostically helpful. Both groups show a significantly longer sleep latency, less total sleep time, and lower sleep efficiency than do normal subjects. In both Cushing's disease and MDD patients, REM latency is significantly shortened and REM density in the first REM period is significantly increased compared with control subjects.47,48

One procedure that holds some promise as a possible tool for helping to distinguish between patients with Cushing's disease and MDD is the administration of CRH to stimulate ACTH secretion.49 At present, however, there is still a substantial overlap in the ACTH secretory responses between the two groups, impairing the test's clinical utility. Further refinements will be required. One already proposed includes taking the pre-CRH plasma Cortisol level as well as the peak ACTH response to CRH into account.50

As research continues to provide greater understanding of HPA axis pathophysiology and its associations with the depressive syndromes, we can anticipate the development of better tools to provide the discriminatory power necessary.


1. Ribeiro SCM, Tandon R, Grunhaus L, Greden JF. The DST as a predictor of outcome in depression: a metaanalysis. AmJ Psychiatry. 1993; 150:1618-1629.

2. Starkman MN, Zelnik TC, Nesse RM, Cameron OG. Anxiety in patients with pheochromocytomas. Arch Intern Med. 1985; 145:248-252.

3. Starkman MN, Cameron OG, Nesse RM, Zelnik T. Peripheral catecholamine levels and the symptoms of anxiety: studies in patients with and without pheochromocytoma. Psychosom Med. 1990; 52:129-142.

4. Cushing H. The basophil adenomas of the pituitary body and their clinical manifestations (pituitary basophilism). Bulletin of the Johns Hopkins Hospital. 1932; 50:137-195.

5. Trethowan WH, Cobb S. Neuropsychiatrie aspects of Cushing's syndrome. Arch Neurol Psychiatry. 1952; 67:283-309.

6. Cohen SI. Cushing's syndrome: a psychiatric study of 29 patients. Br J Psychiatry. 1980; 136:120-124.

7. Starkman MN, Schteingart DE, Schork MA. Depressed mood and other psychiatric manifestations of Cushing's syndrome: relationship to hormone levels. Psychosom Med. 1981; 43:3-18.

8. Haskett RF. Diagnostic categorization of psychiatric disturbance in Cushing's syndrome. Am J Psychiatry. 1985; 142:911-916.

9. Guillemin R, Vargo T, Rossier J, et al. Beta-endorphin and adrenocorticotropin are secreted concomitantly by the pituitary gland. Science. 1977; 30:1367-1369.

10. Gold PW, Chrousos G, Kellner C, et al. Psychiatric implications of basic and clinical studies with corticotropin-releasing factor. Am J Psychiatry. 1984; 141:619-627.

11. Krieger DT. The central nervous system and Cushing's syndrome. Mt Sinai J Med. 1972; 39:416-428.

12. Krieger DT. Cyproheptadine-induced remission of Cushing's disease. N Engl J Med. 1975; 293:893-896.

13. Van Cauter E, Refetoff S. Evidence for two subtypes of Cushing's disease based on the analysis of episodic Cortisol secretion. N Engl J Med. 1985; 312:1343-1349.

14. Krystal A, Krishnan KR, Raitiere M, et al. Differential diagnosis and pathophysiology of Cushing's syndrome and primary affective disorder. J Neuropsychiatry Clin Neurosa. 1990; 2:34-43.

15. Gifford S, Gunderson JG. Cushing's disease as a psychosomatic disorder: a selective review of the clinical and experimental literature and a report of ten cases. Perspect Biol Med. 1970; 13:169-221.

16. Sonino N, Fava GA, Grandi S, Maniero F, Boscaro M. Stressful life events in the pathogenesis of Cushing's syndrome. Clin Endocrinol I Ox ft. 1988; 29:617-623,

17. Starkman MN, Schteingart DE. Neuropsychiatrie manifestations of patients with Cushing's syndrome; relationship to Cortisol and adrenocorticotropic hormone levels. Arch Intern Med. 1981; 141:215-219.

18. Boston Collaborative Drug Surveillance Program. Acute adverse reactions to prednisone in relation to dosage. Clin Pharmacol Ther. 1972; 13:694-698.

19. Hall RC, Popkin MK, Stickney SK, Gardner ER. Presentation of the steroid psychoses. J Nerv Ment Dis. 1979; 167:229-236.

20. Loosen PT, Chambliss B, DeBold CR, Shelton R, Orth DN. Psychiatric phenomenology in Cushing's disease. Pharmacopsychiatry, 1992; 25:192-198.

21. Tucker RP, Weinstein HE, Schteingart DE, Starkman MN. EEG changes and serum Cortisol levels in Cushing's syndrome. Clin Electroencephalogr. 1978; 9:32-37.

22. Whelan TB, Schteingart DE, Starkman MN1 Smith A, Neuropsychological deficits in Cushing's syndrome. J Nerv Ment Dis. 1980; 168:753-757.

23. McEwen BS, Davis PG, Parson B. The brain as a target for steroid hormone action. Annu Rev Neurosci. 1979; 2:65-112.

24. Sarrieau A, Dussaillant M, Agid F, Philibert D, Agid Y, Rostene W. Autoradiographic localization of glucocorticosteroid and progesterone binding sites in the human post-mortem brain. J Steroid Biochem. 1986; 25:717721.

