Spontaneous Thyrotropin and Cortisol Secretion Interactions in ...

1 downloads 0 Views 453KB Size Report
GEORGE P. CHROUSOS, AND SERGIO BERNASCONI. Department of Pediatrics, University of Parma (L.G., M.E.S., A.V., G.M., M.V.), 43100 Parma, Italy;.
0021-972X/97/$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1997 by The Endocrine Society

Vol. 82, No. 11 Printed in U.S.A.

Spontaneous Thyrotropin and Cortisol Secretion Interactions in Patients with Nonclassical 21Hydroxylase Deficiency and Control Children* LUCIA GHIZZONI, GEORGE MASTORAKOS, MARIA E. STREET, ALESSANDRA VOTTERO, GEMMA MAZZARDO, MAURIZIO VANELLI, GEORGE P. CHROUSOS, AND SERGIO BERNASCONI Department of Pediatrics, University of Parma (L.G., M.E.S., A.V., G.M., M.V.), 43100 Parma, Italy; and the Department of Pediatrics, University of Modena (S.B.), 41100 Modena, Italy; Evgenidion Hospital, Endocrine Unit, Athens University Medical School (G.M.), 11528 Athens, Greece; and the Developmental Endocrinology Branch, National Institute of Child Health and Human Development, National Institutes of Health (G.P.C.), Bethesda, Maryland 20892 ABSTRACT Both exogenous and endogenous hypercortisolism result in reduced TSH secretion and mild hypothyroidism. However, little is known about the relation between endogenous TSH and cortisol secretion under physiological or slightly disturbed conditions. To examine this, we evaluated the pulsatility and circadian rhythmicity and time-cross-correlated the 24-h secretory patterns of cortisol and TSH in eight prepubertal children with nonclassical congenital adrenal hyperplasia (NCCAH) and eight age-matched short normal children. In both groups, TSH and cortisol were secreted in a pulsatile and circadian fashion, with a clear nocturnal TSH surge. Although no difference in mean 24-h TSH levels was observed between the two groups, daytime TSH levels were lower in the NCCAH group than in control children (P , 0.05). The cross-correlation analysis of the 24-h

raw data showed that TSH and cortisol were negatively correlated, with a 2.5-h lag time for both groups, with cortisol leading TSH. This correlation might reflect a negative glucocorticoid effect exerted on the hypothalamic-pituitary-thyroid axis under physiological conditions. A significant positive correlation with TSH leading cortisol was observed at 8.5 and 5.5 h lag times for the control and NCCAH groups, respectively. The substantially shorter lag time of this positive correlation in NCCAH children than in controls suggests that in the latter, the nocturnal TSH peak occurs temporally closer to their compromised morning cortisol peak. These data indicate that the hypothalamic-pituitary-adrenal axis has a primarily negative influence on endogenous TSH secretion and that even mild disturbances in cortisol biosynthesis are associated with slight alterations in TSH secretion. (J Clin Endocrinol Metab 82: 3677–3683, 1997)

T

In man, endogenous hypercortisolism (Cushing’s syndrome) is associated with a decrease in mean 24-h plasma TSH caused by a decrease in TSH pulse amplitude (10). Pharmacological doses of dexamethasone rapidly suppressed basal serum TSH levels and completely inhibited the pulsatile secretion of this hormone, but left the TSH response to exogenous TRH unaffected (11). Activation of the stress system is associated with decreased production of TSH, which may be caused by the increased levels of glucocorticoids (9, 12). CRH-stimulated increases in SRIH might also participate in the central component of thyroid axis suppression during stress, which is counteracted by the stimulatory effects of the activated locus ceruleus-norepinephrine system (13). On the other hand, in patients with primary adrenal insufficiency, plasma TSH levels are increased and return to normal after corticosteroid replacement (14 –16). Such findings suggest a direct suprapituitary and/or pituitary effect of glucocorticoids on TSH secretion. In normal subjects, Re et al. (1) found an increase in TSH within 24 h in response to metyrapone, suggesting that physiological plasma cortisol levels may have a suppressive effect on TSH secretion. However, abolition of the circadian cortisol rhythm by metyrapone administration did not lead to disruption of the TSH circadian rhythm (17). In normal men and women it appears that the diurnal rhythms of cortisol and TSH are in reverse phase (18), sug-

