This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hochberg, Z.'e. Articles by Chrousos, G. P. Search for Related Content PubMed PubMed Citation Articles by Hochberg, Z.'e. Articles by Chrousos, G. P.

Endocrine Withdrawal Syndromes

Division of Endocrinology (Z.H.), Meyer Children’s Hospital, Haifa 31096, Israel; and Pediatric and Reproductive Endocrinology Branch (Z.H., K.P., G.P.C.), National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892

Correspondence: Address requests for reprints to: George P. Chrousos, M.D., Pediatric and Reproductive Endocrinology Branch, National Institute of Child Health and Human Development, National Institutes of Health, Building 10, Room 9D42, 10 Center Drive, MSC 1583, Bethesda, Maryland 20892-1583. E-mail: Chrousog{at}mail.nih.gov

	Abstract

Top Abstract I. Introduction II. Glucocorticoids III. Estrogens and Progestins IV. Androgens V. GH VI. Conclusions References

 
Hypersecretion of endogenous hormones or chronic administration of high doses of the same hormones induces varying degrees of tolerance and dependence. Elimination of hormone hypersecretion or discontinuation of hormone therapy may result in a mixed picture of two syndromes: a typical hormone deficiency syndrome and a generic withdrawal syndrome. Thus, hormones with completely different physiological effects may produce similar withdrawal syndromes, with symptoms and signs reminiscent of those observed with drugs of abuse, suggesting shared mechanisms. This review postulates a unified endocrine withdrawal syndrome, with changes in the hypothalamic-pituitary-adrenal axis and the central opioid peptide, in which noradrenergic and dopaminergic systems of the brain act as common links in its pathogenesis. Long-term adaptations to hormones may involve relatively persistent changes in molecular switches, including common intracellular signaling systems, from membrane receptors to transcription factors. The goals of therapy are to ease withdrawal symptoms and to expedite weaning of the patient from the hormonal excess state. Clinicians should resort to the fundamentals of tapering hormones down over time, even in the case of abrupt removal of a hormone-producing tumor. In addition, the prevention of stress and concurrent administration of antidepressants may ameliorate symptoms and signs of an endocrine withdrawal syndrome.

I. Introduction

II. Glucocorticoids

A. Withdrawal syndrome after discontinuation of glucocorticoid therapy

B. Withdrawal syndrome after correction of hypercortisolism in Cushing’s syndrome

C. Possible mechanisms of the glucocorticoid withdrawal syndrome

D. Therapeutic approaches to glucocorticoid withdrawal

III. Estrogens and Progestins

A. Postpartum as a withdrawal syndrome

B. Menopause as a withdrawal syndrome

C. Withdrawal syndrome after interruption of hormone replacement therapy

D. Premenstrual syndrome as a withdrawal phenomenon

E. Possible mechanisms of the estrogen withdrawal syndromes

F. Therapeutic approaches to estrogen withdrawal

IV. Androgens

A. Withdrawal syndrome after discontinuation of replacement therapy

B. Withdrawal syndrome in athletes abusing androgens

C. Withdrawal from physiological androgen levels

D. Possible mechanisms of the androgen withdrawal syndrome

E. Therapeutic approaches to androgen withdrawal

V. GH

A. Withdrawal syndrome after discontinuation of GH therapy

B. Withdrawal syndrome after correction of hypersomatotropism in acromegaly

C. Possible mechanisms of the GH withdrawal syndrome

D. Therapeutic approaches to GH withdrawal

VI. Conclusions

A. Possible pathways for a unified endocrine withdrawal syndrome

B. Therapeutic approaches to endocrine withdrawal

	I. Introduction

Top Abstract I. Introduction II. Glucocorticoids III. Estrogens and Progestins IV. Androgens V. GH VI. Conclusions References

 
DEPENDENCE, OFTEN ASSOCIATED with drugs of abuse, is a biological phenomenon with both psychological and physiological components. During the period of addictive substance use, the body adjusts to a new level of pathological homeostasis, or allostasis. When the drug is abruptly discontinued, this equilibrium is disturbed and the organism reacts both behaviorally and physiologically with a constellation of manifestations collectively called a withdrawal syndrome. Dependence is preceded by a phase of tolerance, which signifies a progressively decreased response to the effect of a drug, necessitating ever-larger doses to achieve the same effect. Tolerance is largely due to compensatory responses of the organism, which attempt to mitigate the drug’s pharmacological actions. Tolerance may result from the functional adjustments of target tissue signal transduction systems and/or from metabolic adjustment associated with increased catabolism and disposition of the drug of abuse taken chronically. The term "addiction" implies both psychological and physiological dependence, with clear adverse behavioral and social consequences, and has been used mainly with regard to drugs of abuse. The severity of a withdrawal syndrome depends on the genetics and developmental history of a patient, on his/her environment, and on the phase the patient has reached (Fig. 1).

View larger version (22K): [in this window] [in a new window]   FIG. 1. Phases in the development of hormonal withdrawal syndromes. Similar to the phases after the consumption of drugs of abuse, chronic hormonal excess production or administration may lead to tolerance for the respective hormone. This may be followed by dependence and, rarely, addiction. Correction of the chronic hormone excess syndrome at each of these phases may lead to withdrawal syndromes. The severity of the withdrawal syndrome depends on the phase and degree of dependence.

Endocrine withdrawal syndromes have often been misinterpreted as symptoms and signs of specific hormone deficiency, after removal of an endocrine gland or after discontinuation of hormonal therapy. However, critical examination of the symptoms and signs during this period frequently shows a mixed picture of two different syndromes that can be differentiated into a typical hormonal deficiency syndrome and a generic withdrawal syndrome. Thus, the mania, hypomania, and depression attributed to the withdrawal of anabolic steroid abuse in athletes has little to do with the symptoms and signs of testicular failure, because these symptoms occur in some patients with hypogonadism. Similarly, the syndrome that follows discontinuation of high-dose glucocorticoids or correction of hypercortisolism in Cushing’s syndrome bears similar but not identical symptoms and signs of adrenal insufficiency.

Interestingly, hormones with completely different physiological effects can produce similar withdrawal syndromes, whereas some of the clinical manifestations that are due to the chronic presence of high hormone levels or withdrawal syndromes are also observed with drugs of abuse. This review postulates that changes of the hypothalamic-pituitary-adrenal (HPA) axis and the central opioid peptide, noradrenergic and dopaminergic systems act as shared features in the pathogenesis of several endocrine withdrawal syndromes. The molecular and cellular bases of endocrine or drug-related addiction and withdrawal syndromes, however, are poorly understood. The best established molecular and cellular mechanisms of short- and long-term adaptation to hormones or allostasis induced by drugs of abuse is potentiation of G protein receptor coupling, up-regulation of cAMP, increased activities of protein kinases, and changes in the expression and activities of several transcription factors.

