- Email: [email protected]
DIFFERENTIAL EFFECTS OF HYDROCORTISONE AND DEXAMETHASONE ON CORTISOL SUPPRESSION IN A CHILD PSYCHIATRIC POPULATION
Psychoneuroendocrinology, Vol. 23, No. 3, pp. 295 – 305, 1998 © 1998 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0306-4530/98 $19.00 + .00
PII: S0306-4530(97)00097-8
DIFFERENTIAL EFFECTS OF HYDROCORTISONE AND DEXAMETHASONE ON CORTISOL SUPPRESSION IN A CHILD PSYCHIATRIC POPULATION Christine C. Gispen-de Wied1, Lucres M.C. Jansen1, Herman J. Wynne3, Walter Matthys1, Rutger J. van der Gaag1, Jos H.H. Thijssen2 and Herman van Engeland1 1
Rudolf Magnus Institute for Neuroscience, Department of Child and Adolescent Psychiatry, Utrecht University, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands 2 Rudolf Magnus Institute for Neuroscience, Department of Endocrinology, Utrecht University, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands 3 Center for Biostatistics, Utrecht University, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands
(Recei6ed 17 January 1997; in final form 29 September 1997)
SUMMARY The suppressive effect of hydrocortisone and dexamethasone on salivary cortisol was investigated in a 2-year study of pituitary–adrenal function in a variety of child psychiatric patients and healthy controls. Symptomatology was assessed using the Child Behavioral Checklist (CBCL). Cortisol day profiles were assessed at 2-h intervals from 0800 to 2000h on three occasions. Dexamethasone and hydrocortisone were administered orally twice at 2000h, the doses being adjusted for bodyweight according to the standard dexamethasone suppression test. Fifty-one patients, including patients with dysthymia, oppositional defiant disorder, pervasive developmental disorder, and attention deficit hyperactivity disorder, and ten age and sex matched controls participated. Basal cortisol levels in patients were generally lower than in controls. Both dexamethasone and hydrocortisone were effective in suppressing salivary cortisol, although dexamethasone was somewhat more potent and its effect lasted longer. Hyporesponsiveness to hydrocortisone, but not to dexamethasone, distinguished patients with dysthymia and oppositional defiant disorder from controls. Responsiveness to hydrocortisone was correlated with the symptom clusters social problems and anxious/depressed. The data support the idea that there exist syndrome aspecific disturbances in feedback activity beyond the level of the pituitary, i.e. at the hypothalamic level, at an early age. From this perspective, hydrocortisone suppression is a useful tool for studying pituitary – adrenal function in children. Behavioral correlates of these disturbances of pituitary– adrenal function should be determined. © 1998 Elsevier Science Ltd. All rights reserved. Keywords—HPA; Cortisol; Dexamethasone; Children; Psychiatry.
INTRODUCTION Pituitary – adrenal function has been extensively investigated in adults with psychiatric Address correspondence and reprint requests to: C.C. Gispen-de Wied, Department of Psychiatry, University Hospital (A01.126), PO Box 85500, 3508 GA Utrecht, Netherlands (Tel.: 31 30 2508459; Fax: 31 30 2505443; E-mail: [email protected]) 295
296
B. Buwalda et al.
disorders, particularly major affective disorders (MAD). In general, MAD are characterized by a state dependent hyperactivity of the hypothalamic pituitary–adrenal (HPA) axis (Gold et al., 1988a,b). This dysfunction of the HPA axis can be detected either by an increase in cortisol output or by disturbances in the feedforward or feedback mechanism, such as occur with the corticotropin releasing hormone test (Chrousos, 1992; Holsboer et al., 1987) or dexamethasone suppression test (Carroll, 1989). However, different malfunctions of the HPA axis may exist independently of each other, in a particular diagnostic group. Therefore, pituitary – adrenal dysfunction may reflect inadequate adaptation to the environment. Pituitary – adrenal function has been investigated to a lesser extent in children. Puig-Antich et al. (1979) were one of the first to describe a state dependent hypercortisolemia in prepubertal children with MAD, similar to that seen in adults. However, the finding could not be confirmed by the same authors in a larger study on cortisol secretion in prepubertal children (Puig-Antich et al., 1989). Several investigators have reported disturbances in dexamethasone suppression in children with MAD (Birmaher et al., 1992; Pfeffer et al., 1989; Weller et al., 1984, 1985). However, as in adults, disturbances of HPA function in children were not restricted to depression, but were found in other child psychiatric disorders either alone or in combination with depression (Dahl et al., 1992) The development of tests for the measurement of cortisol in saliva has made it possible to investigate pituitary – adrenal function in children in a simple non-stressful way, although it should be remembered that cortisol reflects only the final step in the hypothalamic pituitary – adrenal feedback loop. In children, the pituitary–adrenal axis is functional at an adult level from the age of 4 years on (Genazzani et al., 1983), which validates the comparison of pituitary–adrenal function, based on cortisol measurements, in children and adults. Salivary cortisol measurements have made it possible to carry out repeated sampling in order to gain more information about basal cortisol regulation. Hypercortisolemia has been reported in depressed children (Foreman and Goodyer, 1988; Goodyer et al., 1991), but also in children with disruptive behavior (Scarpa Scerbo and Kolko, 1994) and in children with conduct disorder with anxiety (McBurnett et al., 1991). However, the opposite finding, i.e. increased basal cortisol in non-depressive children as compared with depressive children (Casat et al., 1989) or no differences between non-depressive and depressive children (Birmaher et al., 1992) or between depressive children and healthy controls (DeBellis et al., 1996) has been reported as well. Although disturbances in HPA function, `ıf present in childhood, are still thought to be related to factors such as depression, anxiety or emotion (McBurnett et al., 1991; Scarpa Scerbo and Kolko, 1994), it is difficult to determine which factors are of major importance to pituitary–adrenal function. From this perspective we investigated pituitary–adrenal function in a child psychiatric population for differential diagnostic purposes. We determined cortisol profiles from 0800 to 2000h with and without the inhibiting effect of the natural steroid hydrocortisone and the synthetic steroid dexamethasone. Both dexamethasone and hydrocortisone exert their inhibiting effect through the glucocorticoid receptor, which is mainly located in the pituitary, hypothalamus, and hippocampus and which can be distinguished from the mineralocorticoid receptor. The latter receptor type is located particularly in extrahypothalamic structures and is important in maintaining basal function and circadian rhythmicity (de Kloet, 1991). Dexamethasone is active mainly at the pituitary level, whereas hydrocortisone has more affinity for supra-pituitary glucocorticoid receptors (de Kloet, 1991). Therefore, use of the natural steroid hydrocortisone rather than dexamethasone may give a better indication of endogenous cortisol regulation. We had previously found
Hydrocortisone/Dexamethasone Inhibition
297
a differential effect of hydrocortisone and dexamethasone on plasma levels of b-endorphin and adrenocorticotropin hormone (ACTH) in adults with depression (Gispen-de Wied et al., 1987, 1993). The present study was part of a larger study on pituitary–adrenal function in a child psychiatric population, in which basal function and stress challenge were investigated. The aim of our study was 2-fold: to evaluate the reliability of measuring salivary cortisol in children during their routine daily activities and to investigate whether hydrocortisone and dexamethasone have the same effect on pituitary–adrenal function. We hypothesized that children with depressive symptomatology could be distinguished from children with other psychiatric disorders by their HPA function. Furthermore, we hypothesized that hydrocortisone would provide more information about the site of HPA dysfunction than would dexamethasone.
