0021-972X/90/7002-0311$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright© 1990 by The Endocrine Society
Vol. 70, No. 2 Printed in U.S.A.
Abnormal Cortisol Secretion and Responses to Corticotropin-Releasing Hormone in Women with Hypothalamic Amenorrhea* BEVERLY M. K. BILLER, HOWARD J. FEDEROFFf, JAMES I. KOENIG, AND ANNE KLIBANSKI Neuroendocrine Unit, Departments of Medicine and Neurology, Massachusetts General Hospital, Boston, Massachusetts 02114
± 90 nmol/L, respectively). ACTH responses to CRH did not differ between HA patients and normal women. The 24-h mean cortisol was significantly higher (P = 0.006) in the HA patients than in the normal controls (280 ± 50 and 220 ± 50 nmol/L, respectively), due to higher cortisol levels at night. The urinary free cortisol level was significantly higher (P = 0.005) in the HA patients (230 ± 70 nmol/day) than in normal women (150 ± 40 nmol/day). We conclude that women with HA have a blunted cortisol response to CRH administration. In addition, they have hypercortisolism, as demonstrated by elevated 24-h mean serum cortisol levels and urinary free cortisol values. This hypothalamicpituitary-adrenal axis activation in patients with stress or weight loss may be a mechanism in the development of amenorrhea and may relate to other potential adverse effects of HA. (J Clin Endocrinol Metab 70: 311, 1990)
ABSTRACT. Hypothalamic amenorrhea (HA) is a common disorder associated with hypoestrogenemia and has adverse effects. The mechanism of GnRH deficiency in these women is not yet known. To investigate the role of the hypothalamic pituitary-adrenal axis in HA, we studied 10 women [mean age, 29 ± 7 (±SD) yr] with 0.5-13 yr of amenorrhea (mean, 4.3 ± 3.7 yr) related to simple weight loss or psychological stress. We investigated cortisol and ACTH responses to a bolus of ovine CRH, 24-h plasma cortisol levels obtained every 10 min, and urinary free cortisol levels in these patients. Results were compared with those obtained in normal women during all phases of the menstrual cycle. We found that mean basal concentrations of cortisol were significantly higher (P = 0.03) in the HA patients (mean, 210 ± 130 nmol/L) than in the normal women (100 ± 30 nmol/L). The A (peak - basal) cortisol was significantly lower (P = 0.004) in the HA patients than in the normal women (320 ± 100 vs. 440
H
YPOTHALAMIC amenorrhea (HA) is a functional disorder of presumed GnRH deficiency associated with profound estrogen deficiency and adverse effects, including loss of bone density (1). Prevalence rates vary widely, with reports as high as 29% in nurses during their first year of training; however, in most studies approximately 15% of secondary amenorrhea cases are due to HA (2-5). It is well recognized that both physical and psychological stress can result in amenorrhea, but the mechanism by which stress alters GnRH secretion remains unknown. Both the opioid and dopaminergic systems have been implicated as potential mediators of stress-related amenorrhea (6). Recent attention has been directed toward
the role of CRH in stress and reproductive dysfunction. CRH produces a dose-related decrease in GnRH release from the mediobasal hypothalamus in vitro (7). In animal studies the intraventricular administration of CRH leads to a decrease in plasma LH levels (8-10), thought to be due to inhibition of GnRH (11). Rats subjected to hemodynamic stress have increased levels of CRH in portal blood (12), and stress-induced inhibition of pulsatile LH release can be prevented by administration of a CRH antagonist (13). It has been shown that plasma gonadotropin levels are lowered in normal women during continuous infusion of CRH. However, the gonadotropin response to a bolus injection of GnRH is unaffected by continuously infused CRH, suggesting a hypothalamic site of CRH action (14). Several investigators have examined the relationship between the hypothalamic-pituitary-adrenal (HPA) axis and amenorrhea. Women with anorexia nervosa are known to have cortisol excess and abnormal responses to CRH (15-19). However, in these studies the relative roles of weight loss and psychiatric dysfunction in HPA
Received April 17,1989. Address requests for reprints to: Dr. Anne Klibanski, Neuroendocrine Unit, Massachusetts General Hospital, Boston, Massachusetts 02114. * This work was supported in part by Grants HD-21204, DK-39251, and RR-01066 from the NIH. t Present address: Departments of Medicine and Neuroscience, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461.
311
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
312
BILLER
axis alterations have not been defined. The HPA axis has also been studied in women with HA unrelated to anorexia nervosa. In 1965, elevated urinary 17-hydroxycorticosteroid levels were found in women who developed "boarding school" amenorrhea (20), and recently, several investigators have demonstrated hypercortisolism in women with functional HA (21-23). The availability of CRH has proven to be valuable in defining the pathophysiology of hypercortisolemic states. We, therefore, characterized HPA axis function in women with HA due to simple weight loss or stress by evaluating their cortisol and ACTH response to CRH administration. The time dependency and magnitude of hypercortisolism were assessed by 24-h blood sampling for cortisol measured at 10-min intervals and 24-h urinary free cortisol determinations.
Materials and Methods Subjects We studied 10 patients and 38 normal volunteers at the Clinical Research Center at the Massachusetts General Hospital. All gave written informed consent, and the protocols were approved by the Hospital Human Studies Committee. All subjects had complete history and physical examinations, including menstrual and exercise history, and a detailed nutritional assessment by the research dietician using a computerized nutritional data base (24). They also underwent calculation of percent ideal body weight (% IBW) based on the Metropolitan Life Insurance Co. tables and determinations of percent body fat (% fat) by Lange caliper skinfold measurement at four sites (biceps, triceps, subscapular, and suprailiac) (25). Ten subjects were women with functional hypothalamic secondary amenorrhea of at least 6-month duration. Their mean age was 29 ± 7 (±SD) yr. The length of amenorrhea ranged from 0.5-13.0 yr (mean, 4.3 ± 3.7 yr). These women did not have hirsutism, chronic medical illness, anorexia nervosa, clinical depression, alcoholism, or a history of drug abuse. None engaged in regular exercise or took any medications known to affect hormone levels, except one patient on replacement thyroid hormone who had normal thyroid function tests. Six of these women had a history of weight loss in association with the onset of amenorrhea, ranging from 22-81 lb (mean, 42 lb). However, at the time of study they had maintained a stable weight for at least 1 yr. All of these women had profound estrogen deficiency, as documented by undetectable serum estradiol levels and, in those who were administered provera, an absence of withdrawal bleeding. All 10 women had serum PRL, T4, testosterone, androstenedione, dehydroepiandrosterone, LH, and FSH concentrations which were not elevated and a LH/FSH ratio of less than 1. Cranial computed tomography or magnetic resonance imaging scans were normal in all subjects. These women underwent CRH testing, 24-h frequent blood sampling at 10-min intervals, and 24-h urinary cortisol collections, as described below. Twenty normal women with regular menstrual periods and a mean age of 32 ± 8 yr served as normal controls for the CRH
ET AL.
