Spontaneous Growth Hormone (GH) Secretion Is Not Directly Affected ...

0021-972X/04/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 89(11):5488 –5495 Copyright © 2004 by The Endocrine Society doi: 10.1210/jc.2004-0225

COMMENT Spontaneous Growth Hormone (GH) Secretion Is Not Directly Affected by Ghrelin in Either Short Normal Prepubertal Children or Children with GH Neurosecretory Dysfunction LUCIA GHIZZONI, GEORGE MASTORAKOS, ALESSANDRA VOTTERO, MARIANGELA ZIVERI, IOANNIS ILIAS, AND SERGIO BERNASCONI Department of Pediatrics, University of Parma (L.G., A.V., M.Z., S.B.), 43100 Parma, Italy; and Aretaieion Hospital, Athens University Medical School (G.M., I.I.), 11528 Athens, Greece Ghrelin, a specific endogenous ligand for the GH secretagogue receptor, stimulates GH secretion in humans when given in pharmacological amounts. Under physiological conditions, however, it is controversial whether ghrelin affects GH secretion and vice versa. No studies have reported on the relationship between daily ghrelin and GH secretion in children. Therefore, plasma ghrelin and GH concentrations over a 24-h period were studied in 10 prepubertal short normal children (five females and five males) to determine the potential relationship between the secretion of these two hormones. Furthermore, five prepubertal patients (two females and three males) with GH neurosecretory dysfunction (GHNSD) were studied in the same way to assess potential alterations in ghrelin secretion in a condition associated with distinct GH changes. No gender difference in ghrelin spontaneous secretion was detected in either short normal children or GHNSD patients, and in both male and female subjects, ghrelin was secreted in a pulsatile and circadian fashion, with a nocturnal surge. Twenty-four-hour secretion and daytime ghrelin secretion of short normal children were similar to those in GHNSD patients, whereas nighttime hormone secretion in the latter

G

HRELIN IS A 28-amino acid peptide recently characterized from rat stomach extracts as an endogenous ligand for the GH secretagogue receptor (1). It stimulates GH release in vitro from rat dispersed anterior pituitary cells (1) and is a potent GH secretagogue in vivo when administered systemically to rats and humans in pharmacological doses (2–5). Ghrelin has also been shown to stimulate GH release when administered into the central nervous system, suggesting that an important component of the GH stimulatory effect of ghrelin is at the hypothalamic level (2). In fact, although the major anatomical site of origin is the stomach, ghrelin-immunoreactive neurons have been detected in the Abbreviations: ApEn, Approximate entropy; AUCb, area under the curve above baseline; AUCo, area under the curve above zero line; GHNSD, GH neurosecretory dysfunction; SDS, sd score; SMS, somatostatin. JCEM is published monthly by The Endocrine Society (http://www. endo-society.org), the foremost professional society serving the endocrine community.

group was significantly greater than that in short normal children. The cross-correlation of 24-h ghrelin and GH levels revealed significant positive and negative correlations, which were similar in the two groups examined. The positive one, with GH leading ghrelin, might reflect a somatostatin (SMS)mediated inhibitory effect on both GH and ghrelin secretion (low SMS levels are followed by high GH and ghrelin levels, and vice versa). The negative correlation, with ghrelin leading GH, might again reflect the positive effect of ghrelin on SMS, as shown in both animal and human studies. In conclusion, the results of the present study indicate that ghrelin secretion in prepubertal children is pulsatile and is not sexually dimorphic. Although the parallelism of ghrelin and GH dynamics hints at the potential relevance of endogenous ghrelin as a promoter of GH release, our data do not support this hypothesis. We suggest that the interactions of ghrelin and GH are the result of SMS action. SMS inhibits GH secretion not only by a direct effect on the pituitary and by inhibiting hypothalamic GHRH, but also through the suppression of ghrelin release. (J Clin Endocrinol Metab 89: 5488 –5495, 2004)