The withdrawal syndrome after glucocorticoid discontinuation has drawn a great deal of attention over the years and has been studied extensively. We suggest that the symptoms and signs that occur after cessation of administration of several other hormones may also comprise a similar withdrawal syndrome. These other hormones have not been studied as thoroughly as the glucocorticoids, and the following discussion refers primarily to animal studies and clinical inferences from such studies. Whereas some of the observations and suggestions are scientifically evident and valid, based on solid data in both animals and humans, other annotations are tentative or speculative, lacking evidence from well-designed clinical studies. Further clinical research needs to take place before these suggestions are implemented.

	II. Glucocorticoids

Top Abstract I. Introduction II. Glucocorticoids III. Estrogens and Progestins IV. Androgens V. GH VI. Conclusions References

 
A. Withdrawal syndrome after discontinuation of glucocorticoid therapy
Glucocorticoids are widely used in clinical practice to control the activity of autoimmune, inflammatory, and allergic diseases, neoplasms of the hematopoietic system, and other nosological entities. High therapeutic doses of glucocorticoids, when used to control the activity of these diseases, suppress the HPA axis and exert numerous central nervous system (CNS) effects, including anxiety, insomnia, impairment of cognition, and mood changes ranging from euphoria to hypomania, mania, depression, and psychosis (1). Some of these symptoms have been attributed to the suppression of hypothalamic CRH and proopiomelanocortin (POMC)-derived peptides such as ß-endorphin, stimulation of the amygdala, and initial stimulation followed by tolerance and inhibition of the dopaminergic mesocorticolimbic system.

Four aspects of drug withdrawal after cessation of pharmacological high-dose glucocorticoid therapy deserve special attention (2). First, the illness treated by steroids may relapse. Second, the HPA axis and POMC-derived peptide secretion may remain suppressed for a long time. Third, a nonspecific withdrawal syndrome may develop even while patients are receiving physiological replacement doses of glucocorticoids. Fourth, psychological dependence to these hormones often develops. Thus, after abrupt discontinuation of glucocorticoid therapy, patients may develop anorexia, nausea, emesis and weight loss, fatigue, myalgias, arthralgias and headache, abdominal pain, lethargy and postural hypotension, fever, and skin desquamation (Table 1). Interestingly, the syndrome may occur during weaning from pharmacological high-dose therapy, while the patient is on adequate glucocorticoid replacement. This may also happen after the response of the HPA axis to stimuli has returned to normal (3, 4), indicating that long-term tolerance to glucocorticoids has developed and hormone substitution is inadequate to allow the central nervous system or other organs to function properly.

C. Possible mechanisms of the glucocorticoid withdrawal syndrome
The withdrawal syndrome, which patients experience after either discontinuation of glucocorticoid therapy or correction of hypercortisolism in Cushing’s syndrome, has been considered a withdrawal reaction due to established physical dependence on supraphysiological glucocorticoid levels. Several mediators may be considered, and include CRH, vasopressin, POMC, central noradrenergic and dopaminergic systems, cytokines, and prostaglandins (Table 1).

Chronic hypercortisolemia decreases CRH mRNA expression in the rat hypothalamus while increasing it in the central nucleus of the amygdala (9). Acute withdrawal after chronic glucocorticoid administration decreases rat hypothalamic CRH for a further 7 d or more (10), and CRH neurons are the last part of the HPA axis to normalize. This is supported by data on humans, showing that patients with Cushing’s disease show markedly decreased CRH levels in cerebrospinal fluid, suggesting that some central CRH neurons secreting into or spilling CRH into the cerebrospinal fluid have been restrained by long-standing hypercortisolism (11).

CRH in the brain not only activates the HPA axis but also mediates stress-related behavioral effects. Intracerebroventricular administration of high doses of CRH in the rat or monkey enhances fear-related behaviors, decreases exploration, and inhibits sleep, feeding, and sexual activity (12). Similarly, the origin of hypercortisolism in human melancholia has been suggested to have a central origin, predominantly reflecting hypersecretion of central CRH (12, 13). Glucocorticoid-induced hyperactivity of CRH neurons in the amygdala induces arousal, fear response, and anxiety (14). Moreover, an intact CRH system in the brain seems necessary for adequate mesolimbic dopaminergic function. Thus, central CRH hyposecretion in the period of acute glucocorticoid withdrawal may further contribute to anxiety and depression via inadequate stimulation of dopaminergic neurons terminating in the nucleus accumbens.

With regard to suppressed CRH, some investigators feel that hyposecretion of central CRH plays an important role in the pathogenesis of atypical depression. Thus, abrupt glucocorticoid withdrawal in patients with Cushing’s syndrome is associated with profound psychopathology (6), which may be attributed to long-standing hypoactivity of central CRH neurons (6, 11, 12, 13). Because several recent lines of evidence in man and experimental animals suggest that fatigue, hypersomnia, lethargy, and hyperphagia are associated with hyposecretion of hypothalamic CRH (5, 12, 15), symptoms and signs of acute glucocorticoid withdrawal may reflect hypoactivity of central CRH neurons.

Elevated plasma vasopressin levels, accompanied by improved water excretion, are observed in patients with adrenal insufficiency and can be normalized by glucocorticoid replacement (16, 17). In contrast, chronic glucocorticoid administration in man increases water diuresis due to direct renal tubular effect, but also due to central inhibition of vasopressin release, and escape from this inhibition is defined by vasopressin secretion (18). Clinical findings of increased frequency of urination in patients with Cushing’s syndrome and recent experiments showing glucocorticoid-induced suppression of vasopressin expression in human hypothalamic neurons further support these conclusions (19). The number of vasopressin-immunoreactive neurons in the parvocellular paraventricular nucleus of patients with depression is significantly increased (20). Some clinical studies have shown improvement in short- and long-term memory processes, mood, and concentration of depressed patients after vasopressin administration (21, 22). In contrast, oxytocin was found to impair memory performance (23, 24). Behavioral and other changes may occur via a direct effect of vasopressin on central neuronal processes or via potentiation of the central CRH effects. The activation of oxytocin neurons mediates CRH-induced depression in experimental animals (25). One may, therefore, speculate that in humans, glucocorticoid withdrawal symptoms, such as anorexia, nausea, decreased motivation, anxiety, and depression, may also reflect chronic glucocorticoid-induced disturbances of vasopressin and oxytocin neurons.

Substantial clinical and animal evidence supports the view that hypercortisolism can produce depression. In particular, about two thirds of patients with hypercortisolism due to Cushing’s syndrome are clinically depressed, and correction of their hypercortisolism, such as by surgery or medical therapy, usually ameliorates the depression (6, 12, 13, 26, 27). Conversely, a substantial proportion of patients with major depression have abnormalities of the HPA axis, such as an increased apparent frequency of ACTH secretory episodes, elevated urinary free cortisol excretion, elevated CRH levels in cerebrospinal fluid, and relative insensitivity of the HPA axis to the suppressive effect of dexamethasone (12, 27, 28). Recently, the central noradrenergic and dopaminergic systems were considered to play important roles in the pathogenesis of cortisol-induced mood disorders (29, 30, 31, 32, 33, 34).

