The effect of repeated human corticotropin-releasing hormone administration on dexamethasone-suppressed pituitary—adrenocortical activity in healthy subjects

The Effect of Repeated Human CorticotropinReleasing Hormone Administration on Dexamethasone-Suppressed Pituitary-Adrenocortical Activity in Healthy Subjects Klaus Wiedemann and Florian Holsboer

A dexamethasone suppression test (DST) using a dosage of 1.5 mg dexamethasone was administered two times in randomized order" to 10 healthy male subjects. From 2300 hours to 0700 hours subjects were injected repeatedly with either increasing dosages of human corticotropin-releasing hormone (hCRH) or 0.9% saline. In comparison to saline administration, in which cortisol levels remained suppressed, the time course of cortisol concentrations with hCRH stimulation showed a biphasic secretory pattern. According to a criterion level of a minimum of 40 ng/mL plasma for nonsuppression, the majority of the subjects changed their DST status to nonsuppression with hCRH. Adrenocorticotropic hormone secretion also differed significantly between saline and hCRI-1 administration. During stimulation with hCRH, plasma dexamethasone levels were slightly and nonsignificantly reduced in the morning hours. Our results indicate that repeated dosages of hCRH impair the dexamethasone-induced suppression in man and support an involvement of CRH also in mediation of the DST nonsuppression during depre:~sive illness. © 1997 Socie~ of Biological Psychiatry.

Key Words: Corticotropin-releasing hormone, dexamethasone, adrenocorticotropic hormone, cortisol, depression, pituitary B~OL PSYCHIATRY 1997;42;882--888

Introduction The dexamethasone suppression test (DST) has been widely used as a dynamic test of hypothalamic-pituitaryadrenocortical (HPA) dysfunction, especially in patients with depression, in whom HPA alterations are :frequently observed (Carroll et al 1981; Holsboer et al 1982). Today From the Max Planck Institute of Psychiatry, Munich, Germany. Address reprint requests to Dr. Klaus Wiedemann, Max Planck Institute of Psychiatry, Kraepelinstrasse 2-10, 80804 Munich, Germany. Received May 22, 1995; revised September 18, 1996.

© 1997 Society of Biological Psychiatry

the DST and a combination of the DST with subsequent stimulation of cortisol by human corticotropin-releasing hormone (hCRH) are used as neuroendocrine evaluations of an altered stress system activation that may be useful in predicting the patient's clinical course (Holsboer-Trachsler et al 1991; Heuser et al 1996). From clinical studies applying the major regulatory hormone of the pituitaryadrenocorlical axis, hCRH (Vale et al 1981; Spiess et al 1981), a :~uprapituitary mechanism for the observed disturbances has been supported, because cortisol hypersecretion has been associated with a blunted adrenocortico00116-3223/97/$17.00 PII S0006..3223(96100434-9

CRH Administration during DST

tropic hormone (ACTH) response to human or ovine CRH in depressed patients (Holsboer et al 1984b; Gold et al 1984). If these patients are pretreated with dexamethasone, hCRH administration at 1500 hours evokes an escape of plasma cortisol secretion from the dexamethasone-induced suppression of pituitary-adrenocortical activity, whereas in healthy controls the same paradigm does not produce elevated plasma cortisol concentrations (von Bardeleben et al 1985; Heuser et al 1994; Holsboer et al 1987; Holsboer-Trachsler et al 1991). These findings point to a role of CRH in the mediation of the DST nonsuppression phenomenon in patients. Moreover, in patients showing DST nonsuppression, a close association of elevated cortisol levels and reduced plasma dexamethasone levels has been observed (Holsboer et al 1984a, 1986a; Arana et al 1984). This finding led first to the conclusion that inappropriately low dexamethasone levels during the DST are the primary cause of DST nonsuppression. Recent studies investigating dexamethasone pharmacokinetics revealed that the accelerated elimination of dexamethasone appears to be a related phenomenon of DST nonsuppression in depression, because no differences between plasma dexamethasone levels of DST suppressors and nonsuppressors during the resorption and early distribution phase were obvious (Holsboer et al 1986b; Wiedemann and Holsboer 1987, 1990). Therefore, the period of time around midnight is assumed to be the critical period for the suppressive effects of dexamethasone. Thus, it is necessary to investigate the effects of hCRH administration on DST outcome to test the hypothesis whether hCRH itself has an influence on dexamethasone-suppressed pituitaryadrenocortical activity, the pharmacokinetics of dexamethasone, and the DST outcome. As in former experiments, an experimental paradigm was chosen, in which daytime stimulation by repeated administration of 10 lxg hCRH for 10 hours was used, so that the circadian rhythm of cortisol remained preserved (Wiedemann et al 1991). Because the frequency of injections could not be increased due to interference with blood sampling, the repeated dosages were increased to imitate the exaggerated CRH secretion that occurs in depressed patients. The study was conducted to clarify whether hCRH administration during a conventional DST alters ACTH and cortisol secretion or dexamethasone kinetics and also whether it determines the DST outcome.