HERE HAVE been no studies addressing the physiological role of the hypothalamic-pituitary-adrenal axis in the regulation of the hypothalamic-pituitary-thyroid axis in man or animals. Yet, pharmacological studies have suggested an interaction between these axes at many levels. Thus, in both man (1–3) and the rat (4, 5), exogenously administered glucocorticoids inhibited thyroid function. Also, corticosterone stereotaxically infused into the anterior pituitaries of rats significantly suppressed the pituitary TSH response to submaximal TRH stimulation (6). Conversely, adrenalectomy in rats resulted in an increase in pro-TRH messenger ribonucleic acid in paraventricular nucleus neurons, whereas corticosterone and dexamethasone administration to intact rats resulted in a reduction of prepro-TRH messenger ribonucleic acid (7). In addition, intracerebroventricular CRH injections, via stimulation of somatostatin (SRIH), inhibited both TRH and TSH secretion in rats in vivo and/or in vitro (8, 9). Received March 3, 1997. Revision received July 8, 1997. Accepted July 17, 1997. Address all correspondence and requests for reprints to: Lucia Ghizzoni, M.D., Department of Pediatrics, University of Parma, Via Gramsci 14, 43100 Parma, Italy. E-mail: [email protected]. * This work was presented in part at the 35th Annual Meeting of the European Society for Pediatric Endocrinology, Montpellier, France, September 15–18, 1996.

3677

3678

JCE & M • 1997 Vol 82 • No 11

GHIZZONI ET AL.

gesting that the hypothalamic-pituitary-adrenal and thyroid axes may have reciprocal interactions. As most of the studies on the interactions between these axes have been performed in disease states or under experimental conditions using pharmacological doses of glucocorticoids, we investigated the relation between these two axes under physiological or slightly disturbed conditions. To do this, we evaluated the spontaneous cortisol and TSH secretion in short normal children and in children with nonclassical congenital adrenal hyperplasia (NCCAH) due to 21-hydroxylase deficiency (19), who had mild nocturnal cortisol insufficiency (20, 21).

pital diet. We arbitrarily chose the period from 2200 h in the evening to 1000 h in the morning and named it nighttime, because we expected the main circadian peaks of cortisol and TSH during this period. We started the pulsatility study at 1000 h to avoid dividing the morning cortisol surge. Blood samples for measurements of cortisol, TSH, T4, T3, FT4, and FT3 concentrations were kept at room temperature and centrifuged within 24 h. After centrifugation, serum was stored at 220 C until assayed. T4, T3, FT4, and FT3 levels were measured in pooled samples collected during the 12-h daytime and nighttime periods. Bone age was determined by the method of Greulich and Pyle (24).

Hormone assays Commercial kits were used for the estimation of serum cortisol (RIA, Radim, Pomezia, Italy), TSH (immunoradiometric assay, Nichols Institute, San Juan Capistrano, CA), T3, T4, FT3, and FT4 (enzyme-linked immunosorbent assay, Boehringer Mannheim, Mannheim, Germany), and 17-OHP (RIA, Diagnostic Products Corp., Los Angeles, CA) concentrations. The sensitivities of the assays were 2.48 nmol/L for cortisol, 0.04 mU/L for TSH, 0.46 nmol/L for T3, 7.7 nmol/L for T4, 0.46 pmol/L for FT3, 1.3 pmol/L for FT4, and 0.2 nmol/L for 17-OHP. Mean intra- and interassay coefficients of variation were 4.1% and 6.5% for cortisol, 4.4% and 6.8% for TSH, 5.2% and 9.8% for T3, 4.9% and 8% for T4, 4.2% and 5% for FT3, 5.6% and 9.1 for FT4, and 4.6% and 5.1% for 17-OHP, respectively.

Subjects and Methods Subjects This study was approved by the Clinical Research Committee of the Department of Pediatrics at the University of Parma (Parma, Italy). Eight patients with NCCAH and eight normal children were studied. The clinical characteristics of all subjects are summarized in Table 1. All patients were clinically and biologically euthyroid. Integrated concentrations of T4, T3, free T4 (FT4), and free T3 (FT3) from the 12-h daytime and nighttime periods of control and NCCAH children are shown in Table 2. The diagnosis of NCCAH was based on a serum 17-hydroxyprogesterone (17-OHP) level exceeding 45 nmol/L 60 min after an iv bolus dose of 250 mg corticotropin (Synacthen, Ciba-Geigy, Basel, Switzerland). Patients 3 and 7 had no symptoms and were identified while testing family members of affected patients. All other patients presented with or had a history of premature pubarche. Control children were chosen among subjects with familial short stature whose hypothalamic-pituitary function was normal and were matched to NCCAH patients according to baseline adrenal steroid levels, which were all within the normal range for Tanner stage I or II for breast in girls and for testicular size in boys (22) and were similar in the two groups.