Evidence from preclinical investigations indicates disturbances of mesocortical and mesolimbic dopaminergic function in anxiety, anhedonia and, depression (31, 32, 35, 36, 37, 38, 39). Genetically selected Flinders Sensitive Line rats that exhibit behavioral features characteristic of anxiety and depression have markedly decreased dopamine release in the nucleus accumbens (40). Dopamine depletion in the nucleus accumbens and caudate nucleus occurs in rats with learned helplessness, and treatment with dopamine agonists prevents development of this syndrome in this animal model of depression (41, 42). This is supported by human studies, showing that enhanced dopamine release in the nucleus accumbens correlates with reward-related activities, increased psychomotor activation, and decreased anxiety (43, 44). Increases in dopamine release after acute glucocorticoid treatment have been demonstrated in the nucleus accumbens (33), and this could be associated with the euphoric (hypomanic) states seen acutely with high doses of steroids. Blunted responses of dopamine to palatable food have been proposed to serve as a marker of anhedonia (45).

Acute stress-induced dopamine release in the human mesolimbic system associated with marked activation of the HPA axis promotes behavioral activation and results in defensive responses toward the stressful stimulus (30). In contrast, prolonged exposure to stress leads to inhibition of the mesolimbic dopaminergic system, associated with coping failure and cessation of defensive attempts, with subsequent development of depressive signs (30). Similarly, chronic hypercortisolemia inhibits dopaminergic activity in the nucleus accumbens but not in the prefrontal cortex (46, 47). Thus, region-specific glucocorticoid-induced alterations in dopaminergic activity may be an important factor underlying pathophysiological mechanisms of depressive symptoms in chronic hypercortisolemia.

In clinical studies, patients with endogenous depression have decreased levels of dopamine metabolites in cerebrospinal fluid and low values for indices of brain dopamine turnover (32, 48), consistent with an association between depression and inhibition of central dopaminergic systems. After clinical improvement, brain dopamine turnover increases (32). The magnitude of increase in dopamine-2 receptor binding in the striatum and anterior cingulate gyrus, assessed by single-photon emission tomography, correlates with clinical recovery from depression (49, 50). Drugs such as reserpine, methyl dihydroxyphenylalanine (DOPA), and neuroleptic agents that block dopamine receptors can cause depression. This may occur via effects on central dopaminergic systems (32). In contrast, dopamine receptor agonists, such as bromocriptine, pergolide, and roxindole, have antidepressant efficacy similar to tricyclic antidepressants (29, 51, 52). Dopamine antagonists block euphoria induced by amphetamine, which promotes release of dopamine and inhibits its uptake (32).

The interplay between central noradrenergic systems and glucocorticoids in the pathogenesis of mood changes, especially depression and withdrawal syndrome, remains unclear. Both increased and decreased central noradrenergic activity has been described in depression (for reviews, see Refs. 53 and 54). Similarly, antidepressant responses have been linked to increased (53, 55) as well as decreased (54, 56, 57) central noradrenergic function.

As discussed earlier in this review, based on results from animal and human studies, there is enough evidence that, in major melancholic depression, there is marked activation of the central CRH systems that leads to persistent activation of central noradrenergic systems, especially in the locus ceruleus (LC) (58, 59, 60). In contrast, in animals exposed to chronic hypercortisolemia, the activity of the central noradrenergic systems (e.g., those originating in the brainstem or the LC and terminating in the hypothalamic paraventricular nucleus), as well as the CRH system in the paraventricular nucleus, are inhibited, whereas the activity of the CRH system in the amygdala is increased (47, 61, 62). In contrast, adrenalectomy increases norepinephrine (NE) release and turnover in different brain regions, as well as in periphery, and cortisol replacement blunts these changes (63, 64). Furthermore, CRH-knockout mice show decreased adrenomedullary activity and increased sympathetic nervous activity (65).

Chronic hypercortisolemia in patients with Cushing’s syndrome or exogenous administration of glucocorticoids in normal volunteers inhibits peripheral sympathoadrenal activity (66, 67, 68). Furthermore, functional imaging using [¹⁸F]fluorodeoxyglucose as a positron-emitting agent shows that patients with Cushing’s syndrome have decreased cerebral uptake rates for glucose in all brain regions except the striatum, whereas patients with major depression have increased activity in the amygdala and the ventral prefrontal cortex and decreased activity in the dorsal prefrontal cortex, regions that have abundant noradrenergic innervation (69). Lack of proper glucocorticoid replacement in the postoperative period in patients with Cushing’s syndrome is associated with increased incidence in panic behavior (6) together with increased sympathoadrenal activity, which returns to near-normal preoperative levels after appropriate glucocorticoid replacement. These data suggest that acute glucocorticoid withdrawal characterized by anxiety and panic symptomatology in patients with Cushing’s syndrome may be partly due to lack of glucocorticoid restraint characterized by markedly increased central noradrenergic activity. When the functions of CRH and noradrenergic neurons recover, symptoms of anxiety and panic subside.

Thus, whereas patients with major depression can be described as hyperadrenergic and hypercortisolemic, patients with depression due to Cushing’s syndrome can be described as hypoadrenergic and hypercortisolemic and patients after glucocorticoid withdrawal as hypernoradrenergic and hypocortisolemic.

ACTH is synthesized as part of the 241-amino-acid precursor POMC, which is also cleaved to generate the NH₂-terminal peptide, the joining peptide, lipotropin, melanocyte stimulating hormone, and ß-endorphin. CRH and vasopressin stimulate serum levels of all of these peptides (70), and glucocorticoids suppress pituitary and hypothalamic POMC expression (71). Hence, some of the symptoms of Cushing’s syndrome may be related to deficiency of these peptides. Furthermore, recovery of the HPA axis is associated with parallel recovery of POMC-derived peptide secretion (72). It is not surprising, therefore, that some of the CNS symptoms of glucocorticoid dependence and withdrawal are related to those of opiate withdrawal. Brain neurons, neurotransmitters, and their receptors, as well as peripheral signal transduction machinery, adapt to POMC deficiency in the course of Cushing’s syndrome.

During the acute phase of glucocorticoid withdrawal and the flu-like syndrome that characterizes it, plasma IL-6 levels rise markedly and TNF and IL-1ß levels increase. Exogenous administration of IL-6 induces similar manifestations, and a role for this cytokine in the pathogenesis of the flu-like syndrome in adrenal insufficiency was suggested (5). This reaction might also involve the suppression by glucocorticoids of prostanoid and platelet activating factor production, and a sudden increase in their production upon withdrawal of steroid hormones. Indeed, prostaglandins E2 and I2 also induce many of the features of the flu-like syndrome (73).

Interestingly, a dose-dependent increase in plasma cortisol levels is found in habitual smokers after smoking two cigarettes, and complete cessation of smoking is followed by a fall in plasma cortisol levels that is associated with the withdrawal of the nicotine stimulus (74). Many of the nicotine withdrawal symptoms of smokers who try to quit seem to be related to the body’s response to changes in CRH and/or cortisol levels, along with the downstream mesocorticolimbic and POMC-peptide effects. As part of a comprehensive smoking cessation program, one or two im injections of ACTH gel have been shown to help smokers stop and continue to abstain from smoking (74).