JCE & M • 1990 Vol70«No2
tests and were also asked to perform 24-h urine collections. Eighteen additional normal women with regular menstrual periods, 27 ± 5 yr old, provided normative data for the frequent cortisol sampling. None was taking any medications known to affect hormone levels. Of the normal subjects who underwent frequent sampling, seven were in the early follicular phase, five were in the late follicular phase, and six were in the luteal phase, as determined by daily basal body temperature charting and luteal phase progesterone levels above 5 ng/mL. Protocols
CRH tests. Subjects had an indwelling catheter inserted in the antecubital vein at —180 min and were maintained at bed rest throughout the study. Two baseline blood samples were obtained at —15 min and time zero (2000 h), followed by an iv bolus injection of 1 MgAg ovine CRH. This investigational drug was purchased from Bachem (Torrance, CA) and prepared by standard methods (26, 27). Blood samples were subsequently obtained at 15, 30, 60, 90, and 120 min. Samples for ACTH determinations were collected in chilled polypropylene tubes containing EDTA and aprotinin. All blood samples were kept on ice during testing and centrifugation. Plasma was stored at -70 C until ACTH and cortisol RIAs were performed. Frequent sampling for cortisol. Subjects had an indwelling catheter inserted, and blood samples were obtained at 10-min intervals for 24 h beginning at 0800 h. Standard mixed meals composed of 30% protein, 30% fat, and 40% carbohydrate and free of caffeine were provided, and the women were allowed unrestricted activity. Smoking was prohibited. All samples were centrifuged and decanted, and the plasma was stored at —20 C until cortisol assays were performed. All samples from an individual subject were analyzed in the same cortisol assay. Urinary free cortisol. Subjects were asked to collect a 24-h urine sample for volume, free cortisol, and creatinine determinations. The collection was considered adequate if the total creatinine exceeded 15 mg/kg-day. Hormone assays. All assays were performed in duplicate. The RIA for plasma ACTH was performed using standard procedures (28). The ACTH antibody and 125I-labeled ACTH were purchased from IgG Corp. (Nashville, TN) and Radioassay Systems Laboratory (Carson, CA), respectively. The intra- and interassay coefficients of variations (CVs) were 6.4% and 9.8%, respectively. The RIA for cortisol was performed using a kit (RIA Kit Gamma Coat [125I] cortisol, Dade, Baxter Travenol Diagnostics, Cambridge, MA). The intraassay CV was 5.9%, and the interassay CV was 11.4%. Statistics. Data are expressed as the mean ± 1 SD. CRH test results are expressed for both ACTH and cortisol as the mean basal, peak, A (peak minus basal), and area under the concentration-time curve (calculated by a computer program using the trapezoidal method). Frequent sampling for cortisol is expressed as the mean for 24 h and for 8-h intervals consisting of day (0800-1600 h), evening (1600-0000 h) and night (0000-0800 h). Results were compared between the two groups using Student's t test. Correlations were examined with
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
CORTISOL SECRETION IN HYPOTHALAMIC AMENORRHEA linear regression analysis. The results of frequent cortisol sampling in normal women were analyzed separately between subjects in the early follicular, late follicular, and luteal phases of the menstrual cycle. However, the results did not differ statistically between groups, so subsequent comparisons with HA patients were performed using all of the normals.
Results The mean age was comparable between normal controls and women with HA. The HA patients did not differ from normal women in % IBW (98 ±22% us. 104 ± 14%) but had lower (P = 0.02) % body fat (26 ± 8% us. 32 ± 6%).
313
nmol/L, respectively). This elevation of cortisol in the HA patients occurred primarily during the evening (1600-0000 h) and night (0000-0800 h) sampling intervals. During both evening and night, the HA patients had higher (P = 0.008 and P = 0.04) mean cortisol concentrations than normal women (190 ± 50 us. 130 ± 50 and 300 ± 70 us. 240 ± 60 nmol/L, respectively). The daytime (0800-1600 h) values did not differ between the two groups. When the six patients with simple weight loss were analyzed separately from the women with stress-induced amenorrhea, all of the above findings remained significantly different from those in the normal women. Urinary cortisol
CRH tests The time curves of response to CRH tests are shown in Fig. 1. Mean basal concentrations of cortisol were significantly higher (P = 0.03) in the HA patients than in the normal women (210 ± 130 us. 100 ± 30 nmol/L, respectively), as shown on the top half of Fig. 2. The A cortisol (peak - basal), as shown in the bottom half of Fig. 2, was significantly lower (P = 0.004) in the HA patients than in the normal women (320 ± 100 us. 440 ± 90 nmol/L, respectively). The mean basal, peak, A, and integrated (area under the curve) ACTH values did not differ between HA patients and normal women. The means for peak ACTH and A ACTH were lower in the HA patients than in normals (6 ± 3 us. 8 ± 4 and 4 ± 2 us. 5 ± 3 pmol/L, respectively), but did not reach statistical significance (P = 0.2 and P = 0.1, respectively). Frequent serum cortisol sampling The means of plasma cortisol levels sampled every 10 min in HA patients and normal women are shown in Table 1 and Fig. 3. The 24-h mean cortisol (0800-0800 h) was significantly higher (P = 0.006) in the HA patients than in the normal controls (280 ± 50 and 220 ± 50
FiG. 1. Cortisol and ACTH responses to a 1 A*g/kg iv bolus of ovine CRH given at time zero (2000 h). The shaded area represents the mean ± SD of results in normal women, and the line represents the results in HA patients.
The urinary free cortisol level was significantly higher (P = 0.005) in the 8 HA patients from whom adequate urine collections were obtained than in the 14 normal women (230 ± 70 us. 150 ± 40 nmol/day, respectively), as shown in Fig. 4. Discussion Our study demonstrates that women with functional HA due to simple weight loss or stress have elevated basal 2000 h plasma cortisol levels and a blunted cortisol response to CRH administration. We have also demonstrated hypercortisolism in these women, as determined by elevated 24-h mean serum cortisol levels and urinary free cortisol values attributable to increased cortisol secretion during the evening and night rather than during the day. Disorders of the HPA axis have been investigated in other groups of amenorrheic women. Patients with anorexia nervosa have hypercortisolism, a lack of cortisol suppression with dexamethasone, blunted ACTH responses to CRH and elevated cerebrospinal fluid levels of CRH (15-19, 29). Amenorrheic athletes have been characterized as having plasma CRH, plasma ACTH,
800
20
600
15
_j 400 O
10
C/)
g 200
1
J •
1—k—.,
%
5 /
/
1
-15 0
15 30 45 60 75 90 105120
TIME (min)
-15 0
15 30 45 60 75 90 105120
TIME (min)
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
BILLER
314
500 O CO
r
p - 0.03
400
OS
oI
a< l l -
200 < DO
DQ
100
p - 0.004
30
o
r
400 20
CO 01
300
o 10 100
HA NL HA NL FIG. 2. Upper panel, mean (±SD) basal cortisol and ACTH levels at the start of ovine CRH tests in HA patients and normal women (NL). Lower panel, mean (±SD) A (peak - basal) cortisol and ACTH levels after an injection of ovine CRH in patients with HA and normal women (NL).
and urinary cortisol levels no different from those of eumenorrheic athletes (30, 31). However, it has recently been shown that mean morning cortisol levels were higher in amenorrheic than in eumenorrheic athletes (32). In addition, highly trained runners have elevated basal plasma ACTH and cortisol levels and blunted responses of these hormones to CRH (33). It is well established that alterations of the HPA axis can affect gonadal function. Amenorrhea is a known consequence of Cushing's syndrome (34), and exogenous corticosteroids affect the reproductive axis. Glucocorticoids have been reported to inhibit GnRH-induced LH release and cause anovulation in women (35, 36). Central administration of CRH to animals decreases GnRH and LH concentrations (8-11), and infusion of CRH into
ET AL.
