| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
London Center for Pediatric Endocrinology (E.C., C.G.D.B., P.C.H.), University College London, London, W1T 3AA, United Kingdom; Guilford (S.M.P.), Connecticut 06437; Diabetes Research Laboratories (D.R.M.), Radcliffe Infirmary, Oxford OX2 6HE, United Kingdom; and Clinical Pharmacology (A.J.), St. Bartholomew’s School of Medicine and Dentistry, London EC1M 6BQ, United Kingdom
Address all correspondence and requests for reprints to: Evangelia Charmandari, M.D., National Institute of Child Health and Human Development, National Institutes of Health, Pediatric and Reproductive Endocrinology Branch, 10 Center Drive, Building 10, Suite 9D42, Bethesda Maryland 20892-1583. E-mail: . charmane{at}mail.nih.gov
Abstract
In humans, GH and cortisol are secreted in a pulsatile fashion and a mutual bidirectional interaction between the GH/IGF-I axis and hypothalamic-pituitary-adrenal axis has been established. Classic congenital adrenal hyperplasia (CAH) is characterized by a defect in the synthesis of glucocorticoids and often mineralocorticoids, and adrenal hyperandrogenism. Substitution therapy is given to prevent adrenal crises and to suppress the abnormal secretion of androgens and steroid precursors from the adrenal cortex. However, treatment with twice or three times daily oral hydrocortisone does not mimic physiological adrenal rhythms and may influence the activity of the GH/IGF-I axis. We investigated the pattern of GH and cortisol secretion and the synchrony of joint GH-cortisol secretory dynamics in 15 children with classic 21-hydroxylase deficiency (5 males and 10 females; median age 9.5 yr, range 6.1–11.0 yr) and 28 short normal children (23 males and 5 females; median age 7.7 yr, range 4.9–9.3 yr). All subjects were prepubertal. Serum GH and cortisol concentrations were determined at 20-min intervals for 24 h. The irregularity of GH and cortisol secretion was assessed using approximate entropy (ApEn), a scale- and model-independent statistic. The synchrony of joint GH-cortisol secretion was quantified using the cross-ApEn statistic. Cross-correlation analysis of GH and cortisol secretory patterns was computed at various time lags covering the 24-h period. Children with CAH had significantly lower mean 24-h serum cortisol concentrations (6.4 ± 2.2 vs. 10.4 ± 2.6 µg/dl, P < 0.001), ApEn (GH) (0.64 ± 0.13 vs. 0.74 ± 0.17, P = 0.04), ApEn (cortisol) (0.54 ± 0.13 vs. 1.08 ± 0.18, P < 0.001) and cross-ApEn values of paired GH-cortisol secretion (0.78 ± 0.19 vs. 1.05 ± 0.12, P < 0.001) than normal children. There was no difference in mean 24-h GH concentrations between the two groups (4.5 ± 2.9 vs. 4.5 ± 1.9 mU/liter). In children with CAH, a significant positive correlation between GH and cortisol was noted at lag time 0 min (r = 0.299, P < 0.01), peaking at 20 min (r = 0.406, P < 0.0001), whereas in normal children, a significant negative correlation between the two hormones was noted at lag time 0 min (r = -0.312, P < 0.01). The above findings suggest that children with classic CAH have a more regular pattern of GH secretion and a more synchronous joint GH-cortisol secretory dynamics than their normal counterparts. These differences reflect bidirectional interactions between the GH/IGF-I axis and hypothalamic-pituitary-adrenal axis in humans, and are likely to evolve as a result of the exogenous administration of hydrocortisone at fixed doses and at specific time intervals, which leads to a more regular pattern in circulating cortisol concentrations, independent of variations in CRH and ACTH concentrations.
