- Email: [email protected]
Desmopressin augments pituitary–adrenal responsivity to corticotropin-releasing hormone in subjects with chronic fatigue syndrome and in healthy volunteers
Desmopressin Augments Pituitary–Adrenal Responsivity to Corticotropin-Releasing Hormone in Subjects with Chronic Fatigue Syndrome and in Healthy Volunteers Lucinda V. Scott, Sami Medbak, and Timothy G. Dinan Background: Corticotopin-releasing hormone (CRH) and vasopressin (VP) are the two principal neuropeptide regulators of the hypothalamic–pituitary–adrenal axis in man, with VP serving to augment CRH-induced adrenocorticotropic hormone (ACTH) release. Unlike VP, desmopressin (DDAVP), which is a synthetic analogue of VP, when administered alone, has not been shown in healthy subjects to have consistent ACTH-releasing properties. It has been suggested that chronic fatigue syndrome (CFS), characterized by profound fatigue and a constellation of other symptoms, may be caused by a central deficiency of CRH. Methods: We administered 100 mg ovine CRH (oCRH) and 10 mg DDAVP, both alone and in combination, to a group of subjects with CFS, and to a group of healthy volunteers. Our aim was to establish the effect of DDAVP on CRH-induced ACTH release in these two groups. Results: The d-ACTH responses to oCRH were attenuated in the CFS (21.0 6 4.5 ng/L) compared to the control subjects (57.8 6 11.0 ng/L; t 5 3.2, df 5 21, p , .005). The d-cortisol responses were also reduced in the CFS (157.6 6 40.7 nmol/L) compared to the healthy subjects (303.5 6 20.9 nmol/L; t 5 3.1, df 5 21, p , .01). The d-ACTH and d-cortisol responses to DDAVP alone did not differ between the two groups. On administration of both CRH and DDAVP no response differences between the two groups for either ACTH (p 5 .3) or cortisol output (p 5 .87) were established. Comparing the ACTH and cortisol responses to CRH and CRH/DDAVP in only those individuals from each group who had both tests, the cortisol output to the combination was significantly greater in the CFS compared to the healthy group. The ACTH output was also increased in the former group, though this was not significant. From the Department of Psychiatry, Trinity College Medical School, St. James’ Hospital, Dublin, Ireland (LVS); Department of Chemical Pathology, Queen Alexandra Hospital, Portsmouth, United Kingdom (SM); and Department of Psychiatry, Royal College of Surgeons in Ireland, St. Stephen’s Green, Dublin, Ireland (TGD). Address reprint requests to Professor Ted Dinan, Department of Psychiatry, Royal College of Surgeons in Ireland, St. Stephen’s Green, Dublin 2, Ireland. Received December 1, 1997; revised June 22, 1998; accepted June 26, 1998.
© 1999 Society of Biological Psychiatry
Conclusions: DDAVP augments CRH-mediated pituitary–adrenal responsivity in healthy subjects and in patients with CFS. That DDAVP was capable of normalizing the pituitary–adrenal response to oCRH in the CFS group suggests there may be increased vasopressinergic responsivity of the anterior pituitary in CFS and/or that DDAVP may be exerting an effect at an adrenal level. Biol Psychiatry 1999;45:1447–1454 © 1999 Society of Biological Psychiatry Key Words: Chronic fatigue syndrome, corticotropinreleasing hormone, vasopressin, desmopressin, hypothalamic–pituitary–adrenal axis, adrenocorticotropin, cortisol
Introduction
C
orticotropin-releasing hormone (CRH) is the primary regulator of the hypothalamic–pituitary–adrenal axis (HPA) in man. Its main companion regulator is vasopressin (VP), a nonapeptide, produced from the parvicelluar neurons of the paraventricular nucleus (PVN), and the magnocellular neurons of the supraoptic nucleus (Antoni 1993). VP plays a role as an adrenocorticotropic hormone (ACTH)-releasing factor at the recently cloned V1b receptor on the anterior pituitary (Sugimoto et al 1994). Although CRH has been shown to be a more potent secretagogue of ACTH than VP, when VP is administered alone it has significant ACTH-releasing properties in humans (Brostoff et al 1968; Salata et al 1988). The coadministration of the two peptides causes an augmented release of ACTH (Liu et al 1983; DeBold et al 1984; Lamberts et al 1984); this synergism may be associated with a VP-mediated enhancement of CRH cyclic adenosine monophosphate production at the level of the corticotrope (Giguere et al 1982). Advances in our understanding of stress physiology have been facilitated by the availability of both ovine and human forms of the neuropeptide CRH as exogenous activators of the HPA. Abnormalities of CRH-induced 0006-3223/99/$20.00 PII S0006-3223(98)00232-7
1448
L.V. Scott et al
BIOL PSYCHIATRY 1999;45:1447–1454
ACTH release have been demonstrated in a range of psychiatric disorders, including depression (Holsboer et al 1984a, 1984b) and anorexia nervosa (Gold et al 1986). VP has been employed clinically to assess the dynamic integrity of the HPA, but its use is limited by its side effects of nausea, abdominal pain, and flushing. Desmopressin (DDAVP), which has relative specificity for the renal V2 receptor with only mild V1-mediated pressor effects (Sawyer et al 1974), has not been shown to produce a consistent rise in ACTH or cortisol when administered alone to healthy subjects (Andersson et al 1972; Malerbi et al 1993; Gaillard et al 1988; Winkelmann et al 1994; Kasagi et al 1994). Both VP and DDAVP can potentiate the response to stimulation with ovine CRH (oCRH) in healthy individuals (Favrod-Coune et al 1993; Foppiani et al 1996; Ceserini et al 1997). These studies indicated that DDAVP administration is not associated with the hypotensive effects of VP, rendering it a potentially more satisfactory research and diagnostic tool. The last decade has seen a surge in interest in HPA functioning in chronic fatigue syndrome (CFS). For a diagnosis of CFS an individual must have severe fatigue of 6 months duration resulting in a reduction in activity levels of over 50%. Four of the following eight symptoms must also be present over this period: recurrent sore throats, enlarged and tender lymph glands, unrefreshing sleep, arthralgia, myalgia, recurrent headaches, postexertional malaise, and neuropsychological complaints (Fukuda et al 1994). Demitrack et al (1991), in a dynamic assessment of the HPA in CFS, demonstrated a blunted ACTH release but normal cortisol release to CRH administration, this occurring in the setting of low circulating cortisol levels. A subsequent study found both reduced ACTH and reduced cortisol responses (Scott et al 1998a) to CRH stimulation. These and other studies pointed to an overall hypofunctioning of the HPA axis in this disorder (Bearn et al 1995; Dinan et al 1997). DDAVP is known not to release ACTH in healthy subjects, yet the data indicate a potentiating effect on CRH-induced ACTH release. Given that blunted ACTH responses to CRH stimulation in CFS subjects have been previously reported, what impact would the administration of both substances have on pituitary–adrenal activation in this disorder? In addressing this question we administered CRH, DDAVP, and DDAVP/CRH combined to a group of healthy subjects and chronic fatigue patients.
Methods and Materials Subjects Patients were recruited from a chronic fatigue clinic. All subjects fulfilled the Centers for Disease Control and Prevention criteria (Fukuda et al 1994) and none, on the basis of a structured clinical interview (SCID, Spitzer et al 1987), had a comorbid psychiatric disorder according to DSM-III-R criteria (American Psychiatric
Association 1987). The volunteer group was free from any significant medical or psychiatric disorder. Any subject with a history of excess alcohol or illicit drug use was excluded. Two CFS subjects were on hormone replacement therapy (HRT). No healthy or CFS subject was on any other medication known to affect the HPA including an anovulant form of contraceptive, oral, or inhaled steroids. All female subjects were tested in the early follicular phase of the menstrual cycle. Participants refrained from alcohol consumption for 48 hours before the test. A total of 26 individuals participated in the studies; 13 CFS patients and 13 healthy volunteers. Eleven CFS (6 male, 5 female) subjects participated in the three arms of the study. One further female subject had CRH and DDAVP tests. The mean 6 SEM age of the CFS group was 38.9 6 2.5 years. Six healthy subjects had the three tests (5 male, 1 female). Three subjects (2 male, 1 female) had a CRH and the combined test, and one female subject had a DDAVP and combined test. Two male subjects had a DDAVP and CRH test, and 2 subjects (1 male, 1 female) had the CRH/DDAVP test alone. Two female subjects solely had a DDAVP test (CRH, n 5 11; DDAVP, n 5 11; CRH 1 DDAVP, n 5 12). The mean age of the healthy subjects participating in the tests was 39.4 6 3.6 years. The mean weight of the CFS group was 70.0 6 2.0 kg and of the control group was 72.7 6 3.3 kg. The mean 6 SEM duration of illness of the CFS group was 5.0 6 0.8 years. All participants gave written informed consent.