JCE & M • 1990 Vol 70 • No 2
normal women lowers plasma gonadotropin levels (14). The site of this action is thought to be the hypothalamus, as CRH does not prevent GnRH-induced LH or FSH secretion (14). CRH-binding sites are localized in many areas of the brain, including those controlling GnRH and gonadotropin secretion, providing anatomical support for the in vivo findings (37). Physical and psychological stress has been linked to abnormal HPA axis function. Experimental animals subjected to hemorrhage have increased CRH levels in portal blood (12). The adrenal glands of physically stressed humans exhibit changes comparable to those seen after the administration of exogenous ACTH (38). The stress of a competitive oral examination increases plasma ACTH and cortisol levels (39), and patients with depression have hypercortisolism associated with blunted plasma ACTH responses to CRH (40). The possibility that stress could lead to reproductive dysfunction has been considered for many years. In 1965, Shanan et al. (20) reported that 22% of 64 girls who moved from the U.S. to Israel developed secondary amenorrhea associated with higher 24-h urinary 17-hydroxysteroid levels than those who remained eumenorrheic. Recent studies have provided insights into possible mechanisms for stress-related hypercortisolism and amenorrhea. Several investigators have postulated that chronic stress activates the HPA axis via CRH overproduction. The elevated cortisol is attributed to adrenal hyperplasia resulting from increased ACTH levels, and the ACTH response to CRH is blunted due to normal feedback inhibition of cortisol (19, 40). Support for this model includes the finding that continuous infusion of CRH for 24 h results in blunted ACTH and cortisol responses to a bolus of CRH (41). Although only the cortisol response was blunted after CRH injection in our patients, the model of stress-induced CRH excess resulting in gonadotropin suppression may be applicable to women with HA. There are several possible explanations for the dissociation between ACTH and cortisol responses to CRH in our subjects. Patients with HA may have a defect at the level of the adrenal gland or have a hypothalamic defect in the control of adrenal function, Simple (nonanorectic) weight loss is known to be associated with hypothalamic dysfunction (42-44). Responses to CRH may differ in women with simple weight loss compared
TABLE 1. Frequent sampling of plasma cortisol (micrograms per dL)
HA (n = 9) Normal women (n = 18) P values
0800-0800 h
0800-1600 h
1600-0000 h
0000-0800 h
10.1 ± 1.7 8.0 ± 1.7
12.3 ± 2.1 10.3 ± 2.9
7.0 ± 1.9 4.9 ± 1.7
10.8 ± 2.5 8.8 ± 2.2
0.006
NS
0.008
0.04
All values are expressed as the mean ± SD.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
CORTISOL SECRETION IN HYPOTHALAMIC AMENORRHEA
315
800 r
600
400 FIG. 3. Mean (±SD) plasma cortisol levels sampled every 10 min in HA (•—•) and normal women (A—A).
O CO
Ql
O O
200
0
u
8
9
10 11 12 1
AM
2
3
4
5
6
7
8 9
10 11 12
1
2
3
4
MIDNIGHT
NOON
5
6
7
8 AM
TIME (min) 400
p < 0.01
o CO
300
C£ O O^ LJ "o 200 LU E £T>5 LJ_
UJ
? 01
100
HA
NL
FlG. 4. Mean (±SD) urinary free cortisol levels in HA patients and normal women (NL).
to women with amenorrhea due to psychological stress. In addition, a blunted ACTH rise may have been missed due to the large SD around the mean of peak ACTH responses. CRH may stimulate cortisol secretion by pituitaryindependent or paracrine mechanisms. CRH administration stimulates cortisol secretion in patients with isolated
ACTH deficiency (45). Binding sites for CRH have been located in the adrenal glands, and the adrenal medulla can make ACTH (46, 47). In states of CRH excess, the adrenal CRH receptors could be down-regulated, leading to a blunted cortisol response to CRH administration. In addition, extrapituitary sources of ACTH, including leukocytes (48, 49), may modulate cortisol secretion during stress via the immune system. Alternatively, it is possible that hypercortisolism in patients with HA results from an excess of an ACTH secretagogue other than CRH. Arginine vasopressin causes ACTH release from pituitary cells (50) and stimulates both ACTH and cortisol release in humans (51). Thus, it is possible that in our patients with HA, chronic stress leads to the release of an ACTH secretagogue other than CRH, which, in turn, results in cortisol hypersecretion. A bolus of CRH would then be expected to produce a normal rise in ACTH via pituitary CRH receptors which are not down-regulated, but a blunted response of cortisol because of downregulation of adrenal ACTH receptors. This report describes abnormal CRH responses in women with HA. In one prior study, CRH tests were normal in patients with HA, but basal and A cortisol levels were not provided (23). The demonstration of hypercortisolism in patients with functional HA in the absence of illness, anorexia nervosa, or excess exercise extends the findings of several prior studies. HA patients were found to have elevated mean serum cortisol values compared to normal women when blood samples were obtained at 15-min intervals for 3 h (21). Suh et al. (22) showed significantly higher mean 24-h serum cortisol levels in women with HA. In contrast to our study, the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
BILLER
316
cortisol elevations in their HA patients occurred during the daytime. However, in a recent study the same group showed a significantly higher area under the curve of serum cortisol during the night as well as in the daytime in HA patients compared to normal women (23). We have documented hypercortisolism and abnormal CRH responses in patients with hypothalamic amenorrhea due to simple weight loss or stress. This HPA axis activation may be one of the mechanisms of amenorrhea in HA patients and may also contribute to osteoporosis and other adverse effects in this patient group.
Acknowledgments The authors wish to thank Dr. George Chrousos for assistance in obtaining CRH, Dr. Joel Finkelstein for assistance with data analysis, Dr. Daniel I. Spratt for patient referrals, Mr. Arnold Kana for technical assistance, Ms. Ellen Williams for nutritional evaluations, the staff of the Clinical Research Center for expert patient care, and Ms. Rose Mooradian for secretarial help.
References 1. Klibanski A, Biller BMK, Rosenthal DI, Schoenfeld, DA, Saxe V. Effects of prolactin and estrogen deficiency in amenorrheic bone loss. J Clin Endocrinol Metab. 1988;67:124-30. 2. Drew FL. The epidemiology of secondary amenorrhea. J Chronic Dis 1961;14:396-407. 3. Hull MGR, Knuth UA, Murray MAF, Jacobs HS. The practical value of the progestogen challenge test, serum oestradiol estimation or clinical examination in assessment of the oestrogen state and response to clomiphene in amenorrhea. Br J Obstet Gynaecol. 1979;86:799-805. 4. Soules MR, Jelovsek FR, Wiebe RH. Amenorrhea: observations based on the analysis of luteinizing hormone releasing testing. Am J Obstet Gynecol. 1979;135:651-62. 5. McCormick WO. Amenorrhea and other menstrual symptoms in student nurses. J Psychosom Res. 1975;19:131-7. 6. Quigley ME, Sheehan KL, Casper RF, Yen SSC. Evidence for increased dopaminergic and opioid activity in patients with hypothalamic hypogonadotropic amenorrhea. J Clin Endocrinol Metab. 1980;50:949-54. 7. Gambacciani M, Yen SS, Rasmussen DD. GnRH release from the mediobasal hypothalamus: in vitro inhibition by corticotropinreleasing factor. Neuroendocrinology. 1986;43:533-6. 8. Rivier C, Vale W. Influence of corticotropin-releasing factor on reproductive functions in the rat. Endocrinology. 1984;114:914-21. 9. Ono N, Lumpkin MD, Samson WK, McDonald JK, McCann SM. Intrahypothalamic action of corticotropin-releasing-factor (CRH) to inhibit growth hormone on LH release in the rat. Life Sci. 1984;35:1117-23. 10. Olster DH, Ferin M. Corticotropin-releasing hormone inhibits gonadotropin secretion in the ovariectomized rhesus monkey. J Clin Endocrinol Metab. 1987;65:262-7. 11. Petraglia F, Sutton S, Vale W, Plotsky P. Corticotropin-releasing factor decreases plasma luteinizing hormone levels in female rats by inhibiting gonadotropin-releasing hormone release into hypophyseal-portal circulation. Endocrinology. 1987;120:1083-8. 12. Plotsky P, Vale W. Hemorrhage-induced secretion of corticotropinreleasing-factor like immunoreactivity into the rat hypophyseal portal circulation and its inhibition by glucocorticoids. Endocrinology. 1984;114:164-9. 13. Rivier C, Rivier J, Vale W. Stress-induced inhibition of reproductive functions: role of endogenous corticotropin-releasing factor. Science. 1986;231:607-9. 14. Barbarino A, deMarinis L, Tofani A, et al. Corticotropin-releasing hormone inhibition of gonadotropin release and the effects of
ET AL.