Pulse analysis The Pulsar program was used to quantitate the pulse properties of cortisol and TSH time series objectively (25). Samples were analyzed for mean 24- and 12-h serum hormone concentrations, area under the curve above baseline (AUCb), area under the curve above zero line (AUCo), number of significant pulses, mean pulse height, mean pulse amplitude, mean pulse area, mean pulse length, and mean interpulse interval. The cut-off parameters G1–5 were set at 5.78, 2.89, 1.84, 1.27, and 0.89 times the intraassay sd as criteria for accepting peaks 1, 2, 3, 4, and 5 points wide, respectively. The smoothing time was set at half the total profile time, that is 12 h (24 points) and 24 h (48 points) for the 12- and 24-h profiles, respectively.

Protocol At 1000 h, after an overnight fast, an indwelling nonthrombogenic catheter was inserted into an antecubital vein and connected to a portable constant withdrawal pump, according to the method of Kowarski et al. (23). The rate of withdrawal was 4 mL/h, and blood collection tubes were changed every 30 min for 24 h. During this time, children were encouraged to continue normal activity and had a standard hos-

Statistical analysis Values are reported as the mean 6 sem unless otherwise stated. A test for normality was performed on all data. Statistical significance was

TABLE 1. Clinical profile of patients with NCCAH and controls Subject no. and sex

Patients 1F 2F 3M 4M 5F 6F 7M 8M Mean 6 Controls 1M 2M 3M 4M 5M 6M 7F 8F Mean 6

SEM

SEM

HSDS, Height

SD

Age (yr)

Bone age (yr)

9.4 7.25 8.7 8.1 7.58 7.5 7.6 9

10 6.83 5.5 12.1 10 10.5 9.5 10.5

8.14 6 0.28

9.36 6 0.75

9.33 11.08 9.58 6.66 11.91 8.33 7.66 7.41

7 9.5 7 6 10 6.5 7.6 5.75

8.99 6 0.64

7.41 6 0.55

TS B/T

PH

II I I I I II I I

II II I II III III I II

I I I I I I I I

II II I I II I II I

HSDS/CA

HSDS/BA

BMI (kg/m2)

1.05 0 21.53 2 3.03 2.73 0.28 0.48

0.57 0.53 1.7 21.32 0.44 20.3 21.52 21.59

19.23 15.94 16.22 16.76 18.96 21.58 14.1 18.67

1.00 6 0.53

20.18 6 0.42

17.68 6 0.83

23.11 21.72 21.86 20.9 22.04 21.86 22.71 21.96

20.95 20.54 0.43 20.35 20.68 3.78 22.71 0.18

14.7 15.08 15.69 13.93 17.14 13.88 15.52 14.87

22.02 6 0.23

20.1 6 0.64

15.10 6 0.37

score; CA, chronological age; BA, bone age; TS, Tanner stage; B/T, breast/testes; PH, pubic hair; BMI, body mass index.

TSH AND CORTISOL INTERACTION

3679

TABLE 2. Integrated concentrations of thyroid hormones in NCCAH patients and controls Subject no. and sex

Patients 1F 2F 3M 4M 5F 6F 7M 8M Mean 6 SEM Controls 1M 2M 3M 4M 5M 6M 7M 8F Mean 6

SEM

Day

Night

T4 (nmol/L)

T3 (nmol/L)

FT4 (pmol/L)

FT3 (pmol/L)

T4 (nmol/L)

T3 (nmol/L)

FT4 (pmol/L)

FT3 (pmol/L)