hypothalamic arcuate nucleus in rats and the infundibular nucleus in humans (6). Whether pituitary GH release is under the control of ghrelin from the stomach and/or the hypothalamus is presently unknown. Data on the feedback of GH on systemic ghrelin concentrations in humans are also not yet available. In a murine model of acromegaly (PEPCK.hGH transgenic mouse), low levels of systemic ghrelin compared with normal animals were found, possibly due to negative feedback exerted by the high GH levels (7). In contrast, in humans with active acromegaly, circulating systemic ghrelin levels were not decreased (8). Systemic ghrelin levels in subjects with GH deficiency were neither elevated nor modified by 1-yr GH replacement therapy, suggesting that GH does not modulate circulating ghrelin levels (9). Ghrelin, like GH, is secreted in a pulsatile fashion in rats (10). In this animal model, plasma ghrelin and GH levels were not strictly correlated, and cross-approximate entropy (cross-ApEn) did not show a synchronism between these two

5488

Ghizzoni et al. • Comments

J Clin Endocrinol Metab, November 2004, 89(11):5488 –5495

parameters (11). Recently, ghrelin, administered either centrally or peripherally to rats, was shown to exert potent time-dependent stimulation of spontaneous pulsatile GH secretion and to be a functional antagonist of somatostatin (SMS) at the pituitary level (12). No data on the relationship of ghrelin and GH in humans are currently available. The aim of this study was to investigate the interactions between ghrelin and GH under physiological and pathological conditions to search for concordance or discordance between the two hormones. For this reason, we evaluated the 24-h spontaneous secretion of ghrelin and GH and performed a crosscorrelation analysis between the two hormone profiles in short normal prepubertal children and in patients with GH neurosecretory dysfunction (GHNSD). This latter condition refers to slowly growing short children with normal GH response to provocative tests, but impaired spontaneous daily GH secretion due to altered control of somatotroph secretion by neurotransmitters (13–15). In the same children, the orderliness of GH and ghrelin secretion was assessed by estimating the ApEn (16 –18) of each hormone, and the orderliness of interaction in secretion between the two hormones was determined by analyzing the cross-ApEn (16 –18). Subjects and Methods Subjects This study was approved by the clinical research committee of the Department of Pediatrics at University of Parma (Parma, Italy), and informed consent was obtained from the children’s parents. For ethical reasons, a 24-h pulsatility study could not be performed in normal children; therefore, 10 prepubertal children (five males and five females) with familial short stature and normal GH response to pharmacological stimuli (⬎10 ␮g/liter), defined as short normals, and five prepubertal children with GHNSD were selected for the study. Patients with GHNSD had peak GH levels greater than 10 ␮g/liter and 24-h integrated concentrations of GH of 3.2 ␮g/liter or less. Plasma IGF-I levels were within normal levels in short normal children and were below normal levels in GHNSD patients. All subjects were born at term and had normal birth weight and blood thyroid, cortisol, and gonadotropin levels. No evidence of dysmorphic syndromes or psychosocial deprivation was

5489

present in any of the children studied, and none of them was taking any medication. Their clinical characteristics are summarized in Table 1.

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. (19). 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 engage in normal activity. Times of meals and sleep were not purposefully recorded, because although these times are relatively fixed in the hospitals, children rarely comply with both hospital diet and sleep times. Blood samples for measurements of ghrelin and GH concentrations were kept at 4 C and centrifuged within 12 h. After centrifugation, serum was stored at ⫺80 C until assayed. Bone age was determined by the method of Greulich and Pyle (20).

Hormone assays Commercial kits were used for the measurement of serum GH (immunoradiometric assay; Nichols Institute Diagnostics, San Juan Capistrano, CA), and ghrelin (RIA; Phoenix Pharmaceuticals, Belmont, CA). Mean intra- and interassay coefficients of variation were, respectively, 7.8% and 12.4% for ghrelin, and 3.3% and 5.1% for GH.

Statistical analysis Values are reported as the mean ⫾ sem unless otherwise stated. A test for normality was performed on all data. Statistical significance was determined by the Wilcoxon signed rank test or the Wilcoxon rank-sum test, as appropriate. Linear association between two variables was analyzed by linear regression analysis before and after log transformation of the data. The latter was performed to normalize the distribution of the data. The level of significance was set at P ⬍ 0.05.