Methods

Subjects Ten healthy male subjects from 25 to 31 years of age and weighing 6 8 - 8 4 kg were studied after a thorough medical examination including endocrine evaluation of thyroid and

BIOLPSYCHIATRY 1997:42:882--888

883

adrenocortical function. The subjects had not taken any medication for at least 3 months before admission to the study and they had also refrained from alcohol and were not sIeep deprived. Drug abuse was ruled out in all subjects by screening of urine samples. The study protocol was approved by the Ethical Committee of the Max Planck Institute, and written informed consent was obtained from each subject prior to starting the investigation.

Investigational Procedure All subjects remained under video observation in a singlebed, soundproof room. Starting at 2100 hours an intravenous cannula was inserted into a forearm vein and connected with a long catheter placed through a soundproof lock 1:o the adjacent laboratory. At 2300 hours 1.5 mg dexamethasone (E. Merck, Darmstadt, FRG) was administered orally. Lights were turned off at 2300 hours, and all subjects were allowed to sleep until the next morning. Blood samples were drawn every 30 rain :from :2200 hours until 0200 hours, then every 60 rain until 0900 hours, and also once at 1600 hours. All subjects participated in two treatments separated by a time interval of 1 week. During the treatments they received in randomized order: 1. Repeated saline injections (l mL 0.9% saline); or 2. Repeated hCRH injections (Clinalfa, L~tufelfingen, Switzerland) according to the following schedule: 2300 hours: 150 ng hCRH/kg body weight (BW) 2400 hours: 150 ng hCRH/kg BW 0100 hours: 150 ng hCRH/kg BW 0200 hours: 300 ng hCRH/kg BW 0300 hours: 450 ng hCRH/kg BW 0400 hours: 600 ng hCRH/kg BW 0500 hours: 750 ng hCRH/kg BW 0600 hours; 900 ng hCRH/kg BW 0700 hours; 1 p~g hCRH/kg BW In total, 4.45 lxg/kg BW was administered during 1 night (BW varied between 68 and 84 kg, and the total amount of hCRH injected ranged from 303 to 374 Ixg).

Assays Blood samples for cortisol and dexamethasone determinations were heparinized (125 IU NA/K-heparin/mL blood), and for determination of ACTH immunoreactivity ethylenediaminetetraacetic acid (EDTA) (1 mg/mL blood; E. Merck, Darmstadt, FRG) and kallikrein inhibitor (350 KIU/mL blood; Bayer, Leverkusen, FRG) were added to native blood samples. All blood samples were immediately centrifuged at 4000 g, and then the plasma was frozen and stored at - 8 0 ° C until analysis. For cortisol

884

BIOLPSYCHIATRY

K. Wiedemann and F. Holsboer

1997;42:882-888

analysis a commercial radioimmunoassay kit was employed (ICN Biomedicals Inc., Carson, CA). The sensitivity was 7 ng/mL plasma, and intra- and interassay coefficients of variation were below 5%. The conversion factor from ng/mL to nmol/L is 0.276. ACTH was measured using a commercially available immunoradiometric assay (Nichols Institute, San Juan Capistrano, CA). The sensitivity was 4 pg/mL plasma, and intra- and interassay coefficients of variation were below 8%. The conversion factor from pg/mL to pmol/L is 0.2202. Plasma dexamethasone was measured using a method already described in detail (Wiedemann and Holsboer 1987). The antiserum was raised against dexamethasone C-oxime; the cross-reactivities of this antiserum against cortisol and corticosterone are 0.1% and 0.2%, respectively. The radiolabeled tracer was 6,7-H3-dexamethasone. The minimum detectable amount was 0.8 ng/mL plasma, and the intra- and interassay coefficients of variation were below 8% and 9%, respectively.