Methods The tests were performed in random order with at least 3 days separating each test. Subjects and persons performing the test were blinded to the substance being administered. All tests were carried out at 12:30 PM, the subject having fasted following a light breakfeast. At 12:30 PM (time 230 min), a cannula was placed in an anterior forearm vein, and this was kept patent with heprinse. After a 30-min relaxation period the subject received either 100 mg oCRH, 10 mg DDAVP, or 100 mg oCRH/10 mg DDAVP. When the combination was administered, DDAVP was administered as a bolus, and CRH was infused 2 min later over a 1-min period. Plasma samples for ACTH and cortisol estimation were taken at 230, 0, 115, 130, 145, 160, 190, and 1120 min. All subjects remained supine throughout the test. Blood pressure and heart rate were monitored at 15-min intervals.
ACTH Assay ACTH was measured using a commercially available two-site immunoradiometric assay. This is a nonextraction assay supplied by the Nichols Institue, San Juan Capistrano, California (Raff and Findling 1989). Intra- and interassay coefficients of variation were 3% and 6% respectively. The reliable lower limit of detection was 4.4 pmol/L (10 ng/L).
Cortisol Assay Cortisol was measured by an automated system (Immuno-I, Bayer Diagnostics, Newbury, England; Dash et al 1975) using an enzyme immunoassay method. The sensitivity of the assay is 10
Desmopressin and HPA Regulation in CFS
BIOL PSYCHIATRY 1999;45:1447–1454
1449
Table 1. Basal and d-ACTH (ng/L) and Basal and d-Cortisol (nmol/L) Responses to CRH, DDAVP, and CRH/DDAVP in CFS and Healthy Subjects (CONT)
Basal ACTH Basal CORT CRH d-ACTH DDAVP d-ACTH CRH/DDAVP d-ACTH CRH d-CORT DDAVP d-CORT CRH/DDAVP d-CORT
CFS (mean 6 SEM)
CONT (mean 6 SEM)
19.4 6 2.3 316.6 6 21.4 21.0 6 4.5 11.7 6 2.6 89.9 6 14.0 157.6 6 40.7 76.9 6 20.7 338.7 6 23.0
24.3 6 2.1 275.3 6 15.6 57.8 6 11.0 10.4 6 4.1 114.4 6 20.5 303.5 6 20.9 62.1 6 22.1 344.5 6 26.9
nmol/L and it has a between-batch variation of ,5% over the range 50 –1600 nmol/L.
Statistical Analyses The mean of the 230-min and 0-min results were taken as the baseline value. A one-way analysis of variance (ANOVA) was used to examine for differences in basal ACTH and cortisol levels for the three tests in the CFS and healthy group respectively. The delta (d; maximum increase from baseline) ACTH and cortisol values were examined in each of the two groups on the three tests. Student’s t tests were used to compare group means. Responses on each of the three tests were also analyzed as area under the curve (AUC). Statistical significance was taken as p , .05. Results are expressed as mean 6 SEM. In accordance with Todd and Lawrence (1990) we defined a positive response to oCRH as an increase of 50% or greater in the value of ACTH and an increase of 20% or greater in the value of cortisol with reference to baseline values. Chi-square tests examined for differences in responsivity patterns in the CFS and healthy subjects for each test. Pearson product–moment correlations were used where appropriate. Statgraphics (v2.7) was used for the statistical analysis.
Statistic t t t t t t t t
5 5 5 5 5 5 5 5
21.5, df 5 67 1.5, df 5 67 3.2, df 5 1 0.3, df 5 21 0.9, df 5 21 3.1, df 5 21 0.8, df 5 21 0.16, df 5 21
p value .15 .12 ,.005 .7 .3 ,.01 .4 .9
it did not induce significant ACTH release (x2 5 4.4; df 5 1, p , .05), and in 5 it failed to increase cortisol levels (x2 5 6.3, df 5 1, p 5 .01). The d-ACTH response was significantly attenuated in the CFS (21.0 6 4.5 ng/L) compared to healthy subjects (57.8 6 11.0 ng/L; t 5 3.2, df 5 21, p , .005). The d-cortisol response was also significantly lower in the CFS (157.6 6 40.7 nmol/L) compared to the healthy subject group (303.5 6 20.9 nnmol/L; t 5 3.1, df 5 21; p , .01) (see Table 1 and Figures 1 and 2).
Responses to DDAVP Four of 12 CFS subjects had ACTH responses of .50% of baseline values, as did 4 of the 11 healthy comparison subjects, following administration of DDAVP (x2 5 0.03, df 5 1, p 5 .86). Two of the control group had increments in cortisol of 20% or greater, whereas 3 of the 12 CFS subjects had such a response (x2 5 0.15; df 5 1, p 5 .7). The d-ACTH values did not differ between the two
Results Basal ACTH and Cortisol Mean basal ACTH did not differ between the three tests in the healthy (F 5 0.8, df 5 2, p 5 .5) or in the CFS group (F 5 0.5, df 5 2, p 5 .6). Mean basal cortisol levels similarly did not differ between the three tests in the healthy (F 5 0.008, df 5 2, p 5 .99) or CFS subjects (F 5 2.3, df 5 2, p 5 .12). No difference was found between the CFS patients and healthy comparison subjects on the three tests combined for either basal ACTH (19.4 6 2.3 ng/mL vs. 24.2 6 2.1 ng/L; t 5 1.5, df 5 67, p 5 .14) or cortisol measures (316.6 6 21.4 nmol/L vs. 275.3 6 15.6 nmol/L; t 5 1.5, df 5 67, p 5 .12) (see Table 1).
Responses to oCRH A positive ACTH and cortisol response to oCRH was seen in all healthy subjects, but in 4 of the 12 in the CFS group
Figure 1. A summary of the delta ACTH (maximum increment from baseline) responses to 100 mg ovine corticotropin-releasing hormone (CRH), 10 mg desmopressin (DDAVP), and CRH/ DDAVP combined in healthy subject (CONT) and chronic fatigue syndrome (CFS) groups. Mean 6 SEM for each group is shown.
1450
BIOL PSYCHIATRY 1999;45:1447–1454
L.V. Scott et al
groups: CFS, 11.7 6 2.6 ng/L; control subjects, 10.4 6 4.1 ng/L; t 5 0.28, df 5 21, p 5 .8. The d-cortisol response was similar in the two groups: CFS, 76.9 6 20.7 nmol/L; healthy subjects, 62.1 6 22.1 nmol/L; t 5 0.8, df 5 21, p 5 .4 (see Table 1 and Figures 1 and 2).
Responses to oCRH and DDAVP The mean d-ACTH response to CRH/DDAVP was 89.9 6 14.0 ng/L and 114.4 6 20.5 ng/L in the CFS and healthy subjects respectively (t 5 0.9; df 5 21, p 5 .3). The mean d-cortisol response was similarly not found to differ between the two groups: CFS, 338.7 6 23.0 nmol/L; healthy subjects, 344.5 6 26.9 nmol/L; t 5 0.16, df 5 21, p 5 .87 (see Table I and Figures 1 and 2). On comparing the d-ACTH responses to the CRH and CRH/DDAVP tests in CFS and control participants (all subjects who had either or both of these tests were examined), a significant difference is demonstrated between the ACTH responses in the CFS group (CRH: 21.0 6 4.5 ng/L; CRH/DDAVP: 89.9 6 14.0 ng/L; t 5 4.1, df 5 21, p , .005) (Figure 3). The ACTH response to the combination was also significantly greater than that to CRH alone in the control group (CRH: 57.8 6 11.0 ng/L; CRH/DDAVP: 114.4 6 20.5 ng/L; t 5 2.3, df 5 21, p , .05) (Figure 3). The differences in d-cortisol responses on the CRH and CRH/DDAVP test were similarly found to be significantly different in the CFS (CRH: 157.5 6 40.7 nmol/L; CRH/DDAVP: 329.6 6 22.5 nmol/L; t 5 3.5, df 5 20, p , .05) but not in the healthy subject group (CRH: 303.5 6 20.9 nmol/L; CRH/DDAVP: 338.7 6 20.3 nmol/L; t 5 1.2, df 5 21, p 5 .2) (Figure 4). Calculating the differences between the ACTH responses to CRH 1 DDAVP and CRH alone for each individual who did both tests (11 CFS subjects and 9 controls), and comparing both groups, the difference between groups for ACTH output was not significant (CFS: 70.2 6 15.3) but the cortisol response to the CRH 1 DDAVP minus the response to CRH alone was significantly greater in the CFS (196 6 48.2 nmol/L) compared to the healthy cohort (40.9 6 15.3 nmol/L; t 5 2.8; df 5 18, p 5 .01) (see Table 2). The nonresponders to the oCRH test in the CFS group all became responders to the combination of oCRH with DDAVP. No relationship was established between those CFS subjects who did not respond to the oCRH test and those who responded to the DDAVP test (r 5 .48; p 5 .11). Analysis of the data as AUC did not yield results any different from the above. Similarly, the data were analyzed with and without the two participants on HRT with no effect on the significance vs. nonsignificance of the above results.
Figure 2. Summary of the delta cortisol (maximum increment from baseline) responses to 100 mg corticotropin-releasing hormone (CRH), 10 mg desmopressin (DDAVP), and CRH/DDAVP combined in healthy subjects (CONT) and a chronic fatigue syndrome (CFS) patient group. Mean 1 SEM for each group is shown.