METHODS Subjects Over a 2-year period, children admitted for observation to the Department of Child and Adolescent Psychiatry of the University Hospital Utrecht were screened for participation in the study on pituitary–adrenal function. Patients with any endocrine, cardio-pulmonary, or organic brain disease were excluded from participation. Diagnoses were based on extensive psychiatric evaluation including information from developmental history, medical examination, a psychiatric interview and 8 weeks of clinical observation. The existence of comorbidity (Table I) was taken into account with respect to decision making. When consensus on the main diagnosis was reached by two psychiatrists (RJG, WM, or HE), patients were classified according to DSM-IV criteria. Fifty-one patients (37 boys, 14 girls; mean age 10.09 1.5 years) and ten healthy controls (eight boys, two girls; mean age 10.0 9 1.7 years) participated and were assigned to four different groups: Dysthymia (Dysth, n= 8); oppositional defiant disorder (ODD, n =14); pervasive developmental disorder, not otherwise specified (PDDNOS, n =13) and attention deficit hyperactivity disorder (ADHD, n =9). Seven patients were assigned to a rest group because they did not have a well-defined diagnosis. Symptoms were assessed by means of the Child Behavior Checklist (CBCL; Achenbach), which was completed by the parents. CBCL symptom clusters are with-
Table I. Comorbidity in patient groups Comorbidity
Dysth (n= 8)
ODD/CD ADHD Other
3 2 3
ODD/CD (n= 14) PDDNOS (n = 13) ADHD (n=9)
5
1
4
4
4
Other, either enuresis/encopresis, reactive attachment disorder, parent – child problem or tic disorder. Dysth, dysthymia; ODD/CD, oppositional defiant disorder/conduct disorder; PDDNOS, pervasive developmental disorder not otherwise specified; ADHD, attention deficit hyperactivity disorder.
298
B. Buwalda et al.
Table II. Schematic representation of the study design A.
Saliva sampling 0800 – 2000h
Drug administration 2000h
Day 1 (baseline) Day 2 (post-dex)
CCCCCCC CCCCCCC DDDDDDD
0.01 mg/kg bodyweight dex
Day 3 Day 4 (post-hydrocort)
0.7 mg/kg bodyweight hydrocort CCCCCCC
B. Dosage weight scheme Dexamethasone/hydrocortisone
Weight
0.25 mg/12.5 mg 0.5 mg/25 mg 1 mg/50 mg
20–35 kg 36–55 kg 56 – 70 kg
C, Salivary cortisol; D, salivary dexamethasone.
drawal, somatic complaints, anxious/depressed, social problems, thought problems, attention problems, delinquent behavior, aggressive behavior and sexual problems. A score above the 98th percentile is considered pathologic. Control subjects were recruited from two elementary schools. Controls with any symptom cluster score above the 98th percentile were excluded from participation. They were screened for physical abnormalities, by using a short medical checklist completed by the parents. A physical examination was not part of the procedure. The study was approved by the Ethics Committee of the University Hospital. Informed consent was obtained from parents and children. Procedure Cortisol levels were measured at least 3 weeks after patients were admitted to the hospital in order to reduce the effect of hospitalization stress on cortisol production. For all subjects, saliva collection was incorporated into their daily routine, either at the hospital or at home and at school. Parents and caretakers were asked to assist in collecting the samples. Schedules were made for each individual according to their daily activities. Except for extreme exercise, the subjects were not restricted in their activities. Samples were always obtained before meals in order to avoid the influence of food on cortisol output. Subjects were asked to provide seven saliva samples collected every 2 h from 0800 to 2000 h (Table II(a)). To make this easier, saliva production was stimulated with citric acid crystals. Saliva was collected into plastic vials and stored at −20°C until analysis. The dosage of dexamethasone was based on the 1 mg/70 kg bodyweight standard dexamethasone test as used in adults. In previous studies 50 mg/70 kg bodyweight hydrocortisone was regarded equivalent to the 1 mg/70 kg bodyweight dosage of dexamethasone (Gispen-de Wied et al., 1993). Both dexamethasone and hydrocortisone were
Hydrocortisone/Dexamethasone Inhibition
299
administered orally. For practical purposes three dosage/weight categories were used (Table II(b)). Steroid measurement Cortisol concentrations in saliva were measured without extraction using an in house competitive radioimmunoassay with a polyclonal anticortisol-antibody (K7348). [1,23 H(N)]Hydrocortisone (NET 185, NEN-DUPONT, Dreiech, Germany) was used as a tracer following chromatographic verification of its purity. The lower limit of detection was 0.5 nmol/l and inter-assay variation was 11.0, 8.2 and 7.6% at 4.7, 9.7 and 14.0 nmol/l respectively (n = 20). Dexamethasone concentrations in saliva were measured by radioimmunoassay after extraction with diethylether. Dexamethasone was obtained from Sigma (St. Louis, MO). The antiserum against dexamethasone (IgG-Dex-1) was purchased from IgG (Nashville, TN) and the [1,2,4,6,7-3H]dexamethasone was from Amsterdam Life Sciences (Houten, Netherlands). The lower limit of detection was 10 pmol/l and inter-assay variation was 14.6, 6.8, 6.7 and 6% at 41, 204, 243 and 538 pmol/l respectively. Intra-assay variation varied from 11% at 40 pmol/l to 5.1% at 1000 pmol/l (n= 10). For detail see Thijssen et al. (1996). Statistical analysis Single isolated missing values were replaced by the group averages. Differences in salivary cortisol levels were analyzed by full factorial analysis of variance (ANOVA) with three, two or one factors where appropriate. Factors were diagnostic groups with five levels (dysthymia, ODD, PDDNOS and ADHD patients, and controls). Two repeated measures factors were used i.e. times with six levels (measurements at six consecutive times) and tests with three levels (hydrocortisone, dexamethasone and baseline). We used multivariate tests for statistical analysis of data involving repeated measures factors to take into account correlated error structure and the consequent deviation from the compound symmetry condition. Null hypotheses involving individual diagnostic groups were tested using simple t-distributed contrasts (diagnostic group vs. controls). Simple contrasts were also applied to test for pairwise differences between individual levels of the factor tests. Helmert type contrasts were used to explore differences between times. Wherever two or more statistical tests had to support one single inference, the appropriate Bonferroni correction for multiple comparisons was used. Dexamethasone, hydrocortisone, and basal cortisol levels were compared with the three factors (diagnostic groups, times and tests) in the ANOVA. Single basal cortisol levels and single dexamethasone levels were analyzed with two factors i.e. times and diagnostic groups only. To measure the suppressive effect of dexamethasone and hydrocortisone, we computed two new outcome variables on the basis of cortisol levels at 0800h: Supp-D, post-dexamethasone cortisol – baseline cortisol; Supp-HC, post-hydrocortisone cortisol– baseline cortisol. Supp-D and Supp-HC measure the suppression of cortisol corrected for baseline levels at 0800h (D-cortisol). Analysis of variance of Supp-D and Supp-HC involves one factor only (diagnostic groups). Pearson correlation coefficients were calculated for CBCL symptom clusters, dexamethasone levels, dosage dexamethasone, post-dexamethasone cortisol, dosage hydrocortisone, Supp-D and Supp-HC.
300
B. Buwalda et al.
RESULTS Basal cortisol Fig. 1 shows basal cortisol levels for both patients and controls. Basal cortisol levels showed the normal diurnal rhythmicity, i.e high in the morning with a gradual statistically significant (s.s.) decrease during the day (F= 63.2, pB 0.001). Patients and controls differed from each other in that controls showed a more pronounced decrease in saliva cortisol levels than did patients (diagnostic group× time interaction, F= 2.05, pB 0.05). Pairwise simple contrasts with controls were pB 0.01 for Dysth, pB0.05 for ODD, pB 0.01 for PDDNOS and p =0.2 for ADHD). Dexamethasone and hydrocortisone inhibition Fig. 2 shows the effect of dexamethasone and hydrocortisone on cortisol levels for all subjects. Mean cortisol levels between tests differed s.s. (F=28.37, pB 0.001). Both dexamethasone and hydrocortisone suppressed saliva cortisol levels (t= −7.57, pB 0.001 for dexamethasone and t = − 4.61, pB 0.001 for hydrocortisone as compared with basal levels). Statistically significant different time profiles were found for the three tests (test ×time interaction, F =5.57, pB 0.001). Dexamethasone was somewhat more potent than hydrocortisone in decreasing cortisol levels and its effect lasted longer (at time 0800,
Fig. 1. Basal levels of salivary cortisol (mean 9 SEM) in controls (--), and in patients with dysthymia (--), oppositional defiant disorder (--), pervasive developmental disorder (- -) and attention deficit hyperactivity disorder (-"-).
Hydrocortisone/Dexamethasone Inhibition
301
Fig. 2. Basal (mean9SEM) cortisol levels (closed circles), post-dexamethasone cortisol levels (open triangles) and post-hydrocortisone cortisol levels (open circles) in all subjects (n= 61).