Pulse analysis The Pulsar program (21) was used to quantitate the pulse properties of GH and ghrelin time series objectively, as previously described (22, 23). Samples were analyzed for 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, 3,

TABLE 1. Clinical profile of children Children/sex

Short normal 1/M 2/M 3/M 4/M 5/M 6/F 7/F 8/F 9/F 10/F Mean ⫾ GHNSD 1/F 2/F 3/M 4/M 5/M Mean ⫾

SEM

SEM

Age (yr)

Bone age (yr)

H-SDS

HV-SDS/BA

HV-SDS/CA

BMI-SDS

8.6 10.5 10.9 9.4 10.6 9.5 9.9 10.7 10.5 8.7

5.0 7.5 7.5 9.0 8.5 9.5 9.5 10.5 10.5 6.7

⫺1.4 ⫺1.1 ⫺1.6 ⫺1.9 ⫺1.2 ⫺2.3 ⫺2.4 ⫺1.4 ⫺1.3 ⫺1.5

⫺1.1 ⫺1.1 ⫺1.3 ⫺0.8 ⫺1.1 ⫺1.2 ⫺0.6 ⫺0.8 ⫺1.1 ⫺0.4

⫺0.3 ⫺0.5 ⫺0.8 ⫺0.7 ⫺0.6 ⫺1.2 ⫺0.7 ⫺1.0 ⫺1.1 ⫺0.7

⫺0.5 ⫺0.7 ⫺0.3 ⫺1.2 ⫺0.4 ⫺0.6 ⫺0.9 ⫺0.5 ⫺1.0 ⫺0.4

9.9 ⫾ 0.3

8.3 ⫾ 0.6

⫺1.1 ⫾ 0.4

⫺1.0 ⫾ 0.1

⫺0.7 ⫾ 0.1

⫺0.6 ⫾ 0.9

7.6 8.5 8.6 10.6 9.2

6.5 6.5 5.0 9.0 7.5

⫺1.5 ⫺0.8 ⫺2.4 ⫺1.7 ⫺1.8

⫺1.8 ⫺2.1 ⫺2.4 ⫺2.5 ⫺2.3

⫺1.7 ⫺1.9 ⫺2.0 ⫺2.3 ⫺2.0

⫺1.4 ⫺1.2 ⫺0.6 ⫺0.3 ⫺0.6

8.9 ⫾ 1.1

6.9 ⫾ 1.5

⫺1.4 ⫾ 0.8

⫺2.2 ⫾ 0.1

⫺2.0 ⫾ 0.1

⫺0.8 ⫾ 0.4

H-SDS, Height SDS; HV-SDS/BA, height velocity SDS for bone age; HV-SDS/CA, height velocity SDS for chronological age; BMI-SDS, body mass index SDS; M, male; F, female.

5490



2, 1.5, and 1 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. Because no sexual dimorphism was revealed by the analysis of ghrelin pulsatile secretion in prepubertal children, to achieve stronger statistical power, the raw data from male and female children were pooled, and pulse and statistical analyses were performed on the pooled data.

Time-series analyses To search for a time-ordered relationship between GH and ghrelin, we staggered and correlated the arithmetic (raw data) and smoothed (by the moving average technique) values of the concentration-time series of ghrelin with those of GH. Cross-correlation analysis between GH and ghrelin was computed at 30-min time lags, covering the 24-h study period, as previously described (24, 25). Although more frequent sampling or a prolonged sampling period would have increased the power of the cross-correlation analysis, the frequency and time of the sampling period were selected to limit inconvenience to the patients. In the moving average technique used for smoothing the time series values, each element of the series was replaced by the simple average of five adjacent elements, where 5 is the width of the smoothing window. The general purpose of the smoothing technique is to reveal the major patterns or trends in a time series while deemphasizing minor fluctuations (random noise). All of these mathematical analyses were performed using Statistica software for the Windows operating system (26).