140 120 -- 100 ,s ]5 O

o

80 60 40

20 22'00' 24'00' 02'00' 04'00' 06'00' 08'00' I 0'00' 12'00' 14'00'I 6'00 Time (h) I [] Saline • hCRH

900 (I)

750 (:D

Statistical Methods For hormonal secretion of plasma cortisol, ACTH, and dexamethasone concentrations, mean locations (ML) and area under the curve (AUC; calculated according to the trapezoid rule) during the whole night and in the time intervals 2 3 0 0 - 0 4 0 0 hours and 0 4 0 0 - 0 9 0 0 hours were used for confirmatory statistical analysis. For comparisons of AUC and ML between the two treatments in the whole time interval a one-factorial analysis of variance (ANOVA) with repeated-measures design was applied. "Treatment" was the within-subjects factor with two levels corresponding to "saline" and °'hCRH." To compare AUC and ML values between the two treatments and the two intervals ( 2 3 0 0 - 0 4 0 0 hours and 0 4 0 0 - 0 9 0 0 hours), a repeated-measures A N O V A with two within-subjects factors (treatment and time) was performed. For plasma dexamethasone in addition to the AUC and ML variables the plasma concentrations at the time points 0100, 0300, 0500, 0700, and 0900 hours were also compared between the treatments by means of an A N O V A with repeatedmeasures design. To keep the type I error -< alpha = .05 (nominal level of significance), all hypotheses were tested at a corrected level of significance (Bonferroni correction).

Results Effects of hCRH on Postdexamethasone Cortisol Secretion According to our normal database we selected a criterion level for DST nonsuppression of a minimum of 40 ng/mL plasma cortisol at 800 hours (Holsboer et al 1986a). During saline administration all subjects were DST sup-

y

'~

600

'/// "/// "/// ¢//1

45o

"///

300

"///

150

//A -

0

Saline I

hCRH

~ 2300- 0900 h w72"21 2300 - 0400 h 0400 - 0900 h

Figure 1. Mean plasma cortisol levels with saline (()pen squares) and hCRH stimulation (filled squares) for all 10 subjects. T-bars indicate 1 SD. The investigation was started at 2200 hours; 1.5 mg dexamethasone was administered at 2300 hours. Arrows indicate the time points of hCRH injections. The mean plasma concentrations differed significantly from 2330 hours to 0200 hours and from 0500 hours to 0900 hours. The lower part indicates the area values (see Results; expressed as arbitrary units) for saline and hCRH conditions in lhe three ti:me periods. pressors and showed a gradual decline of plasma cortisol concentrations from 2200 hours to 0900 hours (Figure 1). After administration of hCRH, 7 out of 10 subjects changed their DST status to DST nonsuppression. The time course of mean plasma cortisol levels postdexamethasone during stimulation with hCRH showed a biphasic secretory pattern (Figure 1), with a steep increase at 2330 hours immediately after administration of minute amounts of hCRH, with a nadir at 0400 hours despite increased hCRH dosages, and then with a second maximum at 0800 hours. Seven subjects showed an increase of

CRH Administration during DST

E3JOI~PSYCHIATRY

885

1997;42:882 888

cortisol at 0800 hours above 40 ng/mL; also 7 out of 10 subjects displayed an increase at 2330 hours, and 6 subjects showed an increase at both time points. The mean peak height of the two maxima did not differ significantly (first maximum at 2330 hours: 68.0 ng/mL vs. second maximum at 0800 hours: 73.7 ng/mL); however, the amplitude in comparison to saline conditions was greater at 0800 hours than at 2330 hours (67.0 ng/mL vs. 37.9 ng/mL). AUC and ML values for cortisol from 2300 hours to 0900 hours differed significantly between the two test occasions [Wilk's multivariate test of significance; treatment effect: approximately F(2,8) = 31.15, p < .0001; univariate F tests for the simple effects of AUC and ML in the interval 2 3 0 0 - 0 9 0 0 hours: minimum F values with (1,9) df = 8.71, p < .016]. AUC and ML values for cortisol in the time periods from 2300 to 0400 hours (nadir; end of first half of the night) and from 0400 to 0900 hours (second half of the night) also differed significantly between the two occasions [Wilk's multivariate test of significance; treatment effect: approximately F(2,8) = 7.81, p = .013]. Both variables contributed significantly to the treatment effect [univariate F test for simple effects: minimum F value with (1,9) df = 17.09, p .< .0031].

