Psychiatric Annals


Pituitary and Adrenal Changes in Depression

K Ranga Rama Krishnan, MD


Little is known about the mechanisms underlying the HPA abnormalities seen in depression or the relative contribution of each of the different structures involved in the regulation of the HPA axis.


Little is known about the mechanisms underlying the HPA abnormalities seen in depression or the relative contribution of each of the different structures involved in the regulation of the HPA axis.

It is well established that hypothalamicpituitary-adrenal (HPA) axis hyperactivity occurs in patients with major Idepression, especially in those with melancholia or psychotic features.13 Multiple abnormalities of HPA dysfunction in these subtypes of depression include: 1) increased secretion of Cortisol particularly in the evening and the night, 2) increased 24-hour urinary excretion of Cortisol,4,5 and 3) resistance to suppression of Cortisol, β-endorphin, and adrenocorticotrophin (ACTH) by dexamethasone, i.e., depressed patients tend to show an early escape of ACTH, β-endorphin Cortisol secretion from the usual 24-hour suppression of these hormones by dexamethasone,1,6,7 and 4) hypersecretion of corticotrophin-releasing factor (CRF) as evidenced by increased concentrations of CRF in cerebrospinal fluid.

A number of studies have suggested a potential role for hypercortisolemia in the pathophysiology of cognitive and neuroanatomical changes seen in depression.8*10 However, little is known about the mechanisms underlying the HPA abnormalities seen in depression or the relative contribution of each of the different structures involved in the regulation of the HPA axis (adrenal cortex, anterior pituitary, hypothalamus, and limbic structures such as the hippocampus). In order to understand the mechanisms, each of the different components of the HPA axis has to be studied intensively.

Cells in a portion of the hypothalamus, called the paraventricular nucleus, produce a hormone called corticotrophin-releasing hormone (CRF). This hormone acts on specific cells in the pituitary gland, called the corticotroph cells, to secrete adrenocorticotrophic hormone (ACTH). This hormone acts on the adrenal cortex and causes cells in the cortex to secrete the hormone, Cortisol. The hypothalamus is regulated by structures in the iimbic system, especially the hippocampus. Cortisol secreted by the adrenal cortex inhibits the HPA axis at the level of the pituitary corticotroph, hypothalamus, and the hippocampus. This feedback regulation is believed to be abnormal in depression.


A number of studies using high doses of ACTH have demonstrated an increased adrenocortical response to ACTH in depression.11-13 This increased response could be due to either a hypertrophy (increased size of the adrenocortical cells) or hyperplasia (increased number of adrenocortical cells) of the adrenal cortex, or to an increased sensitivity of the ACTH receptors in the adrenal cortex.

When ACTH is given chronically to animals, the cells in the adrenal cortex are known to increase in size and number. This is reflected as an increase in the size of the adrenal gland. In a preliminary study, we have shown that there is no difference in the adrenocortical response to low-dose ACTH between depressed patients and control subjects, suggesting that adrenocortical sensitivity to ACTH is not altered in depression.14 This suggests that there is probably adrenocortical enlargement in depression.

Zis and Zis15 reported that postmortem adrenal gland weights are increased in suicide victims when compared to controls. With the advent of recently developed high-resolution computed tomography equipment, it is possible to visualize the adrenal gland with sufficient clarity to assess whether the adrenal gland is indeed enlarged in depression. An initial study by Amsterdam et al11 using computed tomography reported that the adrenal gland is bigger in depressed patients.

We undertook a more extensive study using high-resolution computed tomography to test the hypothesis that the adrenal gland is enlarged in depression.16 The examination consisted of contiguous 5-mm sections through the adrenal gland without intravenous contrast while the patient is breath-holding. A GE 9800 scanner was used. The size of the adrenal gland was rated as either enlarged or normal by two radiologists (Dr. Reed Dunnick and Dr. Richard Leder) blind to the diagnosis of the subjects. There was complete concordance between the raters. The volume of the adrenal gland was estimated by a stereological method.

A total of 38 depressed subjects were studied. The patients met criteria for major depression. Severity of depression was assessed by the Montgomery-Asberg Depression Rating Scale. Subjects were excluded if any of the subjects had factors known to alter HPA function. In addition, 26 patients underwent a standard dexamethasone suppression test (DST). The DST was conducted as follows: 1 mg of dexamethasone was given orally at 9:00 PM and blood was drawn for plasma Cortisol the following day at 3:00 PM and 10:00 PM.

