| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Studies |
Departments of Biobehavioral Health (J.W.F., D.E.R., E.G.) and Health Policy and Administration (S.F.), College of Health and Human Development, and Department of Pediatrics (J.W.F., H.E.K., M.R.D.), College of Medicine, The Pennsylvania State University, University Park, Pennsylvania 16802; and Kaunas University of Medicine, Institute of Endocrinology (R.K.), Kaunas, Lithuania
Address correspondence and requests for reprints to: Jordan W. Finkelstein, M.D., The Pennsylvania State University, E 315 HHDEV, University Park, Pennsylvania 16802. E-mail: jwf3{at}psu.edu
Abstract
The aim of this study was to investigate hypothalamic-pituitary-adrenal (HPA) function in children with GH deficiency. Ninety-four patients were evaluated for GH deficiency and cortisol (F) deficiency using clinical criteria and L-dopa and insulin-induced hypoglycemia stimulation tests. They were assigned to three diagnostic groups: organic GH deficient (OGHD), idiopathic GH deficient (IGHD), and not GH-deficient (NGHD). Time series, cross-sectional, regression analysis revealed statistically significantly elevated F [>828 nmol/L (30 µg/dL)] in the OGHD group vs. the NGHD group. The value for F in the IGHD group was not different from the NGHD group. This finding suggests that dysregulation of the HPA axis is present in most children with OGH deficiency and significantly less often in children with IGH deficiency or without GH deficiency. Anatomical disruption of the control pathways for the HPA axis or stress may cause the dysregulation.
THERE ARE MANY stimulation tests to study the function of the hypothalamic-pituitary-somatotropin axis. These include GHRH, glucagon, pyridostigmine, exercise, insulin-induced hypoglycemia, clonidine, L-dopa, and arginine, among others (1). Insulin-induced hypoglycemia (ITT), although it offers some risks, is considered the gold standard (2, 3) and offers the advantage of simultaneously evaluating the integrity of the hypothalamic-pituitary-adrenal (HPA) axis.
There are few detailed, systematic studies of plasma ACTH or cortisol (F) responses in children with short stature after administration of any of the stimuli mentioned above. Most studies report means or maximum F values to ITT, without evaluating other comparisons or providing measures of variability (3, 4, 5, 6, 7).
The aim of this study was to identify possible differences in HPA axis functioning between children with organic GH deficiency (OGHD) and children with idiopathic GH deficiency (IGHD) and children who were not GH deficient (NGHD). Particular attention was paid to children with organic lesions in whom the diagnosis of GH deficiency could be established with the highest degree of certainty. To our surprise, the children with organic GH deficiency had significantly higher F responses than the two other groups of children.
Materials and Methods
Subjects
We studied 94 consecutive patients (56 males and 38 females) for whom there were clinical indications of possible GH deficiency. Their mean age was 11.1 yr (males) and 9.7 yr (females), with a range of 6.4–14.1 yr. All these subjects had height or growth rate and bone age greater than 2 SD below the mean for age. All were completely prepubertal.
Of the 94 children and adolescents, 19 children were found to have idiopathic isolated GH deficiency (maximum GH concentration, after stimulation, <7.5 µg/L). Twelve patients had organic causes for GH deficiency (congenital malformation, tumor, radiation, chemotherapy, infiltrative disorders), of whom three had TSH deficiency and were on T4 replacement at the time of testing. None were diagnosed with ACTH deficiency or had diabetes insipidus. Subjects were not selected for this study because they had normal ACTH responses. None were taking sex steroids. Sixteen patients were eventually diagnosed as having idiopathic or familial short stature. Another 19 patients were diagnosed with constitutional delay in growth and development. Six patients were diagnosed with Turner’s syndrome. The rest of the 22 patients had miscellaneous disorders associated with growth failure that could not be classified in the categories mentioned above.
Procedure
The L-dopa and ITT stimulation tests were started between 0800 and 0900 h after an overnight fast. No patients were pretreated with sex steroids. After a baseline sample, 0.5 g L-dopa was administered orally. Blood samples were harvested at 60, 90, and 120 min. Then, insulin (0.05–0.1 U/kg body weight) was administered iv and blood samples were drawn at 150, 180, 210, and 240 min.