16

14 E ~b

'-r" [-. O <

During saline plasma ACTH plasma concentrations showed a regular decline until morning (Figure 2). After administration of hCRH, an increase in plasma ACTH concentrations during the morning hours was observed, in contrast to the cortisol secretory profile after 2300 hours, only one marked secretory peak was detected, and it preceded the second cortisol surge. Around midnight no significant treatment effect was observable, and in both conditions plasma A C T H concentrations varied between 9 and 12 pg/mL. Both AUC and ML values for ACTH secretion from 2300 to 0900 hours differed significantly between the two test occasions [Wilk's multivariate test: of significance: treatment effect: approximately F(2,8) = 8.73, p = .01; univariate F tests for the simple effects of both AUC and ML values: minimum F value with (1,9) df = 19.04, p < .002]. Comparison of the AUC and ML values between the two treatments and the two intervals 2300--0400 hours and 0 4 0 0 - 0 9 0 0 hours with A N O V A revealed that both variables contributed significantly to a main treatment effect as well as to a time by treatment interaction effect [Wilk's multivariate test of significance; treatment effect: approximately F(2,8) = 10.80, p = .005; time by treatment effect: approximately F(2,8) = 6.05, p = .0251]. Separate analysis of the simple treatment effects in each

10

22~00' 24~00'02'00' 04'00' 06~00' 08'00' 10'03' 12'00' 14'00'160C Time (h) ]

U Saline

IhCRl-'

I

® 220 t

~ 18o] "o c

c

Effects of hCRH Stimulation on Postdexamethasone ACTH Secretion

12

140

100

zs

60

Saline

hCRH

r77"1 2300 - 0900 h v///4 2300 - 0400 h 0400 - 0900 h Figure 2. Mean plasma ACTH levels with saline (open squares) and hCRH stimulation (filled squares) for all 10 subjects. T-bars indicate 1 SD. The investigation was started at 2200 hours. Arrows indicate the time points of hCRH injections. The mean plasma concentrations differed significantly from 0500 hours to 0900 hours (p < .05). The lower part indicate~,; the area values (see Results; expressed as arbitrary units) tbr saline and hCRH conditions in the three time periods. time interval revealed significant differences between the two treatments only in the interval 0 4 0 0 - 0 9 0 0 hours [univariate F tests; minimum F value with (1,9) df = 17.09, p < .003].

Effects of hCRH Stimulation on Plasma Dexamethasone Levels The time courses of plasma dexamethasone levels showed a slight difference between the two test occasions with a reduction of plasma dexamethasone levels during hCRH stimulation beginning at 0500 hours (Figure 3), but neither

886

BIOLPSYCHIATRY

K. Wiedemann and F. Holsboer

t997;42:882-888

24

o

|

|

i

i

i

i

1

i

i

|

i

|

i

i

i

i

i

|

t

2200 2400 0200 0400 0600 0800 1000 1200 1400 1600 Time (h)

[

O Saline I hCRH

I

240

o

180

~E

120

=

60

Saline

hCRH

2300-0900 h ~//A 2 3 0 0 - 0 4 0 0 h I 0400-0900 h v/J

Figure 3. Mean plasma dexamethasone concentrations with saline (open squares) and hCRH (filled squares) stimulation. T-bars indicate 1 SD. At 2300 hours dexamethasone was administered. Arrows indicate the time points of hCRH injections. The lower part indicates the area values (see Results; expressed as arbitrary units) for saline and hCRH conditions in the three time periods.

the mean concentrations at 0100, 0300, 0500, 0700, and 0900 hours nor the AUC or ML values from 231)0 to 0900 hours differed significantly between the two test occasions. At 0700 hours cortisol and dexamethasone levels with hCRH stimulation showed a trend in the direction of a negative correlation (r = -.56, p = .093, Spearman's rank correlation).