JCE & M • 1990 Vol70«No2
opioid blockade. J Clin Endocrinol Metab. 1989;68:523-8. 15. Walsh T, Katz JL, Levin J, Kream J, Fukushima D, Weiner H, Zumoff B. The production rate of cortisol declines during recovery from anorexia nervosa. J Clin Endocrinol Metab. 1981;53:203-5. 16. Landon J, Greenwood FC, Stamp TC, Wynn V. The plasma sugar, free fatty acid, cortisol, and growth hormone response to insulin, and the comparison of this procedure with other tests of pituitary and adrenal function. J Clin Invest. 1966;45:437-49. 17. Boyar RM, Hellman LD, Roffwang H, et al. Cortisol secretion and metabolism in anorexia nervosa. N Engl J Med. 1977;296:190-3. 18. Gold PW, Gwirtsman H, Avgerinos C, et al. Abnormal hypothalamic-pituitary-adrenal function in anorexia nervosa. N Engl J Med. 1986;314:1335-42. 19. Hotta M, Shibasaki T, Masuda A, et al. The responses of plasma adrenocorticotropin and cortisol to corticotropin-releasing hormone and cerebrospinal fluid immunoreactive CRH in anorexia nervosa patients. J Clin Endocrinol Metab. 1986;62:319-24. 20. Shanan J, Brzezinsk A, Sulman F, Sharon M. Active coping behavior, anxiety, and cortical steroid excretion in the prediction of transient amenorrhea. Behav Sci. 1965;10:461-5. 21. Boesgaard S, Hagen C, Andersen AN, Djursing H, Fenger M. Cortisol secretion in patients with normoprolactinemic amenorrhea. Acta Endocrinol (Copenh) 1988;118:544-50. 22. Suh BY, Liu JH, Berga SL, Quigly ME, Laughlin GA, Yen SS. Hypercortisolism in patients with functional hypothalamic amenorrhea. J Clin Endocrinol Metab. 1988;66:733-9. 23. Berga SL, Mortola JF, Girton L, et al. Neuroendocrine aberrations in women with functional hypothalamic amenorrhea. J Clin Endocrinol Metab. 1989:68:301-8. 24. Dennis B, Ernst N, Hjortland M, Tillotson J, Grambsch V, The NHLBI nutrient data systems. J Am Diet Assoc. 1980;77:641-7. 25. Jackson AS, Pollock ML. Practical assessment of body composition. Phys Sports Med. 1985; 13:56. 26. Chrousos GP, Schulte HM, Oldfield EH, et al. Corticotropinreleasing factor: basic and clinical studies. Physchopharmacol Bull. 1983;19:416-21. 27. Gold PW, Chrousos G, Kellner C, et al. Psychiatric implications of basic and clinical studies with corticotropin-releasing factors. Am J Psychiatry. 1984;141:619-27. 28. Nicholson WE, Davis DR, Sherrel BJ, Orth DN. Rapid radioimmunoassay for corticotropin in unextracted human plasma. Clin Chem. 1984;30:259-65. 29. Kaye WH, Gwirtsman HE, George DT, et al. Elevated cerebrospinal fluid levels of immunoreactive corticotropin-releasing hormone in anorexia nervosa: relation to state of nutrition, adrenal function, and intensity of depression. J Clin Endocrinol Metab. 1987;64:2038. 30. Villaneuva AL, Schlosser C, Hopper B, Liu JH, Hoffman DI, Rebar RW. Increased cortisol production in women runners. J Clin Endocrinol Metab. 1986;63:133-6. 31. Hohtari H, Elovainio R, Salminen K, Laatikainen T. Plasma corticotropin-releasing hormone, corticotropin, and endorphins at rest and during exercise in eumenorrheic and amenorrheic athletes. Fertil Steril. 1988;50:233-8. 32. Ding JH, Sheckter CB, Drinkwater BL, Soules MR, Bremner WJ. High serum cortisol levels in exercise-associated amenorrhea. Ann Intern Med. 1988;108:530-4. 33. Luger A, Deuster PA, Kyle SB, et al. Acute hypothalamic-pituitaryadrenal responses to the stress of treadmill exercise. N Engl J Med. 1987;316:1309-15. 34. Cushing H. The basophil adenomas of the pituitary body and their clinical manifestations. Bull Johns Hopkins Hosp. 1932;50:137-95. 35. Sakakura M, Takebe K, Nahagawa S. Inhibition of luteinizing hormone secretion induced by synthetic LRH by long-term treatment with glucocorticoids in human subjects. J Clin Endocrinol Metab. 1975;40:774-9. 36. Cunningham GR, Caperton Jr EM, Goldzieher JW. Antiovulatory activity of synthetic corticoids. J Clin Endocrinol Metab. 1975;40:265-7. 37. DeSouza EB, Insel TR, Perrin MH, Rivier J, Vale WW, Kuhar MJ. Corticotropin-releasing factor receptors are widely distributed within the rat central nervous system: an autoradiographic study.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
CORTISOL SECRETION IN HYPOTHALAMIC AMENORRHEA J Neurosci. 1985;5:3189-202. 38. Symington T, Duguid WP, Davidson JN. Effect of exogenous corticotropin on the histochemical pattern of the human adrenal cortex and a comparison with the changes during stress. J Clin Endocrinol Metab. 1955;16:580-98. 39. Meyerhoff JL, Oleshansky MA, Mougey EH. Psychologic stress increases plasma levels of prolactin, cortisol, and POMC-derived peptides in man. Psychosom Med. 1988;50:295-303. 40. Gold PW, Loriaux DL, Ray A, et al. Responses to corticotropinreleasing hormone in the hypercortisolism of depression and Cushing's disease. N Engl J Med. 1986;4:1329-35. 41. Schulte HM, Chrousos GP, Gold PW, et al. Continuous administration of synthetic ovine corticotropin-releasing factor in man. J Clin Invest. 1985;75:1781-5. 42. Kapen S, Sternthal E, Braverman L. A "pubertal" 24 hour luteinizing hormone (LH) secretory pattern following weight loss in the absence of anorexia nervosa. Psychosom Med. 1981;43:177-82. 43. Vigersky RA, Andersen AE, Thompson RH, Loriaux DL. Hypothalamic dysfunction in secondary amenorrhea associated with simple weight loss. N Engl J Med. 1977;297:1141-5. 44. Warren MP, Jewelewicz R, Dyrenfurth I, Ans R, Khalaf S, Vande Wiele RL. The significance of weight loss in the evaluation of pituitary response to LH-RH in women with secondary amenor-
317
rhea. J Clin Endocrinol Metab. 1975;40:601-ll. 45. Fehm HL, Holl R, Spath-Schwalbe E, Born J, Voight KH. Ability of corticotropin releasing hormone to stimulate cortisol secretion independent from pituitary adrenocorticotropin. Life Sci. 1988;42:679-86. 46. Dave JR, Eiden LE, Eskay RL. Corticotropin-releasing factor binding to peripheral tissue and activation of the adenylate cyclaseadenosine 3',5'-monophosphate system. Endocrinology. 1985;116: 2152-9. 47. Evans CHJ, Erdelyi E, Weber E, Barchas JD. Identification of proopiomelanocortin-derived peptides in the human adrenal medulla. Science. 1983;221:957-60. 48. Smith EM, Meyer JW, Blalock JE. Virus-induced corticosterone in hypophysectomized mice: a possible lymphoid-adrenal axis. Science. 1982;218:1311-2. 49. Blalock JE, Smith EM, Meyer WJ. The pituitary-adrenocortical axis and the immune system. Clin Endocrinol Metab. 1985;14: 1021-38. 50. Vale W, Rivier C. Substances modulating the secretion of ACTH by cultured anterior pituitary cells. Fed Proc. 1977;36:2094-99. 51. DeBold CR, Sheldon WR, DeCherney GS, et al. Arginine vasopressin potentiates adrenocorticotropin release induced by ovine corticotropin-releasing factor. J Clin Invest. 1984;73:533-8.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
Vol. 70, No. 2 Printed in U.S.A.