Figure 3. Delta ACTH responses in chronic fatigue syndrome (CFS) patients following 100 mg corticotropin-releasing hormone (CRH) administration (n 5 12), and following the combination of 100 mg corticotropin-releasing hormone and 10 mg desmopressin (CRH/DDAVP) (n 5 11). Delta ACTH responses in healthy subjects (CONT) following a CRH test (n 5 11) and CRH/DDAVP test (n 5 12). Mean 6 SEM for each group is illustrated.
Desmopressin and HPA Regulation in CFS
Influence of Basal ACTH and Cortisol Basal ACTH did not influence the d-ACTH in either the CFS or control group in any of the three tests. In the CFS group a negative relationship was established between the basal cortisol and d-cortisol on the CRH test (r 5 2.57, df 5 22, p 5 .05) but not on the DDAVP or CRH/DDAVP test. An inverse relationship was established for basal cortisol and d-cortisol output for the CRH/DDAVP test (r 5 2.76, df 5 22, p , .005) in the healthy subjects, and there was a trend for a similar relationship on the CRH test (r 5 2.53, df 5 20, p 5 .09). For those who had both a CRH and CRH/DDAVP test no relationship was established between basal cortisol and delta ACTH in either the CFS (r 5 2.18, p 5 .57) or healthy group (r 5 .06, p 5 .87) following CRH, or in the CFS (r 5 2.5, p 5 .13) or healthy group (r 5 2.45, p 5 .21) following the combination test.
Age and Gender No influence of age or gender was established for CFS or healthy subjects on basal or d-ACTH and cortisol responses on any of the three tests.
Figure 4. Delta cortisol responses in chronic fatigue syndrome (CFS) patients following 100 mg corticotropin-releasing hormone (CRH) administration (n 5 12) and following the combination of 100 mg CRH and 10 mg desmopressin (DDAVP). Delta cortisol responses in healthy subjects (CONT) following a CRH test and following a CRH/DDAVP test. Mean 1 SEM for each group is illustrated.
BIOL PSYCHIATRY 1999;45:1447–1454
1451
Table 2. Differences in d-ACTH and d-Cortisol Responses for CRH/DDAVP Test Minus d-ACTH and d-Cortisol Responses to CRH Alone for Those Individuals in the CFS Group (n 5 11) and Those in the Healthy Group (CONT: n 5 9) Who Had Both Tests
ACTH (ng/mL) Cortisol (nmol/L)
CONT
CFS
48.7 6 16.6 40.9 6 15.3
70.2 6 15.3 196 6 48.2
t 5 0.9, p 5 .3 t 5 2.8, p 5 .01
Side Effects Five of the CFS and 6 of the control group experienced transient facial flushing following oCRH infusion. Two in each group had this experience on testing with DDAVP. The coadministration of DDAVP with CRH did not increase the number of subjects who experienced the facial flushing with CRH alone. No headaches were reported, and no differences in pressor responses were recorded in any subject on each of the three tests.
Discussion We have demonstrated a blunted ACTH response to CRH stimulation in CFS subjects as has been previously reported (Demitrack et al 1991; Scott et al 1998a). The coadministration of CRH and DDAVP, however, produces both ACTH and cortisol responses indistinguishable from normal healthy subjects. The augmentation of CRHinduced ACTH release in healthy humans, by DDAVP, is consistent with other recent similar studies (Foppiani et al 1996; Ceserini et al 1997). Foppiani et al (1996) tested 6 healthy young women, also in the early phase of the menstrual cycle, but with the testing conducted in the morning, and showed that DDAVP was able to significantly enhance the ACTH response to oCRH. Previous studies of DDAVP-induced ACTH release in healthy subjects have been inconsistent (Malerbi et al 1993; Ceserini et al 1997; Andersson et al 1972; Rado and Juhos 1976) but have largely concluded (Winkelman et al 1994; Kasagi et al 1994) that DDAVP is not a pituitary– adrenal stimulant in healthy individuals. What is suggested from some of these studies is that interindividual variation may exist (Rado and Juhos 1976; Malerbi et al 1996); Malerbi and colleagues found that 2 of 15 subjects responded with an increase in cortisol above baseline of 58% and 69% respectively. Our study would suggest that an even greater proportion of the healthy population responds to DDAVP administration. Low basal cortisol levels have been documented in CFS (Demitrack et al 1991; Cleare et al 1995), but this is not a consistent finding (Bearn et al 1995; Dinan et al 1997). In the latter and present studies nonsignificantly higher basal cortisol levels were found, but no relationship was estab-
1452
BIOL PSYCHIATRY 1999;45:1447–1454
lished between basal cortisol and ACTH output in either the CFS or healthy group following stimulation with CRH or the CRH/DDAVP combination. The normal cortisol response to stimulation with CRH and DDAVP combined, found in CFS subjects, was surprising. Although several studies (Demitrack et al 1991; Dinan et al 1997) have demonstrated a normal cortisol output in the context of reduced ACTH, that is, an increased ACTH:CORT ratio, suggesting a heightened sensitivity of the adrenocorticotropin receptor, recent work by our group points to a diminished adrenocortical secretory reserve, as tested by exogenous ACTH stimulation (Scott et al 1998b). That the ACTH and CORT responses to oCRH stimulation were both reduced in this study supports this possibility. In healthy subjects cortisol release following CRH stimulation was not significantly augmented by DDAVP, unlike the finding in CFS, and despite an increase in ACTH release in this group. This suggests that in health, the ACTH release to 100 mg oCRH stimulation alone provides almost maximal stimulation of the adrenal cortex, thereby producing a ceiling effect on cortisol output. Alternatively, DDAVP may be acting directly at an adrenal level, although it has not been previously shown (Perraudin et al 1993), unlike VP (Guillon et al 1995), to have this effect. Our finding that, on analysis of individual responses to CRH/DDAVP minus the reponse to CRH alone, for those subjects who had both tests, cortisol output but not ACTH was significantly greater in the CFS compared to the healthy group suggests that DDAVP in this group may be acting directly at an adrenal level. An alternative interpretation is that there may be heightened sensitivity of the adrenoceptor to circulating ACTH in the CFS group, but that this only occurs above a certain ACTH threshold that is not reached following CRH alone, but is manifested following the combination of CRH and DDAVP. Increased vasopressinergic control of pituitary ACTH release is one possible explanation as to why ACTH and cortisol responses to the combination of CRH and DDAVP are equivalent in healthy and CFS subjects, when the response to oCRH in CFS is clearly blunted. In animal models of chronic stress (Whitnall 1989; Rabadan-Diehl et al 1995) in which there is an increased expression of VP there is an associated up-regulation of the pituitary V1b receptor; this stands in contrast to the more common receptor down-regulatory response in the setting of elevated peptide/transmitter levels. In the case of heightened V1b receptor responsivity in CFS, one may have anticipated a more uniform elevation of ACTH in CFS subjects when DDAVP was administered alone. The lack of uniform response to DDAVP may also be a reflection of DDAVP being a less than optimal probe of V1b receptor function. Repeat studies employing vasopressin, or indeed greater doses of DDAVP, would provide more useful
L.V. Scott et al
indication of altered activity. Furthermore, that DDAVP/ ACTH release was not more robust suggests that the greater response to the CRH/DDAVP compared to the CRH alone, in CFS compared to healthy subjects, more likely reflects an increased VP activity exerting its effect through a potentiation of CRH-induced ACTH release rather than a general increase in V1b receptor number and function in CFS. The coadministration of DDAVP with CRH may be overcoming a defect at postreceptor level; CRH with DDAVP may stimulate CRH-induced proopiomelanocortin messenger (mRNA) and consequently ACTH production. Demitrack et al (1991) originally proposed that the HPA abnormalities found in CFS may relate to a mild adrenal insufficiency due to a deficiency of CRH. In the same study a single evening sample of cerebrospinal fluid CRH levels did not show low levels; more optimal information will derive from 24-hour profiles. With regard to VP Bakheit and co-workers (1993) demonstrated low levels of circulating VP in 8 subjects with postviral fatigue syndrome. Increases in numerical response and group variability of VP levels were found in subjects with fibromyalgia (FM), this condition having a well-defined symptomatic and neuroendocrinological overlap with CFS (Crofford et al 1994). These studies employed paradigms that stimulated the magnocellular as opposed to parvicellular VP system, limiting conclusions about parvicellular VP activity that may be drawn. Based on these and other studies, Demitrack (1997) proposed that CFS and FM represent different forms of understimulation of the HPA, with both expressing low hypothalamic CRH, but FM being characterized by increased exposure of the corticotropes to VP, and CFS patients having reduced VP levels. This study does not support the likelihood of low VP levels in CFS. Similarities have also been drawn between CFS and atypical depression. Chrousos and Gold (1992) have proposed that both are manifestations of decreased stress system activity due to low circulating levels of CRH. In major depression elevated CRH levels are found (Nemeroff et al 1984), and hypothalamic neurons in such patients have been shown to overexpress VP (Purba et al 1996). More detailed studies of central CRH and VP levels in both CFS and atypical depression may forward the current hypothesis of a stress hyporesponsivity in both conditions. A number of animal models contribute some insight into this likelihood; for example, the Lewis rat, which is hyporesponsive to both inflammatory and noninflammatory stressors, is characterized by a deficiency of CRH mRNA (Sternberg et al 1989). This is associated with a substantial elevation in both plasma and hypothalamic VP levels (Patchev et al 1992). Moreover, behaviorally, the CRH-deficient and stress-hyporesponsive Lewis rat is
Desmopressin and HPA Regulation in CFS
characterized by low mobility and limited arousal (Calogero et al 1992), paralleling features of fatigue and anergia in CFS. In summary, this study demonstrates that the blunted ACTH and cortisol response to CRH in CFS may be overcome by the combined administration of DDAVP with CRH. This may relate to increased vasopressinergic regulation of the HPA in CFS. DDAVP administration may have a therapeutic benefit in CFS by enhancing ACTH release or by directly impacting on cortisol output at an adrenal level. Furthermore, this study questions the understanding that DDAVP is not a consistent activator of the HPA; it suggests that there are large interindividual differences in response that may be pertinent to its future use in a clinical and diagnostic setting.