ApEn analysis The serial data were evaluated by ApEn and cross-ApEn algorithms using empirical statistics based on Monte Carlo procedure 1 (16 –18). We set the input parameters of m (length vector) at 1 and r (tolerance factor of the datasets) at 20% sd of the individuals’ time series. ApEn and cross-ApEn results are indicators of hormonal secretion disorderliness and synchronicity, respectively; thus, the higher the regularity of secretion and the synchronicity between two variables, the lower ApEn and cross-ApEn values. Comparisons of ApEn and cross-ApEn were made using an unpaired t test.

Results GH 24-h profile and pulsatility in short normal children (Fig. 1A) and GHNSD patients (Fig. 1B)

Twenty-four-hour, 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 ghrelin values (⫾ sem) of short normal children and GHNSD patients are reported in Table 2. Because no differences between males and females were detected, data from the two groups were pooled and analyzed together. Most of the 24-h parameters examined were significantly lower in GHNSD patients than in short normal children, except for the number, frequency, and length of pulses and the interpulse interval. Nocturnal AUCo, peak height, and mean GH concentration were significantly higher than the corresponding daytime values in both groups of children studied. No differences in the daytime parameters of hormone secretion were detected between the two groups. In contrast, nocturnal AUCo, peak height, and mean GH concentrations were significantly lower in GHNSD patients than in short normal children. Ghrelin 24-h profile and pulsatility in short normal children (Fig. 1A) and GHNSD patients (Fig. 1B)

Twenty-four-hour, daytime (1000 –2200 h), and nighttime (2200 –1000 h) mean, mean AUCb, AUCo, peak characteris-

FIG. 1. Twenty-four-hour serum GH (gray area) and ghrelin (black area) concentrations in short normal children (A) and GHNSD patients (B). The gray and black areas delineate the lower and upper quartiles, respectively, for each hormone. The 24-h secretory profiles of the two hormones have been superimposed for better comparison.

tics (height, amplitude, area, and length), interpeak interval, and number of peak ghrelin values (⫾ sem) of short normal children and GHNSD patients are reported in Table 3. Because no differences between males and females were detected, data from the two groups were pooled and analyzed together. All 24-h and daytime parameters examined were not different between the two groups. In short normal children, AUCo and mean ghrelin values were significantly higher at night than during the day, whereas in GHNSD patients, AUCb and peak area significantly increased at night compared with daytime values. Nighttime AUCb, peak height, and mean ghrelin values and number of hormone peaks were significantly higher in GHNSD patients than in short normal children. The remaining 12-h nighttime parameters examined were not significantly different between the two groups. Times of meal consummation and sleep initiation were not recorded. However, ghrelin levels 30 min before and 0, 30, and 60 min after the time when trays were brought to the patients’ rooms were compared, and no statistically significant differences were found among the values (12 h, 304.2 pg/ml; 1230 h, 254.4 pg/ml; 1300 h, 311.9;



5491

TABLE 2. Properties of pulsatile GH secretion in prepubertal short normal children and GHNSD patients 24 h

AUCb (␮g/liter) AUCo (␮g/liter) Pulse length (h) Pulse height (␮g/liter) Pulse amplitude (␮g/liter) Pulse area (␮g/liter䡠min) Mean (␮g/liter) Interpulse interval (h) No. of pulses Frequency (h) a b

Day

Night

Short normal children

GHNSD patients


GHNSD patients


GHNSD patients

95.1 ⫾ 9.4 113.5 ⫾ 10.5 1.2 ⫾ 0.2 10.6 ⫾ 1.3 9.8 ⫾ 1.3 11.2 ⫾ 2.1 4.7 ⫾ 0.4 3.1 ⫾ 0.2 8.2 ⫾ 0.7 0.3 ⫾ 0.2

49.9 ⫾ 8.0a 61.4 ⫾ 6.9a 1.6 ⫾ 0.1 5.5 ⫾ 1.2a 5.0 ⫾ 1.3a 4.9 ⫾ 1.5a 1.9 ⫾ 0.5a 2.4 ⫾ 0.2 9.7 ⫾ 1.0 0.4 ⫾ 0.4