Twelve of the 38 subjects had an enlarged gland. None of the controls had an enlarged adrenal gland. The volume of the adrenal gland in patients was significantly larger than in controls. Of the 26 depressed patients who had the DST, 18 were nonsuppressors as defined by plasma Cortisol concentrations greater than 5 µ-g/dl at either 3:00 PM or 10:00 PM. There was no correlation between maximal plasma Cortisol concentration and adrenal gland volume, measured before dexamethasone suppression. Adrenal gland volume was larger in men than in women. Age was not significantly related to adrenal volume. Adrenal volumes were higher in depressed patients than in control subjects after covarying for age and sex. Adrenal gland volume was not related to the severity of depression nor to duration of the episode.16

The demonstration of enlarged adrenal glands in depression confirms previous studies. Enlargement of the adrenal gland could reflect either adrenocortical hypertrophy or adreno-medullary enlargement. However, adrenal gland enlargement is more likely to reflect adrenocortical hypertrophy because:

* The adrenal medulla constitutes less than 10% of the adrenal gland. To account for the increase, the medulla would have to increase several times in size.

* The adrenal medullary cells that are derived from the neural crest are not known to increase in size or multiply.

* Adrenocortical cells can both multiply and increase in size.

Enlargement of the adrenal gland in depression probably reflects chronic and persistent activation of the HPA axis. The lack of a significant relationship between postdexamethasone Cortisol and adrenal size is not surprising because postdexamethasone Cortisol concentrations and adrenal size probably do not reflect the same time domains. The adrenal gland size probably reflects chronic activation of the HPA axis whereas postdexamethasone Cortisol concentrations are likely to reflect the immediate changes in the HPA axis. The finding of an enlarged adrenal gland in depression may have important implications for diagnosis and management in depression. Future studies are needed to assess the development of adrenal hypertrophy and the relationship between adrenal size and clinical features such as recovery and relapse.


A number of studies have demonstrated an increase in ACTH secretion by the corticotroph in patients with depression. The putative causes for this increase in ACTH secretion are believed to be corticotrophin-releasing factor and vasopressin. Preliminary studies by Holsboer et al17 indicate that depressed patients have higher ACTH and Cortisol responses to corticotrophin-releasing factor (CRF) after dexamethasone compared to normal subjects. Lisansky et al18 and von Bardeleben et al19 have also shown that the ACTH response to CRF after metyrapone administration is increased in depressed patients compared to normal subjects. (Metyrapone is used to inhibit Cortisol synthesis and therefore eliminate feedback.) These studies raise the possibility that either the corticotroph is sensitized to CRF or that there is hypertrophy and hyperplasia of the corticotroph cells of the pituitary.

Animal studies indicate that the pituitary corticotroph can exhibit both hypertrophy and hyperplasia following chronic CRF administration.20,21 Thus there is a strong possibility that the pituitary gland will be increased in size in depression, reflecting the increased secretion of CRF and activity of the HPA axis. We decided to examine the size of the pituitary gland using magnetic resonance imaging (MRI). Prior to the advent of MRI, it was not possible to easily assess the size of the pituitary gland. Pituitary volume can be estimated using sagittal and/or coronal images through the pituitary gland.


We estimated the size of the pituitary gland in a group of 19 depressed patients and 19 controls. The pituitary gland was assessed using a GE 1.5 Tesla Signa MRI system.22 An axial localizing sequence was first obtained and a midsagittal section was then specified. T1 weighted midsagittal and parasagittal sections were obtained on either side of the midsagittal section. Coronal sections were graphically prescribed from the midsagittal section such that at least one passed through the coronary stalk. To ensure consistency of location, the coronal planes were obtained at right angles to the anterior commissure to posterior commissure line. Pituitary height, pituitary length, and cross-sectional area of the pituitary gland were obtained from the midsagittal slice using programs on the GE Signa on an off-line console. Pituitary width was assessed on the coronal slice that passed through the pituitary stalk. Two raters measured these variables independently in all subjects.

The pituitary optic chiasma distance was also measured. The volume of the pituitary was estimated as area X width, and as 0.5 X height X length x width. This was based on studies conducted by Gonzalez et al.23 Depressed patients (43 ± 10 mmp 2) had a larger midsagittal area than controls (32 ±9 mmp 2) (¿=-3.62, p<0.0009). Pituitary volume was also increased in patients (577.5 ± 167 mmp 3) compared to controls (408.4 ±172 mmp 3) (t = 2.85, p<0.007).

This study clearly indicates that the pituitary gland is enlarged in patients with depression. Width, length, and height were all increased in depressed patients. The pituitary enlargement is probably anterior rather than posterior because:

* Posterior pituitary constitutes less than 20% of pituitary and has to more than double in size to account for the increase seen in depression.

* The posterior pituitary is of neural origin and there are no known conditions of posterior pituitary enlargement (other than nonendocrine tumors).