Lucinda Scott is supported by the Linbury Trust.
References American Psychiatric Association (1987): Diagnostic and Statistical Manual of Mental Disorders, 3rd ed rev. Washington, DC: American Psychiatric Press. Andersson KE, Arner B, Hedner P, Mulder JL (1972): Effects of 8-lysine vasopressin and synthetic analogues on release of ACTH. Acta Endocrinol 69:640 – 648. Antoni FA (1993): Vasopressinergic control of pituitary adrenocorticotropin secretion comes of age. Front Neuroendocrinol 14:76 –122. Bakheit AM, Behan PO, Watson WS, Morton JJ (1993): Abnormal arginine vasopression secretion and water metabolism in patients with post-viral fatigue syndrome. Acta Neurol Scand 87:234 –238. Bearn J, Allain T, Coskeran P, Munro N, Butler J, McGregor A (1995): Neuroendocine responses to D-fenfluramine and insulin-induced hypoglycaemia in chronic fatigue syndrome. Biol Psychiatry 37:245–252. 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. Calogero A, Sternberg E, Bagdy G (1992): Neurotransmitterinduced hypothalamic-pituitary-adrenal axis responsiveness is defective in inflammatory disease-susceptible Lewis rats: In vivo and in vitro studies suggesting globally defective hypothalamic secretion of corticotropin-releasing hormone. Neuroendocrinology 55:600 – 608. Ceserini G, Freddi M, Paccotti P, Valenti G, Merchenthaler I (1997): Effects of ovine corticotropin-releasing hormone injection and desmopressin coadministration on galanin and adrenocorticotropin plasma levels in normal women. J Clin Endocrinol Metab 82:607– 610. Chrousos GP, Gold PW (1992): The concepts of stress and stress system disorders. JAMA 267:1244 –1252. Cleare AJ, Bearn J, Allain T, et al (1995): Contrasting neuroendocrine responses in depression and chronic fatigue syndrome. J Affect Disord 35:283–289.
BIOL PSYCHIATRY 1999;45:1447–1454
1453
Crofford L, Pillemer S, Kalageros K, et al (1994): Hypothalamic pituitary-adrenal axis perturbations in patients with fibromyalgia. Arthritis Rheumatol 37:1583–1592. Dash R, England B, Rees Midgeley A, Niswende G (1975): A specific, nonchromatographic radioimmunoassay for human plasma cortisol steroids. Steroids 26:647– 661. DeBold CR, Sheldon WR, DeCherney GS, et al (1984): Arginine vasopressin potentiates adrenocorticotropin release induced by ovine corticotropin-releasing factor. J Clin Invest 73:533– 538. Demitrack MA (1997): Neuroendocrine correlates of chronic fatigue syndrome: A brief review. J Psychiatr Res 31:69 – 82. Demitrack MA, Dale JK, Laue L, et al (1991): Evidence for impaired activation of the pituitary-adrenal axis in chronic fatigue syndrome. J Clin Endocrinol Metab 73:1224 –1234. Dinan TG, Majeed T, Lavelle E, Scott LV, Berti C, Behan P (1997): Serotonin-mediated activation of the hypothalamicpituitary adrenal axis in chronic fatigue syndrome. Psychoneuroendocrinology 22:261–267. Favrod-Coune C, Raux-Demay M, Proeschel M, et al (1993): Potentiation of the classic ovine corticotropin-releasing hormone test by the combined administration of small doses of lysine vasopressin. Clin Endocrinol 38:405– 410. Foppiani L, Sessarego P, Valenti S, et al (1996): Lack of effect of desmopressin on ACTH and cortisol responses to ovine corticotropin-releasing hormone in anorexia nervosa. Eur J Clin Invest 26:879 – 883. Fukuda K, Straus SE, Hickie I, Sharpe M, Dobbins JG, Komaroff A, and the International Chronic Fatigue Syndrome Study Group (1994): The chronic fatigue syndrome: A comprehensive approach to its treatment and study. Ann Intern Med 12:953–959. Gaillard RC, Riondal AM, Ling N, Muller AF (1988): Corticotropin-releasing factor activity of CRF 41 in normal man is potentiated by angiotensin II and vasopressin but not by desmopressin. Life Sci 43:1935–1944. Giguere V, Labrie F, Cote C, Coy DH, Sueiras-Diaz J, Schally AV (1982): Stimulation of cyclic AMP accumulation and corticotropin release by synthetic ovine corticotropin-releasing factor in rat: Site of glucocorticoid action. Proc Natl Acad Sci USA 79:34966 –34969. Gold P, Gwirstman H, Averginos P, et al (1986): Abnormal hypothalamic-pituitary-adrenal function in anorexia nervosa: Pathophysiological mechanisms in normal and underweight patients. N Engl J Med 314:1335–1342. Guillon G, Trueba M, Joubert D, et al (1995): Vasopressin stimulates steroid secretion in human adrenal glands: Comparison with angiotensin II effect. Endocrinology 136:1285– 1295. Holsboer F, von Bardeleben U, Gerken A, Stalla GK, Muller OA (1984a): Blunted corticotropin and normal cortisol response to human corticotropin-releasing factor (hCRF) in depression. N Engl J Med 311:1127. Holsboer F, Gerken A, Steiger A, Benkert O, Muller OA, Stalla GK (1984b): Corticotropin-releasing factor induced pituitaryadrenal responses in depression. Lancet i:55. Kasagi Y, Suda T, Horiba N, et al: DDAVP test for the diagnosis of ACTH-dependant Cushings syndrome. In: Proceedings of the 76th Annual Meeting of the Endocrine Society, Washington DC. Bethesda, MD: Endocrine Society, 1994: p 99.
1454
BIOL PSYCHIATRY 1999;45:1447–1454
Lamberts S, Verleun T, Oosterom R, de Jong F, Hackeng W (1984): Corticotropin-releasing factor (ovine) and vasopressin exert synergistic effect on adrenocorticotropin release in man. J Clin Endocrinol Metab 57:298 –303. Liu JH, Muse K, Contreras P, et al (1983): Augmentation of ACTH-releasing factor (CRF) by vasopressin in women. J Clin Endocrinol Metab 57:1087–1089. Malerbi DA, Mendonca B, Liberman B, et al (1993): The desmopressin stimulation test in the differential diagnosis of Cushings disease. Clin Endocrinol 38:463– 472. Nemeroff CB, Widerlov E, Bissette G, et al (1984): Elevated concentrations of corticotropin-releasing factor-like immunoreactivity in depressed patients. Science 226:1342–1344. Patchev V, Kalogeras K, Zelazowski P, Wilder R, Chrousos G (1992): Increased plasma concentration, hypothalamic content, and in vitro release of arginine vasopressin in inflammatory disease-prone hypothalamic corticotropinhormone deficient Lewis rats. Endocrinology 131:1453– 1457. Perraudin V, Delarue C, Lefevre H, Contesse V, Kuhn JM, Vaudry H (1993): Vasopressin stimulates cortisol secretion from human adrenocortical tissue through activation of V1 receptors. J Clin Endocrinol Metab 76:1522–1528. Purba J, Hoogendijk W, Hofman M, Swaab D (1996): Increased numbers of vasopressin and oxytocin expressing neurons in the paraventricular nucleus of the hypothalamus in depression. Arch Gen Psychiatry 53:137–143. Rabadan-Diehl C, Lolait S, Aguilera G (1995): Regulation of pituitary V1b receptor mRNA during stress in the rat. J Neuroendocrinol 7:903–910. Rado JS, Juhos E (1976): Effect of 1-deamino-8-D-arginine vasopressin (DDAVP) on plasma cortisol (hydrocortisone ). J Clin Pharmacol 16:333–337. Raff H, Findling J (1989): A new immunoradiometric assay for
L.V. Scott et al
corticotropin, evaluated in normal subjects and patients with Cushings syndrome. Clin Chem 35:595– 600. Salata RA, Jarrett DB, Verbalis JG, Robinson AG (1988): Vasopressin stimulation of adrenocorticotropin hormone (ACTH) in humans. J Clin Invest 81:766 –774. Sawyer WH, Acosta M, Balaspiri L, Judd J, Manning M (1974): Structural changes in the arginine vasopressin molecule. Endocrinology 94:1106 –1115. Scott LV, Medbak S, Dinan TG (1998a): Blunted adrenocorticotropin and cortisol responses to corticotropin releasing hormone stimulation in chronic fatigue syndrome. Acta Psychiatr Scand 97:450 – 457. Scott LV, Medbak S, Dinan TG (1998b): The low dose ACTH test in chronic fatigue syndrome and in health. Clin Endocrinol 48:733–737. Spitzer R, Williams J, Gibbon M (1987): Structured Clinical Interview for DSM III-R. New York: New York State Psychiatric Institute, Biometrics Research. Sternberg E, Hill J, Chrousos G, et al (1989): Inflammatory mediator-induced hypothalamic-pituitary-adrenal axis activation is defective in streptococcal cell wall arthritis susceptible Lewis rats. Proc Natl Acad Sci USA 86:4771– 4775. Sugimoto Y, Sato M, Mochizuki S, et al (1994): Molecular cloning and functional characterization of a cDNA encoding the human V1b vasopressin receptor. J Biol Chem 269: 27088 –27092. Todd BK, Lawrence C (1990): The Cushing syndrome: An update on diagnostic tests. Ann Intern Med 112:434 – 444. Whitnall MH (1989): Stress selectively activates the vasopressincontaining subpopulations of corticotropin-releasing hormone neurons. Neuroendocrinology 50:702–707. Winkelmann W, Heppner C, Schmitt HW, Deuss U, Nieke D, Kaulen D (1994): The desmopressin stimulation test in patients with ACTH hypersecretion. Eur J Endocrinol 130(suppl 2):151.