119.94 141.57 122.52 120.72 93.56 88.16 88.80 96.65

2.55 2.77 2.37 2.34 2.31 2.66 2.61 2.29

19.17 18.79 15.70 17.50 17.11 15.95 15.44 17.88

8.20 8.58 6.80 8.11 8.34 9.09 8.90 7.63

114.67 147.49 106.56 113.25 88.41 107.07 118.40 96.52

2.49 2.77 2.49 2.29 2.75 2.72 2.69 2.41

18.79 19.69 15.18 17.50 17.50 16.21 16.85 18.27

8.00 9.15 6.88 8.00 9.12 9.18 8.21 7.57

108.99 6 6.98

2.49 6 0.06

17.19 6 0.49

8.20 6 0.25

111.55 6 6.20

2.58 6 0.06

17.50 6 0.5

8.26 6 0.29

110.03 82.49 102.31 90.34 99.99 136.93 95.36 128.05

2.18 2.10 2.24 2.34 2.21 2.66 2.21 2.60

18.53 18.53 16.47 14.80 17.37 20.46 17.76 18.79

7.71 7.00 8.20 7.06 8.49 8.30 7.81 7.54

108.1 102.05 100.38 98.32 110.93 123.29 92.79 136.55

1.92 2.34 2.24 2.35 2.26 2.60 2.41 2.69

17.24 20.97 15.95 15.31 18.40 19.81 17.37 18.79

7.12 7.89 8.00 7.57 8.20 8.04 7.89 7.86

105.69 6 6.57

2.32 6 0.07

17.84 6 0.59

7.76 6 0.19

109.05 6 5.11

2.35 6 0.08

17.98 6 0.67

7.82 6 0.11

Normal ranges: T4, 58 –151 nmol/L; T3, 1.2–2.7 nmol/L; FT4, 11.8 –24.6 pmol/L; FT3, 5.4 –9.3 pmol/L. determined by the Wilcoxon signed rank test or the Wilcoxon rank sum test, as appropriate. P , 0.05 was considered significant.

Correlation analysis To search for a time-ordered relation between TSH and cortisol, we staggered and correlated the arithmetic or log-transformed values of the concentration-time series of TSH with those of cortisol. As circadian periodicity in the TSH and cortisol series might result in significant correlation between them, reflecting the relative phases of the two circadian rhythms, we looked for the presence of positive or negative correlations on a finer time scale between the detrended series (using a five-point moving average) and the first difference series (D series: second value minus first, third minus second, etc.). These transformations mitigate the effect of baseline shifts and the circadian component. All of the above correlations were performed using Statistica software for IBM computers (26). Cross-correlation analysis between cortisol and TSH was computed at various time lags covering the 24-h study period, as previously described (21).

Results TSH 24-h profile and pulsatility in control and NCCAH children (Fig. 1)

Daytime (1000 –2200 h) and nighttime (2200 –1000 h) mean, mean AUCb, AUCo, peak characteristics (height, amplitude, area, and length), interpeak interval, and number of peak TSH values (6sem) in the control and NCCAH groups of children are reported in Table 3. In both groups, mean and mean AUCo values were higher at night than during the day (P , 0.05). In control children, the nighttime mean pulse height and number of pulses were also higher than those during the day (P , 0.05). The rest of the peak characteristics did not show a day/night difference in this group. In NCCAH children, nighttime mean AUCb, pulse height, area, and length were higher than those during the day (P , 0.05). The daytime mean, mean AUCo, and peak length values were significantly lower in NCCAH than in control children

(P , 0.05). The daytime mean AUCb value was lower in NCCAH children, without, however, reaching statistical significance. There was no difference observed in the nighttime secretory parameters of TSH between the two groups studied. The mean TSH values and the secretory characteristics analyzed over the 24 h did not differ between the two groups (data not shown). No difference was observed in the 12-h pooled (daytime and nighttime) T4,T3, FT4, and FT3 values (Table 2). Cortisol 24-h profile and pulsatility in control and NCCAH children

Daytime (1000 –2200 h) and nighttime (2200 –1000 h) mean, mean AUCb, AUCo, peak characteristics (height, amplitude, area, and length), interpeak interval, and number of peak cortisol values (6sem) in the control and NCCAH groups of children are reported in Table 4. In the control group, mean, mean AUCb, AUCo, and peak height, amplitude, and area values were higher at night than during the day (P , 0.05). In the NCCAH group, most of these values were lower at night than during the day; however, none of them was significantly different. There was no difference in all daytime secretory characteristics between the NCCAH and control children. However, daytime mean AUCb and most of the mean peak secretory characteristics (height, amplitude, area, and number) were higher in NCCAH than in control children, without, however, reaching statistical significance. The nighttime mean peak characteristics (height, amplitude, area, and length) were significantly lower in NCCAH than in control children (P , 0.05). The nighttime mean, mean AUCb, AUCo, and interpeak interval were lower in NCCAH than in control children; however, none of them was significantly different.