Abnormal Cortisol Secretion and Responses to Corticotropin-Releasing Hormone in Women with Hypothalamic Amenorrhea* BEVERLY M. K. BILLER, HOWARD J. FEDEROFFf, JAMES I. KOENIG, AND ANNE KLIBANSKI Neuroendocrine Unit, Departments of Medicine and Neurology, Massachusetts General Hospital, Boston, Massachusetts 02114
± 90 nmol/L, respectively). ACTH responses to CRH did not differ between HA patients and normal women. The 24-h mean cortisol was significantly higher (P = 0.006) in the HA patients than in the normal controls (280 ± 50 and 220 ± 50 nmol/L, respectively), due to higher cortisol levels at night. The urinary free cortisol level was significantly higher (P = 0.005) in the HA patients (230 ± 70 nmol/day) than in normal women (150 ± 40 nmol/day). We conclude that women with HA have a blunted cortisol response to CRH administration. In addition, they have hypercortisolism, as demonstrated by elevated 24-h mean serum cortisol levels and urinary free cortisol values. This hypothalamicpituitary-adrenal axis activation in patients with stress or weight loss may be a mechanism in the development of amenorrhea and may relate to other potential adverse effects of HA. (J Clin Endocrinol Metab 70: 311, 1990)
ABSTRACT. Hypothalamic amenorrhea (HA) is a common disorder associated with hypoestrogenemia and has adverse effects. The mechanism of GnRH deficiency in these women is not yet known. To investigate the role of the hypothalamic pituitary-adrenal axis in HA, we studied 10 women [mean age, 29 ± 7 (±SD) yr] with 0.5-13 yr of amenorrhea (mean, 4.3 ± 3.7 yr) related to simple weight loss or psychological stress. We investigated cortisol and ACTH responses to a bolus of ovine CRH, 24-h plasma cortisol levels obtained every 10 min, and urinary free cortisol levels in these patients. Results were compared with those obtained in normal women during all phases of the menstrual cycle. We found that mean basal concentrations of cortisol were significantly higher (P = 0.03) in the HA patients (mean, 210 ± 130 nmol/L) than in the normal women (100 ± 30 nmol/L). The A (peak - basal) cortisol was significantly lower (P = 0.004) in the HA patients than in the normal women (320 ± 100 vs. 440
H
YPOTHALAMIC amenorrhea (HA) is a functional disorder of presumed GnRH deficiency associated with profound estrogen deficiency and adverse effects, including loss of bone density (1). Prevalence rates vary widely, with reports as high as 29% in nurses during their first year of training; however, in most studies approximately 15% of secondary amenorrhea cases are due to HA (2-5). It is well recognized that both physical and psychological stress can result in amenorrhea, but the mechanism by which stress alters GnRH secretion remains unknown. Both the opioid and dopaminergic systems have been implicated as potential mediators of stress-related amenorrhea (6). Recent attention has been directed toward
the role of CRH in stress and reproductive dysfunction. CRH produces a dose-related decrease in GnRH release from the mediobasal hypothalamus in vitro (7). In animal studies the intraventricular administration of CRH leads to a decrease in plasma LH levels (8-10), thought to be due to inhibition of GnRH (11). Rats subjected to hemodynamic stress have increased levels of CRH in portal blood (12), and stress-induced inhibition of pulsatile LH release can be prevented by administration of a CRH antagonist (13). It has been shown that plasma gonadotropin levels are lowered in normal women during continuous infusion of CRH. However, the gonadotropin response to a bolus injection of GnRH is unaffected by continuously infused CRH, suggesting a hypothalamic site of CRH action (14). Several investigators have examined the relationship between the hypothalamic-pituitary-adrenal (HPA) axis and amenorrhea. Women with anorexia nervosa are known to have cortisol excess and abnormal responses to CRH (15-19). However, in these studies the relative roles of weight loss and psychiatric dysfunction in HPA
Received April 17,1989. Address requests for reprints to: Dr. Anne Klibanski, Neuroendocrine Unit, Massachusetts General Hospital, Boston, Massachusetts 02114. * This work was supported in part by Grants HD-21204, DK-39251, and RR-01066 from the NIH. t Present address: Departments of Medicine and Neuroscience, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461.
311
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
312
BILLER
axis alterations have not been defined. The HPA axis has also been studied in women with HA unrelated to anorexia nervosa. In 1965, elevated urinary 17-hydroxycorticosteroid levels were found in women who developed "boarding school" amenorrhea (20), and recently, several investigators have demonstrated hypercortisolism in women with functional HA (21-23). The availability of CRH has proven to be valuable in defining the pathophysiology of hypercortisolemic states. We, therefore, characterized HPA axis function in women with HA due to simple weight loss or stress by evaluating their cortisol and ACTH response to CRH administration. The time dependency and magnitude of hypercortisolism were assessed by 24-h blood sampling for cortisol measured at 10-min intervals and 24-h urinary free cortisol determinations.
Materials and Methods Subjects We studied 10 patients and 38 normal volunteers at the Clinical Research Center at the Massachusetts General Hospital. All gave written informed consent, and the protocols were approved by the Hospital Human Studies Committee. All subjects had complete history and physical examinations, including menstrual and exercise history, and a detailed nutritional assessment by the research dietician using a computerized nutritional data base (24). They also underwent calculation of percent ideal body weight (% IBW) based on the Metropolitan Life Insurance Co. tables and determinations of percent body fat (% fat) by Lange caliper skinfold measurement at four sites (biceps, triceps, subscapular, and suprailiac) (25). Ten subjects were women with functional hypothalamic secondary amenorrhea of at least 6-month duration. Their mean age was 29 ± 7 (±SD) yr. The length of amenorrhea ranged from 0.5-13.0 yr (mean, 4.3 ± 3.7 yr). These women did not have hirsutism, chronic medical illness, anorexia nervosa, clinical depression, alcoholism, or a history of drug abuse. None engaged in regular exercise or took any medications known to affect hormone levels, except one patient on replacement thyroid hormone who had normal thyroid function tests. Six of these women had a history of weight loss in association with the onset of amenorrhea, ranging from 22-81 lb (mean, 42 lb). However, at the time of study they had maintained a stable weight for at least 1 yr. All of these women had profound estrogen deficiency, as documented by undetectable serum estradiol levels and, in those who were administered provera, an absence of withdrawal bleeding. All 10 women had serum PRL, T4, testosterone, androstenedione, dehydroepiandrosterone, LH, and FSH concentrations which were not elevated and a LH/FSH ratio of less than 1. Cranial computed tomography or magnetic resonance imaging scans were normal in all subjects. These women underwent CRH testing, 24-h frequent blood sampling at 10-min intervals, and 24-h urinary cortisol collections, as described below. Twenty normal women with regular menstrual periods and a mean age of 32 ± 8 yr served as normal controls for the CRH
ET AL.