IN HUMANS, GROWTH hormone and cortisol are secreted in a pulsatile fashion and physiological peak concentrations are observed at different times. The relationship between GH and cortisol is dynamic and a mutual bidirectional interaction between the GH/IGF-I axis and hypothalamic-pituitary-adrenal (HPA) axis has been established (1, 2, 3, 4, 5, 6). Significant chronic hypercortisolism results in reduced GH secretion and growth suppression (4, 5, 6) as well as attenuation of GH response to exogenous stimuli (2, 3), and small increases in cortisol concentrations exert a stimulatory effect on GH secretion by increasing GH mRNA levels and enhancing GH gene expression (7, 8). In idiopathic ACTH deficiency, appropriate glucocorticoid replacement is necessary to reestablish a normal pattern of GH release in response to provocation (9), while in GH deficiency, the low basal and stimulated cortisol concentrations are normalized following exogenous GH administration (1). In addition, beyond the previously reported alterations in the amplitude and/or frequency of GH pulses in response to alterations in cortisol secretion, individual nodal changes in the regularity of GH secretion have been observed secondary to alterations in the regularity of cortisol secretion (10).
Congenital adrenal hyperplasia (CAH) due to 21- hydroxylase deficiency is an autosomal recessive condition in which deletions or mutations of the cytochrome P450 21-hydroxylase gene result in glucocorticoid and often mineralocorticoid deficiency. This leads to increased secretion of ACTH by the anterior pituitary, adrenal hyperplasia, and accumulation of steroid precursors before the enzymatic defect (11). Glucocorticoid and mineralocorticoid substitution is given to prevent adrenal crises and to suppress the abnormal secretion of androgens and steroid precursors from the adrenal cortex. Although conventional therapy with oral hydrocortisone twice or three times daily does not mimic physiological adrenal rhythms, it does provide a useful paradigm to assess the interrelation between serum GH and cortisol concentrations when cortisol pulsatility is perturbed.
Subjects and Methods
Subjects
Fifteen children with salt-wasting 21-hydroxylase deficiency (5 males and 10 females; median age 9.5 yr, range 6.1–11.0 yr) and 28 short normal children (23 males and 5 females; median age 7.7 yr, range 4.9–9.3 yr) attending the London Center for Pediatric Endocrinology were studied prospectively. All subjects were prepubertal and their clinical characteristics are summarized in Table 1
.
|
-fludrocortisone given once daily. Patients were excluded from the study if there was evidence of central precocious puberty or other associated endocrine disorder. No subject had clinical or biochemical evidence of hepatic or renal disease and none was receiving medications known to influence GH or corticosteroid-binding globulin secretion or to induce hepatic enzymes. All normal children were growing at an appropriate growth rate, and GH provocation testing had excluded GH deficiency. The study was approved by the University College London Hospitals Committee on the Ethics of Human Research. Written informed consent was obtained in all cases by a parent, and assent was given by children older than 7 yr.
Materials and Methods
All children were admitted to the Endocrine Unit, Middlesex Hospital, 1 d before the study, and standard anthropometric measurements, including Tanner pubertal staging (12, 13), were obtained by a single trained observer. An indwelling venous catheter for blood sampling was inserted at least 12 h before sampling to allow a period of adaptation. Children with CAH received their oral hydrocortisone tablets at 0900 h and 2100 h or at 0800 h, 1500 h, and 2200 h, depending on whether they were on twice (n = 9) or three times (n = 6) daily regimen, respectively. 9-Fludrocortisone was given with the morning hydrocortisone tablets. Standard hospital meals were delivered at 0800 h, 1230 h, and 1730 h.
On the day of the study, serum GH and cortisol concentrations were determined at 20-min intervals for 24 h. An additional blood sample for measurement of IGF-I concentrations was drawn at 0800 h. Blood samples were centrifuged, separated, and stored at -80 C until assayed.
Assays
Cortisol. Serum total cortisol was measured using the Coat-A-Count RIA (Diagnostic Products, Los Angeles, CA). This is a solid-phase RIA with a sensitivity of 0.2 µg/dl. The within-assay coefficients of variation (CV) were 5.7% and 2.6% at serum concentrations of 1.0 µg/dl and 20 µg/dl, respectively, and the between-assay CV were 6.3% and 4.5% at serum concentrations of 5.0 µg/dl and 10.0 µg/dl, respectively.