21.6 ⫾ 5.4 32.8 ⫾ 6.6b 1.6 ⫾ 0.1 7.3 ⫾ 1.3b 6.2 ⫾ 1.2 4.7 ⫾ 1.2 2.9 ⫾ 0.5b 1.7 ⫾ 0.8 4.3 ⫾ 0.3 0.2 ⫾ 0.1

15.1 ⫾ 4.6 18.7 ⫾ 4.9b 1.6 ⫾ 0.2 4.5 ⫾ 1.3b 4.2 ⫾ 1.3 3.6 ⫾ 1.5 1.6 ⫾ 0.4b 2.4 ⫾ 0.2 5.0 ⫾ 1.0 0.4 ⫾ 0.1

34.4 ⫾ 7.2 76.3 ⫾ 9.5 1.5 ⫾ 0.1 16.1 ⫾ 3.1 10.4 ⫾ 2.0 8.0 ⫾ 1.4 6.4 ⫾ 0.8 4.9 ⫾ 2.1 4.2 ⫾ 0.4 0.2 ⫾ 0.1

21.3 ⫾ 6.3 39.3 ⫾ 3.4a 1.5 ⫾ 0.2 6.7 ⫾ 1.0a 4.7 ⫾ 1.3 4.2 ⫾ 1.9 3.4 ⫾ 0.2a 2.1 ⫾ 0.1 5.5 ⫾ 0.6 0.5 ⫾ 0.05

P ⬍ 0.05, short normal children vs. GHNSD. P ⬍ 0.05, night vs. day.

TABLE 3. Properties of pulsatile ghrelin secretion in prepubertal short normal children and GHNSD patients 24 h

AUCb (pg/ml) AUCo (pg/ml) Pulse length (h) Pulse height (pg/ml) Pulse amplitude (pg/ml) Pulse area (pg/ml䡠min) Mean (pg/ml) Interpulse interval (h) No. of pulses Frequency (h) a b

Day

Night


GHNSD patients


GHNSD patients


GHNSD patients

1209.8 ⫾ 561.4 6415.6 ⫾ 855.7 6.6 ⫾ 1.7 406.6 ⫾ 60.0 201.1 ⫾ 38.7 273.8 ⫾ 101.9 267.1 ⫾ 35.7 3.7 ⫾ 1.1 5.0 ⫾ 1.3 0.2 ⫾ 0.05

1509.0 ⫾ 436.9 8093.9 ⫾ 1143.6 3.4 ⫾ 0.9 474.4 ⫾ 83.4 200.5 ⫾ 25.4 139.6 ⫾ 31.9 337.2 ⫾ 48.0 2.2 ⫾ 0.8 7.0 ⫾ 2.2 0.3 ⫾ 0.1

348.2 ⫾ 172.6 2785.1 ⫾ 211.1a 4.6 ⫾ 0.6 281.2 ⫾ 48.4 130.4 ⫾ 32.0 130.3 ⫾ 42.2 210.1 ⫾ 24.2a 1.7 ⫾ 0.8 2.7 ⫾ 1.0 0.2 ⫾ 0.1

417.1 ⫾ 265.8a 4536.9 ⫾ 1273.5 3.0 ⫾ 0.6 428.8 ⫾ 54.6 147.8 ⫾ 41.6 70.1 ⫾ 7.2a 292.7 ⫾ 24.7 1.8 ⫾ 1.0 3.2 ⫾ 1.4 0.5 ⫾ 0.2

211.1 ⫾ 73.1b 4741.5 ⫾ 533.4 2.7 ⫾ 0.9 373.1 ⫾ 108.1b 150.0 ⫾ 91.1 83.7 ⫾ 62.7 325.0 ⫾ 52.8b 4.9 ⫾ 2.1 2.7 ⫾ 0.7b 0.2 ⫾ 0.1

1052.9 ⫾ 286.7 5903.5 ⫾ 1614.6 3.3 ⫾ 0.2 448.8 ⫾ 129.9 160.5 ⫾ 26.1 119.9 ⫾ 18.9 1044.7 ⫾ 517.8 1.7 ⫾ 0.2 6.7 ⫾ 0.7 0.6 ⫾ 0.1

P ⬍ 0.05, night vs. day. P ⬍ 0.05, short normal children vs. GHNSD.