JCE & M • 1990 Vol70«No2
tests and were also asked to perform 24-h urine collections. Eighteen additional normal women with regular menstrual periods, 27 ± 5 yr old, provided normative data for the frequent cortisol sampling. None was taking any medications known to affect hormone levels. Of the normal subjects who underwent frequent sampling, seven were in the early follicular phase, five were in the late follicular phase, and six were in the luteal phase, as determined by daily basal body temperature charting and luteal phase progesterone levels above 5 ng/mL. Protocols
CRH tests. Subjects had an indwelling catheter inserted in the antecubital vein at —180 min and were maintained at bed rest throughout the study. Two baseline blood samples were obtained at —15 min and time zero (2000 h), followed by an iv bolus injection of 1 MgAg ovine CRH. This investigational drug was purchased from Bachem (Torrance, CA) and prepared by standard methods (26, 27). Blood samples were subsequently obtained at 15, 30, 60, 90, and 120 min. Samples for ACTH determinations were collected in chilled polypropylene tubes containing EDTA and aprotinin. All blood samples were kept on ice during testing and centrifugation. Plasma was stored at -70 C until ACTH and cortisol RIAs were performed. Frequent sampling for cortisol. Subjects had an indwelling catheter inserted, and blood samples were obtained at 10-min intervals for 24 h beginning at 0800 h. Standard mixed meals composed of 30% protein, 30% fat, and 40% carbohydrate and free of caffeine were provided, and the women were allowed unrestricted activity. Smoking was prohibited. All samples were centrifuged and decanted, and the plasma was stored at —20 C until cortisol assays were performed. All samples from an individual subject were analyzed in the same cortisol assay. Urinary free cortisol. Subjects were asked to collect a 24-h urine sample for volume, free cortisol, and creatinine determinations. The collection was considered adequate if the total creatinine exceeded 15 mg/kg-day. Hormone assays. All assays were performed in duplicate. The RIA for plasma ACTH was performed using standard procedures (28). The ACTH antibody and 125I-labeled ACTH were purchased from IgG Corp. (Nashville, TN) and Radioassay Systems Laboratory (Carson, CA), respectively. The intra- and interassay coefficients of variations (CVs) were 6.4% and 9.8%, respectively. The RIA for cortisol was performed using a kit (RIA Kit Gamma Coat [125I] cortisol, Dade, Baxter Travenol Diagnostics, Cambridge, MA). The intraassay CV was 5.9%, and the interassay CV was 11.4%. Statistics. Data are expressed as the mean ± 1 SD. CRH test results are expressed for both ACTH and cortisol as the mean basal, peak, A (peak minus basal), and area under the concentration-time curve (calculated by a computer program using the trapezoidal method). Frequent sampling for cortisol is expressed as the mean for 24 h and for 8-h intervals consisting of day (0800-1600 h), evening (1600-0000 h) and night (0000-0800 h). Results were compared between the two groups using Student's t test. Correlations were examined with
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
CORTISOL SECRETION IN HYPOTHALAMIC AMENORRHEA linear regression analysis. The results of frequent cortisol sampling in normal women were analyzed separately between subjects in the early follicular, late follicular, and luteal phases of the menstrual cycle. However, the results did not differ statistically between groups, so subsequent comparisons with HA patients were performed using all of the normals.
Results The mean age was comparable between normal controls and women with HA. The HA patients did not differ from normal women in % IBW (98 ±22% us. 104 ± 14%) but had lower (P = 0.02) % body fat (26 ± 8% us. 32 ± 6%).
313
nmol/L, respectively). This elevation of cortisol in the HA patients occurred primarily during the evening (1600-0000 h) and night (0000-0800 h) sampling intervals. During both evening and night, the HA patients had higher (P = 0.008 and P = 0.04) mean cortisol concentrations than normal women (190 ± 50 us. 130 ± 50 and 300 ± 70 us. 240 ± 60 nmol/L, respectively). The daytime (0800-1600 h) values did not differ between the two groups. When the six patients with simple weight loss were analyzed separately from the women with stress-induced amenorrhea, all of the above findings remained significantly different from those in the normal women. Urinary cortisol
CRH tests The time curves of response to CRH tests are shown in Fig. 1. Mean basal concentrations of cortisol were significantly higher (P = 0.03) in the HA patients than in the normal women (210 ± 130 us. 100 ± 30 nmol/L, respectively), as shown on the top half of Fig. 2. The A cortisol (peak - basal), as shown in the bottom half of Fig. 2, was significantly lower (P = 0.004) in the HA patients than in the normal women (320 ± 100 us. 440 ± 90 nmol/L, respectively). The mean basal, peak, A, and integrated (area under the curve) ACTH values did not differ between HA patients and normal women. The means for peak ACTH and A ACTH were lower in the HA patients than in normals (6 ± 3 us. 8 ± 4 and 4 ± 2 us. 5 ± 3 pmol/L, respectively), but did not reach statistical significance (P = 0.2 and P = 0.1, respectively). Frequent serum cortisol sampling The means of plasma cortisol levels sampled every 10 min in HA patients and normal women are shown in Table 1 and Fig. 3. The 24-h mean cortisol (0800-0800 h) was significantly higher (P = 0.006) in the HA patients than in the normal controls (280 ± 50 and 220 ± 50
FiG. 1. Cortisol and ACTH responses to a 1 A*g/kg iv bolus of ovine CRH given at time zero (2000 h). The shaded area represents the mean ± SD of results in normal women, and the line represents the results in HA patients.
The urinary free cortisol level was significantly higher (P = 0.005) in the 8 HA patients from whom adequate urine collections were obtained than in the 14 normal women (230 ± 70 us. 150 ± 40 nmol/day, respectively), as shown in Fig. 4. Discussion Our study demonstrates that women with functional HA due to simple weight loss or stress have elevated basal 2000 h plasma cortisol levels and a blunted cortisol response to CRH administration. We have also demonstrated hypercortisolism in these women, as determined by elevated 24-h mean serum cortisol levels and urinary free cortisol values attributable to increased cortisol secretion during the evening and night rather than during the day. Disorders of the HPA axis have been investigated in other groups of amenorrheic women. Patients with anorexia nervosa have hypercortisolism, a lack of cortisol suppression with dexamethasone, blunted ACTH responses to CRH and elevated cerebrospinal fluid levels of CRH (15-19, 29). Amenorrheic athletes have been characterized as having plasma CRH, plasma ACTH,
800
20
600
15
_j 400 O
10
C/)
g 200
1
J •
1—k—.,
%
5 /
/
1
-15 0
15 30 45 60 75 90 105120
TIME (min)
-15 0
15 30 45 60 75 90 105120
TIME (min)
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
BILLER
314
500 O CO
r
p - 0.03
400
OS
oI
a< l l -
200 < DO
DQ
100
p - 0.004
30
o
r
400 20
CO 01
300
o 10 100
HA NL HA NL FIG. 2. Upper panel, mean (±SD) basal cortisol and ACTH levels at the start of ovine CRH tests in HA patients and normal women (NL). Lower panel, mean (±SD) A (peak - basal) cortisol and ACTH levels after an injection of ovine CRH in patients with HA and normal women (NL).
and urinary cortisol levels no different from those of eumenorrheic athletes (30, 31). However, it has recently been shown that mean morning cortisol levels were higher in amenorrheic than in eumenorrheic athletes (32). In addition, highly trained runners have elevated basal plasma ACTH and cortisol levels and blunted responses of these hormones to CRH (33). It is well established that alterations of the HPA axis can affect gonadal function. Amenorrhea is a known consequence of Cushing's syndrome (34), and exogenous corticosteroids affect the reproductive axis. Glucocorticoids have been reported to inhibit GnRH-induced LH release and cause anovulation in women (35, 36). Central administration of CRH to animals decreases GnRH and LH concentrations (8-11), and infusion of CRH into
ET AL.