Discussion The major findings of our investigation are that repeated intravenous administration of hCRH during a conventional

DST procedure impairs the suppressive action of dexamethasone on ACTH and cortisol secretion. Around midnight, a pronounced release of cortisol without a preceding ACTH surge occurred. The dexamethasone elimination was accelerated in some of the subjects; however, no significant differences between the saline and hCRH conditions were found. The large cortisol increase at midnight after the injection of minute amounts of hCRH in the absence of comparable increases in plasma ACTH concentrations suggests a direct adrenal effect of CRH. Extracts of human adrenals have revealed the existence of ACTH and proopiomelanocorticotropin-derived peptides in the medulla (Evans et al 1983), and it has been reported that short loop interactions between the medulla and the cortex may modulate the secretion of cortisol from the cortex (Charlton 1990). Sympathetic nerve effects could also influence the release of adrenocortical steroids without a preceding pituitary ACTH surge. Stimulation of the splanchnic nerve in conscious calves results in an abrupt increase in adrenal cortisol output that is preceded by a release of CRH from adrenal glands (Edwards and Jones 1988). Another explanation, however, might depend on one potential weakness of our study that the frequency of blood sampling every 30 min was too low, and therefore short-term increases of ACTH were not seen. The most plausible explanation for the hCRH effects (luring the morning hours and the concomitant release of ACTH and cortisol is that the suppression of endogenously secreted CRH is compensated by the repeated injections of intravenous hCRH; however, other neuropeptides play a role as synergizing secretagogues for ACTH as well. CRH is the major regulatory hormone of the HPA system, but administration of vasopressin also stimulates and modulates the release of ACTH in man, such as in isolated pituitaries and anterior pituitary cells (Gillies and Lowry 1982; Brostoff et al 1968; Salata et al 1988; Fleischer and Vale 1968; Kjaer 1993). Vasopressin has been demonstrated to override the dexamethasone-induced suppression if it is administered concomitantly with CRH 1700 hours after dexamethasone (yon Bardeleben et al 1985). From studies in rats it has been concluded that the negative :feedback of dexamethasone on ACTH secretion of ACTH is produced mainly by a reduction of the vasopressin secretion and only to a lesser degree by its effect on CRH release (Tsukuda et al 1983). Therefore, the feedback inhibition of the HPA system by dexamethasone may be mediated partially by a vasopressin suppression. In man, the time around midnight defines the period of maximum sensitivity to the effects of dexamethasone (Nichols et al 1965; Krieger et al 1971). In the daytime, larger dosages of dexamethasone are needed to obtain

CRH Administration during DST

BIOLPSYCHIATRY

887

[997;42:882--888

comparable suppressive effects. This may be possibly related to circadian changes of corticosteroid receptors at different feedback levels. Because of its limited binding to the corticosteroid-binding globulin, dexamethasone acts mainly at pituitary glucocorticoid receptors (Miller et al 1992). Therefore, circadian variations of corticosteroid receptor concentrations in the brain (De Kloet and Reul 1987) are unlikely to account for these effects. In this context it seems noteworthy that CRH prevents the early onset of glucocorticoid feedback by corticosterone in rat anterior pituitary tissue (Shipston and Antoni 1992). The dexamethasone pharmacokinetics observed in our study also support an involvement of this mechanism because the first hCRH bolus was injected 4 hours before the plasma dexamethasone levels reached their maximum. Therefore, this early interaction between hCRH and dexamethasone at the pituitary level appears to be responsible for the observed escape phenomenon. Comparisons of the dexamethasone half-life in depressed patients revealed that DST nonsuppressors show an accelerated elimination of dexamethasone from the circulation in comparison with DST suppressors (Wiedemann and Holsboer 1987, 1990; Maguire et al 1987; Johnson et al 1987). Because in these studies no differences were obvious during the resorption and early distribution phase of DST suppressors and nonsuppressors, it was suggested that the acceleration in dexamethasone half-life would be a coherent phenomenon of the underlying endocrine disturbance. This assumption is supported by the observation that rapid changes of the [)ST suppres-