© 1999 Society of Biological Psychiatry
Conclusions: DDAVP augments CRH-mediated pituitary–adrenal responsivity in healthy subjects and in patients with CFS. That DDAVP was capable of normalizing the pituitary–adrenal response to oCRH in the CFS group suggests there may be increased vasopressinergic responsivity of the anterior pituitary in CFS and/or that DDAVP may be exerting an effect at an adrenal level. Biol Psychiatry 1999;45:1447–1454 © 1999 Society of Biological Psychiatry Key Words: Chronic fatigue syndrome, corticotropinreleasing hormone, vasopressin, desmopressin, hypothalamic–pituitary–adrenal axis, adrenocorticotropin, cortisol
Introduction
C
orticotropin-releasing hormone (CRH) is the primary regulator of the hypothalamic–pituitary–adrenal axis (HPA) in man. Its main companion regulator is vasopressin (VP), a nonapeptide, produced from the parvicelluar neurons of the paraventricular nucleus (PVN), and the magnocellular neurons of the supraoptic nucleus (Antoni 1993). VP plays a role as an adrenocorticotropic hormone (ACTH)-releasing factor at the recently cloned V1b receptor on the anterior pituitary (Sugimoto et al 1994). Although CRH has been shown to be a more potent secretagogue of ACTH than VP, when VP is administered alone it has significant ACTH-releasing properties in humans (Brostoff et al 1968; Salata et al 1988). The coadministration of the two peptides causes an augmented release of ACTH (Liu et al 1983; DeBold et al 1984; Lamberts et al 1984); this synergism may be associated with a VP-mediated enhancement of CRH cyclic adenosine monophosphate production at the level of the corticotrope (Giguere et al 1982). Advances in our understanding of stress physiology have been facilitated by the availability of both ovine and human forms of the neuropeptide CRH as exogenous activators of the HPA. Abnormalities of CRH-induced 0006-3223/99/$20.00 PII S0006-3223(98)00232-7
1448
L.V. Scott et al
BIOL PSYCHIATRY 1999;45:1447–1454
ACTH release have been demonstrated in a range of psychiatric disorders, including depression (Holsboer et al 1984a, 1984b) and anorexia nervosa (Gold et al 1986). VP has been employed clinically to assess the dynamic integrity of the HPA, but its use is limited by its side effects of nausea, abdominal pain, and flushing. Desmopressin (DDAVP), which has relative specificity for the renal V2 receptor with only mild V1-mediated pressor effects (Sawyer et al 1974), has not been shown to produce a consistent rise in ACTH or cortisol when administered alone to healthy subjects (Andersson et al 1972; Malerbi et al 1993; Gaillard et al 1988; Winkelmann et al 1994; Kasagi et al 1994). Both VP and DDAVP can potentiate the response to stimulation with ovine CRH (oCRH) in healthy individuals (Favrod-Coune et al 1993; Foppiani et al 1996; Ceserini et al 1997). These studies indicated that DDAVP administration is not associated with the hypotensive effects of VP, rendering it a potentially more satisfactory research and diagnostic tool. The last decade has seen a surge in interest in HPA functioning in chronic fatigue syndrome (CFS). For a diagnosis of CFS an individual must have severe fatigue of 6 months duration resulting in a reduction in activity levels of over 50%. Four of the following eight symptoms must also be present over this period: recurrent sore throats, enlarged and tender lymph glands, unrefreshing sleep, arthralgia, myalgia, recurrent headaches, postexertional malaise, and neuropsychological complaints (Fukuda et al 1994). Demitrack et al (1991), in a dynamic assessment of the HPA in CFS, demonstrated a blunted ACTH release but normal cortisol release to CRH administration, this occurring in the setting of low circulating cortisol levels. A subsequent study found both reduced ACTH and reduced cortisol responses (Scott et al 1998a) to CRH stimulation. These and other studies pointed to an overall hypofunctioning of the HPA axis in this disorder (Bearn et al 1995; Dinan et al 1997). DDAVP is known not to release ACTH in healthy subjects, yet the data indicate a potentiating effect on CRH-induced ACTH release. Given that blunted ACTH responses to CRH stimulation in CFS subjects have been previously reported, what impact would the administration of both substances have on pituitary–adrenal activation in this disorder? In addressing this question we administered CRH, DDAVP, and DDAVP/CRH combined to a group of healthy subjects and chronic fatigue patients.
Methods and Materials Subjects Patients were recruited from a chronic fatigue clinic. All subjects fulfilled the Centers for Disease Control and Prevention criteria (Fukuda et al 1994) and none, on the basis of a structured clinical interview (SCID, Spitzer et al 1987), had a comorbid psychiatric disorder according to DSM-III-R criteria (American Psychiatric
Association 1987). The volunteer group was free from any significant medical or psychiatric disorder. Any subject with a history of excess alcohol or illicit drug use was excluded. Two CFS subjects were on hormone replacement therapy (HRT). No healthy or CFS subject was on any other medication known to affect the HPA including an anovulant form of contraceptive, oral, or inhaled steroids. All female subjects were tested in the early follicular phase of the menstrual cycle. Participants refrained from alcohol consumption for 48 hours before the test. A total of 26 individuals participated in the studies; 13 CFS patients and 13 healthy volunteers. Eleven CFS (6 male, 5 female) subjects participated in the three arms of the study. One further female subject had CRH and DDAVP tests. The mean 6 SEM age of the CFS group was 38.9 6 2.5 years. Six healthy subjects had the three tests (5 male, 1 female). Three subjects (2 male, 1 female) had a CRH and the combined test, and one female subject had a DDAVP and combined test. Two male subjects had a DDAVP and CRH test, and 2 subjects (1 male, 1 female) had the CRH/DDAVP test alone. Two female subjects solely had a DDAVP test (CRH, n 5 11; DDAVP, n 5 11; CRH 1 DDAVP, n 5 12). The mean age of the healthy subjects participating in the tests was 39.4 6 3.6 years. The mean weight of the CFS group was 70.0 6 2.0 kg and of the control group was 72.7 6 3.3 kg. The mean 6 SEM duration of illness of the CFS group was 5.0 6 0.8 years. All participants gave written informed consent.
Methods The tests were performed in random order with at least 3 days separating each test. Subjects and persons performing the test were blinded to the substance being administered. All tests were carried out at 12:30 PM, the subject having fasted following a light breakfeast. At 12:30 PM (time 230 min), a cannula was placed in an anterior forearm vein, and this was kept patent with heprinse. After a 30-min relaxation period the subject received either 100 mg oCRH, 10 mg DDAVP, or 100 mg oCRH/10 mg DDAVP. When the combination was administered, DDAVP was administered as a bolus, and CRH was infused 2 min later over a 1-min period. Plasma samples for ACTH and cortisol estimation were taken at 230, 0, 115, 130, 145, 160, 190, and 1120 min. All subjects remained supine throughout the test. Blood pressure and heart rate were monitored at 15-min intervals.
ACTH Assay ACTH was measured using a commercially available two-site immunoradiometric assay. This is a nonextraction assay supplied by the Nichols Institue, San Juan Capistrano, California (Raff and Findling 1989). Intra- and interassay coefficients of variation were 3% and 6% respectively. The reliable lower limit of detection was 4.4 pmol/L (10 ng/L).