JCE & M • 1990 Vol 70 • No 2
normal women lowers plasma gonadotropin levels (14). The site of this action is thought to be the hypothalamus, as CRH does not prevent GnRH-induced LH or FSH secretion (14). CRH-binding sites are localized in many areas of the brain, including those controlling GnRH and gonadotropin secretion, providing anatomical support for the in vivo findings (37). Physical and psychological stress has been linked to abnormal HPA axis function. Experimental animals subjected to hemorrhage have increased CRH levels in portal blood (12). The adrenal glands of physically stressed humans exhibit changes comparable to those seen after the administration of exogenous ACTH (38). The stress of a competitive oral examination increases plasma ACTH and cortisol levels (39), and patients with depression have hypercortisolism associated with blunted plasma ACTH responses to CRH (40). The possibility that stress could lead to reproductive dysfunction has been considered for many years. In 1965, Shanan et al. (20) reported that 22% of 64 girls who moved from the U.S. to Israel developed secondary amenorrhea associated with higher 24-h urinary 17-hydroxysteroid levels than those who remained eumenorrheic. Recent studies have provided insights into possible mechanisms for stress-related hypercortisolism and amenorrhea. Several investigators have postulated that chronic stress activates the HPA axis via CRH overproduction. The elevated cortisol is attributed to adrenal hyperplasia resulting from increased ACTH levels, and the ACTH response to CRH is blunted due to normal feedback inhibition of cortisol (19, 40). Support for this model includes the finding that continuous infusion of CRH for 24 h results in blunted ACTH and cortisol responses to a bolus of CRH (41). Although only the cortisol response was blunted after CRH injection in our patients, the model of stress-induced CRH excess resulting in gonadotropin suppression may be applicable to women with HA. There are several possible explanations for the dissociation between ACTH and cortisol responses to CRH in our subjects. Patients with HA may have a defect at the level of the adrenal gland or have a hypothalamic defect in the control of adrenal function, Simple (nonanorectic) weight loss is known to be associated with hypothalamic dysfunction (42-44). Responses to CRH may differ in women with simple weight loss compared
TABLE 1. Frequent sampling of plasma cortisol (micrograms per dL)
HA (n = 9) Normal women (n = 18) P values
0800-0800 h
0800-1600 h
1600-0000 h
0000-0800 h
10.1 ± 1.7 8.0 ± 1.7
12.3 ± 2.1 10.3 ± 2.9
7.0 ± 1.9 4.9 ± 1.7
10.8 ± 2.5 8.8 ± 2.2
0.006
NS
0.008
0.04
All values are expressed as the mean ± SD.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
CORTISOL SECRETION IN HYPOTHALAMIC AMENORRHEA
315
800 r
600
400 FIG. 3. Mean (±SD) plasma cortisol levels sampled every 10 min in HA (•—•) and normal women (A—A).
O CO
Ql
O O
200
0
u
8
9
10 11 12 1
AM
2
3
4
5
6
7
8 9
10 11 12
1
2
3
4
MIDNIGHT
NOON
5
6
7
8 AM
TIME (min) 400
p < 0.01
o CO
300
C£ O O^ LJ "o 200 LU E £T>5 LJ_
UJ
? 01
100
HA
NL
FlG. 4. Mean (±SD) urinary free cortisol levels in HA patients and normal women (NL).
to women with amenorrhea due to psychological stress. In addition, a blunted ACTH rise may have been missed due to the large SD around the mean of peak ACTH responses. CRH may stimulate cortisol secretion by pituitaryindependent or paracrine mechanisms. CRH administration stimulates cortisol secretion in patients with isolated
ACTH deficiency (45). Binding sites for CRH have been located in the adrenal glands, and the adrenal medulla can make ACTH (46, 47). In states of CRH excess, the adrenal CRH receptors could be down-regulated, leading to a blunted cortisol response to CRH administration. In addition, extrapituitary sources of ACTH, including leukocytes (48, 49), may modulate cortisol secretion during stress via the immune system. Alternatively, it is possible that hypercortisolism in patients with HA results from an excess of an ACTH secretagogue other than CRH. Arginine vasopressin causes ACTH release from pituitary cells (50) and stimulates both ACTH and cortisol release in humans (51). Thus, it is possible that in our patients with HA, chronic stress leads to the release of an ACTH secretagogue other than CRH, which, in turn, results in cortisol hypersecretion. A bolus of CRH would then be expected to produce a normal rise in ACTH via pituitary CRH receptors which are not down-regulated, but a blunted response of cortisol because of downregulation of adrenal ACTH receptors. This report describes abnormal CRH responses in women with HA. In one prior study, CRH tests were normal in patients with HA, but basal and A cortisol levels were not provided (23). The demonstration of hypercortisolism in patients with functional HA in the absence of illness, anorexia nervosa, or excess exercise extends the findings of several prior studies. HA patients were found to have elevated mean serum cortisol values compared to normal women when blood samples were obtained at 15-min intervals for 3 h (21). Suh et al. (22) showed significantly higher mean 24-h serum cortisol levels in women with HA. In contrast to our study, the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
BILLER
316
cortisol elevations in their HA patients occurred during the daytime. However, in a recent study the same group showed a significantly higher area under the curve of serum cortisol during the night as well as in the daytime in HA patients compared to normal women (23). We have documented hypercortisolism and abnormal CRH responses in patients with hypothalamic amenorrhea due to simple weight loss or stress. This HPA axis activation may be one of the mechanisms of amenorrhea in HA patients and may also contribute to osteoporosis and other adverse effects in this patient group.
Acknowledgments The authors wish to thank Dr. George Chrousos for assistance in obtaining CRH, Dr. Joel Finkelstein for assistance with data analysis, Dr. Daniel I. Spratt for patient referrals, Mr. Arnold Kana for technical assistance, Ms. Ellen Williams for nutritional evaluations, the staff of the Clinical Research Center for expert patient care, and Ms. Rose Mooradian for secretarial help.
References 1. Klibanski A, Biller BMK, Rosenthal DI, Schoenfeld, DA, Saxe V. Effects of prolactin and estrogen deficiency in amenorrheic bone loss. J Clin Endocrinol Metab. 1988;67:124-30. 2. Drew FL. The epidemiology of secondary amenorrhea. J Chronic Dis 1961;14:396-407. 3. Hull MGR, Knuth UA, Murray MAF, Jacobs HS. The practical value of the progestogen challenge test, serum oestradiol estimation or clinical examination in assessment of the oestrogen state and response to clomiphene in amenorrhea. Br J Obstet Gynaecol. 1979;86:799-805. 4. Soules MR, Jelovsek FR, Wiebe RH. Amenorrhea: observations based on the analysis of luteinizing hormone releasing testing. Am J Obstet Gynecol. 1979;135:651-62. 5. McCormick WO. Amenorrhea and other menstrual symptoms in student nurses. J Psychosom Res. 1975;19:131-7. 6. Quigley ME, Sheehan KL, Casper RF, Yen SSC. Evidence for increased dopaminergic and opioid activity in patients with hypothalamic hypogonadotropic amenorrhea. J Clin Endocrinol Metab. 1980;50:949-54. 7. Gambacciani M, Yen SS, Rasmussen DD. GnRH release from the mediobasal hypothalamus: in vitro inhibition by corticotropinreleasing factor. Neuroendocrinology. 1986;43:533-6. 8. Rivier C, Vale W. Influence of corticotropin-releasing factor on reproductive functions in the rat. Endocrinology. 1984;114:914-21. 9. Ono N, Lumpkin MD, Samson WK, McDonald JK, McCann SM. Intrahypothalamic action of corticotropin-releasing-factor (CRH) to inhibit growth hormone on LH release in the rat. Life Sci. 1984;35:1117-23. 10. Olster DH, Ferin M. Corticotropin-releasing hormone inhibits gonadotropin secretion in the ovariectomized rhesus monkey. J Clin Endocrinol Metab. 1987;65:262-7. 11. Petraglia F, Sutton S, Vale W, Plotsky P. Corticotropin-releasing factor decreases plasma luteinizing hormone levels in female rats by inhibiting gonadotropin-releasing hormone release into hypophyseal-portal circulation. Endocrinology. 1987;120:1083-8. 12. Plotsky P, Vale W. Hemorrhage-induced secretion of corticotropinreleasing-factor like immunoreactivity into the rat hypophyseal portal circulation and its inhibition by glucocorticoids. Endocrinology. 1984;114:164-9. 13. Rivier C, Rivier J, Vale W. Stress-induced inhibition of reproductive functions: role of endogenous corticotropin-releasing factor. Science. 1986;231:607-9. 14. Barbarino A, deMarinis L, Tofani A, et al. Corticotropin-releasing hormone inhibition of gonadotropin release and the effects of
ET AL.