1000 and 1200h p B0.05 for dexamethasone and at time 0800h pB 0.05 for hydrocortisone). Differences between patients and controls. Table III shows the baseline corrected cortisol levels (Supp-D and Supp-HC) produced by dexamethasone and hydrocortisone in patients and controls. The effect of hydrocortiTable III. Cortisol suppression ( mean 9 SD) by dexamethasone ( Supp-D) and hydrocortisone ( Supp-HC) at 0800 h Controls Number Weight Dose dexamethasone Dose H-cortisone Supp-D Supp-HC
Dysth.
ODD/CD
PDDNOS
10 8 14 13 kg 37.0 98.6 30.296.9 35.196.8 33.798 mg 0.490.1 0.3 9 0.1 0.490.1 0.39 0.1 mg 209 6.4 14.194.4 179 6.2 16.79 6.2 D −10.69 2 −3.39 2.1 −4.89 2 −6.79 1.7 D −8.79 1.8 1.89 2.3* −1.39 2* −4.191.2
ADHD 9 31.195.3 0.39 0.1 15.3 95.5 −6.19 2.3 −4.59 2.2
Dysth, dysthymia; ODD/CD, oppositional defiant disorder/conduct disorder; PDDNOS, pervasive developmental disorder not otherwise specified; ADHD, attention deficit hyperactivity disorder. * pB0.05 Different from controls.
302
B. Buwalda et al.
Fig. 3. Salivary dexamethasone (mean 9 SEM) in controls (--), and in patients with dysthymia (--), oppositional defiant disorder (--), pervasive developmental disorder (- -) and attention deficit hyperactivity disorder (-"-).
sone on cortisol levels, but not that of dexamethasone, distinguished the patient groups from controls (ANOVA F =3.59, pB 0.02 and F= 1.64, p= 0.2 respectively). This was due to the s.s. difference between the response of patients with dysthymia and patients with ODD and the response of controls, with pairwise simple contrasts with controls, p=0.001 for Dysth, p =0.006 for ODD, p= 0.1 for PDDNOS and p = 0.1 for ADHD (Fig. 3). The symptom clusters social problems and anxious/depressed were positively correlated with Supp-HC (r =0.41; p =0.005 and r= 0.37; p= 0.01 respectively). The patients with dysthymia had the highest scores for both symptom clusters. No correlations were found between symptom cluster scores and Supp-D. Sali6ary dexamethasone Fig. 3 shows the saliva levels of dexamethasone in patients and controls. In both patients and controls dexamethasone levels were high in the morning and decreased during the day (overall time effect F = 22.3, pB 0.001). No differences in time profile were found between patient groups and controls (diagnostic group× time interaction, F= 1.17, p = 0.3). Interestingly, dexamethasone was detectable in saliva 24 h after ingestion. No correlation was found between saliva dexamethasone levels and post-dexamethasone cortisol levels in the morning (r = −0.16, p = 0.2), and neither was there a correlation
Hydrocortisone/Dexamethasone Inhibition
303
between the dose dexamethasone used and the salivary dexamethasone levels in the morning (r = −0.006, p =0.9). A negative correlation was found between the dose of dexamethasone used and Supp-D (r= − 0.41, p = 0.001). A similar negative correlation was found between the dose of hydrocortisone and Supp-HC (r= − 0.25, p= 0.05).
DISCUSSION In this study basal cortisol levels were generally lower in patients than in controls. The cortisol levels of both patients and controls varied within the range of average cortisol levels during the day as reported by Kirschbaum and Hellhammer (1989) for adults. The fact that we did not find hypercortisolemia in the depressed children may have been due to their classification as dysthymics or the existence of a fairly amount of comorbidity in this group of patients. Hypercortisolemia, `ıf present, may be found in patients with MAD (Foreman and Goodyer, 1988; Goodyer et al., 1991), but may be absent as well (DeBellis et al., 1996). To our knowledge, there are no studies on basal cortisol levels in dysthymic children. Emslie et al. (1987) report less disturbance on the DST in children with dysthymia than in children with MAD, indicating that both syndromes share similarities and dissimilarities in pathophysiology. Our study included children with a variety of psychiatric disorders, all of which are common in child psychiatry. Interestingly, the basal cortisol levels of the patients were similar, irrespective of the diagnosis. Although the groups were small, this indicates that, for a differential diagnosis basal cortisol levels may not be the most informative parameter of disrupted pituitary–adrenal function. A possible reason for the slightly increased levels of salivary cortisol in our control group may be the novelty of the experiment for these children as compared to the institutionally organized life of the hospitalized children. Such novelty stress has been reported in children (Tennes and Kreye, 1985). Both dexamethasone and hydrocortisone were effective in suppressing salivary cortisol in healthy control children and children with various psychiatric disorders. To our knowledge this is the first study to show an effect on pituitary–adrenal function of the natural steroid hydrocortisone in children. Provided only small amounts of this steroid are used, so as not to interfere with endogenous production the day following ingestion, hydrocortisone has a suppressive effect, similar to that of dexamethasone. Dexamethasone, however, is somewhat more potent and its effect lasts longer, due to the difference in half-life (Peterson, 1959). Since we administered dexamethasone and hydrocortisone according to bodyweight, the maximum suppressive effect was best evaluated by determining cortisol suppression corrected for baseline levels in the morning. With this approach, the dexamethasone-induced decrease in cortisol levels, although less in both dysthymics and ODD, could not distinguish patients from controls. These findings are consistent with reports in the literature that dexamethasone does not consistently suppress cortisol secretion, irrespective of basal cortisol levels (Dahl et al., 1992; Kaneko et al., 1993). Surprisingly, a significant hyporesponsiveness to hydrocortisone was found in children with dysthymia and patients with oppositional defiant disorder. Only few studies have incorporated the latter group of patients in studies on dexamethasone suppression and generally no disturbances were found (Doherty et al., 1986; Livingston et al., 1984; Petty et al., 1985). Our data correspond with the data reported by Birmaher et al. (1992). They showed in a 24-h study of cortisol before and after administration of dexamethasone, that the first effect of dexamethasone is to block pituitary corticotropin release, resulting in an
304
B. Buwalda et al.
absence of cortisol for several hours. Thereafter cortisol levels start to rise, but they remain low, indicating interference with corticotropin synthesis. The latter process is a gene-mediated intracellular process and is probably less dependent on the actual steroid plasma level (De Kloet, 1991). Using a recently developed assay for dexamethasone in saliva (Thijssen et al., 1996), we found no difference in dexamethasone levels between patients and controls. Moreover, there was no relation between saliva dexamethasone levels and post-dexamethasone cortisol levels or between saliva dexamethasone levels and the dose of dexamethasone used. Therefore, levels of dexamethasone in saliva measured the day following dexamethasone ingestion may not be very helpful in interpreting the suppressive effect of the steroid. These findings contradict the findings of Arana et al. (1984), Carr et al. (1984) and Casat et al. (1989) but are in line with reports by Naylor et al. (1990) and Birmaher et al. (1992). Studies of the two steroid receptor types in the brain (mineralocorticoid and glucocorticoid), provide increasing evidence that the regulation of HPA function in experimental animals reflects a balance between glucocorticoid activity at mineralocorticoid and glucocorticoid receptors in the hippocampus, hypothalamus, and pituitary (De Kloet, 1991). The affinity of dexamethasone for the glucocorticoid receptor is highest in the pituitary, whereas hydrocortisone has more affinity for the glucocorticoid and mineralocorticoid receptors in the hypothalamus. Therefore, by using the natural steroid hydrocortisone, it is possible to assess the natural feedback activity at a level beyond the pituitary, i.e. hypothalamus, which is less sensitive to dexamethasone. Obviously, impairment in feedback activity is not specific for depression, and can be found in other child psychiatric disorders as well. In our study, symptom clusters such as social problems and anxious/depressed on the CBCL were related to hyporesponsiveness to hydrocortisone. No such relation between symptom clusters and the suppressive effect of dexamethasone was found. In conclusion, both basal cortisol levels and cortisol levels after ingestion of either dexamethasone or hydrocortisone showed the expected rhythmicity and suppression without much fluctuation. Therefore, salivary cortisol seems a stable measure of pituitary– adrenal function in children during their routine daily activities, and it is not necessary to take extra precautions, but to control food intake and extreme exercise. Pituitary-adrenal dysfunction can be objectified in child psychiatric disorders, the malfunctions not being restricted to depression. The use of hydrocortisone in low doses may be preferred over dexamethasone, since it may mimic more closely endogenous cortisol regulation. However, the aspecificity of these disturbances in feedback activity may reflect a general maladaptation of patients with psychiatric disorders to the environment rather than that these disturbances are intrinsic to a specific syndrome. Behavioral and symptomatological correlates should be determined for these HPA disturbances.