Cortisol Assay Cortisol was measured by an automated system (Immuno-I, Bayer Diagnostics, Newbury, England; Dash et al 1975) using an enzyme immunoassay method. The sensitivity of the assay is 10
Desmopressin and HPA Regulation in CFS
BIOL PSYCHIATRY 1999;45:1447–1454
1449
Table 1. Basal and d-ACTH (ng/L) and Basal and d-Cortisol (nmol/L) Responses to CRH, DDAVP, and CRH/DDAVP in CFS and Healthy Subjects (CONT)
Basal ACTH Basal CORT CRH d-ACTH DDAVP d-ACTH CRH/DDAVP d-ACTH CRH d-CORT DDAVP d-CORT CRH/DDAVP d-CORT
CFS (mean 6 SEM)
CONT (mean 6 SEM)
19.4 6 2.3 316.6 6 21.4 21.0 6 4.5 11.7 6 2.6 89.9 6 14.0 157.6 6 40.7 76.9 6 20.7 338.7 6 23.0
24.3 6 2.1 275.3 6 15.6 57.8 6 11.0 10.4 6 4.1 114.4 6 20.5 303.5 6 20.9 62.1 6 22.1 344.5 6 26.9
nmol/L and it has a between-batch variation of ,5% over the range 50 –1600 nmol/L.
Statistical Analyses The mean of the 230-min and 0-min results were taken as the baseline value. A one-way analysis of variance (ANOVA) was used to examine for differences in basal ACTH and cortisol levels for the three tests in the CFS and healthy group respectively. The delta (d; maximum increase from baseline) ACTH and cortisol values were examined in each of the two groups on the three tests. Student’s t tests were used to compare group means. Responses on each of the three tests were also analyzed as area under the curve (AUC). Statistical significance was taken as p , .05. Results are expressed as mean 6 SEM. In accordance with Todd and Lawrence (1990) we defined a positive response to oCRH as an increase of 50% or greater in the value of ACTH and an increase of 20% or greater in the value of cortisol with reference to baseline values. Chi-square tests examined for differences in responsivity patterns in the CFS and healthy subjects for each test. Pearson product–moment correlations were used where appropriate. Statgraphics (v2.7) was used for the statistical analysis.
Statistic t t t t t t t t
5 5 5 5 5 5 5 5
21.5, df 5 67 1.5, df 5 67 3.2, df 5 1 0.3, df 5 21 0.9, df 5 21 3.1, df 5 21 0.8, df 5 21 0.16, df 5 21
p value .15 .12 ,.005 .7 .3 ,.01 .4 .9
it did not induce significant ACTH release (x2 5 4.4; df 5 1, p , .05), and in 5 it failed to increase cortisol levels (x2 5 6.3, df 5 1, p 5 .01). The d-ACTH response was significantly attenuated in the CFS (21.0 6 4.5 ng/L) compared to healthy subjects (57.8 6 11.0 ng/L; t 5 3.2, df 5 21, p , .005). The d-cortisol response was also significantly lower in the CFS (157.6 6 40.7 nmol/L) compared to the healthy subject group (303.5 6 20.9 nnmol/L; t 5 3.1, df 5 21; p , .01) (see Table 1 and Figures 1 and 2).
Responses to DDAVP Four of 12 CFS subjects had ACTH responses of .50% of baseline values, as did 4 of the 11 healthy comparison subjects, following administration of DDAVP (x2 5 0.03, df 5 1, p 5 .86). Two of the control group had increments in cortisol of 20% or greater, whereas 3 of the 12 CFS subjects had such a response (x2 5 0.15; df 5 1, p 5 .7). The d-ACTH values did not differ between the two
Results Basal ACTH and Cortisol Mean basal ACTH did not differ between the three tests in the healthy (F 5 0.8, df 5 2, p 5 .5) or in the CFS group (F 5 0.5, df 5 2, p 5 .6). Mean basal cortisol levels similarly did not differ between the three tests in the healthy (F 5 0.008, df 5 2, p 5 .99) or CFS subjects (F 5 2.3, df 5 2, p 5 .12). No difference was found between the CFS patients and healthy comparison subjects on the three tests combined for either basal ACTH (19.4 6 2.3 ng/mL vs. 24.2 6 2.1 ng/L; t 5 1.5, df 5 67, p 5 .14) or cortisol measures (316.6 6 21.4 nmol/L vs. 275.3 6 15.6 nmol/L; t 5 1.5, df 5 67, p 5 .12) (see Table 1).
Responses to oCRH A positive ACTH and cortisol response to oCRH was seen in all healthy subjects, but in 4 of the 12 in the CFS group
Figure 1. A summary of the delta ACTH (maximum increment from baseline) responses to 100 mg ovine corticotropin-releasing hormone (CRH), 10 mg desmopressin (DDAVP), and CRH/ DDAVP combined in healthy subject (CONT) and chronic fatigue syndrome (CFS) groups. Mean 6 SEM for each group is shown.
1450
BIOL PSYCHIATRY 1999;45:1447–1454
L.V. Scott et al
groups: CFS, 11.7 6 2.6 ng/L; control subjects, 10.4 6 4.1 ng/L; t 5 0.28, df 5 21, p 5 .8. The d-cortisol response was similar in the two groups: CFS, 76.9 6 20.7 nmol/L; healthy subjects, 62.1 6 22.1 nmol/L; t 5 0.8, df 5 21, p 5 .4 (see Table 1 and Figures 1 and 2).
Responses to oCRH and DDAVP The mean d-ACTH response to CRH/DDAVP was 89.9 6 14.0 ng/L and 114.4 6 20.5 ng/L in the CFS and healthy subjects respectively (t 5 0.9; df 5 21, p 5 .3). The mean d-cortisol response was similarly not found to differ between the two groups: CFS, 338.7 6 23.0 nmol/L; healthy subjects, 344.5 6 26.9 nmol/L; t 5 0.16, df 5 21, p 5 .87 (see Table I and Figures 1 and 2). On comparing the d-ACTH responses to the CRH and CRH/DDAVP tests in CFS and control participants (all subjects who had either or both of these tests were examined), a significant difference is demonstrated between the ACTH responses in the CFS group (CRH: 21.0 6 4.5 ng/L; CRH/DDAVP: 89.9 6 14.0 ng/L; t 5 4.1, df 5 21, p , .005) (Figure 3). The ACTH response to the combination was also significantly greater than that to CRH alone in the control group (CRH: 57.8 6 11.0 ng/L; CRH/DDAVP: 114.4 6 20.5 ng/L; t 5 2.3, df 5 21, p , .05) (Figure 3). The differences in d-cortisol responses on the CRH and CRH/DDAVP test were similarly found to be significantly different in the CFS (CRH: 157.5 6 40.7 nmol/L; CRH/DDAVP: 329.6 6 22.5 nmol/L; t 5 3.5, df 5 20, p , .05) but not in the healthy subject group (CRH: 303.5 6 20.9 nmol/L; CRH/DDAVP: 338.7 6 20.3 nmol/L; t 5 1.2, df 5 21, p 5 .2) (Figure 4). Calculating the differences between the ACTH responses to CRH 1 DDAVP and CRH alone for each individual who did both tests (11 CFS subjects and 9 controls), and comparing both groups, the difference between groups for ACTH output was not significant (CFS: 70.2 6 15.3) but the cortisol response to the CRH 1 DDAVP minus the response to CRH alone was significantly greater in the CFS (196 6 48.2 nmol/L) compared to the healthy cohort (40.9 6 15.3 nmol/L; t 5 2.8; df 5 18, p 5 .01) (see Table 2). The nonresponders to the oCRH test in the CFS group all became responders to the combination of oCRH with DDAVP. No relationship was established between those CFS subjects who did not respond to the oCRH test and those who responded to the DDAVP test (r 5 .48; p 5 .11). Analysis of the data as AUC did not yield results any different from the above. Similarly, the data were analyzed with and without the two participants on HRT with no effect on the significance vs. nonsignificance of the above results.
Figure 2. Summary of the delta cortisol (maximum increment from baseline) responses to 100 mg corticotropin-releasing hormone (CRH), 10 mg desmopressin (DDAVP), and CRH/DDAVP combined in healthy subjects (CONT) and a chronic fatigue syndrome (CFS) patient group. Mean 1 SEM for each group is shown.
Figure 3. Delta ACTH responses in chronic fatigue syndrome (CFS) patients following 100 mg corticotropin-releasing hormone (CRH) administration (n 5 12), and following the combination of 100 mg corticotropin-releasing hormone and 10 mg desmopressin (CRH/DDAVP) (n 5 11). Delta ACTH responses in healthy subjects (CONT) following a CRH test (n 5 11) and CRH/DDAVP test (n 5 12). Mean 6 SEM for each group is illustrated.
Desmopressin and HPA Regulation in CFS
Influence of Basal ACTH and Cortisol Basal ACTH did not influence the d-ACTH in either the CFS or control group in any of the three tests. In the CFS group a negative relationship was established between the basal cortisol and d-cortisol on the CRH test (r 5 2.57, df 5 22, p 5 .05) but not on the DDAVP or CRH/DDAVP test. An inverse relationship was established for basal cortisol and d-cortisol output for the CRH/DDAVP test (r 5 2.76, df 5 22, p , .005) in the healthy subjects, and there was a trend for a similar relationship on the CRH test (r 5 2.53, df 5 20, p 5 .09). For those who had both a CRH and CRH/DDAVP test no relationship was established between basal cortisol and delta ACTH in either the CFS (r 5 2.18, p 5 .57) or healthy group (r 5 .06, p 5 .87) following CRH, or in the CFS (r 5 2.5, p 5 .13) or healthy group (r 5 2.45, p 5 .21) following the combination test.
Age and Gender No influence of age or gender was established for CFS or healthy subjects on basal or d-ACTH and cortisol responses on any of the three tests.