Glucose (G) was estimated by means of a G oxidase method. Plasma F and GH levels were determined by RIA in the Hershey Medical Center (Hershey, PA) core endocrine laboratory.
Statistical analysis
Data for G, F, and GH were pooled for each subject by three
diagnostic groups: NGHD, idiopathic isolated GH deficient (IGHD), and
organic GH deficient (OGHD). Pooling combined all data at all time
points, for each variable (G, F, and GH), for each diagnostic group
(see Table 1
). Pooling allowed us to
maximize data for use in analysis (degrees of freedom, 751), thus
increasing substantially the power of our next analyses. To the pooled
data, we next applied a cross-sectional, time series analysis using the
method of Fuller and Batisse (8). In this regression model
the concentration of F was the dependent (outcome) variable, and the
independent (predictor) variables were gender, diagnosis, the
concentration of G and GH, and each of the seven sampling points in
time, excluding the basal fasting time point. The comparison groups
were the NGHD patients. Some simple comparisons of means were estimated
using ANOVA.
|
Results
The concentration of each hormone pooled across all the data
within each of the diagnostic groups is shown in Table 1
. ANOVA showed
that the mean concentration of GH is significantly lower in the IGHD
(2.21 µg/L) and OGHD (1.46 µg/L) groups when each was compared with
the NGHD (6.64 µg/L) group (P < 0.01). The mean
concentration of GH was not significantly different between the IGHD
and OGHD groups by ANOVA.
The time series, cross-sectional analysis used the NGHD patients as the
group to which both GHD groups were compared. The concentration of F
was the dependent variable, whereas the independent variables are shown
in Table 2. Patients with OGH deficiency
had a significantly higher pooled concentration of F than did NGHD
patients (ß = 2.97, P < 0.02). The IGHD
patients did not have a significantly higher pooled concentration of F
than the NGHD patients. The patients with OGH deficiency seemed to fall
into two groups—those with maximum F greater than 828 nmol/L and those
with maximum F less than 828 nmol/L. The value of 828 nmol/L was based
on the bimodal distribution of the maximum concentration of F and the
closeness of 828 nmol/L to 2 SD above the mean
for the OGHD group. Seven (58%) patients had F maxima above 828 nmol/L
whereas five (42%) patients had maxima between 441.0 nmol/L and 744.9
nmol/L. There was no specific etiology or other characteristic of the
seven patients with the highest F (>828 nmol/L). There was no
significant difference between the concentration of F between NGHD and
IGHD patients (ß = 0.59, P = 0.57). However, 7
of the 19 (37%) patients with IGH deficiency had maximum F greater
than 828 nmol/L. The pooled concentration of G significantly predicted
the concentration of F (ß = 0.10, P < 0.006).
The fall in G after insulin was 2.37 ± 0.61 mmol/L in the GHD
groups compared with 2.45 ± 0.82 mmol/L in the NGHD group
(P = not significant by ANOVA). This fall was
50% and indicates that the stimulus for release of ACTH and GH
during hypoglycemia was adequate. The concentration of GH did not
predict the concentration of F (ß = 0.09, P =
0.15). This lack of effect is most likely the result of colinearity of
this variable (GH) with other independent variables. The effect of the
time of sampling was significant for the points at the end of the
L-dopa administration (point 5) and following the
administration of insulin (points 6–8, ß = 5.58–11.4,
P < 0.0001–0.008), clearly showing that these time
points significantly predicted the concentration of F. There were no
significant time effects during points 2–4 following
L-dopa administration, indicating that these
three time points did not statistically predict the concentration of F.