REFERENCES Arana, G. W., Workman, R. J. and Baldessarini, R. J. (1984) Association between low plasma levels of dexamethasone and elevated levels of cortisol in psychiatric patients given dexamethasone. Am J Psychiatry 141, 619–1620. Birmaher, B., Ryan, N. D., Dahl, R., Rabinovich, H., Ambrosini, P., Williamson, D. E., Novacenko, H., Nelson, B., Lo, E. S. and Puig Antich, J. (1992) Dexamethasone suppression test in children with major depressive disorder. J Am Acad Child Adolesc Psychiatry, 31, 291 – 297. Carr, D. B., Wool, C., Lydiard, R. B., Fisher, J., Gelenberg, A. and Klerman, G. (1984) Rate-sensitive inhibition of ACTH release in depression. Am J Psychiatry 141, 590 – 592.
Hydrocortisone/Dexamethasone Inhibition
305
Caroll, B. J. (1989) Diagnostic validity and laboratory studies: Rules of the game, In: Robbins, L. N., Barrett, J. E. (Eds), The Validity of Psychiatric Diagnosis, Raven Press, New York, pp. 229–245. Casat, C. D., Arana, G. D. and Powel, K. (1989) The DST in children and adolescents with major depressive disorder. Am J Psychiatry, 146, 503 – 507. Chrousos, G. P. (1992) Regulation and dysregulation of the hypothalamic – pituitary – adrenal axis—The corticotropin-releasing hormone perspective. Neuroendocrinology 21, 833 – 858. Dahl, R. E., Kaufman, J., Ryan, N. D., Perel, J., al Shabbout, M., Birmaher, B., Nelson, B. and Puig Antich, J. (1992) The dexamethasone suppression test in children and adolescents: a review and a controlled study. Biol Psychiatry 32, 109 – 126. DeBellis, M. D., Dahl, E. R., Perel, J., Birmaher, B., Al Shabbout, M., Williamson, D. E., Nelson, B. and Ryan, N. D. (1996) Nocturnal ACTH, cortisol, growth hormone, and prolactin secretion in prepubertal depression. J Am Acad Child Adolesc Psychiatry 35, 1130 – 1138. De Kloet, E. R. (1991) Brain corticosteroid receptor balance and homeostatic control. Front Neuroendocrinol 12, 95–164. Doherty, M., Madansky, D., Kraft, J., Carter-Ake, L., Rosenthal, P. and Coughlin, B. (1986) Cortisol dynamics and test performance of the dexamethasone suppression test in 97 psychiatrically hospitalized children aged 3– 16 years. J Am Acad Child Psychiatry 25, 400 – 408. Emslie, G., Weinberg, W., Rush, A. J., Weissenburger, J. and Parkin-Feigenbaum, L. (1987) Depression and dexamethasone suppression testing in children and adolescents. J Child Neurol 2, 31–37. Foreman, D. M. and Goodyer, I. M. (1988) Salivary cortisol hypersecretion in juvenile depression. J Child Psychol Psychiatry 29, 311 – 320. Genazzani, A. R., Facchinetti, F., Pintor, C., Puggioni, R., Parrini, D., Petraglia, F., Bagnoli, F. and Corda, R. (1983) Proopiocortin-related peptide plasma levels throughout prepuberty and puberty. J Clin Endocrinol Metab 57, 56–61. Gispen-de Wied, C. C., D’Haenen, H., Verhoeven, W. M. A., Wynne, H. J., Westenberg, H. G. M., Thijssen, J. H. H. and Van Ree, J. M. (1993) Inhibition of the pituitary-adrenal axis with dexamethasone and cortisol in depressed patients and healthy subjects: a dose response study. Psychoneuroendocrinology 18, 191– 204. Gispen-de Wied, C. C., Westenberg, H. G. M., Thijssen J. H. H. and Van Ree, J. M. (1987) The dexamethasone and cortisol suppression test in depression: beta-endorphin as a useful marker. Psychoneuroendocrinology 12, 355– 366. Gold, P. W., Goodwin, F. K. and Chrousos, G. P. (1988a) Clinical and biochemical manifestations of depression. Relation to the neurobiology of stress (1). N Engl J Med 319, 348 – 353. Gold, P. W., Goodwin, F. K. and Chrousos, G. P. (1988b) Clinical and biochemical manifestations of depression. Relation to the neurobiology of stress (2). N Engl J Med 319, 413 – 420. Goodyer, I., Herbert, J., Moor, S. and Altham, P. (1991) Cortisol hypersecretion in depressed school-aged children and adolescents. Psychiatry Res 37, 237 – 244. Holsboer, F., Gerken, A., Stalla, G. K. and Muller, O. A. (1987) Blunted aldosterone and ACTH release after human CRH administration in depressed patients. Am J Psychiatry 144, 229 – 231. Kaneko, M., Hoshino, Y., Hashimoto, S., Okano, T. and Kumashiro, H. (1993) Hypothalamic– pituitary–adrenal axis function in children with attention-deficit hyperactivity disorder. J Autism De6 Disord 23(1), 59–65. Kirschbaum, C. and Hellhammer, D. H. (1989) Salivary cortisol in psychobiological research: an overview. Neuropsychobiology 22, 150 – 169. Livingston, R., Reis, C. and Ringkahl, I. (1984) Abnormal dexamethasone suppression test results in depressed and nondepressed children. Am J Psychiatry 141, 106 – 108. McBurnett, K., Lahey, B. B., Frick, P. J., Risch, C., Loeber, R., Hart, E. L., Christ, M. A. and Hanson, K. S. (1991) Anxiety, inhibition, and conduct disorder in children: II Relation to salivary cortisol. J Am Acad Child Adolesc Psychiatry 30, 192 – 196. Naylor, M. W., Greden, J. F. and Alessi, N. E. (1990) Plasma dexamethasone levels in children given the dexamethasone suppression test. Biol Psychiatry 27, 592 – 600. Petty, L., Asarnow, J., Carlson, G. and Lesser, L. (1985) The dexamethasone suppression test in depressed, dysthymic, and nondepressed children. Am J Psychiatry 142, 631 – 633. Peterson, R. E. (1959) Metabolism of adrenocorticoids in man. Ann NY Acad Sci 82, 846. Pfeffer, C., Stokes, P. and Weiner, A. et al. (1989) Psychopathology and plasma cortisol responses to dexamethasone in prepubertal psychiatric inpatients. Biol Psychiatry 26, 677 – 689.
306
B. Buwalda et al.
Puig-Antich, J., Chambers, W. and Halpern, F. (1979) Cortisol hypersecretion in prepubertal major depressive illness: a preliminary report. Psychoneuroendocrinology 4, 191 – 197. Puig-Antich, J., Dahl, R., Ryan, N., Novacenko, H., Goetz, D., Goetz, R., Twomey, J. and Klepper, T. (1989) Cortisol secretion in prepubertal children with major depressive disorder. Episode and recovery. Arch Gen Psychiatry 46, 801 – 809. Scarpa Scerbo, A. and Kolko, D. J. (1994) Salivary testosterone and cortisol in disruptive children: relationship to aggressive, hyperactive, and internalizing behaviors. J Am Acad Child Adolesc Psychiatry 33(8), 1174–1184. Tennes, K. and Kreye, M. (1985) Children’s adrenocortical response to classroom activities and tests in elementary school. Psychosom Med 47, 451 – 460. Thijssen, J. H. H., Gispen-de Wied, C. C., Van Heeswijk, G. M. and Veeman, W. (1996) The determination of dexamethasone in saliva. Clin Chem 42(8), 1238 – 1242. Weller, E., Weller, R., Fristad, M. and Preskorn, S. (1984) The dexamethasone suppression test in hospitalized prepubertal children. Am J Psychiatry 141, 290 – 291. Weller, E., Weller, R., Fristad, M., Preskorn, S. and Teare, M. (1985) The dexamethasone suppression test in hospitalized prepubertal children. J Clin Psychiatry 46, 511 – 513.