3680

JCE & M • 1997 Vol 82 • No 11

GHIZZONI ET AL.

groups of children (P , 0.05), with cortisol leading TSH by this time interval. Similar coefficients of correlation were observed from 2–2.5 h and from 2– 4 h lag times for control and NCCAH children, respectively. In addition, a slightly significant positive correlation was observed over time between cortisol and TSH concentrations, peaking at lag time 13 and 14 h for the control and NCCAH groups of children (P , 0.05), respectively, with cortisol leading TSH by these time intervals (Fig. 2b). A strongly significant positive correlation was observed over time between cortisol and TSH concentrations, peaking at lag times 28.5 and 25.5 h for the control and NCCAH groups of children, respectively, with TSH leading cortisol by these time intervals (Fig. 2c). In addition, a significant negative correlation was observed over time between cortisol and TSH concentrations, peaking at lag time 216.5 and 217.5 h for the control and NCCAH groups of children, respectively, with TSH leading cortisol by these time intervals (Fig. 2d). When the cross-correlation analysis was performed between the cortisol and TSH D and detrended values, no significant correlation was found. Comparison of TSH secretory profiles in control and NCCAH children after 3-h lagging of NCCAH TSH time series

FIG. 1. Twenty-four-hour serum TSH concentrations in control (A) and NCCAH (B) children. The solid line represents the median of values, the dotted lines represent the minimum and maximum values, and the shaded area delineates the lower and upper quartiles. Such a plot (Box-Plot) was chosen to best show outliers and asymmetric behavior.

The mean cortisol values and the secretory characteristics analyzed over 24 h did not differ between the two groups (data not shown). Data for pulsatile cortisol secretion were concordant with those reported previously (23). Correlation analyses

The correlation analysis at lag 0 of raw and D values, after logarithmic transformation and detrending, showed a positive, but not significant, correlation between cortisol and TSH in both control and NCCAH groups of children (data not shown). The graphs depicting the mean coefficients of correlation from the cross-correlation analyses over the 24 h between the cortisol and TSH raw values of both groups are shown in Fig. 2. A strongly significant negative correlation over time (Fig. 2a) was observed between cortisol and TSH concentrations, peaking at lag time 2.5 h for both the control and NCCAH

To verify whether the daytime differences in TSH secretory parameters between the two groups were due to the distinct secretory profile of TSH in the NCCAH group, we analyzed the TSH daytime and nighttime time series of the NCCAH children after shifting them backward by 3 h with respect to control children. We did so, based on the results of the cross-correlation analysis showing an absolute 3-h difference between the lag time of the stronger positive correlation of TSH over cortisol time series in both the control (8.5 h) and NCCAH (5.5 h) groups. Specifically, the 1000 – 2200 and 2200 –1000 h TSH values of the control children were compared with the 1300 – 0100 and 0100 –1300 h values of the patients. In contrast to the previous analysis, there was no difference in the daytime secretory characteristics of TSH between the two groups. Nighttime TSH secretory parameters remained similar in the two groups. Discussion

The results of the present study indicate that there are no quantitative differences in the 24-h TSH secretory parameters between prepubertal normal and NCCAH children. Both groups demonstrated the expected increase in TSH levels during the night. However, in addition to the nighttime baseline and peak height TSH increases in both groups, the nocturnal TSH increase in NCCAH patients was mainly due to an augmentation of the peak area and length, whereas in control children it was caused by an increase in the number of secretory peaks. In the daytime, NCCAH children exhibited lower TSH levels than controls. The evaluation of spontaneous cortisol release confirmed the defective nighttime steroid production previously reported in NCCAH children (20, 21). Daytime mean cortisol AUCb, but not AUCo, and most of the peak secretory characteristics (height, amplitude,

TSH AND CORTISOL INTERACTION

3681

TABLE 3. Properties of pulsatile TSH release in NCCAH and control children 24 h

AUCb (mU/L) AUCo (mU/L) Pulse ht (mU/L) Pulse length (h) Pulse amplitude (mU/L) Pulse area (mU/Lzmin) Mean (mU/L) Interpulse interval (h) No. of pulses Frequency (h)

Day

Night

Controls

Patients

Controls

Patients

Controls

Patients

10.19 6 1.24 49.74 6 4.1 2.60 6 0.17 1.83 6 0.21 0.94 6 0.07 1.10 6 0.11 2.15 6 0.2 3.17 6 0.35 8.00 6 0.84 0.35 6 0.03