JCE & M • 1990 Vol70«No2
opioid blockade. J Clin Endocrinol Metab. 1989;68:523-8. 15. Walsh T, Katz JL, Levin J, Kream J, Fukushima D, Weiner H, Zumoff B. The production rate of cortisol declines during recovery from anorexia nervosa. J Clin Endocrinol Metab. 1981;53:203-5. 16. Landon J, Greenwood FC, Stamp TC, Wynn V. The plasma sugar, free fatty acid, cortisol, and growth hormone response to insulin, and the comparison of this procedure with other tests of pituitary and adrenal function. J Clin Invest. 1966;45:437-49. 17. Boyar RM, Hellman LD, Roffwang H, et al. Cortisol secretion and metabolism in anorexia nervosa. N Engl J Med. 1977;296:190-3. 18. Gold PW, Gwirtsman H, Avgerinos C, et al. Abnormal hypothalamic-pituitary-adrenal function in anorexia nervosa. N Engl J Med. 1986;314:1335-42. 19. Hotta M, Shibasaki T, Masuda A, et al. The responses of plasma adrenocorticotropin and cortisol to corticotropin-releasing hormone and cerebrospinal fluid immunoreactive CRH in anorexia nervosa patients. J Clin Endocrinol Metab. 1986;62:319-24. 20. Shanan J, Brzezinsk A, Sulman F, Sharon M. Active coping behavior, anxiety, and cortical steroid excretion in the prediction of transient amenorrhea. Behav Sci. 1965;10:461-5. 21. Boesgaard S, Hagen C, Andersen AN, Djursing H, Fenger M. Cortisol secretion in patients with normoprolactinemic amenorrhea. Acta Endocrinol (Copenh) 1988;118:544-50. 22. Suh BY, Liu JH, Berga SL, Quigly ME, Laughlin GA, Yen SS. Hypercortisolism in patients with functional hypothalamic amenorrhea. J Clin Endocrinol Metab. 1988;66:733-9. 23. Berga SL, Mortola JF, Girton L, et al. Neuroendocrine aberrations in women with functional hypothalamic amenorrhea. J Clin Endocrinol Metab. 1989:68:301-8. 24. Dennis B, Ernst N, Hjortland M, Tillotson J, Grambsch V, The NHLBI nutrient data systems. J Am Diet Assoc. 1980;77:641-7. 25. Jackson AS, Pollock ML. Practical assessment of body composition. Phys Sports Med. 1985; 13:56. 26. Chrousos GP, Schulte HM, Oldfield EH, et al. Corticotropinreleasing factor: basic and clinical studies. Physchopharmacol Bull. 1983;19:416-21. 27. Gold PW, Chrousos G, Kellner C, et al. Psychiatric implications of basic and clinical studies with corticotropin-releasing factors. Am J Psychiatry. 1984;141:619-27. 28. Nicholson WE, Davis DR, Sherrel BJ, Orth DN. Rapid radioimmunoassay for corticotropin in unextracted human plasma. Clin Chem. 1984;30:259-65. 29. Kaye WH, Gwirtsman HE, George DT, et al. Elevated cerebrospinal fluid levels of immunoreactive corticotropin-releasing hormone in anorexia nervosa: relation to state of nutrition, adrenal function, and intensity of depression. J Clin Endocrinol Metab. 1987;64:2038. 30. Villaneuva AL, Schlosser C, Hopper B, Liu JH, Hoffman DI, Rebar RW. Increased cortisol production in women runners. J Clin Endocrinol Metab. 1986;63:133-6. 31. Hohtari H, Elovainio R, Salminen K, Laatikainen T. Plasma corticotropin-releasing hormone, corticotropin, and endorphins at rest and during exercise in eumenorrheic and amenorrheic athletes. Fertil Steril. 1988;50:233-8. 32. Ding JH, Sheckter CB, Drinkwater BL, Soules MR, Bremner WJ. High serum cortisol levels in exercise-associated amenorrhea. Ann Intern Med. 1988;108:530-4. 33. Luger A, Deuster PA, Kyle SB, et al. Acute hypothalamic-pituitaryadrenal responses to the stress of treadmill exercise. N Engl J Med. 1987;316:1309-15. 34. Cushing H. The basophil adenomas of the pituitary body and their clinical manifestations. Bull Johns Hopkins Hosp. 1932;50:137-95. 35. Sakakura M, Takebe K, Nahagawa S. Inhibition of luteinizing hormone secretion induced by synthetic LRH by long-term treatment with glucocorticoids in human subjects. J Clin Endocrinol Metab. 1975;40:774-9. 36. Cunningham GR, Caperton Jr EM, Goldzieher JW. Antiovulatory activity of synthetic corticoids. J Clin Endocrinol Metab. 1975;40:265-7. 37. DeSouza EB, Insel TR, Perrin MH, Rivier J, Vale WW, Kuhar MJ. Corticotropin-releasing factor receptors are widely distributed within the rat central nervous system: an autoradiographic study.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
CORTISOL SECRETION IN HYPOTHALAMIC AMENORRHEA J Neurosci. 1985;5:3189-202. 38. Symington T, Duguid WP, Davidson JN. Effect of exogenous corticotropin on the histochemical pattern of the human adrenal cortex and a comparison with the changes during stress. J Clin Endocrinol Metab. 1955;16:580-98. 39. Meyerhoff JL, Oleshansky MA, Mougey EH. Psychologic stress increases plasma levels of prolactin, cortisol, and POMC-derived peptides in man. Psychosom Med. 1988;50:295-303. 40. Gold PW, Loriaux DL, Ray A, et al. Responses to corticotropinreleasing hormone in the hypercortisolism of depression and Cushing's disease. N Engl J Med. 1986;4:1329-35. 41. Schulte HM, Chrousos GP, Gold PW, et al. Continuous administration of synthetic ovine corticotropin-releasing factor in man. J Clin Invest. 1985;75:1781-5. 42. Kapen S, Sternthal E, Braverman L. A "pubertal" 24 hour luteinizing hormone (LH) secretory pattern following weight loss in the absence of anorexia nervosa. Psychosom Med. 1981;43:177-82. 43. Vigersky RA, Andersen AE, Thompson RH, Loriaux DL. Hypothalamic dysfunction in secondary amenorrhea associated with simple weight loss. N Engl J Med. 1977;297:1141-5. 44. Warren MP, Jewelewicz R, Dyrenfurth I, Ans R, Khalaf S, Vande Wiele RL. The significance of weight loss in the evaluation of pituitary response to LH-RH in women with secondary amenor-
317
rhea. J Clin Endocrinol Metab. 1975;40:601-ll. 45. Fehm HL, Holl R, Spath-Schwalbe E, Born J, Voight KH. Ability of corticotropin releasing hormone to stimulate cortisol secretion independent from pituitary adrenocorticotropin. Life Sci. 1988;42:679-86. 46. Dave JR, Eiden LE, Eskay RL. Corticotropin-releasing factor binding to peripheral tissue and activation of the adenylate cyclaseadenosine 3',5'-monophosphate system. Endocrinology. 1985;116: 2152-9. 47. Evans CHJ, Erdelyi E, Weber E, Barchas JD. Identification of proopiomelanocortin-derived peptides in the human adrenal medulla. Science. 1983;221:957-60. 48. Smith EM, Meyer JW, Blalock JE. Virus-induced corticosterone in hypophysectomized mice: a possible lymphoid-adrenal axis. Science. 1982;218:1311-2. 49. Blalock JE, Smith EM, Meyer WJ. The pituitary-adrenocortical axis and the immune system. Clin Endocrinol Metab. 1985;14: 1021-38. 50. Vale W, Rivier C. Substances modulating the secretion of ACTH by cultured anterior pituitary cells. Fed Proc. 1977;36:2094-99. 51. DeBold CR, Sheldon WR, DeCherney GS, et al. Arginine vasopressin potentiates adrenocorticotropin release induced by ovine corticotropin-releasing factor. J Clin Invest. 1984;73:533-8.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 April 2015. at 12:06 For personal use only. No other uses without permission. . All rights reserved.