sor status are accompanied by simultaneous changes in dexamethasone kinetics (Holsboer-Trachsler et al 1988). Injection of cortisol acetate to rabbits results in a sustained hypercortisolemia with a simultaneous acceleration of dexamethasone kinetics (Stoll et al 1990). Our present observations point in the same direction; however, no significant alterations in the dexamethasone kinetics could be found. Whereas earlier studies indicate that the pharmacodynamic effects of dexamethasone (Wiedemann and Holsboer 1987; Gupta et al 1992) are temporally distinct from its plasma kinetics, we observed a transient suppressive effect of dexamethasone in all subjects around 0400 hours that coincides with the maximum plasma concentrations. This observation may possibly be related to direct inhibitory effects of dexamethasone on the corticotrophs of the pituitary (Childs et al 1986). Increasing the dosage of hCRH can overcome this effect, and A C T H and cortisol levels can be increased to concentrations regularly observed in subjects without dexamethasone. Our results confirm that the HPA system underlies circadian changes in its sensitivity not only against dexamethasone but also against hCRH, which appears to also play an important role in the mediation of the DST nonsuppression phenomenon. An endogenous hypersecretion of CRH in conjunction with other corticotrophic secretagogues, such as vasopressin (Purba e t al 1996), may explain why a bolus injection of 100 I~g hCRH is effective in overcoming the DST suppression in patients but not in health)' controls (Holsboer et al 1992; Heuser et al 1996).

References Arana AA, Workman RJ, Baldessarini RJ (1984): Association between low plasma levels of de×amethasone cortisol in psychiatric patients given dexamethasone. Am J Psychiatry 141:1619. Brostoff J, James VHT, Landon J (1968): Plasma corticosteroid and growth hormone response to lysine vasopressin in man. J Clin Endocrinol Metab 28:511-518. Carroll BJ, Feinberg M, Greden JF, et al (1981): A specific laboratory ,test for the diagnosis of melancholia. Arch Gen Psychiatry 38:15-22. Charlton B (1990): Adrenal cortical innervation and glucocorticold secretion. J Endocrinol 126:5-8. Childs G, Morell JL, Niendorf A, Aguilera G (1986): Cytochemical studies of corticotropin-releasing factor (CRF) receptors in anterior lobe corticotropes: Binding, glucocorticoid regulation, and endocytosis of [Biotinyl-Serl]CRF. Endocrinology 119:2129-2142, De Kloet ER, Reul JMHM (1987): Feedback action and tonic influence of corticosteroids on brain function: A concept arising from the heterogeneity of brain receptor systems. Psychoneuroendocrinology 12:83-105. Edwards AV, Jones CT (1988): Secretion of corticotropin

releasing factor from the adrenal during splanchnic nerve stimulation in conscious calves. J Physiol 400:89--100. Evans CJ, Erdelyi E, Weber E, Barchas JD (1983): Identification of proopiomelanocortin-derived peptides in the human adrenal rnedulla. Science 221:957-960. Fleischer N, Vale W (1968): Inhibition of vasopressin-induced ACTH release from the pituitary by glucocorticoids in vitro. Endocrinology 83:1232-1236. Gillies GE, Lowry PJ (1982): Corticotropin-releasing hormone and its vasopressin component. Front Neuroendocrinol 7:4575. Gold PW, Chrousos GP, Kellner CH, et al (1984): Psychiatric implications of basic and clinical studies v,ith corticotropinreleasing factor. Am J Psychiatry 141:619--627. Gupta SK, Ritchie JC, Ellinwood EH, Wiedemann K, Holsboer F (1992): Modeling the pharmacokinetics and pharmacodynamics of dexamethasone in depressed patients. Eur J Clin Pharmacol 43:51-55. Heuser 1, Yassouridis A, Holsboer F (1994): The combined dexamethasone/CRH test: A refined laboratory test for psychiatric disorders. J Psychiatr Res 28:341-356. Heuser IJE, Schweiger U, Gotthardt U, et al (t996): Pituitary-