8.22 6 0.64 45.87 6 1.74 2.37 6 0.08 2.02 6 0.25 0.85 6 0.05 1.15 6 0.17 1.9 6 0.06 3.64 6 0.42 7.00 6 0.68 0.29 6 0.03

2.51 6 0.76 21.84 6 1.84 2.27 6 0.19 1.73 6 0.17 0.69 6 0.14 0.75 6 0.16 1.97 6 0.19 3.58 6 0.63 2.87 6 0.44 0.25 6 0.03

1.28 6 0.31 15.95 6 1.72b 1.94 6 0.09 1.16 6 0.20b 0.62 6 0.09 0.41 6 0.05 1.45 6 0.04b 3.12 6 0.69 2.83 6 0.47 0.24 6 0.04

4.34 6 0.60 27.94 6 2.61a 3.03 6 0.22a 1.60 6 0.13 0.99 6 0.12 1.00 6 0.21 2.34 6 0.21a 2.81 6 0.34 4.12 6 0.58a 0.36 6 0.04a

3.51 6 0.56a 28.53 6 1.68a 2.99 6 0.22a 1.73 6 0.24a 0.88 6 0.13 0.99 6 0.15a 2.35 6 0.13a 2.22 6 0.46 3.66 6 0.61 0.31 6 0.05

AUCb, Area under the curve above baseline; AUCo, area under the curve above zero line. a P , 0.05, day vs. night. b P , 0.05, patients vs. controls. TABLE 4. Properties of pulsatile cortisol release in NCCAH and control children 24 h Controls

Day Patients

Controls

Night Patients

Controls

Patients

AUCb (nmol/L) 2530.88 6 259.65 2538.34 6 413.32 695.81 6 87.45 943.79 6 191.92 1501.95 6 118.64 974.6 6 291.13 AUCo (nmol/L) 7891.87 6 597.00 7841.46 6 1185.28 3530.21 6 301.57 3800.61 6 565.03 4646.38 6 305.53a 3546.23 6 519.66 a Pulse ht (nmol/L) 467.82 6 35.73 469.82 6 73.69 395.45 6 33.73 436.54 6 72.05 520.56 6 19.99 392.73 6 64.17b Pulse length (h) 1.69 6 0.13 1.45 6 0.09 1.49 6 0.09 1.35 6 0.14 1.48 6 0.38 1.22 6 0.11b Pulse amplitude (nmol/L) 232.58 6 30.24 230.24 6 45.02 142.84 6 16.19 188.19 6 39.32 259.85 6 10.92a 155.52 6 31.83b Pulse area (nmol/Lzmin) 278.17 6 49.47 233.27 6 51.37 145.85 6 20.67 156.29 6 35.69 248.98 6 14.58a 142.07 6 38.96b Mean (nmol/L) 329.47 6 24.60 328.12 6 47.06 293.83 6 26.70 315.05 6 46.31 387.41 6 25.87a 299.43 6 43.05 Interpulse interval (h) 2.85 6 0.17 2.44 6 0.13 2.92 6 0.13 2.19 6 0.14 2.42 6 0.12 2.23 6 0.16 No. of pulses 8.62 6 0.56 9.85 6 0.50 4.25 6 0.36 5.28 6 0.71 4.87 6 0.22 5 6 0.36 Frequency (h) 0.36 6 0.02 0.42 6 0.02 0.36 6 0.03 0.46 6 0.05 0.42 6 0.02 0.44 6 0.03 a

AUCb, Area under the curve above baseline; AUC0, area under the curve above zero line. a P , 0.05, day vs. night. b P , 0.05, patients vs. controls.

area, and number) were higher in NCCAH than in control children, without, however, reaching statistical significance. The relation between the 24-h concentration-time series of plasma cortisol and TSH raw values, as demonstrated by cross-correlation analyses, was almost similar in the two groups. The highest negative correlation was observed when cortisol preceded TSH values by a 2.5-h lag in both groups of children, suggesting that cortisol might negatively regulate TSH secretion even under physiological conditions. This reciprocal relation might reflect a direct negative effect of glucocorticoids on TRH and TSH. Alternatively, glucocorticoids might exert their negative regulation indirectly through inhibition of the noradrenergically mediated stimulation of TRH (27). Furthermore, CRH-activated SRIH release also might be responsible for TRH and TSH inhibition. It has been shown that glucocorticoid secretion during stress is associated with TSH inhibition due to CRH-induced SRIH increase (13). It should be pointed out that this negative correlation of cortisol over TSH was sustained over a longer period of time in NCCAH than in control children. This probably reflects the greater variability in cortisol secretion in this group due to different levels of 21-hydroxylase deficiency (28). A strong positive correlation between the two hormones was observed at 28.5 and 25.5 h lag times for the control and NCCAH groups of children, respectively, with TSH leading cortisol, indicating that sustained secretion of TSH precedes a cortisol surge by these time intervals. The shorter interval