D. Therapeutic approaches to glucocorticoid withdrawal
Gradual tapering of high-dose glucocorticoid therapy has become the standard of practice. However, the glucocorticoid withdrawal syndrome, which develops after correction of endogenous hypercortisolism, is largely ignored or considered as a separate entity. In attempting to minimize postoperative withdrawal symptoms and signs, the clinician is faced with two options: the first is to normalize cortisol secretion before surgery, employing medical suppression of steroidogenesis. This has to be done gradually, or else withdrawal symptoms might ensue. The second option is to reinstitute high-dose glucocorticoid replacement therapy after surgery and taper it off gradually. It sounds reasonable, although untested, to resume a dose that would result in pretreatment urinary free cortisol levels as the basis for tapering off. The disadvantage of this therapy is the very likely prolongation of Cushing’s symptoms and signs, as well as adrenal suppression. These preventive options await well-designed clinical studies. With a decrease in CRH and the central dopaminergic and POMC-peptide systems, the rationale is there for correcting these derangements gradually in cases of severe withdrawal syndrome.

	III. Estrogens and Progestins

Top Abstract I. Introduction II. Glucocorticoids III. Estrogens and Progestins IV. Androgens V. GH VI. Conclusions References

 
Estrogens are potent stimuli to the HPA axis and the LC/NE system. Postpartum, menopause, and the premenstrual syndrome are all associated with decreasing estrogen and withdrawal syndrome-like manifestations (75, 76) (Table 2). These may include hot flushes and autonomic hyperactivity, but also fatigue, irritability, anxiety and depression, and even psychosis. Withdrawal symptoms and signs do not resemble those of estrogen hormonal deficiency, as they manifest in young women with Turner syndrome or hypogonadotropic hypogonadism.

The withdrawal syndrome that follows labor and delivery has much more to it than a reaction to the birth process. Bloch et al. (78) investigated the possible role of changes in gonadal steroid levels in postpartum blues, or depression, by simulating two hormonal conditions related to pregnancy and parturition in euthymic women. The supraphysiological gonadal steroid levels of pregnancy and withdrawal from these high levels to a hypogonadal state were simulated in women with or without a history of postpartum depression. Hypogonadism was induced with a GnRH agonist, adding back supraphysiological doses of estradiol and progesterone for 8 wk, and then withdrawing both steroids under double-blind conditions. Five of eight women with a history of postpartum depression (62.5%), but none of eight women in the comparison group, developed significant mood symptoms during the withdrawal period. These investigations suggested that women with a history of postpartum depression are differentially sensitive to mood-destabilizing effects of gonadal steroids. Mild depression, or "the blues," occurs in as many as 60% of women, whereas an additional 13% develop full-fledged depression (79). When women with histories of puerperal psychosis or major depression were treated with high-dose oral estrogen immediately after delivery, they remained healthy and required no treatment with psychotropic medications during the 1-yr follow-up period as compared with an expected 35–60% recurrence rate (80).

Changes in levels of endorphins may be involved in the pathophysiology of psychiatric postpartum estrogen withdrawal syndrome (81). Studies of various endorphins indicate a possible relation between levels of endorphins and depressive symptoms. In addition, some studies employing naloxone suggested a relation between a blockade in the action of endorphins and the development of a syndrome of dysphoric symptoms similar to the depressive features manifested premenstrually and in the postpartum. Estrogen and endorphin levels have been shown to covary. During the postpartum and the premenstrual periods, levels of both change rapidly and substantially.

B. Menopause as a withdrawal syndrome
Women who suffer hot flushes soon after menopause have lower estrogen levels than those who do not have hot flushes (82), indicating that this may be a symptom of estrogen deficiency or withdrawal, and, indeed, it can be ameliorated by estrogen replacement therapy. Several lines of evidence support a possible withdrawal syndrome. For example, some postmenopausal symptoms are self-limited and distinct from those of pure estrogen deficiency. The more severe withdrawal symptoms of autonomic hyperactivity, hot flushes, and hyperemic coronary flow occur when menopause is instituted abruptly by surgery or antiestrogens, such as clomiphene citrate or tamoxifen (83, 84), but not in congenital forms of hypogonadism and less so in slowly progressing premature ovarian failure, when the signs and symptoms of estrogen deficiency include amenorrhea, endometrial and breast atrophy, and osteoporosis.

Another important symptom of menopause is climacteric depression; however, no convincing evidence has linked depression with menopause yet (85). Its duration and characteristics are reminiscent of those seen after correction of hypercortisolism in Cushing’s syndrome, and we propose that it reflects an estrogen and CRH withdrawal syndrome. In both cases, recovery with hormonal replacement therapy is not instantaneously beneficial.

C. Withdrawal syndrome after interruption of hormone replacement therapy
Hormone replacement therapy has been widely prescribed only in the past two decades, and the indications for treatment and the risk/benefit ratio are still disputed. Estrogens are psychoactive: they cause mood changes, and their use has powerful psychological effects. Reports of women with supraphysiological estradiol concentrations may represent tolerance and withdrawal (85). Menopausal-like symptoms sometimes evolve despite high serum levels of estradiol, perhaps as tolerance to estrogen. Hormone replacement therapy is often withdrawn abruptly for a variety of clinical indications, but little attention is given to symptoms and signs of withdrawal. In addition to physiological dependence, hormone replacement therapy promotes feelings of well-being, which may contribute to psychological dependence (85). A dramatic case report described a 51-yr-old woman with no previous psychiatric history who amputated her hand in a "psychotiform" state after discontinuation of her contraceptive medication. The patient stabilized under a combined therapy with estrogen-progestin substitution (86).

D. Premenstrual syndrome as a withdrawal phenomenon
On a smaller scale, each menstrual cycle is terminated with a miniature withdrawal syndrome during the late luteal phase. Premenstrual irritability, fatigue, and mood changes are common. The syndrome, called also late luteal or premenstrual dysphoric disorder (PMDD), is associated with the cyclically recurring increase and decrease in levels of estrogen and progesterone. The key characteristics of PMDD, with clear onset and offset of symptoms closely linked to the menstrual cycle and the prominence of symptoms of anger, irritability, and internal tension, were contrasted with those of known mood and anxiety disorders (87). PMDD displays a distinct clinical picture that, in the absence of treatment, is remarkably stable from cycle to cycle and over time. Normal or near-normal functioning of the HPA axis, biological characteristics generally related to the serotonin system, and a genetic component unrelated to major depression are further features of PMDD that separate it from other mood disorders (87, 88). These symptoms are often abrogated after eliminating cyclic hormone levels by administering estrogen and progesterone daily (89). Women who suffer a premenstrual withdrawal syndrome are more likely to develop late pregnancy depression and anxiety, as well as postpartum blues and depression (90). This would suggest a common mechanism for the various facets of estrogen and progesterone withdrawal syndromes.