. .
PII: S0306-4530(97)00097-8
DIFFERENTIAL EFFECTS OF HYDROCORTISONE AND DEXAMETHASONE ON CORTISOL SUPPRESSION IN A CHILD PSYCHIATRIC POPULATION Christine C. Gispen-de Wied1, Lucres M.C. Jansen1, Herman J. Wynne3, Walter Matthys1, Rutger J. van der Gaag1, Jos H.H. Thijssen2 and Herman van Engeland1 1
Rudolf Magnus Institute for Neuroscience, Department of Child and Adolescent Psychiatry, Utrecht University, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands 2 Rudolf Magnus Institute for Neuroscience, Department of Endocrinology, Utrecht University, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands 3 Center for Biostatistics, Utrecht University, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands
(Recei6ed 17 January 1997; in final form 29 September 1997)
SUMMARY The suppressive effect of hydrocortisone and dexamethasone on salivary cortisol was investigated in a 2-year study of pituitary–adrenal function in a variety of child psychiatric patients and healthy controls. Symptomatology was assessed using the Child Behavioral Checklist (CBCL). Cortisol day profiles were assessed at 2-h intervals from 0800 to 2000h on three occasions. Dexamethasone and hydrocortisone were administered orally twice at 2000h, the doses being adjusted for bodyweight according to the standard dexamethasone suppression test. Fifty-one patients, including patients with dysthymia, oppositional defiant disorder, pervasive developmental disorder, and attention deficit hyperactivity disorder, and ten age and sex matched controls participated. Basal cortisol levels in patients were generally lower than in controls. Both dexamethasone and hydrocortisone were effective in suppressing salivary cortisol, although dexamethasone was somewhat more potent and its effect lasted longer. Hyporesponsiveness to hydrocortisone, but not to dexamethasone, distinguished patients with dysthymia and oppositional defiant disorder from controls. Responsiveness to hydrocortisone was correlated with the symptom clusters social problems and anxious/depressed. The data support the idea that there exist syndrome aspecific disturbances in feedback activity beyond the level of the pituitary, i.e. at the hypothalamic level, at an early age. From this perspective, hydrocortisone suppression is a useful tool for studying pituitary – adrenal function in children. Behavioral correlates of these disturbances of pituitary– adrenal function should be determined. © 1998 Elsevier Science Ltd. All rights reserved. Keywords—HPA; Cortisol; Dexamethasone; Children; Psychiatry.
INTRODUCTION Pituitary – adrenal function has been extensively investigated in adults with psychiatric Address correspondence and reprint requests to: C.C. Gispen-de Wied, Department of Psychiatry, University Hospital (A01.126), PO Box 85500, 3508 GA Utrecht, Netherlands (Tel.: 31 30 2508459; Fax: 31 30 2505443; E-mail: [email protected]) 295
296
B. Buwalda et al.
disorders, particularly major affective disorders (MAD). In general, MAD are characterized by a state dependent hyperactivity of the hypothalamic pituitary–adrenal (HPA) axis (Gold et al., 1988a,b). This dysfunction of the HPA axis can be detected either by an increase in cortisol output or by disturbances in the feedforward or feedback mechanism, such as occur with the corticotropin releasing hormone test (Chrousos, 1992; Holsboer et al., 1987) or dexamethasone suppression test (Carroll, 1989). However, different malfunctions of the HPA axis may exist independently of each other, in a particular diagnostic group. Therefore, pituitary – adrenal dysfunction may reflect inadequate adaptation to the environment. Pituitary – adrenal function has been investigated to a lesser extent in children. Puig-Antich et al. (1979) were one of the first to describe a state dependent hypercortisolemia in prepubertal children with MAD, similar to that seen in adults. However, the finding could not be confirmed by the same authors in a larger study on cortisol secretion in prepubertal children (Puig-Antich et al., 1989). Several investigators have reported disturbances in dexamethasone suppression in children with MAD (Birmaher et al., 1992; Pfeffer et al., 1989; Weller et al., 1984, 1985). However, as in adults, disturbances of HPA function in children were not restricted to depression, but were found in other child psychiatric disorders either alone or in combination with depression (Dahl et al., 1992) The development of tests for the measurement of cortisol in saliva has made it possible to investigate pituitary – adrenal function in children in a simple non-stressful way, although it should be remembered that cortisol reflects only the final step in the hypothalamic pituitary – adrenal feedback loop. In children, the pituitary–adrenal axis is functional at an adult level from the age of 4 years on (Genazzani et al., 1983), which validates the comparison of pituitary–adrenal function, based on cortisol measurements, in children and adults. Salivary cortisol measurements have made it possible to carry out repeated sampling in order to gain more information about basal cortisol regulation. Hypercortisolemia has been reported in depressed children (Foreman and Goodyer, 1988; Goodyer et al., 1991), but also in children with disruptive behavior (Scarpa Scerbo and Kolko, 1994) and in children with conduct disorder with anxiety (McBurnett et al., 1991). However, the opposite finding, i.e. increased basal cortisol in non-depressive children as compared with depressive children (Casat et al., 1989) or no differences between non-depressive and depressive children (Birmaher et al., 1992) or between depressive children and healthy controls (DeBellis et al., 1996) has been reported as well. Although disturbances in HPA function, `ıf present in childhood, are still thought to be related to factors such as depression, anxiety or emotion (McBurnett et al., 1991; Scarpa Scerbo and Kolko, 1994), it is difficult to determine which factors are of major importance to pituitary–adrenal function. From this perspective we investigated pituitary–adrenal function in a child psychiatric population for differential diagnostic purposes. We determined cortisol profiles from 0800 to 2000h with and without the inhibiting effect of the natural steroid hydrocortisone and the synthetic steroid dexamethasone. Both dexamethasone and hydrocortisone exert their inhibiting effect through the glucocorticoid receptor, which is mainly located in the pituitary, hypothalamus, and hippocampus and which can be distinguished from the mineralocorticoid receptor. The latter receptor type is located particularly in extrahypothalamic structures and is important in maintaining basal function and circadian rhythmicity (de Kloet, 1991). Dexamethasone is active mainly at the pituitary level, whereas hydrocortisone has more affinity for supra-pituitary glucocorticoid receptors (de Kloet, 1991). Therefore, use of the natural steroid hydrocortisone rather than dexamethasone may give a better indication of endogenous cortisol regulation. We had previously found
Hydrocortisone/Dexamethasone Inhibition
297
a differential effect of hydrocortisone and dexamethasone on plasma levels of b-endorphin and adrenocorticotropin hormone (ACTH) in adults with depression (Gispen-de Wied et al., 1987, 1993). The present study was part of a larger study on pituitary–adrenal function in a child psychiatric population, in which basal function and stress challenge were investigated. The aim of our study was 2-fold: to evaluate the reliability of measuring salivary cortisol in children during their routine daily activities and to investigate whether hydrocortisone and dexamethasone have the same effect on pituitary–adrenal function. We hypothesized that children with depressive symptomatology could be distinguished from children with other psychiatric disorders by their HPA function. Furthermore, we hypothesized that hydrocortisone would provide more information about the site of HPA dysfunction than would dexamethasone.
METHODS Subjects Over a 2-year period, children admitted for observation to the Department of Child and Adolescent Psychiatry of the University Hospital Utrecht were screened for participation in the study on pituitary–adrenal function. Patients with any endocrine, cardio-pulmonary, or organic brain disease were excluded from participation. Diagnoses were based on extensive psychiatric evaluation including information from developmental history, medical examination, a psychiatric interview and 8 weeks of clinical observation. The existence of comorbidity (Table I) was taken into account with respect to decision making. When consensus on the main diagnosis was reached by two psychiatrists (RJG, WM, or HE), patients were classified according to DSM-IV criteria. Fifty-one patients (37 boys, 14 girls; mean age 10.09 1.5 years) and ten healthy controls (eight boys, two girls; mean age 10.0 9 1.7 years) participated and were assigned to four different groups: Dysthymia (Dysth, n= 8); oppositional defiant disorder (ODD, n =14); pervasive developmental disorder, not otherwise specified (PDDNOS, n =13) and attention deficit hyperactivity disorder (ADHD, n =9). Seven patients were assigned to a rest group because they did not have a well-defined diagnosis. Symptoms were assessed by means of the Child Behavior Checklist (CBCL; Achenbach), which was completed by the parents. CBCL symptom clusters are with-
Table I. Comorbidity in patient groups Comorbidity
Dysth (n= 8)
ODD/CD ADHD Other
3 2 3
ODD/CD (n= 14) PDDNOS (n = 13) ADHD (n=9)
5
1
4
4
4
Other, either enuresis/encopresis, reactive attachment disorder, parent – child problem or tic disorder. Dysth, dysthymia; ODD/CD, oppositional defiant disorder/conduct disorder; PDDNOS, pervasive developmental disorder not otherwise specified; ADHD, attention deficit hyperactivity disorder.