Figure 4. Delta cortisol responses in chronic fatigue syndrome (CFS) patients following 100 mg corticotropin-releasing hormone (CRH) administration (n 5 12) and following the combination of 100 mg CRH and 10 mg desmopressin (DDAVP). Delta cortisol responses in healthy subjects (CONT) following a CRH test and following a CRH/DDAVP test. Mean 1 SEM for each group is illustrated.
BIOL PSYCHIATRY 1999;45:1447–1454
1451
Table 2. Differences in d-ACTH and d-Cortisol Responses for CRH/DDAVP Test Minus d-ACTH and d-Cortisol Responses to CRH Alone for Those Individuals in the CFS Group (n 5 11) and Those in the Healthy Group (CONT: n 5 9) Who Had Both Tests
ACTH (ng/mL) Cortisol (nmol/L)
CONT
CFS
48.7 6 16.6 40.9 6 15.3
70.2 6 15.3 196 6 48.2
t 5 0.9, p 5 .3 t 5 2.8, p 5 .01
Side Effects Five of the CFS and 6 of the control group experienced transient facial flushing following oCRH infusion. Two in each group had this experience on testing with DDAVP. The coadministration of DDAVP with CRH did not increase the number of subjects who experienced the facial flushing with CRH alone. No headaches were reported, and no differences in pressor responses were recorded in any subject on each of the three tests.
Discussion We have demonstrated a blunted ACTH response to CRH stimulation in CFS subjects as has been previously reported (Demitrack et al 1991; Scott et al 1998a). The coadministration of CRH and DDAVP, however, produces both ACTH and cortisol responses indistinguishable from normal healthy subjects. The augmentation of CRHinduced ACTH release in healthy humans, by DDAVP, is consistent with other recent similar studies (Foppiani et al 1996; Ceserini et al 1997). Foppiani et al (1996) tested 6 healthy young women, also in the early phase of the menstrual cycle, but with the testing conducted in the morning, and showed that DDAVP was able to significantly enhance the ACTH response to oCRH. Previous studies of DDAVP-induced ACTH release in healthy subjects have been inconsistent (Malerbi et al 1993; Ceserini et al 1997; Andersson et al 1972; Rado and Juhos 1976) but have largely concluded (Winkelman et al 1994; Kasagi et al 1994) that DDAVP is not a pituitary– adrenal stimulant in healthy individuals. What is suggested from some of these studies is that interindividual variation may exist (Rado and Juhos 1976; Malerbi et al 1996); Malerbi and colleagues found that 2 of 15 subjects responded with an increase in cortisol above baseline of 58% and 69% respectively. Our study would suggest that an even greater proportion of the healthy population responds to DDAVP administration. Low basal cortisol levels have been documented in CFS (Demitrack et al 1991; Cleare et al 1995), but this is not a consistent finding (Bearn et al 1995; Dinan et al 1997). In the latter and present studies nonsignificantly higher basal cortisol levels were found, but no relationship was estab-
1452
BIOL PSYCHIATRY 1999;45:1447–1454
lished between basal cortisol and ACTH output in either the CFS or healthy group following stimulation with CRH or the CRH/DDAVP combination. The normal cortisol response to stimulation with CRH and DDAVP combined, found in CFS subjects, was surprising. Although several studies (Demitrack et al 1991; Dinan et al 1997) have demonstrated a normal cortisol output in the context of reduced ACTH, that is, an increased ACTH:CORT ratio, suggesting a heightened sensitivity of the adrenocorticotropin receptor, recent work by our group points to a diminished adrenocortical secretory reserve, as tested by exogenous ACTH stimulation (Scott et al 1998b). That the ACTH and CORT responses to oCRH stimulation were both reduced in this study supports this possibility. In healthy subjects cortisol release following CRH stimulation was not significantly augmented by DDAVP, unlike the finding in CFS, and despite an increase in ACTH release in this group. This suggests that in health, the ACTH release to 100 mg oCRH stimulation alone provides almost maximal stimulation of the adrenal cortex, thereby producing a ceiling effect on cortisol output. Alternatively, DDAVP may be acting directly at an adrenal level, although it has not been previously shown (Perraudin et al 1993), unlike VP (Guillon et al 1995), to have this effect. Our finding that, on analysis of individual responses to CRH/DDAVP minus the reponse to CRH alone, for those subjects who had both tests, cortisol output but not ACTH was significantly greater in the CFS compared to the healthy group suggests that DDAVP in this group may be acting directly at an adrenal level. An alternative interpretation is that there may be heightened sensitivity of the adrenoceptor to circulating ACTH in the CFS group, but that this only occurs above a certain ACTH threshold that is not reached following CRH alone, but is manifested following the combination of CRH and DDAVP. Increased vasopressinergic control of pituitary ACTH release is one possible explanation as to why ACTH and cortisol responses to the combination of CRH and DDAVP are equivalent in healthy and CFS subjects, when the response to oCRH in CFS is clearly blunted. In animal models of chronic stress (Whitnall 1989; Rabadan-Diehl et al 1995) in which there is an increased expression of VP there is an associated up-regulation of the pituitary V1b receptor; this stands in contrast to the more common receptor down-regulatory response in the setting of elevated peptide/transmitter levels. In the case of heightened V1b receptor responsivity in CFS, one may have anticipated a more uniform elevation of ACTH in CFS subjects when DDAVP was administered alone. The lack of uniform response to DDAVP may also be a reflection of DDAVP being a less than optimal probe of V1b receptor function. Repeat studies employing vasopressin, or indeed greater doses of DDAVP, would provide more useful
L.V. Scott et al
indication of altered activity. Furthermore, that DDAVP/ ACTH release was not more robust suggests that the greater response to the CRH/DDAVP compared to the CRH alone, in CFS compared to healthy subjects, more likely reflects an increased VP activity exerting its effect through a potentiation of CRH-induced ACTH release rather than a general increase in V1b receptor number and function in CFS. The coadministration of DDAVP with CRH may be overcoming a defect at postreceptor level; CRH with DDAVP may stimulate CRH-induced proopiomelanocortin messenger (mRNA) and consequently ACTH production. Demitrack et al (1991) originally proposed that the HPA abnormalities found in CFS may relate to a mild adrenal insufficiency due to a deficiency of CRH. In the same study a single evening sample of cerebrospinal fluid CRH levels did not show low levels; more optimal information will derive from 24-hour profiles. With regard to VP Bakheit and co-workers (1993) demonstrated low levels of circulating VP in 8 subjects with postviral fatigue syndrome. Increases in numerical response and group variability of VP levels were found in subjects with fibromyalgia (FM), this condition having a well-defined symptomatic and neuroendocrinological overlap with CFS (Crofford et al 1994). These studies employed paradigms that stimulated the magnocellular as opposed to parvicellular VP system, limiting conclusions about parvicellular VP activity that may be drawn. Based on these and other studies, Demitrack (1997) proposed that CFS and FM represent different forms of understimulation of the HPA, with both expressing low hypothalamic CRH, but FM being characterized by increased exposure of the corticotropes to VP, and CFS patients having reduced VP levels. This study does not support the likelihood of low VP levels in CFS. Similarities have also been drawn between CFS and atypical depression. Chrousos and Gold (1992) have proposed that both are manifestations of decreased stress system activity due to low circulating levels of CRH. In major depression elevated CRH levels are found (Nemeroff et al 1984), and hypothalamic neurons in such patients have been shown to overexpress VP (Purba et al 1996). More detailed studies of central CRH and VP levels in both CFS and atypical depression may forward the current hypothesis of a stress hyporesponsivity in both conditions. A number of animal models contribute some insight into this likelihood; for example, the Lewis rat, which is hyporesponsive to both inflammatory and noninflammatory stressors, is characterized by a deficiency of CRH mRNA (Sternberg et al 1989). This is associated with a substantial elevation in both plasma and hypothalamic VP levels (Patchev et al 1992). Moreover, behaviorally, the CRH-deficient and stress-hyporesponsive Lewis rat is
Desmopressin and HPA Regulation in CFS
characterized by low mobility and limited arousal (Calogero et al 1992), paralleling features of fatigue and anergia in CFS. In summary, this study demonstrates that the blunted ACTH and cortisol response to CRH in CFS may be overcome by the combined administration of DDAVP with CRH. This may relate to increased vasopressinergic regulation of the HPA in CFS. DDAVP administration may have a therapeutic benefit in CFS by enhancing ACTH release or by directly impacting on cortisol output at an adrenal level. Furthermore, this study questions the understanding that DDAVP is not a consistent activator of the HPA; it suggests that there are large interindividual differences in response that may be pertinent to its future use in a clinical and diagnostic setting.
Lucinda Scott is supported by the Linbury Trust.