The concept of PMDD as a withdrawal syndrome was recently challenged (91). These authors suggested that the link between the decline in hormonal levels and the symptoms of PMDD was much more like that of a menstrual zeitgeber or synchronizer than a hormonal withdrawal state, because extending or truncating the luteal phase did not immediately affect the course of symptoms. The concept of Schmidt et al. (91), however, is compatible with the idea that women with a history of postpartum depression are differentially sensitive to mood-destabilizing effects of gonadal steroids.

E. Possible mechanisms of the estrogen withdrawal syndromes
Sex hormones have long been known to exert powerful effects on brain functions including mood, behavior, and neuroendocrine regulation (92, 93). Postpartum estrogen withdrawal occurs in the first week after parturition when there is a substantial fall in plasma estrogen levels along with many a decrease in other hormones (75, 94). Numerous clinical studies support the notion that cyclic patterns of epileptic seizures and mood disturbances may also result from rapidly changing sex hormone levels, as seen during specific phases of the menstrual cycle (95, 96, 97).

Several mechanisms may explain the pathogenesis of estrogen withdrawal syndrome. From numerous animal studies, it has been postulated that some of the antidepressant effects of estrogens and progesterone reflect their action on central CRH, opioid peptidergic, catecholaminergic, and serotoninergic neurons (75, 98, 99, 100). Ovarian steroids also modulate the activity of the central dopaminergic system in rats (101). A case of postpartum psychosis with abnormal movements was shown to result from dopamine hypersensitivity that was unmasked by withdrawal of endogenous estrogens (102).

In terms of dopamine synthesis, chronic administration of estrogen inhibits tyrosine hydroxylase in various dopaminergic brain areas of hypophysectomized rats (99). The inhibitory effect of estradiol on dopamine synthesis is consistent with other observations that chronic treatment with estrogens causes a decrease in the limbic content of dopamine in rats (101, 102, 103). Mesolimbic dopaminergic activity, as characterized by dopamine release and reuptake, fluctuates with alterations in endogenous gonadal steroid levels. In rat proestrus, nucleus accumbens dopaminergic activity is significantly decreased (104, 105). Estrogens, on the other hand, significantly increase the density of 5-hydroxytryptamine 2A binding sites in the rat anterior frontal, cingulate and primary olfactory cortex, and in the nucleus accumbens, all of which are areas of the brain involved in the control of emotion, cognition, and behavior (93). Furthermore, elevated estrogen levels at the estrous phase of the rat cycle significantly decrease extracellular hypothalamic serotonin levels (14). Whether similar mechanisms operate in humans remains to be determined. However, the response of patients with PMDD to selective serotonin reuptake inhibitors (87) suggests a role for serotonin in its pathophysiological mechanism (Table 2).

Following earlier evidence that hot flushes are triggered within the hypothalamus by ₂-adrenergic receptors on noradrenergic neurons (106), it has been shown that clonidine, an ₂-agonist, is sometimes effective in reducing autonomic hyperfunction during menopause (107).

POMC-related peptides are increased within several minutes of subjective menopausal flushes (108), but, on the other hand, sex steroids are able to increase ß-endorphin and ß- lipotropin secretion in postmenopausal women, with a concomitant relief of climacteric symptoms (109). Another alternative common pathway for various steroid withdrawal syndromes is vasopressin. Within 24 h of parturition, rodent vasopressin mRNA levels increase (110). Should human experiments support a similar effect, exposure and then withdrawal of estrogen and testosterone could mimic this increase.

Abrupt discontinuation of progesterone in the rat, or its decline after chronic administration or pregnancy termination, increases anxiety and sensitivity for convulsions (111, 112). This is supported by human data, as both anxiety and seizures are part of the premenstrual and postpartum syndromes, for which the role of progesterone withdrawal has been repeatedly discussed (111, 112, 113). This might be mediated by progesterone’s -aminobutyric acid (GABA)_A-modulator metabolite 3-OH-5-pregnan-20-one, which is abruptly decreased after progesterone withdrawal in the rat (111, 112, 113, 114). Manipulation of GABA_A receptor by altered progesterone levels may also affect cross-tolerance with sedative drugs whose abrupt withdrawal may also cause premenstrual-like symptoms (112).

From the studies described above, it can be concluded that acutely altered central opioid peptide, catecholamine, and serotonin levels due to ovarian steroid manipulations might be an important contribution to the symptoms of ovarian steroid withdrawal syndromes.

F. Therapeutic approaches to estrogen withdrawal
The currently recommended treatment for postpartum and climacteric depression is antidepressants. Also, in PMDD, at least 60% of patients respond to selective serotonin reuptake inhibitors (87). Recognizing the possible mechanism of hormone withdrawal may lead to a different rationale and approach. Lower-dose hormone replacement therapy could diminish dependence. Postpartum administration of high-dose estrogen and tapering off minimize the psychiatric estrogen withdrawal syndrome (80). The autonomic syndrome of hot flushes in postmenopause responds partly to hormone replacement therapy and also to androgens, thus suggesting a common mechanism with the androgen withdrawal syndrome. The latter may in fact be mediated by aromatization of androgen to estrogen. Initiation of hormone replacement therapy immediately after oophorectomy may prevent the withdrawal syndrome that would otherwise follow surgery. An alternative therapeutic approach would be to use drugs with CNS effects, such as some of the selective antidepressants, to modify the neurotransmitter abnormalities postulated to be associated with the withdrawal symptoms.

	IV. Androgens

Top Abstract I. Introduction II. Glucocorticoids III. Estrogens and Progestins IV. Androgens V. GH VI. Conclusions References

 
A. Withdrawal syndrome after discontinuation of replacement therapy
Traditionally, replacement therapy in patients with testicular failure of any etiology utilizes long-acting preparations, the most common of which is testosterone enanthate. The pharmacokinetics of a dose given every 3–4 wk is such that it provides supraphysiological serum levels over the initial week and subphysiological levels on wk 3–4 (115). A few days before each injection, emotional lability and mood swings may occur (116, 117). This cyclic withdrawal syndrome is easily prevented by administration of lower doses of testosterone enanthate every 1 or 2 wk or continuous administration with skin patches or gels, which eliminates a major fluctuation in testosterone serum levels (118, 119). In this respect, the androgen withdrawal syndrome and its therapy are analogous to the premenstrual syndrome and its management.

B. Withdrawal syndrome in athletes abusing androgens
The abuse of anabolic steroids by athletes and body builders is, for obvious reasons, poorly documented (120). Generally, doses of abused steroids may be up to 100 times greater than therapeutic replacement doses. Both psychological and physical dependence occur, and withdrawal symptoms are seen frequently. The analogy of androgen withdrawal syndrome with that of drugs of abuse was made over a decade ago (121). Taken in large doses, these compounds have severe psychological and behavioral side effects, including aggressive and violent behavior (122). Problems with drug withdrawal and dependence may result in decreased sexual drive, but also in a flu-like syndrome that mimics in many ways the glucocorticoid withdrawal syndrome (123). Fatigue, muscle and joint pain, headache, and insomnia are followed by a second phase of depression (124), a condition that appears to be more common than previously realized (125). As many as 23% of users reported major mood syndromes: mania, hypomania, and depression probably occur as a result of the episodic nature of use and discontinuation of anabolic steroids (117). Androgen withdrawal is often associated with the desire to resume steroid consumption ("craving") (124). These symptoms and signs are obviously unrelated to specific symptoms and signs of androgen deficiency, as they manifest in patients with hyper- or hypogonadotrophic hypogonadism.