In a subsequent study we examined the relationship between the size of the pituitary gland and the severity of nonsuppression with the DST.24 Twenty-four patients were studied; there were no controls. Pituitary volume was estimated from 3-mm sagittal slices using a stereological method. A standard DST was also obtained. One mg of dexamethasone was administered at 11:00 PM. The next day blood was drawn for measurement of plasma Cortisol at 3:00 PM and 10:00 PM. Both age and sex were related to the size of the pituitary gland.

A significant positive relationship between plasma Cortisol at 3:00 PM and pituitary size was found after covarying for age and sex (f=3.32, p<0.04). Similarly, postdexamethasone Cortisol at 10:00 PM was related to pituitary size after covarying for age and sex (/"=6.27, p<0.02). The highest postdexamethasone Cortisol concentrations were also related to pituitary volume (f=3.24, p<0.04), after covarying for age and sex.

It must be noted that pituitary enlargement alone does not mean increased corticotroph function. The anterior pituitary has cells that secrete other trophic hormones. Changes in pituitary size may reflect changes in other pituitary hormone function. For example, in patients with bulimia and anorexia the pituitary is smaller and related to cessation of menses (altered gonadotrophs function).

The presence of pituitary enlargement and adrenocortical enlargement in depression raises several questions. Many investigators have described a putative role for Cortisol in the etiopathogenesis of depression. A major problem in confirming an etiological role for corticosteroid changes with the development of symptoms in depression has been the lack of studies investigating HPA dysfunction in temporal sequence to the development of affective symptoms.

Early studies did suggest that persistent HPA axis dysfunction even after clinical recovery (as manifested by nonsuppression on the DST) predicted earlier relapse.25 More recent studies by Kathol26 have noted that depressed patients have persistent elevation in 24-hour urinary free Cortisol secretion even after recovery and this may represent a trait for the illness. Further, they showed that, in addition to the residual changes, depressed patients had intermittent changes in the HPA function that may precede and/or be related to the development of depressive symptomatology. This suggests that hyperresponsive adrenal cortex or pituitary corticotroph may render a subject more vulnerable to recurrence of depression. The time it takes for the adrenal cortex and the pituitary gland to return to normal may be related to the period of vulnerability for the recurrence of depression. This may have potential clinical implications in terms of duration of maintenance treatment for depression.


It has been suggested that the hippocampus may be involved not only in the pathophysiology of hypercortisolemia but also in the pathophysiology of depression. Sapolsky27 has shown that the hippocampus is critically involved in the feedback regulation of Cortisol. He has shown that, in animals, prolonged secretion of Cortisol secondary to stress can lead to loss of hippocampal neurons. This leads to a loss of feedback inhibition and a further increase in secretion of Cortisol. Thus a feed-forward cascade develops. It is hypothesized that there will be loss of hippocampal neurons in depressed patients who hypersecrete Cortisol.

We recently studied the size of the amygdalo-hippocampal complex (AHC)18 in 19 patients (mean age 46.7 ±20.4 years) meeting DSM-in criteria for major depression and 30 controls (mean age 56.6 ±19.1 years) using MRI. Severity of depression as assessed by the Montgomery-Asbury Depression Rating Scale was 234. Mean age of onset of depression was 34 ±17 years. Coronal T1 weighted images were obtained at right angles to the AC-PC line. The amygdalo-hippocampal (AHC) volume was estimated from these scans. The volume of the AHC was assessed on each side separately. A stereological method was used. The volume of the AHC was similar to that reported in the literature. A standard DST was obtained as described above. There was no difference in AHC volume between depressed patients and controls. AHC volume was not related to postdexamethasone Cortisol concentrations. AHC volume is not sufficiently sensitive. It is possible that a more sensitive assessment of hippocampal volume might demonstrate subtle changes in hippocampal size in depression. At this time, this study should not be taken as a definitive rejection of Sapolsky's27 hypothesis.

MRI studies have demonstrated structural changes in the neuroendocrine system in depression and have helped clarify the pathophysiology underlying HPA axis hyperactivity in depression. Future studies using a combination of neuroendocrine and MRI assessments may help us understand the role of endocrine changes in the pathophysiology of depressive symptoms and relapse in patients with affective disorder.


1. Carroll BJ. The dexamethasone suppression test for melancholia. Br J Psychiatry. 1982; 140:292-304.

2. Sachar EJ, Hellman L, Fukushima DK, Gailager RF. Cortisol production in depressive illness. Arch Gen Psychiatry. 1970; 23:289-298.

3. Linkowski P, Mendlewicz J, LeClerq R, et al. The 24-hour profile of ACTH and Cortisol in major depressive illness. J Clin Endocrinol Metab. 1985; 61:429-438.

4. Carroll BJ, Curtis GC, Davies BM. Mendels J, Sugarman AA. Urinary free Cortisol excretion in depression. Psychol Med. 1976; 6:43-50.