between the secretory phases of these two hormones observed in the group of NCCAH children suggests that in these patients the nocturnal TSH peak occurs temporally closer to the morning cortisol peak than in controls. This is also evident with simple inspection of the secretory hormone profiles (Fig. 3). This temporally delayed TSH secretion in NCCAH children might be due either to the lower nighttime cortisol levels and/or to a glucocorticoid-unopposed norepinephrine-induced TRH stimulation. Furthermore, the nighttime low cortisol secretion in the NCCAH children might also be responsible for the lower daytime TSH release via mild nocturnal cortisol deficiency-induced CRH hypersecretion and stimulation of SRIH resulting in rearrangement of the TSH circadian rhythmicity. In fact, the quantitative daytime difference in TSH secretion disappeared when the TSH levels of the control group were compared with those of the NCCAH group shifted backward by 3 h. The latter is the absolute temporal difference between the lag time of the positive correlation of TSH over cortisol in the two groups. The slightly higher levels of cortisol secretory parameters during the day in NCCAH children than in control children might also reflect and/or explain the difference in daytime TSH secretion in the patient group. In conclusion, NCCAH children have a different circadian pattern of TSH secretion from control children. The mathematically demonstrated relations between TSH and cortisol concentrations over time suggest that endogenous glucocorticoids exert suppressive regulatory influences on TSH se-

3682

GHIZZONI ET AL.

FIG. 2. Collective graphs depicting the cross-correlation analyses of mean coefficients of correlation over the 24-h period between serum cortisol and TSH concentrations in control (A) and NCCAH (B) children. The gray area includes 0 6 2 SEM calculated from the individual values of rk for all children at the lag time k and indicates the limits of significance (P , 0.05). a, b, c, and d indicate the significant correlations.

cretion even under physiological conditions. Slight disturbances in cortisol biosynthesis are associated with slight alterations of TSH secretion. References 1. Re RN, Kourides IA, Ridgway EC, Weintraub BD, Maloof F. 1976 The effect of glucocorticoid administration on human pituitary secretion of thyrotropin and prolactin. J Clin Endocrinol Metab. 43:338 –347. 2. Tolis G, Friesen HG, Bowers CY, McKenzie JM. 1974 Glucocorticoids and Thyrotropin releasing hormone (TRH) secretion. Neuroendocrinology. 15:245–248. 3. Brabant A, Brabant G, Schuermeyer T, et al. 1989 The role of glucocorticoids in the regulation of thyrotropin. Acta Endocrinol (Copenh). 121:95–100. 4. Wilber JE, Utiger RD. 1969 The effect of glucocorticoids on thyrotropin secretion. J Clin Invest. 48:2096 –2103. 5. Brown MR, Hedge GA. 1974 Effects of glucocorticoids on TRH and TSH secretion: dose and time considerations. Am. J Physiol. 227:289 –294. 6. Pamenter RW, Hedge GA. 1980 Inhibition of thyrotropin secretion by physiological levels of corticosterone. Endocrinology. 106:162–166. 7. Kakucska I, Qi Y, Lechan RM. 1995 Changes in adrenal status affect hypothalamic thyrotropin-releasing hormone gene expression in parallel with corticotropin-releasing hormone. Endocrinology. 136:2795–2802. 8. Rivier C, Vale W. 1985 Involvement of corticotropin-releasing factor and somatostatin in stress-induced inhibition of growth hormone secretion in the rat. Endocrinology. 117:2478 –2882.