C. Withdrawal from physiological androgen levels
Orchiectomy or GnRH analog therapy, such as that given to patients with prostate cancer, often result in hot flushes that resemble the postmenopausal syndrome (126). The symptoms are alleviated by either androgen or estrogen therapy (127), suggesting a role for androgen-derived estrogen.

D. Possible mechanisms of the androgen withdrawal syndrome
Depending on their structure, endogenous androgens and androgenic compounds may be reduced to dihydrotestosterone or aromatized to estrogen in target tissues, including the human brain (128). During androgen abuse, estrogen levels in target tissues may approach those of ovulating females, with gonadotropins suppressed via a negative feedback mechanism by both androgen and estrogen effects at the hypothalamic-pituitary level (129). Symptoms and signs of androgen overdose and withdrawal may be partly accounted for by the relative conversion to these derivatives. In the rat, a further metabolic pathway has been suggested: the conversion of testosterone to 3-androstanediol has been shown to mediate brain effects of androgen, decreasing GABA-stimulated chloride influx in cortical synapto-neurosomes and muscimol binding in the hippocampus (130).

Effects of androgenic steroids on central aminergic systems (Table 3) may explain some of the withdrawal symptoms. In rats, an increase of noradrenergic activity in certain brain areas has been observed immediately after castration (131, 132), whereas testosterone administration decreases hypothalamic NE turnover (133). Bilateral implants of testosterone or dihydrotestosterone in the preoptic area (suprachiasmatic nucleus) of orchidectomized rats decreased dopamine turnover without influencing NE turnover (134). An acute hypernoradrenergic state could at least partially explain some of the anabolic-androgenic steroid withdrawal symptoms, but relevant clinical studies are not available. In humans, elevated plasma levels of NE correlate positively with insomnia, anorexia, and depressed mood, whereas in rats, abnormalities in the monoamines, along with their cotransmitters, may cause many forms of eating disorders (135, 136). In contrast, castrated animals have increased basal and amphetamine-stimulated dopamine release in the mesolimbic dopaminergic system (137). In humans, we can only hypothesize that acute anabolic-androgenic steroid withdrawal may be associated with decreased central dopaminergic activity, reflected by frequent occurrence of depressive symptoms.

Male hot flushes may reflect an effect of aromatized androgen deficiency with a similar mechanism to those of estrogen withdrawal syndromes. The severity and frequency of sweating was reduced with naloxone, as can be seen in female climacteric flushing. Injection of testosterone significantly reduced the frequency and severity of the attacks, which, however, were unexpectedly unaltered by estrogen treatment (141, 142).

E. Therapeutic approaches to androgen withdrawal
The anabolic steroid withdrawal syndrome is preventable by refraining from any abuse of androgens. Yet, the incentive for their use seems to overpower reason, and certain athletes and body builders continue steroid abuse despite the adverse consequences of these agents. The acute flu-like syndrome has been ameliorated by administration of clonidine (123), tranquilizers, and analgesics. The antidepressant fluoxetine has also been effective in treating the androgen withdrawal syndrome (125). A more rational approach would be substitution therapy and tapering of the dose, or use of chorionic gonadotropin treatment. This should be limited to individuals who can be certain of unequivocal termination of drug abuse. This is done under the understanding that withdrawal from androgen overdose is associated with hypogonadotropic hypogonadism (129), which warrants replacement therapy until recovery of the hypothalamic-pituitary-gonadal axis.

	V. GH

Top Abstract I. Introduction II. Glucocorticoids III. Estrogens and Progestins IV. Androgens V. GH VI. Conclusions References

 
Some of the generalizations about a common withdrawal syndrome apply weakly to GH. Yet, several lines of evidence support tolerance to GH and a withdrawal syndrome.

A. Withdrawal syndrome after discontinuation of GH therapy
When treatment of patients with GH deficiency is discontinued in growing children, these patients manifest the symptoms and signs of their original disease, primarily growth deceleration, within a short period of time. Children with idiopathic short stature (143), intrauterine growth retardation (144, 145, 146), or Noonan’s syndrome (147), with no underlying hormone deficiency, also develop deceleration of growth after discontinuation of GH therapy (Table 4). On the other hand, a single study of patients with Russell-Silver syndrome did not observe such deceleration of growth (148). The most obvious presentation of withdrawal symptoms after exogenous GH therapy in non-GH-deficient children is "catch-down growth," i.e., deceleration of growth despite normal GH and IGF-I levels (143), indicating tolerance to GH. Apparently, the withdrawal deceleration of growth is time dependent and more pronounced after prolonged GH treatment. The dose appears to be of no major consequence, whereas the daily schedule of injections might be important for the development of tolerance (149). The withdrawal complex also includes a decline in resting cardiac output (143), an increase in fat mass, a decrease in metabolic rate, and negative balances of nitrogen, phosphorus, sodium, and potassium, reported after withdrawal of GH in children with GH deficiency (150). A mild shortening of night sleep has been reported after withdrawal of GH therapy in children with idiopathic short stature (151).

Doping with GH has become an increasing problem in sports during the last 10 yr. GH has a reputation of being fairly effective among abusers, although the few controlled studies that have been performed with administration of supraphysiological GH doses to athletes have shown no significant positive effects of GH from the point of view of a doping agent (152). No study on GH withdrawal among abusers has been reported so far.

B. Withdrawal syndrome after correction of hypersomatotropism in acromegaly
We are uncertain whether patients with acromegaly develop tolerance to GH. But, similarly to findings observed with withdrawing GH therapy, body composition and metabolism are affected after removal of a GH-producing adenoma. Within 2 wk of surgery, patients lose weight markedly, due to a decrease in body water and cell mass, and recover gradually over the next month (153). We speculate, although it remains unproven, that such patients develop a withdrawal syndrome with negative nitrogen and electrolyte balance, as observed in children during the GH withdrawal syndrome.

C. Possible mechanisms of the GH withdrawal syndrome
The mechanism through which chronic exposure to either exogenous or endogenous GH leads to tolerance, dependence, and a withdrawal syndrome is unclear and does not involve detectable suppression of hormone secretion, as with other endocrine withdrawal syndromes (Table 4). During the nadir of growth velocity and after prolonged drug therapy of children with no GH deficiency, serum GH levels are normal, as are serum IGF-I and IGF-binding protein-3 levels (143). GH therapy for as long as 12 months does not interfere with the endogenous pulsatile secretion of GH (154). If it exists, the mechanism of tolerance to GH may lie in the peripheral target tissues. Chondroprogenitor cells in the growth plate may wear off in the face of continuous high-level GH, in the absence of the unique pulsatile pattern of serum GH (Table 4). Subcutaneous administration of daily GH results in an unphysiological serum GH profile, with peak levels at 4 h and a slow decline over the course of 12–24 h. This pattern can be regarded as continuous administration of GH, rather then the physiological pulses, which occur with a frequency of about eight per day. As previously observed in short-term studies, alternate day therapy, which in a non-GH deficient child would allow for normal GH pulsatility in the interval day, resulted in zero or minimal catch-down of growth (155). Moreover, GH doses commonly used therapeutically often stimulate IGF-I to supraphysiological serum levels. The mechanism seems, therefore, to rest with diminished GH and IGF-I action at the target cells of the growing bone and other tissues and organs.