5. Sachar E, Hellman L, Roffwarg H, Halpern F, Fukushima D, Gallagher D. Disrupted 24-hour patterns of Cortisol secretion in psychotic depression. Arch Gen Psychiatry. 1973; 28:19-24.

6. Reus Vl1 Joseph MS. Dallman MF, ACTH levels after the dexamethasone suppression test in depression. N Engl J Med. 1982; 306:228-239.

7. Sherman B, Pfohl B, Winokur G. Circadian analysis of plasma Cortisol levels before and after dexamethasone administration in depressed patients. ArcA Gen Psyckiatry. 1984; 41:271-275.

8. Kellner CH, Rubinow DR, Gold PW, Post RM. Relationship of Cortisol hypersecretion to bTain CT scan alterations. Psychiatry Res. 1983; 8:191-197.

9. Rao VJ1 Krishnan KRR, GoIi V, et al, Neuroanatomical changes and hypothalamo-pituitary-adrenal axis abnormalities. Biol Psychiatry. 1989; 26:729-732.

10. Rubinow DR, Post RM, Savard R, Gold PW. Cortisol hypersecretion and cognitive impairment in depression. Arch Gen Psychiatry. 1984; 41:179-283.

11. Amsterdam JD, Winokur A, Abelman E, Lucki I, Rickels K. Cosyntropin (ACTH 1-24) stimulation test in depressed patients and healthy subjects. Am J Psychiatry. 1983; 140:907-909.

12. Jaeckle RS, Kathol RG, Lopez JF, Meiler WE, Krummel SJ. Enhanced adrenal sensitivity to exogenous cosyntropin (ACTH 1-24) stimulation in major depression. Arch Gen Psychiatry. 1987; 44:233-240.

13. Kalin NH, Dawson G, Tariot P, et al. Function of the adrenal cortex in patients with major depression. Psychiatry Res. 1987; 22:1 17-125.

14. Krishnan KRR, Ritchie JC, Saunders WB, Nemeroff CB, Carroll BJ. Adrenocortical sensitivity to low dose ACTH administration in depressed patients. Biol Psychiatry. 1990; 27:930-933.

15. Zis KD, Zis A. Increased adrenal weight in victims of violent suicide. Am J Psychiatry. 1987; 144:1214-1215.

16. Nemeroff CB, Krishnan KRR, Reed D, Leder R, Beam C, Dunnick NR. Adrenal gland enlargement in major depression: a computed tomographic study. Arch Gen Psychiatry. 1992; 49:384-387.

17. Holsboer F, von Bardeleben U, Weidemann K, Muller OA, Stalla GK Serial assessment of corticotropinreleasing hormone after dexamethasone in depression: implication for pathophysiology of DST nonsuppression. Biol Psychiatry. 1987; 22:228-234.

18. Lisansky J, Peake GT, Strassman RJ, et al. Augmented pituitary corticotropin response to a threshold dosage of human corticotropin-releasing hormone in depressives pretreated with metyrapone. Arch Gen Psychiatry. 1989; 46:641-643.

19. von Bardeleben U, Stalla GK, Muller OA, Holsboer F. Blunting of ACTH response to human CRF in depressed patients is avoided by metyrapone treatment. Biol Psychiatry. 1988; 24:782-785.

20. Westlund KN, Aguilera G, Childs GV. Quantification of morphological changes in pituitary corticotrophs produced by in vivo corticotropin-releasing factor stimulation and adrenalectomy, Endocrinology. 1985; 116:439-445.

21. Gertz BJ, Contreras LN, McComb DJ, Kovacs K1 Tyrell B, Dallman MF. Chronic administration of corticotropin-releasing factor increases pituitary corticotroph number. Endocrinology. 1987; 120:381-388.

22. Krishnan KRR, Doraiswamy PM, Lurie SN, et al. Pituitary size in depression. J Clin Endocrinol Metab. 1991; 72:256-259.

23. Gonzalez JG, Elizando G, Salvidor D. Pituitary gland growth during normal pregnancy: an in vivo study using magnetic resonance imaging. Am J Med. 1988; 85:217-220.

24. Axelson DA, Doraiswamy PM, Boyko OB, et al. In vivo assessment of pituitary volume with magnetic resonance imaging and systematic stereology: relationship to dexamethasone suppression test results in patients. Psychiatry Res. 1992;44:63-70.

25. Coryell W, Zimmerman M. The dexamethasone suppression test and ECT outcome. Biol Psychiatry. 1983; 18:21-27.

26. Kathol RG. Aetiologic implications of corticosteroid changes in affective disorder. Psychiatr Med. 1985; 3:135-162.

27. Sapolsky P, Krey L, McEwen B. The neuroendocrinology of stress and aging: the glucocorticoid cascade hypothesis. EndocrRev. 1986; 7:284-301.


Sign up to receive

Journal E-contents