References American Psychiatric Association (1987): Diagnostic and Statistical Manual of Mental Disorders, 3rd ed rev. Washington, DC: American Psychiatric Press. Andersson KE, Arner B, Hedner P, Mulder JL (1972): Effects of 8-lysine vasopressin and synthetic analogues on release of ACTH. Acta Endocrinol 69:640 – 648. Antoni FA (1993): Vasopressinergic control of pituitary adrenocorticotropin secretion comes of age. Front Neuroendocrinol 14:76 –122. Bakheit AM, Behan PO, Watson WS, Morton JJ (1993): Abnormal arginine vasopression secretion and water metabolism in patients with post-viral fatigue syndrome. Acta Neurol Scand 87:234 –238. Bearn J, Allain T, Coskeran P, Munro N, Butler J, McGregor A (1995): Neuroendocine responses to D-fenfluramine and insulin-induced hypoglycaemia in chronic fatigue syndrome. Biol Psychiatry 37:245–252. 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. Calogero A, Sternberg E, Bagdy G (1992): Neurotransmitterinduced hypothalamic-pituitary-adrenal axis responsiveness is defective in inflammatory disease-susceptible Lewis rats: In vivo and in vitro studies suggesting globally defective hypothalamic secretion of corticotropin-releasing hormone. Neuroendocrinology 55:600 – 608. Ceserini G, Freddi M, Paccotti P, Valenti G, Merchenthaler I (1997): Effects of ovine corticotropin-releasing hormone injection and desmopressin coadministration on galanin and adrenocorticotropin plasma levels in normal women. J Clin Endocrinol Metab 82:607– 610. Chrousos GP, Gold PW (1992): The concepts of stress and stress system disorders. JAMA 267:1244 –1252. Cleare AJ, Bearn J, Allain T, et al (1995): Contrasting neuroendocrine responses in depression and chronic fatigue syndrome. J Affect Disord 35:283–289.
BIOL PSYCHIATRY 1999;45:1447–1454
1453
Crofford L, Pillemer S, Kalageros K, et al (1994): Hypothalamic pituitary-adrenal axis perturbations in patients with fibromyalgia. Arthritis Rheumatol 37:1583–1592. Dash R, England B, Rees Midgeley A, Niswende G (1975): A specific, nonchromatographic radioimmunoassay for human plasma cortisol steroids. Steroids 26:647– 661. DeBold CR, Sheldon WR, DeCherney GS, et al (1984): Arginine vasopressin potentiates adrenocorticotropin release induced by ovine corticotropin-releasing factor. J Clin Invest 73:533– 538. Demitrack MA (1997): Neuroendocrine correlates of chronic fatigue syndrome: A brief review. J Psychiatr Res 31:69 – 82. Demitrack MA, Dale JK, Laue L, et al (1991): Evidence for impaired activation of the pituitary-adrenal axis in chronic fatigue syndrome. J Clin Endocrinol Metab 73:1224 –1234. Dinan TG, Majeed T, Lavelle E, Scott LV, Berti C, Behan P (1997): Serotonin-mediated activation of the hypothalamicpituitary adrenal axis in chronic fatigue syndrome. Psychoneuroendocrinology 22:261–267. Favrod-Coune C, Raux-Demay M, Proeschel M, et al (1993): Potentiation of the classic ovine corticotropin-releasing hormone test by the combined administration of small doses of lysine vasopressin. Clin Endocrinol 38:405– 410. Foppiani L, Sessarego P, Valenti S, et al (1996): Lack of effect of desmopressin on ACTH and cortisol responses to ovine corticotropin-releasing hormone in anorexia nervosa. Eur J Clin Invest 26:879 – 883. Fukuda K, Straus SE, Hickie I, Sharpe M, Dobbins JG, Komaroff A, and the International Chronic Fatigue Syndrome Study Group (1994): The chronic fatigue syndrome: A comprehensive approach to its treatment and study. Ann Intern Med 12:953–959. Gaillard RC, Riondal AM, Ling N, Muller AF (1988): Corticotropin-releasing factor activity of CRF 41 in normal man is potentiated by angiotensin II and vasopressin but not by desmopressin. Life Sci 43:1935–1944. Giguere V, Labrie F, Cote C, Coy DH, Sueiras-Diaz J, Schally AV (1982): Stimulation of cyclic AMP accumulation and corticotropin release by synthetic ovine corticotropin-releasing factor in rat: Site of glucocorticoid action. Proc Natl Acad Sci USA 79:34966 –34969. Gold P, Gwirstman H, Averginos P, et al (1986): Abnormal hypothalamic-pituitary-adrenal function in anorexia nervosa: Pathophysiological mechanisms in normal and underweight patients. N Engl J Med 314:1335–1342. Guillon G, Trueba M, Joubert D, et al (1995): Vasopressin stimulates steroid secretion in human adrenal glands: Comparison with angiotensin II effect. Endocrinology 136:1285– 1295. Holsboer F, von Bardeleben U, Gerken A, Stalla GK, Muller OA (1984a): Blunted corticotropin and normal cortisol response to human corticotropin-releasing factor (hCRF) in depression. N Engl J Med 311:1127. Holsboer F, Gerken A, Steiger A, Benkert O, Muller OA, Stalla GK (1984b): Corticotropin-releasing factor induced pituitaryadrenal responses in depression. Lancet i:55. Kasagi Y, Suda T, Horiba N, et al: DDAVP test for the diagnosis of ACTH-dependant Cushings syndrome. In: Proceedings of the 76th Annual Meeting of the Endocrine Society, Washington DC. Bethesda, MD: Endocrine Society, 1994: p 99.
1454
BIOL PSYCHIATRY 1999;45:1447–1454
Lamberts S, Verleun T, Oosterom R, de Jong F, Hackeng W (1984): Corticotropin-releasing factor (ovine) and vasopressin exert synergistic effect on adrenocorticotropin release in man. J Clin Endocrinol Metab 57:298 –303. Liu JH, Muse K, Contreras P, et al (1983): Augmentation of ACTH-releasing factor (CRF) by vasopressin in women. J Clin Endocrinol Metab 57:1087–1089. Malerbi DA, Mendonca B, Liberman B, et al (1993): The desmopressin stimulation test in the differential diagnosis of Cushings disease. Clin Endocrinol 38:463– 472. Nemeroff CB, Widerlov E, Bissette G, et al (1984): Elevated concentrations of corticotropin-releasing factor-like immunoreactivity in depressed patients. Science 226:1342–1344. Patchev V, Kalogeras K, Zelazowski P, Wilder R, Chrousos G (1992): Increased plasma concentration, hypothalamic content, and in vitro release of arginine vasopressin in inflammatory disease-prone hypothalamic corticotropinhormone deficient Lewis rats. Endocrinology 131:1453– 1457. Perraudin V, Delarue C, Lefevre H, Contesse V, Kuhn JM, Vaudry H (1993): Vasopressin stimulates cortisol secretion from human adrenocortical tissue through activation of V1 receptors. J Clin Endocrinol Metab 76:1522–1528. Purba J, Hoogendijk W, Hofman M, Swaab D (1996): Increased numbers of vasopressin and oxytocin expressing neurons in the paraventricular nucleus of the hypothalamus in depression. Arch Gen Psychiatry 53:137–143. Rabadan-Diehl C, Lolait S, Aguilera G (1995): Regulation of pituitary V1b receptor mRNA during stress in the rat. J Neuroendocrinol 7:903–910. Rado JS, Juhos E (1976): Effect of 1-deamino-8-D-arginine vasopressin (DDAVP) on plasma cortisol (hydrocortisone ). J Clin Pharmacol 16:333–337. Raff H, Findling J (1989): A new immunoradiometric assay for
L.V. Scott et al
corticotropin, evaluated in normal subjects and patients with Cushings syndrome. Clin Chem 35:595– 600. Salata RA, Jarrett DB, Verbalis JG, Robinson AG (1988): Vasopressin stimulation of adrenocorticotropin hormone (ACTH) in humans. J Clin Invest 81:766 –774. Sawyer WH, Acosta M, Balaspiri L, Judd J, Manning M (1974): Structural changes in the arginine vasopressin molecule. Endocrinology 94:1106 –1115. Scott LV, Medbak S, Dinan TG (1998a): Blunted adrenocorticotropin and cortisol responses to corticotropin releasing hormone stimulation in chronic fatigue syndrome. Acta Psychiatr Scand 97:450 – 457. Scott LV, Medbak S, Dinan TG (1998b): The low dose ACTH test in chronic fatigue syndrome and in health. Clin Endocrinol 48:733–737. Spitzer R, Williams J, Gibbon M (1987): Structured Clinical Interview for DSM III-R. New York: New York State Psychiatric Institute, Biometrics Research. Sternberg E, Hill J, Chrousos G, et al (1989): Inflammatory mediator-induced hypothalamic-pituitary-adrenal axis activation is defective in streptococcal cell wall arthritis susceptible Lewis rats. Proc Natl Acad Sci USA 86:4771– 4775. Sugimoto Y, Sato M, Mochizuki S, et al (1994): Molecular cloning and functional characterization of a cDNA encoding the human V1b vasopressin receptor. J Biol Chem 269: 27088 –27092. Todd BK, Lawrence C (1990): The Cushing syndrome: An update on diagnostic tests. Ann Intern Med 112:434 – 444. Whitnall MH (1989): Stress selectively activates the vasopressincontaining subpopulations of corticotropin-releasing hormone neurons. Neuroendocrinology 50:702–707. Winkelmann W, Heppner C, Schmitt HW, Deuss U, Nieke D, Kaulen D (1994): The desmopressin stimulation test in patients with ACTH hypersecretion. Eur J Endocrinol 130(suppl 2):151.