298
B. Buwalda et al.
Table II. Schematic representation of the study design A.
Saliva sampling 0800 – 2000h
Drug administration 2000h
Day 1 (baseline) Day 2 (post-dex)
CCCCCCC CCCCCCC DDDDDDD
0.01 mg/kg bodyweight dex
Day 3 Day 4 (post-hydrocort)
0.7 mg/kg bodyweight hydrocort CCCCCCC
B. Dosage weight scheme Dexamethasone/hydrocortisone
Weight
0.25 mg/12.5 mg 0.5 mg/25 mg 1 mg/50 mg
20–35 kg 36–55 kg 56 – 70 kg
C, Salivary cortisol; D, salivary dexamethasone.
drawal, somatic complaints, anxious/depressed, social problems, thought problems, attention problems, delinquent behavior, aggressive behavior and sexual problems. A score above the 98th percentile is considered pathologic. Control subjects were recruited from two elementary schools. Controls with any symptom cluster score above the 98th percentile were excluded from participation. They were screened for physical abnormalities, by using a short medical checklist completed by the parents. A physical examination was not part of the procedure. The study was approved by the Ethics Committee of the University Hospital. Informed consent was obtained from parents and children. Procedure Cortisol levels were measured at least 3 weeks after patients were admitted to the hospital in order to reduce the effect of hospitalization stress on cortisol production. For all subjects, saliva collection was incorporated into their daily routine, either at the hospital or at home and at school. Parents and caretakers were asked to assist in collecting the samples. Schedules were made for each individual according to their daily activities. Except for extreme exercise, the subjects were not restricted in their activities. Samples were always obtained before meals in order to avoid the influence of food on cortisol output. Subjects were asked to provide seven saliva samples collected every 2 h from 0800 to 2000 h (Table II(a)). To make this easier, saliva production was stimulated with citric acid crystals. Saliva was collected into plastic vials and stored at −20°C until analysis. The dosage of dexamethasone was based on the 1 mg/70 kg bodyweight standard dexamethasone test as used in adults. In previous studies 50 mg/70 kg bodyweight hydrocortisone was regarded equivalent to the 1 mg/70 kg bodyweight dosage of dexamethasone (Gispen-de Wied et al., 1993). Both dexamethasone and hydrocortisone were
Hydrocortisone/Dexamethasone Inhibition
299
administered orally. For practical purposes three dosage/weight categories were used (Table II(b)). Steroid measurement Cortisol concentrations in saliva were measured without extraction using an in house competitive radioimmunoassay with a polyclonal anticortisol-antibody (K7348). [1,23 H(N)]Hydrocortisone (NET 185, NEN-DUPONT, Dreiech, Germany) was used as a tracer following chromatographic verification of its purity. The lower limit of detection was 0.5 nmol/l and inter-assay variation was 11.0, 8.2 and 7.6% at 4.7, 9.7 and 14.0 nmol/l respectively (n = 20). Dexamethasone concentrations in saliva were measured by radioimmunoassay after extraction with diethylether. Dexamethasone was obtained from Sigma (St. Louis, MO). The antiserum against dexamethasone (IgG-Dex-1) was purchased from IgG (Nashville, TN) and the [1,2,4,6,7-3H]dexamethasone was from Amsterdam Life Sciences (Houten, Netherlands). The lower limit of detection was 10 pmol/l and inter-assay variation was 14.6, 6.8, 6.7 and 6% at 41, 204, 243 and 538 pmol/l respectively. Intra-assay variation varied from 11% at 40 pmol/l to 5.1% at 1000 pmol/l (n= 10). For detail see Thijssen et al. (1996). Statistical analysis Single isolated missing values were replaced by the group averages. Differences in salivary cortisol levels were analyzed by full factorial analysis of variance (ANOVA) with three, two or one factors where appropriate. Factors were diagnostic groups with five levels (dysthymia, ODD, PDDNOS and ADHD patients, and controls). Two repeated measures factors were used i.e. times with six levels (measurements at six consecutive times) and tests with three levels (hydrocortisone, dexamethasone and baseline). We used multivariate tests for statistical analysis of data involving repeated measures factors to take into account correlated error structure and the consequent deviation from the compound symmetry condition. Null hypotheses involving individual diagnostic groups were tested using simple t-distributed contrasts (diagnostic group vs. controls). Simple contrasts were also applied to test for pairwise differences between individual levels of the factor tests. Helmert type contrasts were used to explore differences between times. Wherever two or more statistical tests had to support one single inference, the appropriate Bonferroni correction for multiple comparisons was used. Dexamethasone, hydrocortisone, and basal cortisol levels were compared with the three factors (diagnostic groups, times and tests) in the ANOVA. Single basal cortisol levels and single dexamethasone levels were analyzed with two factors i.e. times and diagnostic groups only. To measure the suppressive effect of dexamethasone and hydrocortisone, we computed two new outcome variables on the basis of cortisol levels at 0800h: Supp-D, post-dexamethasone cortisol – baseline cortisol; Supp-HC, post-hydrocortisone cortisol– baseline cortisol. Supp-D and Supp-HC measure the suppression of cortisol corrected for baseline levels at 0800h (D-cortisol). Analysis of variance of Supp-D and Supp-HC involves one factor only (diagnostic groups). Pearson correlation coefficients were calculated for CBCL symptom clusters, dexamethasone levels, dosage dexamethasone, post-dexamethasone cortisol, dosage hydrocortisone, Supp-D and Supp-HC.
300
B. Buwalda et al.
RESULTS Basal cortisol Fig. 1 shows basal cortisol levels for both patients and controls. Basal cortisol levels showed the normal diurnal rhythmicity, i.e high in the morning with a gradual statistically significant (s.s.) decrease during the day (F= 63.2, pB 0.001). Patients and controls differed from each other in that controls showed a more pronounced decrease in saliva cortisol levels than did patients (diagnostic group× time interaction, F= 2.05, pB 0.05). Pairwise simple contrasts with controls were pB 0.01 for Dysth, pB0.05 for ODD, pB 0.01 for PDDNOS and p =0.2 for ADHD). Dexamethasone and hydrocortisone inhibition Fig. 2 shows the effect of dexamethasone and hydrocortisone on cortisol levels for all subjects. Mean cortisol levels between tests differed s.s. (F=28.37, pB 0.001). Both dexamethasone and hydrocortisone suppressed saliva cortisol levels (t= −7.57, pB 0.001 for dexamethasone and t = − 4.61, pB 0.001 for hydrocortisone as compared with basal levels). Statistically significant different time profiles were found for the three tests (test ×time interaction, F =5.57, pB 0.001). Dexamethasone was somewhat more potent than hydrocortisone in decreasing cortisol levels and its effect lasted longer (at time 0800,
Fig. 1. Basal levels of salivary cortisol (mean 9 SEM) in controls (--), and in patients with dysthymia (--), oppositional defiant disorder (--), pervasive developmental disorder (- -) and attention deficit hyperactivity disorder (-"-).
Hydrocortisone/Dexamethasone Inhibition
301
Fig. 2. Basal (mean9SEM) cortisol levels (closed circles), post-dexamethasone cortisol levels (open triangles) and post-hydrocortisone cortisol levels (open circles) in all subjects (n= 61).