GH. GH was measured using an immunoradiometric assay (Immunotech, Marseille, France). This assay utilizes mouse monoclonal antibodies that recognize the 22-kDa GH monomer. The sensitivity of the assay was 0.1mIU/liter. The intraassay CV were 0.66%, 1.5%, and 1.29% at serum concentrations of 7.29, 13.57, and 55.96 mIU/liter, respectively. The interassay CV were 13.43%, 14.03%, and 13.14% at serum concentrations of 6.53, 12.14, and 51.39 mIU/liter, respectively.
IGF-I. IGF-I concentrations were measured using an immunoradiometric assay (IGF-I IRMA, Nichols Institute Diagnostics, San Juan Capistrano, CA) with a sensitivity of 6 ng/ml, intraassay variations of 4.6% and 4.1% at serum concentrations of 61.0 and 547.9 ng/ml, respectively, and interassay variations of 15.8% and 9.3% at serum concentrations of 60.1 and 594.3 ng/ml, respectively.
Analysis of hormone pulsatility
Approximate entropy (ApEn) analysis. To quantify serial irregularity we used ApEn, a scale- and model-independent statistic that has provided new insights both mathematically and in various biological applications (14, 15, 16, 17, 18). ApEn is complementary to pulse detection algorithms widely employed to appraise hormone secretion time series. It evaluates both dominant and subordinant patterns in data and detects changes in underlying episodic behavior not reflected in peak occurrences or amplitudes. In addition, it provides a direct barometer of feedback system alterations in many coupled, theoretical mathematical systems (15, 19). ApEn assigns a nonnegative number to a time series (14), with higher values indicating greater degree of irregularity of hormone secretion and lower values indicating greater regularity of secretion. The value is derived from the logarithmic likelihood that runs of patterns that are close (within r) for m contiguous observations remain close (within the same tolerance width r) on the next incremental comparison. Profiles were analyzed with m = 1 and r = 20% of the SD of the individual subject time series. These values produce good statistical reproducibility in theoretical (14) and endocrine practice (17, 18).
Cross-ApEn analysis. To quantify asynchrony (conditional irregularity), we used cross-ApEn (16). Cross-ApEn can be employed to compare sequences from two distinct yet intertwined variables in a network, herein applied to the time series of GH and cortisol circulating concentrations. Larger cross-ApEn values indicate greater asynchrony of joint hormone secretion, whereas smaller values suggest greater joint signal synchrony. The precise mathematical definition of cross-ApEn is thematically similar to that of ApEn (16, 17).
Cross-correlation analysis. To search for a time-ordered relationship between GH and cortisol, analysis of correlations between the absolute values of the time series of each hormone was computed at 20-min intervals at various time lags covering the 24-h period of the study. All data were stationarized and processed as a 3-point moving average to reduce assay noise as previously described (20, 21). Cross-correlation was computed after lagging (shifting) the concentration time series of cortisol relative to the concentration time series of GH. If rt is the coefficient of correlation between GH and cortisol at a lag time t for one subject, then the mean rt for all subjects was considered significant when it exceeded zero by more than two SE of the mean. The SE was calculated from the individual values of rt for all individuals at the lag time t.
Statistical analysis
Nonnormally distributed data were logarithmically transformed before analysis. Comparisons between the two groups were performed using the t test.
Results
The biochemical and time series analysis findings are shown in Table 2.
|
Twenty-four-hour GH and cortisol concentrations.
Mean 24-h serum cortisol concentrations were lower in children with CAH than in normal subjects (P < 0.001) (Fig. 1). There was no difference in mean 24-h GH concentrations between children with CAH and normal children (Fig. 2). IGF-I concentrations were higher in children with CAH than in the normal subjects (P = 0.012); however, no difference was noted between the two groups when IGF-I values were corrected for height (Table 2
).
|
|
ApEn analysis.
Children with CAH had significantly lower ApEn (GH) (P = 0.04) and ApEn (cortisol) (P < 0.001) values than their normal counterparts (Table 2). Cross-ApEn values of paired GH-cortisol secretion, with cortisol leading GH or GH leading cortisol, were also lower in children with CAH than in normal children (P < 0.001).
No difference in ApEn (GH), ApEn (cortisol), or cross-ApEn values was noted between males and females in each group. Also, no difference in these parameters was noted between children with CAH receiving hydrocortisone twice daily and those receiving hydrocortisone three times daily.