888

BIOL PSYCHIATRY 1997;42:882--888

adrenal-system regulation and psychopathology during amitriptyline treatment in elderly depressed patients and normal comparison subjects. Am J Psychiat©' 153:93-9cL Holsboer F, Liebl R, Hofschuster E (1982): Repeated dexamethasone suppression test during depressive illness. J Affect Disord 4:93-101. Holsboer F, Haack D, Gerken A, Vecsei P (1984a): Plasma dexamethasone concentrations and differential glucocorticoid suppression responses in depressives and controls. Biol Psvchiatr3" t 9:281-291. Holsboer F, yon Bardeleben U, Gerken A, Stalla GK, Mtiller OA (1984b): Blunted corticotropin and normal cortisol response to human corticotropin-releasing factor (h-CRF) in depression. N Engl J Med 311:1127. Holsboer F, Philipp M, Steiger A, Gerken A (1986a): Multisteroid analysis after DST in depressed patients--A controlled study. J A~'ect Disord 10:241-249. Holsboer F, Wiedemann K, Boll E (1986b): Shortened dexamethasone half-life in depressed dexamethasone nonsuppressors. Arch Gen Psychiato, 43:813- 815. Holsboer F, von Bardeleben U, Wiedemann K, Mtiller OA, Stalla GK (1987): Serial assessment of corticotropin-releasing hormone response after dexamethasone in depression--Implications for pathophysiology of DST nonsuppression. Biol Psychiatry 22:228 234. Holsboer F, Spengler D, Heuser I (1992): The role of corticotropin-releasing hormone in the pathogenesis of Cushing's disease, anorexia nervosa, alcoholism, affective disorders and dementia. Prog Brain Res 93:385-417. Holsboer-Trachsler E, Wiedemann K, Holsboer F (1988): Serial partial sleep deprivation in depression--Clinical effects and dexamethasone suppression test results. Neuropsychobiology 19:73-78. Holsboer-Trachsler E, Stohler R, Hatzinger M (1991 ): Repeated administration of the combined dexamethasone/hCRH stimulation test during treatment of depression. Psychiato, Res 38:163-171. Johnson GF, Hunt G, Kerr K, Caterson I (1987): The plasma dexamethasone "window" and the dexamethasone suppression test in depression. Biol Psychiato, 22:1033-1035. Kjaer A (1993): Vasopressin as a neuroendocrine regulator of anterior pituitary hormone secretion. Acta Endocrinol 129: 489-496. Krieger DT, Allen W, Rizzo F, Krieger HP (1971): Characterization of the normal temporal pattern on plasma corticosteroid levels. J Clin Endocrinol 32:266-284. Maguire KP, Schweitzer I, Biddle N, Bridge S, "Filler JWG

K. Wiedemann and F. Holsboer

(1987): The dexamethasone suppression test: Importance of dexamethasone concentrations. Biol Psychiatry 22:957-967. Miller AH, Spencer RL, Pulera M, Kang S, McEwen BS, Stein M (1992): Adrenal steroid receptor activation in rat brain and pituitary following dexamethasone: hnplications for the dexamethasone suppression test. Biol Psychiat~. 32:850--869. Nichols T, Nugent CA, Tyler FH (1965): Diurnal variation in suppression of adrenal function by glucocorticoids. J Clin Endocrinol 25:343-349. Purba JS, Hoogendijk WJG, 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~ Salata RA, Jarrett DB, Verbalis JG, Robinson AG (1988): Vasopressin stimulation of adrenocorticotrepin hormone (ACTIq[) in humans. J Clin lnvest 81:766-774. Shipston ME, Antoni FA (1992): Inactivation of early glucocorticoid feedback by corticotropin-releasing factor in vitro. Endocrinology 130:2213-2218. Spiess J, Rivier J, Rivier C, Vale W (1981): Primary structure of corticol:ropin-releasing factor from ovine hypothalamus. Proc Natl Acad Sci USA 78:6517-6521. Stoll PM, Schluger JH, Stokes PE (1990): Sustained hypercortisolemia decreases plasma dexamethasone half-life in rabbits. Biol P,sychiato' 27:170A. Tsukuda T, Nakai Y, Koh T, Tsujii S, Imura H 11983): Plasma adrenocorticotropin and cortisol responses to intravenous injection of corticotropin-releasing factor in the morning and evening. J Clin Endocrinol Metab 57:869-871. 'gale W, Spiess J, Rivier C, Rivier J (1981): Chm;acterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and 13-endorphin. Science 213: 1394-1397. yon Bardeleben U, Holsboer F, Stalla GK, Mtiller OA (1985): Combined administration of human corticotropin-releasing factor and lysine vasopressin induces cortisol escape from dexamethasone suppression in healthy subjects. Life Sci 37:16113-1618. Wiedemann K, Holsboer F (1987): Plasma dexamethasone kinetics during the DST after oral and intravenous administration of the test drug. Biol Psychiatr)" 22:1340--1348. Wiedemann K, Holsboer F (1990): The effect of dosage upon plasma cortisol and dexamethasone during the DST. J Affect Disord 19:133-137. Wiedemann K, yon Bardeleben U, Holsboer F (1991): Influence of human corticotropin-releasing hormone and adrenocorticotropin upon spontaneous growth hormone secretion. Neuroendo,:rinology 54:462-468.