1000 and 1200h p B0.05 for dexamethasone and at time 0800h pB 0.05 for hydrocortisone). Differences between patients and controls. Table III shows the baseline corrected cortisol levels (Supp-D and Supp-HC) produced by dexamethasone and hydrocortisone in patients and controls. The effect of hydrocortiTable III. Cortisol suppression ( mean 9 SD) by dexamethasone ( Supp-D) and hydrocortisone ( Supp-HC) at 0800 h Controls Number Weight Dose dexamethasone Dose H-cortisone Supp-D Supp-HC
Dysth.
ODD/CD
PDDNOS
10 8 14 13 kg 37.0 98.6 30.296.9 35.196.8 33.798 mg 0.490.1 0.3 9 0.1 0.490.1 0.39 0.1 mg 209 6.4 14.194.4 179 6.2 16.79 6.2 D −10.69 2 −3.39 2.1 −4.89 2 −6.79 1.7 D −8.79 1.8 1.89 2.3* −1.39 2* −4.191.2
ADHD 9 31.195.3 0.39 0.1 15.3 95.5 −6.19 2.3 −4.59 2.2
Dysth, dysthymia; ODD/CD, oppositional defiant disorder/conduct disorder; PDDNOS, pervasive developmental disorder not otherwise specified; ADHD, attention deficit hyperactivity disorder. * pB0.05 Different from controls.
302
B. Buwalda et al.
Fig. 3. Salivary dexamethasone (mean 9 SEM) in controls (--), and in patients with dysthymia (--), oppositional defiant disorder (--), pervasive developmental disorder (- -) and attention deficit hyperactivity disorder (-"-).
sone on cortisol levels, but not that of dexamethasone, distinguished the patient groups from controls (ANOVA F =3.59, pB 0.02 and F= 1.64, p= 0.2 respectively). This was due to the s.s. difference between the response of patients with dysthymia and patients with ODD and the response of controls, with pairwise simple contrasts with controls, p=0.001 for Dysth, p =0.006 for ODD, p= 0.1 for PDDNOS and p = 0.1 for ADHD (Fig. 3). The symptom clusters social problems and anxious/depressed were positively correlated with Supp-HC (r =0.41; p =0.005 and r= 0.37; p= 0.01 respectively). The patients with dysthymia had the highest scores for both symptom clusters. No correlations were found between symptom cluster scores and Supp-D. Sali6ary dexamethasone Fig. 3 shows the saliva levels of dexamethasone in patients and controls. In both patients and controls dexamethasone levels were high in the morning and decreased during the day (overall time effect F = 22.3, pB 0.001). No differences in time profile were found between patient groups and controls (diagnostic group× time interaction, F= 1.17, p = 0.3). Interestingly, dexamethasone was detectable in saliva 24 h after ingestion. No correlation was found between saliva dexamethasone levels and post-dexamethasone cortisol levels in the morning (r = −0.16, p = 0.2), and neither was there a correlation
Hydrocortisone/Dexamethasone Inhibition
303
between the dose dexamethasone used and the salivary dexamethasone levels in the morning (r = −0.006, p =0.9). A negative correlation was found between the dose of dexamethasone used and Supp-D (r= − 0.41, p = 0.001). A similar negative correlation was found between the dose of hydrocortisone and Supp-HC (r= − 0.25, p= 0.05).
DISCUSSION In this study basal cortisol levels were generally lower in patients than in controls. The cortisol levels of both patients and controls varied within the range of average cortisol levels during the day as reported by Kirschbaum and Hellhammer (1989) for adults. The fact that we did not find hypercortisolemia in the depressed children may have been due to their classification as dysthymics or the existence of a fairly amount of comorbidity in this group of patients. Hypercortisolemia, `ıf present, may be found in patients with MAD (Foreman and Goodyer, 1988; Goodyer et al., 1991), but may be absent as well (DeBellis et al., 1996). To our knowledge, there are no studies on basal cortisol levels in dysthymic children. Emslie et al. (1987) report less disturbance on the DST in children with dysthymia than in children with MAD, indicating that both syndromes share similarities and dissimilarities in pathophysiology. Our study included children with a variety of psychiatric disorders, all of which are common in child psychiatry. Interestingly, the basal cortisol levels of the patients were similar, irrespective of the diagnosis. Although the groups were small, this indicates that, for a differential diagnosis basal cortisol levels may not be the most informative parameter of disrupted pituitary–adrenal function. A possible reason for the slightly increased levels of salivary cortisol in our control group may be the novelty of the experiment for these children as compared to the institutionally organized life of the hospitalized children. Such novelty stress has been reported in children (Tennes and Kreye, 1985). Both dexamethasone and hydrocortisone were effective in suppressing salivary cortisol in healthy control children and children with various psychiatric disorders. To our knowledge this is the first study to show an effect on pituitary–adrenal function of the natural steroid hydrocortisone in children. Provided only small amounts of this steroid are used, so as not to interfere with endogenous production the day following ingestion, hydrocortisone has a suppressive effect, similar to that of dexamethasone. Dexamethasone, however, is somewhat more potent and its effect lasts longer, due to the difference in half-life (Peterson, 1959). Since we administered dexamethasone and hydrocortisone according to bodyweight, the maximum suppressive effect was best evaluated by determining cortisol suppression corrected for baseline levels in the morning. With this approach, the dexamethasone-induced decrease in cortisol levels, although less in both dysthymics and ODD, could not distinguish patients from controls. These findings are consistent with reports in the literature that dexamethasone does not consistently suppress cortisol secretion, irrespective of basal cortisol levels (Dahl et al., 1992; Kaneko et al., 1993). Surprisingly, a significant hyporesponsiveness to hydrocortisone was found in children with dysthymia and patients with oppositional defiant disorder. Only few studies have incorporated the latter group of patients in studies on dexamethasone suppression and generally no disturbances were found (Doherty et al., 1986; Livingston et al., 1984; Petty et al., 1985). Our data correspond with the data reported by Birmaher et al. (1992). They showed in a 24-h study of cortisol before and after administration of dexamethasone, that the first effect of dexamethasone is to block pituitary corticotropin release, resulting in an
304
B. Buwalda et al.
absence of cortisol for several hours. Thereafter cortisol levels start to rise, but they remain low, indicating interference with corticotropin synthesis. The latter process is a gene-mediated intracellular process and is probably less dependent on the actual steroid plasma level (De Kloet, 1991). Using a recently developed assay for dexamethasone in saliva (Thijssen et al., 1996), we found no difference in dexamethasone levels between patients and controls. Moreover, there was no relation between saliva dexamethasone levels and post-dexamethasone cortisol levels or between saliva dexamethasone levels and the dose of dexamethasone used. Therefore, levels of dexamethasone in saliva measured the day following dexamethasone ingestion may not be very helpful in interpreting the suppressive effect of the steroid. These findings contradict the findings of Arana et al. (1984), Carr et al. (1984) and Casat et al. (1989) but are in line with reports by Naylor et al. (1990) and Birmaher et al. (1992). Studies of the two steroid receptor types in the brain (mineralocorticoid and glucocorticoid), provide increasing evidence that the regulation of HPA function in experimental animals reflects a balance between glucocorticoid activity at mineralocorticoid and glucocorticoid receptors in the hippocampus, hypothalamus, and pituitary (De Kloet, 1991). The affinity of dexamethasone for the glucocorticoid receptor is highest in the pituitary, whereas hydrocortisone has more affinity for the glucocorticoid and mineralocorticoid receptors in the hypothalamus. Therefore, by using the natural steroid hydrocortisone, it is possible to assess the natural feedback activity at a level beyond the pituitary, i.e. hypothalamus, which is less sensitive to dexamethasone. Obviously, impairment in feedback activity is not specific for depression, and can be found in other child psychiatric disorders as well. In our study, symptom clusters such as social problems and anxious/depressed on the CBCL were related to hyporesponsiveness to hydrocortisone. No such relation between symptom clusters and the suppressive effect of dexamethasone was found. In conclusion, both basal cortisol levels and cortisol levels after ingestion of either dexamethasone or hydrocortisone showed the expected rhythmicity and suppression without much fluctuation. Therefore, salivary cortisol seems a stable measure of pituitary– adrenal function in children during their routine daily activities, and it is not necessary to take extra precautions, but to control food intake and extreme exercise. Pituitary-adrenal dysfunction can be objectified in child psychiatric disorders, the malfunctions not being restricted to depression. The use of hydrocortisone in low doses may be preferred over dexamethasone, since it may mimic more closely endogenous cortisol regulation. However, the aspecificity of these disturbances in feedback activity may reflect a general maladaptation of patients with psychiatric disorders to the environment rather than that these disturbances are intrinsic to a specific syndrome. Behavioral and symptomatological correlates should be determined for these HPA disturbances.