Cross-correlation analysis.
In children with CAH, a significant positive correlation between GH and cortisol was noted at lag time 0 min (r = 0.299, P < 0.01), peaking at 20 min (r = 0.406, P < 0.0001), with cortisol leading GH by these time intervals. Also, a significant negative correlation between the two hormones was observed over time, peaking at lag time 7 h (r = -0.351, P < 0.0001), with cortisol leading GH by this time interval (Fig. 3A). In normal children, a significant negative correlation between GH and cortisol was noted at lag time 0 min (r = -0.312, P < 0.01). A significant positive correlation between the two hormones was observed over time, peaking at lag time 6 h 40 min (r = 0.401, P < 0.0001), with GH leading cortisol by this time interval (Fig. 3B).
|
Our findings demonstrate clear differences in the regularity of GH secretion as well as the synchrony of joint GH-cortisol secretory dynamics between children with classic 21-hydroxylase deficiency and healthy subjects of similar age. Children with CAH had more regular GH secretion and more synchronous joint GH-cortisol secretory dynamics than their normal counterparts. It is likely that these alterations evolved as a result of the exogenous administration of hydrocortisone at fixed doses and at specific time intervals, which led to a more regular and more predictable pattern in circulating cortisol concentrations, independent of variations in CRH and ACTH concentrations.
In humans, circadian rhythmicity and pulsatile hormone secretion are the physiological hallmarks of the HPA axis (22, 23). The pulsatile nature of cortisol secretion, its 24-h rhythmicity, and the coupled ultradian-circadian dynamics are regulated by the strength and timing of the foregoing within the HPA axis multisite interactions (24, 25, 26). This concept of integrative within-axis feedback control in healthy subjects may explain the disruption of the pulsatile, entropic, and circadian rhythmic properties of cortisol secretion in subjects with autonomous tumoral secretion of ACTH (27, 28, 29). Disruption of the neuroendocrine mechanisms that coordinate the pulsatile and 24-h rhythmic release of cortisol results in alterations in feed-forward- and/or feedback-dependent regulation within the same interactive CRH/AVP-ACTH-adrenal axis (19, 30). Patients with Cushing’s disease, for example, demonstrate a more irregular pattern of ACTH and cortisol secretion, which is restored following transsphenoidal surgery and removal of the pituitary adenoma (27, 28, 29); patients with acromegaly or prolactinoma display a 4-fold disruption of basal, pulsatile, rhythmic, and entropic secretion of GH and PRL, respectively (31, 32, 33); in the elderly, the progressive fall in the individual orderliness of cortisol, FSH, LH, and T is associated with disruption in the synchrony of joint ACTH-cortisol, FSH-LH, LH-PRL, and LH-T secretion (17, 34, 35, 36, 37), further indicating a variable disruption in the feedback and feed-forward interconnections among the neuroendocrine glands within the same axis. Although not examined in the present study, we hypothesize that children with CAH may also demonstrate alterations in the regularity of ACTH secretion compared with normal subjects.
In addition to within-axis feed-forward- and feedback-dependent regulation of hormone release, a number of neuroendocrine mechanisms also operate to sustain the synchrony of the joint two-axes hormone secretion. Disruption of the mechanisms coordinating the pulsatile secretion of a hormone results in disruption of other interconnected hormone systems. In Cushing’s disease, the autonomous disordered secretion of ACTH is associated with a more irregular secretion of both GH and PRL (10), while in older female subjects the more irregular secretion of GH is associated with a more asynchronous joint GH-cortisol secretion, compared with their male counterparts (38, 39).