25. Sapolsky RM, McEwen BS. Stress, glucocorticoids, and their role in degenerative changes in the aging hippocampus. In: Crook T, Bartus R, Ferris S, Gershon S, eds. Treatment Development Strategies for Alzheimer's Disease. Madison, Conn: Mark Powley; 1986.

26. Sapolsky RM, Uno H, Reber Finch CE. Hippocampal damage associated with prolonged glucocorticoid exposure in primates. J Neurosci. 1990; 10:2897-2902.

27. Starkman MN, Gebarski SS, Berent S, Schteingart DE. Hippocampal formation volume, memory dysfunction, and Cortisol levels in patients with Cushing's syndrome. Biol Psychiatry. 1992; 32:756-765.

28. DeLeon MJ, George AE, Slylopoulos LA, Smith G1 Miller DC. Early marker for Alzheimer's disease: The atrophic hippocampus. Lancet. 1989; 2(8664):672-673. Letter to the Editor.

29. DeLeon MJ, McRae T, Tsai JR, et al. Abnormal Cortisol response in Alzheimer's disease linked to hippocampal atrophy. Lancet. 1988; 2(86071:391-392.

30. Momose KJ, Kjellberg RN1 Kliman B. High incidence of cortical atrophy of the cerebral and cerebellar hemispheres in Cushing's disease. Radiology. 1971; 99:341348.

31. Starkman MN, Schteingart DE, Schork MA. Cushing's syndrome after treatment: changes in Cortisol and ACTH levels, and amelioration of the depressive syndrome. Psychiatry Res. 1986; 17:177-188.

32. Carroll BJ. Psychiatric disorders and steroids. In: Usdin E, Hamburg DA, Barchas JD, eds. Neuroregulators and Psychiatric Disorders. New York, NY: Oxford University Press; 1972.

33. Jeffcoate WJ, Silverstone JT, Edwards CR, Besser GM. Psychiatric manifestations of Cushing's syndrome: response to lowering of plasma Cortisol. Q J Med. 1979; 48:465-472.

34. Kelly WF1 Checkley SA, Bender DA, Mashiter K, Cushing's syndrome and depression - a prospective study of 26 patients. Br J Psychiatry 1983; 142:16-19.

35. Majewska M, Starkman MN. Actions of steroids on neuron: role in personality, mood, stress and disease. Integrative Psychiatry. 1987; 5:258-273.

36. Reus VI. Pituitary-adrenal disinhibition as the independent variable in the assessment of behavioral symptoms. Biol Psychiatry. 1982; 17:317-326.

37. Bunney WE Jr1 Fawcett JA, Davis JM1 Gifford S. Further evaluation of urinary 17-hydroxycorticosteroids in suicidal patients. Arch Gen Psychiatry. 1969; 21:138-150.

38. Ostroff R, Giller E1 Bonese K, Ebersole E, Harkness L, Mason J. Neuroendocrine risk factors of suicidal behavfor. Am J Psy chiatry. 1982; 139:1323-1325.

39. Gemer RH, Wilkins JN. Cushing's syndrome Cortisol in patients with depression, mania, or anorexia nervosa and in normal subjects. Am J Psyckiatry. 1983; 140:9294.

40. Green AR1 Curzon G. Decrease of 5-hydrcxytiryptamine in the brain provoked by hydrocortisone and its prevention by allopurinol. Nature. 1968; 220:1095-1097.

41. Schubert D, LaCorbiere M, Klier FG1 Steinbach JH. The modulation of neurotransmitter synthesis by steroid hormones and insulin. Brain Res. 1980; 190:67-79.

42. Mobley PL, Suiser F. Adrenal corticoiàs regulate sensitivity of noradrenaline receptor-coupled adenylate cyclase in brain. Nature. 1980; 286:608-609.

43. DeWied D. Pituitary adrenal system hormones and behavior. Acta Endocrinologica. 1977; 85(suppl):9-18.

44. Jacquet YF. Dual actions of morphine on the central nervous system: parallel actions of beta-endorphin and ACTH. Ann N Y Acad Sci. 1982; 398:272-290.

45. Gold PW1 Chrousos G1 Kellner C, et al. Psychiatric implications of basic and clinical studies with corticotropin-releasing factor. Am J Psychiatry. 1984; 141:619-627.

46. Reus VT, Miner C. Evidence for physiologic effects of hyper-cortisolemia in psychiatric patients. Psychiatry Res. 1985; 14,47-55.

47. Shipley JE, Schteingart DE, Tandon R, Starkman MN. Sleep architecture and sleep apnea in patients with Cushing's disease. Sleep. 1992; 15:514-518.

48. Shipley JE, Schteingart DE1 Tandon R, et al. EEG sleep in Cushing's disease and Cushing's syndrome: comparison with patients with major depressive disorder. Biol Psyckiatry. 1992; 32:146-155.

49. Gold PW1 Loriaux DL1 Roy A1 et al. Responses to corticotropin-releasing hormone in the hypercortisolism of depression and Cushing's disease. N Engl J Med. 1986;314:1329-1335.

50. Gold PW, Kling MA1 Calabrese JR1 et al. Corticotropin disease. N Engl J Med. 1987; 316:218-219. Letter.


Frequency of Psychiatric Symptoms in 35 Patients with Cushing's Syndrome


Frequency of Abnormal Findings in the Mental Status Examination in 35 Patients with Cushing's Syndrome


Sign up to receive

Journal E-contents