D. Therapeutic approaches to GH withdrawal
Although untested, two approaches come to mind to prevent GH withdrawal symptoms and signs. The first is to taper off GH therapy gradually, as is the usual approach after high-dose glucocorticoid therapy. The second is to prevent GH dependence. A possible mode may be to prevent daily exposure to GH, by administering it on alternate days (79). Whereas the growth-promoting effect of such a therapy is expected to be smaller, the prevention of a catch-down growth may turn out to pay off in the long run.

To minimize postoperative withdrawal symptoms and signs in acromegaly, one could attempt to gradually normalize GH secretion before surgery by octreotide. This approach has been shown by some, but not all, to also minimize surgical and postsurgical mortality (156, 157).

	VI. Conclusions

Top Abstract I. Introduction II. Glucocorticoids III. Estrogens and Progestins IV. Androgens V. GH VI. Conclusions References

 
A. Possible pathways for a unified endocrine withdrawal syndrome
The symptoms and signs of endocrine withdrawal syndromes are different from those of the respective endocrine deficiency syndromes, and, thus, the mechanisms appear distinct. Several of the endocrine withdrawal syndromes share common symptoms and signs and, therefore, may share common pathophysiological mechanisms. In others, common pathways are activated in opposite directions. Based on the mechanisms proposed for each of these hormones, we now offer a hypothetical unified concept that applies to hormone withdrawal (see Fig. 3). Fear and anxiety occur due to marked decreases in levels of steroid hormones such as glucocorticoids, estrogens, androgens, and possibly thyroid hormones, whereas the withdrawal syndrome that follows decreases in levels of GH emerges with a different pattern. Euphoria has been reported in glucocorticoid, estrogen, androgen, GH, and thyroid hormone overdosing. Indeed, these hormones interrelate in many ways. To mention a few examples, glucocorticoids affect GH secretion and GH receptor expression, and GH affects cortisol production and metabolism (158). Sex steroids have numerous effects on glucocorticoids and the lactogenic hormones. Also, some commonalties could be due to the psychological aspects of addictive behavior.

View larger version (24K): [in this window] [in a new window]   FIG. 3. Interrelations between the stress system [CRH/arginine vasopressin (AVP) neurons and LC-NE neurons] with the POMC-related opioid peptide system, mesocorticolimbic dopaminergic system (MCLS), and the amygdala. Acutely, glucocorticoids, estrogens, androgens, thyroid hormones, and GH may activate the MCLS and POMC-related opioid peptide system. Tolerance and dependence to these actions may develop though. Upon withdrawal of the hormone, the POMC-related peptide system and MCLS are suppressed, producing dysphoria and anxiety; the latter via amygdala activation (, stimulation; ----, inhibition).

View larger version (41K): [in this window] [in a new window]   FIG. 2. Withdrawal syndromes of several hormones share common symptoms and signs (A) and mechanisms (B). Many of these symptoms and signs are also manifestations of withdrawal syndromes after discontinuation of drugs of abuse. Endogenous opioid peptides as well as the central dopaminergic systems may play a central role in the pathogenesis of several of these syndromes.

The molecular and cellular bases of endocrine or drug-related addiction and withdrawal syndromes are only partly understood. Table 5 summarizes the mechanisms and the time courses of acute and chronic hormone action, tolerance, dependence, and short-term and long-lasting withdrawal (162). Relatively short-term dependence and addiction to drugs of abuse result from adaptations in specific target cells caused by prolonged exposure to a supraphysiological level of a hormone or a drug of abuse. These adaptive mechanisms may explain the apparent contradiction between suppressed POMC in Cushing’s disease and increased POMC in climacteric hot flushes. The best established mechanism of adaptation is up-regulation of cAMP pathway. Acute opiate exposure inhibits neuronal cAMP pathway, whereas chronic exposure leads to a compensatory cAMP up-regulation (162). Adaptation in membrane receptors, membrane receptor-G protein coupling, protein kinase A activity, and other components of this signal transduction pathway has also been shown to be involved in the case of opiates and cocaine (162, 166).

Figure 2 displays the overlaps in symptoms and signs of withdrawal syndromes as well as the postulated common pathways. It is obvious that withdrawal from glucocorticoids, estrogen, and androgen share many of the symptoms and signs with clinical manifestations of withdrawal from drugs of abuse. We suggest, therefore, that changes of the opioid peptide systems and the mesolimbic dopaminergic system act as a link in the pathogenesis of all four withdrawal syndromes. Central aminergic receptors, CRH, and vasopressin also stand out as common pathways for the withdrawal syndromes of all three steroid hormones (Fig. 3).

B. Therapeutic approaches to endocrine withdrawal
The goals of therapy are to ease withdrawal symptoms, as well as to expedite weaning of the patient from the hormonal excess or abstinence from the hormone or drug of abuse. Should one accept the concept of endocrine withdrawal syndromes, there is no reason to repeat the historic trials and errors from drug-abuse treatment. In endocrine withdrawal syndromes, one can use a more rational approach than abrupt cessation of hormonal effect. Until we gain a better understanding of the molecular and cellular mechanisms of dependence and withdrawal, clinicians must resort to the fundamentals of tapering hormones down over time, as we have been doing with glucocorticoids for many years. In the case of abrupt removal of a hormone-producing tumor, a similar strategy must be developed that will allow for gradually diminishing hormonal levels. This can be accomplished before surgery or attempted thereafter. To prove the efficacy of such new strategies, the endocrine community should conduct controlled studies. The analogy made to drugs of abuse and the possibly shared mechanisms suggest that prevention of stress may ameliorate withdrawal symptoms and that antidepressants may be a helpful aid in moderate stress (166). Withdrawal syndromes that develop after natural decreases of physiological hormonal levels may be easier to treat. Gonadal hormonal replacement therapy can be started shortly before gonadectomy to prevent endocrine withdrawal syndromes.

	Footnotes

 
Abbreviations: CNS, Central nervous system; GABA, -aminobutyric acid; HPA, hypothalamic-pituitary-adrenal; LC, locus ceruleus; NE, norepinephrine; PMDD, premenstrual dysphoric disorder; POMC, proopiomelanocortin.