Our findings support previous studies on the bidirectional interaction of interconnected hormone systems and provide additional evidence that variable disruption in the time-delayed feedback and feed-forward interconnections among the neuroendocrine glands would further disrupt the synchrony of joint hormone release within the same or separate axes. The significant differences in the regularity of GH secretion as well as the synchrony of joint GH-cortisol secretory dynamics between children with classic CAH and normal subjects most likely reflect differences in the regularity of cortisol secretion between the two groups and indicate that the nonpulsatile, more regular 24-h cortisol concentrations in children with CAH may play an important role in the regulation of GH secretion, further influencing the synchrony of other interconnected hormone systems. The fact that no difference in the amount of spontaneous GH release was noted between these two prepubertal groups does not preclude the possibility that a more regular pattern of GH secretion at puberty and/or young adulthood may be associated with a reduction in the amplitude and/or frequency of GH pulses and may therefore be an additional factor that contributes to the compromised final height of these patients (40, 41). In addition, it is important to consider that individuals with short stature, who grow at an age-appropriate growth velocity and have demonstrated a normal GH response at provocation testing, may have decreased spontaneous 24-h GH secretion, compared with their average height counterparts, and may thus not be the ideal control group in terms of 24-h GH secretory mass, although they still qualify as an appropriate control group for comparisons related to the regularity of GH secretion as well as the synchrony of joint GH-cortisol secretion.
In addition to the synchrony of joint GH-cortisol secretion, the administration of oral hydrocortisone in children with classic CAH played an important role in determining the GH-cortisol interrelation, as evidenced by the cross-correlation analysis findings in the two groups of children: In normal subjects, the negative correlation between GH and cortisol at lag time 0 min concurs with previous observations (42) and most likely reflects the elevated CRH concentrations, which lead to spontaneous endogenous cortisol secretion, and at the same time they inhibit GH secretion probably by increasing somatostatin release (43, 44). The positive correlation between the two hormones observed over time is probably owing to a direct effect of cortisol on the transcriptional activity of GH gene at the anterior pituitary level, given that the latter has been shown to contain glucocorticoid response elements in its regulatory region (45). In contradistinction, children with CAH displayed a significant positive correlation between GH and cortisol at lag time 0 min, which reached a peak at 20 min. This can be explained by the fact that peak cortisol concentrations attained following administration of oral hydrocortisone in subjects with CAH do not reflect elevated CRH concentrations but rather result in suppression of CRH/ACTH and therefore increased GH secretion (43, 44). The latter is also supported by the demonstration that pretreatment with hydrocortisone augments the GH response to GHRH (46), while the time lag of 20 min is consistent with the pharmacokinetics of oral hydrocortisone. The significant negative correlation between GH and cortisol observed over time is likely to reflect the gradual decline in circulating cortisol concentrations 4–6 h after administration of hydrocortisone (47, 48) and the concomitant elevation in ACTH and CRH concentrations (43, 44).
We conclude that children with classic CAH have a more regular pattern of GH secretion and a more synchronous joint GH-cortisol secretory dynamics than their normal counterparts. These differences reflect bidirectional interactions between the GH/IGF-I axis and HPA axis in humans and are likely to evolve as a result of the exogenous administration of hydrocortisone at fixed doses and specific time intervals, which leads to a more regular pattern in circulating cortisol concentrations, independent of variations in CRH and ACTH concentrations. Our findings also reinforce the importance of cross-ApEn and cross-correlation function in detecting changes in bivariate data behavior.
Acknowledgments
Footnotes
Abbreviations: ApEn, Approximate entropy; CAH, congenital adrenal hyperplasia; CV, coefficients of variation; HPA, hypothalamic- pituitary-adrenal.
Received October 31, 2001.
Accepted February 13, 2002.
References
This article has been cited by other articles:
 |
|
 |
|
  X. Chen, K. H. Chon, and I. C. Solomon Chemical activation of pre-Botzinger complex in vivo reduces respiratory network complexity Am J Physiol Regulatory Integrative Comp Physiol, May 1, 2005; 288(5): R1237 - R1247. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. O. van Aken, A. M. Pereira, M. Frolich, J. A. Romijn, H. Pijl, J. D. Veldhuis, and F. Roelfsema Growth hormone secretion in primary adrenal Cushing's syndrome is disorderly and inversely correlated with body mass index Am J Physiol Endocrinol Metab, January 1, 2005; 288(1): E63 - E70. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Caldirola, L. Bellodi, A. Caumo, G. Migliarese, and G. Perna Approximate Entropy of Respiratory Patterns in Panic Disorder Am J Psychiatry, January 1, 2004; 161(1): 79 - 87. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