The association between F and time 5 (just before administration of
insulin) suggests that L-dopa also has some
stimulatory effect on F. This has been confirmed in another study of F
in GHD children by our group (9).
|
The aim of this study was to investigate thoroughly HPA function in children and adolescents, particularly among those subjects who had organic causes of GH deficiency. Evaluation of the integrity of the HPA axis in relation to the etiology of GH deficiency has not been rigorously pursued in children, with the exception of one study (3). That study of 22 children with organic lesions of the brain and skull identified five patients assessed as having ACTH deficiency based on inadequate F responses to insulin hypoglycemia. As in all previous studies in children, detailed F data were not presented, so that we cannot compare our results to theirs. In our study, the 12 patients with organic causes of GH deficiency were found to have significantly higher concentrations of F than the NGHD or the IGHD groups. Seven (58%) of the OGHD patients had F maxima greater than 828 nmol/L (range, 858-1214 nmol/L). There does not seem to be any study in which the F response to any of the stimuli mentioned above is defined as "elevated." We have defined F concentrations greater than 828 nmol/L as elevated (see above). The remaining five patients had F maxima between 441.4 and 745.0. Patients with IGH deficiency did not have significantly higher concentrations of F than the NGHD group, but 38% percent of the IGHD patients had maximum F greater than 828 nmol/L. Maximum F greater than 828 nmol/L was detected in only 20% of the patients without GH deficiency.
Insulin hypoglycemia tests were reported by Erturk et al. (10) in 193 adults, most of whom had organic acquired lesions in the hypothalamus or pituitary and whose GH status was not described. We cannot, therefore, contrast that study with our results. That study suggested one criterion for the diagnosis of ACTH deficiency—maximum F concentration after hypoglycemia of less than 469 nmol/L. Only one of our patients had F below this criterion (maximum F, 438.7).
Unfortunately, we do not have measures of the concentration of ACTH in our subjects, but in a group of 14 other patients tested by us for GH deficiency, 4 were found to have idiopathic, isolated GH deficiency. The mean F concentrations in these GHD patients were significantly higher than in the NGHD patients. However, there was no difference in the mean concentration of ACTH between the two groups (9). Those findings and the data in this study support the idea that there is some degree of HPA axis dysregulation in some patients with GH deficiency regardless of etiology.
The most probable cause for this phenomenon is some disruption of the brain anatomical pathways regulating the HPA axis, allowing it to be removed from the usual control factors present in an intact HPA system. However, the five patients with OGH deficiency with F maxima less than 828 nmol/L also had similar organic lesions. Perhaps the location of their lesions differed from those with higher concentrations of F. It is possible that the patients in the GHD group very high responding group (F, >828 nmol/L) were experiencing more stress than the other patients without GH deficiency. Unfortunately, we have no data to support this possibility, but dysregulation of the HPA axis has been reported by others for a range of psychiatric, endocrine, and inflammatory disorders (for a review see Ref. 11). Patients with GH deficiency with either organic or nonorganic causes of pituitary dysfunction may experience major stresses, and some may react with sustained high concentrations of F as suggested in the review (11) and in response to the general adaptation syndrome (12).
There have been reports that many children diagnosed with GH deficiency as children who were retested after their growth had ceased were found to have normal GH responses (13, 14, 15, 16, 17). It may be that these patients had high F concentrations that were stress induced during childhood, which inhibited the release and action of GH. Perhaps the stressful conditions no longer existed at the time of retesting when they were older, allowing a normal release of F and GH. No F data are available for those children who have been retested.
Alterations in hormone function during stress includes suppression of the somatotropic, thyroid, and the reproductive axes both centrally and peripherally via activation of the HPA axis and the sympathetic nervous system (11). Our patients with OGH deficiency had GH deficiency, three had TSH deficiency [one of whom had the highest F response, 1214 nmol/L (44.4 µg/dL)], one had possible ACTH deficiency, but none was diagnosed with gonadotropin deficiency primarily because of their young age. F is not the only hormone that has been reported to be elevated in OGH deficiency. Eight children with OGH deficiency showed elevated 24-h mean TSH when compared with four children diagnosed with "constitutional delay" (18).
We hope to follow these subjects to monitor their longer-term course. We encourage other investigators to examine any data they may have in regard to dysregulated HPA axis function in patients with various etiologies of GH deficiency.
Received November 19, 1999.
Revised August 14, 2000.
Accepted February 19, 2001.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