	References

Top Abstract I. Introduction II. Glucocorticoids III. Estrogens and Progestins IV. Androgens V. GH VI. Conclusions References

 

1. Wolkowitz OM 1994 Prospective controlled studies of the behavioral and biological effects of exogenous corticosteroids. Psychoneuroendocrinology 19:233–255[CrossRef][Medline]
2. Dixon RB, Christy NP 1980 On the various forms of corticosteroid withdrawal syndrome. Am J Med 68:224–230[CrossRef][Medline]
3. Amatruda TTJ, Hollingsworth DR, D’Esopo ND 1960 A study of the mechanism of the steroid withdrawal syndrome: evidence of integrity of the hypothalamic-pituitary-adrenal axis. J Clin Endocrinol Metab 20:339
4. Amatruda Jr TT, Hurst MM, D’Esopo ND 1965 Certain endocrine and metabolic facets of the steroid withdrawal syndrome. J Clin Endocrinol Metab 25:1207–217[Medline]
5. Papanicolaou DA, Tsigos C, Oldfield EH, Chrousos GP 1996 Acute glucocorticoid deficiency is associated with plasma elevations of interleukin-6: does the latter participate in the symptomatology of the steroid withdrawal syndrome and adrenal insufficiency? J Clin Endocrinol Metab 81:2303–2306[Abstract]
6. Dorn LD, Burgess ES, Friedman TC, Dubbert B, Gold PW, Chrousos GP 1997 The longitudinal course of psychopathology in Cushing’s syndrome after correction of hypercortisolism. J Clin Endocrinol Metab 82:912–919[Abstract/Free Full Text]
7. Suzuki K, Nonaka K, Ichihara K, Fukumoto Y, Miyazaki A, Yamada Y, Itoh Y, Seino Y, Moriwaki K, Tarui S 1986 Hypercalcemia in glucocorticoid withdrawal. Endocrinol Jpn 33:203–209[Medline]
8. Suzuki K, Nonaka K, Ichihara K, Fukumoto Y, Seino Y, Moriwaki K, Tarui S 1986 Hyperphosphatemia as a detectable laboratory manifestation of glucocorticoid withdrawal syndrome. Bone Miner 1:51–58[Medline]
9. Makino S, Gold PW, Schulkin J 1994 Effects of corticosterone on CRH mRNA and content in the bed nucleus of the stria terminalis; comparison with the effects in the central nucleus of the amygdala and the paraventricular nucleus of the hypothalamus. Brain Res 657:141–149[CrossRef][Medline]
10. Harbuz MS, Nicholson SA, Gillham B, Lightman SL 1990 Stress responsiveness of hypothalamic corticotrophin-releasing factor and pituitary pro-opiomelanocortin mRNAs following high-dose glucocorticoid treatment and withdrawal in the rat. J Endocrinol 127:407–415[Abstract]
11. Kling MA, Roy A, Doran AR, Calabrese JR, Rubinow DR, Whitfield Jr HJ, May C, Post RM, Chrousos GP, Gold PW 1991 Cerebrospinal fluid immunoreactive corticotropin-releasing hormone and adrenocorticotropin secretion in Cushing’s disease and major depression: potential clinical implications. J Clin Endocrinol Metab 72:260–271[Abstract]
12. Gold PW, Chrousos GP 1998 The endocrinology of melancholic and atypical depression: relation to neurocircuitry and somatic consequences. Proc Assoc Am Phys 111:22–34
13. Dorn LD, Burgess ES, Dubbert B, Simpson SE, Friedman T, Kling M, Gold PW, Chrousos GP 1995 Psychopathology in patients with endogenous Cushing’s syndrome: ’atypical’ or melancholic features. Clin Endocrinol (Oxf) 43:433–442[Medline]
14. Shepard JD, Barron KW, Myers DA 2000 Corticosterone delivery to the amygdala increases corticotropin-releasing factor mRNA in the central amygdaloid nucleus and anxiety-like behavior. Brain Res 861:288–295[CrossRef][Medline]
15. Opp M, Obal Jr F, Krueger JM 1989 Corticotropin-releasing factor attenuates interleukin 1-induced sleep and fever in rabbits. Am J Physiol 257(3 Pt 2):R528–R535
16. Raff H 1987 Glucocorticoid inhibition of neurohypophysial vasopressin secretion. Am J Physiol 252(4 Pt 2):R635–R644
17. Oelkers W 1989 Hyponatremia and inappropriate secretion of vasopressin (antidiuretic hormone) in patients with hypopituitarism. N Engl J Med 321:492–496[Abstract]
18. Dingman JF, Despointes RH 1960 Adrenal steroid inhibition of vasopressin release from the neurohypophysis of normal subjects and patients with Addison’s disease. J Clin Invest 39:1851–1863
19. Erkut ZA, Pool C, Swaab DF 1998 Glucocorticoids suppress corticotropin-releasing hormone and vasopressin expression in human hypothalamic neurons. J Clin Endocrinol Metab 83:2066–2073[Abstract/Free Full Text]
20. Purba JS, Hoogendijk WJ, Hofman MA, Swaab DF 1996 Increased number of vasopressin- and oxytocin-expressing neurons in the paraventricular nucleus of the hypothalamus in depression. Arch Gen Psychiatry 53:137–143[Abstract]
21. Mattes JA, Pettinati HM, Stephens S, Robin SE, Willis KW 1990 A placebo-controlled evaluation of vasopressin for ECT-induced memory impairment. Biol Psychiatry 27:289–303[CrossRef][Medline]
22. Van Londen L, Goekoop JG, Zwinderman AH, Lanser JB, Wiegant VM, De Wied D 1998 Neuropsychological performance and plasma cortisol, arginine vasopressin and oxytocin in patients with major depression. Psychol Med 28:275–284[CrossRef][Medline]
23. Fehm-Wolfsdorf G, Born J, Voigt KH, Fehm HL 1984 Human memory and neurohypophyseal hormones: opposite effects of vasopressin and oxytocin. Psychoneuroendocrinology 9:285–292[CrossRef][Medline]
24. Van Ree JM, Hijman R, Jolles J, De Wied D 1985 Vasopressin and related peptides: animal and human studies. Prog Neuropsychopharmacol Biol Psychiatry 9:551–559[CrossRef][Medline]
25. Olson BR, Drutarosky MD, Stricker EM, Verbalis JG 1991 Brain oxytocin receptors mediate corticotropin-releasing hormone- induced anorexia. Am J Physiol 260(2 Pt 2):R448–R452
26. Starkman MN, Schteingart DE, Schork AA 1981 Depressed mood and other psychiatric manifestations of Cushing’s syndrome: relationship to hormone levels. Psychosom Med 43:3–18[Abstract/Free Full Text]
27. Chrousos GP, Gold PW 1992 The concepts of stress and stress system disorders. Overview of physical and behavioral homeostasis. JAMA 267:1244–1252[Abstract]
28. Holsboer F, Barden N 1996 Antidepressants and hypothalamic-pituitary-adrenocortical regulation. Endocr Rev 17:187–205[CrossRef][Medline]
29. Berridge KC, Robinson TE 1998 What is the role of dopamine in reward: hedonic impact, reward learning, or incentive salience? Brain Res Brain Res Rev 28:309–369[CrossRef][Medline]
30. Cabib S, Puglisi-Allegra S 1996 Stress, depression and the mesolimbic dopamine system. Psychopharmacology (Berl) 128:331–342