REFERENCES Arana, G. W., Workman, R. J. and Baldessarini, R. J. (1984) Association between low plasma levels of dexamethasone and elevated levels of cortisol in psychiatric patients given dexamethasone. Am J Psychiatry 141, 619–1620. Birmaher, B., Ryan, N. D., Dahl, R., Rabinovich, H., Ambrosini, P., Williamson, D. E., Novacenko, H., Nelson, B., Lo, E. S. and Puig Antich, J. (1992) Dexamethasone suppression test in children with major depressive disorder. J Am Acad Child Adolesc Psychiatry, 31, 291 – 297. Carr, D. B., Wool, C., Lydiard, R. B., Fisher, J., Gelenberg, A. and Klerman, G. (1984) Rate-sensitive inhibition of ACTH release in depression. Am J Psychiatry 141, 590 – 592.
Hydrocortisone/Dexamethasone Inhibition
305
Caroll, B. J. (1989) Diagnostic validity and laboratory studies: Rules of the game, In: Robbins, L. N., Barrett, J. E. (Eds), The Validity of Psychiatric Diagnosis, Raven Press, New York, pp. 229–245. Casat, C. D., Arana, G. D. and Powel, K. (1989) The DST in children and adolescents with major depressive disorder. Am J Psychiatry, 146, 503 – 507. Chrousos, G. P. (1992) Regulation and dysregulation of the hypothalamic – pituitary – adrenal axis—The corticotropin-releasing hormone perspective. Neuroendocrinology 21, 833 – 858. Dahl, R. E., Kaufman, J., Ryan, N. D., Perel, J., al Shabbout, M., Birmaher, B., Nelson, B. and Puig Antich, J. (1992) The dexamethasone suppression test in children and adolescents: a review and a controlled study. Biol Psychiatry 32, 109 – 126. DeBellis, M. D., Dahl, E. R., Perel, J., Birmaher, B., Al Shabbout, M., Williamson, D. E., Nelson, B. and Ryan, N. D. (1996) Nocturnal ACTH, cortisol, growth hormone, and prolactin secretion in prepubertal depression. J Am Acad Child Adolesc Psychiatry 35, 1130 – 1138. De Kloet, E. R. (1991) Brain corticosteroid receptor balance and homeostatic control. Front Neuroendocrinol 12, 95–164. Doherty, M., Madansky, D., Kraft, J., Carter-Ake, L., Rosenthal, P. and Coughlin, B. (1986) Cortisol dynamics and test performance of the dexamethasone suppression test in 97 psychiatrically hospitalized children aged 3– 16 years. J Am Acad Child Psychiatry 25, 400 – 408. Emslie, G., Weinberg, W., Rush, A. J., Weissenburger, J. and Parkin-Feigenbaum, L. (1987) Depression and dexamethasone suppression testing in children and adolescents. J Child Neurol 2, 31–37. Foreman, D. M. and Goodyer, I. M. (1988) Salivary cortisol hypersecretion in juvenile depression. J Child Psychol Psychiatry 29, 311 – 320. Genazzani, A. R., Facchinetti, F., Pintor, C., Puggioni, R., Parrini, D., Petraglia, F., Bagnoli, F. and Corda, R. (1983) Proopiocortin-related peptide plasma levels throughout prepuberty and puberty. J Clin Endocrinol Metab 57, 56–61. Gispen-de Wied, C. C., D’Haenen, H., Verhoeven, W. M. A., Wynne, H. J., Westenberg, H. G. M., Thijssen, J. H. H. and Van Ree, J. M. (1993) Inhibition of the pituitary-adrenal axis with dexamethasone and cortisol in depressed patients and healthy subjects: a dose response study. Psychoneuroendocrinology 18, 191– 204. Gispen-de Wied, C. C., Westenberg, H. G. M., Thijssen J. H. H. and Van Ree, J. M. (1987) The dexamethasone and cortisol suppression test in depression: beta-endorphin as a useful marker. Psychoneuroendocrinology 12, 355– 366. Gold, P. W., Goodwin, F. K. and Chrousos, G. P. (1988a) Clinical and biochemical manifestations of depression. Relation to the neurobiology of stress (1). N Engl J Med 319, 348 – 353. Gold, P. W., Goodwin, F. K. and Chrousos, G. P. (1988b) Clinical and biochemical manifestations of depression. Relation to the neurobiology of stress (2). N Engl J Med 319, 413 – 420. Goodyer, I., Herbert, J., Moor, S. and Altham, P. (1991) Cortisol hypersecretion in depressed school-aged children and adolescents. Psychiatry Res 37, 237 – 244. Holsboer, F., Gerken, A., Stalla, G. K. and Muller, O. A. (1987) Blunted aldosterone and ACTH release after human CRH administration in depressed patients. Am J Psychiatry 144, 229 – 231. Kaneko, M., Hoshino, Y., Hashimoto, S., Okano, T. and Kumashiro, H. (1993) Hypothalamic– pituitary–adrenal axis function in children with attention-deficit hyperactivity disorder. J Autism De6 Disord 23(1), 59–65. Kirschbaum, C. and Hellhammer, D. H. (1989) Salivary cortisol in psychobiological research: an overview. Neuropsychobiology 22, 150 – 169. Livingston, R., Reis, C. and Ringkahl, I. (1984) Abnormal dexamethasone suppression test results in depressed and nondepressed children. Am J Psychiatry 141, 106 – 108. McBurnett, K., Lahey, B. B., Frick, P. J., Risch, C., Loeber, R., Hart, E. L., Christ, M. A. and Hanson, K. S. (1991) Anxiety, inhibition, and conduct disorder in children: II Relation to salivary cortisol. J Am Acad Child Adolesc Psychiatry 30, 192 – 196. Naylor, M. W., Greden, J. F. and Alessi, N. E. (1990) Plasma dexamethasone levels in children given the dexamethasone suppression test. Biol Psychiatry 27, 592 – 600. Petty, L., Asarnow, J., Carlson, G. and Lesser, L. (1985) The dexamethasone suppression test in depressed, dysthymic, and nondepressed children. Am J Psychiatry 142, 631 – 633. Peterson, R. E. (1959) Metabolism of adrenocorticoids in man. Ann NY Acad Sci 82, 846. Pfeffer, C., Stokes, P. and Weiner, A. et al. (1989) Psychopathology and plasma cortisol responses to dexamethasone in prepubertal psychiatric inpatients. Biol Psychiatry 26, 677 – 689.
306
B. Buwalda et al.
Puig-Antich, J., Chambers, W. and Halpern, F. (1979) Cortisol hypersecretion in prepubertal major depressive illness: a preliminary report. Psychoneuroendocrinology 4, 191 – 197. Puig-Antich, J., Dahl, R., Ryan, N., Novacenko, H., Goetz, D., Goetz, R., Twomey, J. and Klepper, T. (1989) Cortisol secretion in prepubertal children with major depressive disorder. Episode and recovery. Arch Gen Psychiatry 46, 801 – 809. Scarpa Scerbo, A. and Kolko, D. J. (1994) Salivary testosterone and cortisol in disruptive children: relationship to aggressive, hyperactive, and internalizing behaviors. J Am Acad Child Adolesc Psychiatry 33(8), 1174–1184. Tennes, K. and Kreye, M. (1985) Children’s adrenocortical response to classroom activities and tests in elementary school. Psychosom Med 47, 451 – 460. Thijssen, J. H. H., Gispen-de Wied, C. C., Van Heeswijk, G. M. and Veeman, W. (1996) The determination of dexamethasone in saliva. Clin Chem 42(8), 1238 – 1242. Weller, E., Weller, R., Fristad, M. and Preskorn, S. (1984) The dexamethasone suppression test in hospitalized prepubertal children. Am J Psychiatry 141, 290 – 291. Weller, E., Weller, R., Fristad, M., Preskorn, S. and Teare, M. (1985) The dexamethasone suppression test in hospitalized prepubertal children. J Clin Psychiatry 46, 511 – 513.
. .