JCE & M • 1997 Vol 82 • No 11

FIG. 3. Twenty-four-hour serum TSH (gray area) and cortisol (black area) concentrations in control (A) and NCCAH (B) children. The black and gray areas delineate the lower and upper quartiles for each hormone. The arrows indicate the distance between the major secretory peaks of TSH and cortisol. The 24-h secretory profiles of the two hormones have been superimposed for better comparison. 9. Armario A, Marti O, Gavalda A, Giralt M, Jolin T. 1993 Effects of chronic immobilization stress on GH and TSH secretion in the rat: response to hypothalamic regulatory factors. Psychoneuroendocrinology. 18:405– 413. 10. Adriaanse R, Brabant G, Endert E, Wiersinga WM. 1994 Pulsatile thyrotropin secretion in patients with Cushing’s syndrome. Metabolism. 43:782–785. 11. Brabant G, Brabant A, Ranft U, et al. 1987 Circadian and pulsatile thyrotropin secretion in euthyroid man under the influence of thyroid hormone and glucocorticoid administration. J Clin Endocrinol Metab. 65:83– 88. 12. Cizza G, Brady LS, Pacak K, Blackman MR, Gold PW, Chrousos GP. 1995 Stress-induced inhibition of the hypothalamic-pituitary-thyroid axis is attenuated in the aged Fischer 344/N male rat. Neuroendocrinology. 62:506 –513. 13. Chrousos GP, Gold PW. 1992 The concepts of stress and stress system disorders. JAMA. 267:1244 –1252. 14. Hangaard J, Andersen M, Grodum E, Koldkjaer O, Hagen C. 1996 Pulsatile thyrotropin secretion in patients with Addison’s disease during variable glucocorticoid therapy. J Clin Endocrinol Metab. 81:2502–2507. 15. Ismail AA, Burr WA, Walker PL. 1989 Acute changes in serum thyrotrophin in treated Addison’s disease. Clin Endocrinol (Oxf). 30:225–230. 16. Topliss DJ, White EL, Stockigt JR. 1980 Significance of thyrotropin excess in untreated primary adrenal insufficiency. J Clin Endocrinol Metab. 50:52–56. 17. Salvador J, Wilson DW, Harris PE, et al. 1985 Relationships between the circadian rhythms of TSH, prolactin and cortisol in surgically treated microprolactinoma patients. Clin Endocrinol (Oxf). 22:265–272.

TSH AND CORTISOL INTERACTION 18. Salvador J, Dieguez C, Scanlon MF. 1988 The circadian rhythms of thyrotrophin and prolactin secretion. Chronobiol Int. 5:85–93. 19. New MI, Ghizzoni L, Speiser PW. 1996 Update on congenital adrenal hyperplasia. In: Lifshitz F, ed. Pediatric endocrinology, 3rd ed. New York: Marcel Dekker; 305–320. 20. Ghizzoni L, Bernasconi S, Virdis R., et al. 1994 Dynamics of 24-hour pulsatile cortisol, 17-hydroxyprogesterone, and androstenedione release in prepubertal patients with nonclassic 21-hydroxylase deficiency and normal prepubertal children. Metabolism. 43:1– 6. 21. Ghizzoni L, Mastorakos G, Vottero A, Magiakou MA, Chrousos GP, Bernasconi S. 1996 Spontaneous cortisol and growth hormone secretion interactions in patients with nonclassic 21-hydroxylase deficiency (NCCAH) and control children. J Clin Endocrinol Metab. 81:482– 487. 22. Tanner JM, Whitehouse RH. 1976 Clinical longitudinal standards for height, weight, height velocity, weight velocity and stages of puberty. Arch Dis Child.

3683

51:170 –179. 23. Kowarski A, Thompson RG, Migeon CJ, et al. 1971 Determination of integrated concentrations and true secretion rates of human growth hormone. J Clin Endocrinol Metab. 32:356 –360. 24. Greulich WW, Pyle SJ. 1959 Radiographic atlas of skeletal development of the hand and wrist, 2nd Ed. Stanford: Stanford University Press. 25. Merriam GR, Wachter KW. 1982 Algorhythms for the study of episodic hormone secretion. Am J Physiol. 243:E310 –E318. 26. Statistica software for the Windows operating system. 1994 StatSoft Inc., Tulsa, OK. 27. Stratakis CA, Chrousos GP. 1995 Neuroendocrinology and pathophysiology of the stress system. Ann NY Acad Sci. 771:1–18. 28. Speiser PW, Dupont J, Zhu D, et al. 1992 Disease expression and molecular genotype in congenital adrenal hyperplasia due to 21-hydroxylase deficiency. J Clin Invest. 90:584 –595.