1330 h, 287.4; 1800 h, 249.2 pg/ml; 1830 h, 255.8 pg/ml; 1900 h, 242.2 pg/ml; 1930 h, 283.1 pg/ml). Linear regression analysis of auxological parameters and ghrelin and GH secretion

The relationships between daytime, nighttime, and 24-h parameters of ghrelin and GH secretion and chronological and bone ages, height sd score (SDS), height velocity SDS, and body mass index SDS were evaluated using linear regression analysis. No significant relationship was found between the auxological variables and any of the daytime, nighttime, and 24-h ghrelin secretory parameters in either short normal children or GHNSD patients. A positive correlation was observed between height SDS of short normal children and mean 24-h GH secretion, 24-h GH AUCb, and nighttime GH peak length. No significant relationship was found between the auxological parameters of GHNSD patients and any of the daytime, nighttime, and 24-h GH secretory parameters. Analyses of ghrelin and GH 24-h time-series in short normal children

Cross-correlation analysis of the raw values. The mean coefficient of correlation for each time point of the 24-h raw values of the two hormones submitted to cross-correlation analysis is given in Fig. 2A. There is a strongly significant positive correlation over time between ghrelin and GH, peaking at a

1.5 h lag time, with GH leading ghrelin. There is also a negative correlation over time, peaking at a 4 h lag time, with ghrelin leading GH. Cross-correlation analysis of the smoothed values. The graph depicting the mean coefficient of correlation for each time point over 24 h of the smoothed values of the two hormones submitted to cross-correlation analysis is depicted in Fig. 2B. There is a strongly significant positive correlation over time between ghrelin and GH, peaking at a 0 h lag time. There is also a negative correlation over time, peaking at a 4.5 h lag time, with ghrelin leading GH. Analyses of ghrelin and GH 24-h time-series in GHNSD patients

Cross-correlation analysis of the raw values. The mean coefficient of correlation for each time point of the 24-h raw values of the two hormones submitted to cross-correlation analysis is shown in Fig. 3A. There is a significant positive correlation over time between ghrelin and GH, peaking at a 7 h lag time, with GH leading ghrelin. There is also, a significant negative correlation over time, with ghrelin leading GH, peaking at a 7 h lag time. Cross-correlation analysis of the smoothed values. The graph depicting the mean coefficient of correlation for each time point over 24 h of the smoothed values of the two hormones submitted to cross-correlation analysis is shown in Fig. 3B.

5492



FIG. 2. Collective graphs depicting the cross-correlation analyses of mean coefficients of correlation over the 24-h period between serum GH and ghrelin raw (A) and smoothed (B) concentrations in prepubertal short normal children. The area between the dotted lines includes 0 ⫾ 2 SEM calculated from the individual values of rk for all children at lag time k and indicates the limits of significance (P ⬍ 0.05). Arrows indicate the significant correlations.

There is a significant positive correlation over time between ghrelin and GH, peaking at a 7 h lag time, with GH leading ghrelin. There is also a negative correlation over time, peaking at a 7 h lag time, with ghrelin leading GH. ApEn and cross-ApEn analyses

The regularity of ghrelin secretion was similar in the two groups of children studied, whereas GH secretion was quantitatively more irregular in GHNSD patients than in short normal children, as indicated by ApEn (Table 4). Cross-ApEn analysis revealed that ghrelin-GH secretory patterns of short normal children were more synchronized than those of GHNSD patients (Table 4). Discussion

The results of the present study indicate that ghrelin is secreted in a pulsatile fashion with diurnal variation in both

prepubertal short normal children and patients with GHNSD. Specifically, in both groups there was a nighttime increase in plasma ghrelin levels due to an elevation in the mean hormone levels and the total AUC in short normal children and to the augmentation of some of the peak components of the hormone secretion, such as peak area and AUCb, in GHNSD patients. Twenty-four-hour secretion and daytime ghrelin secretion in short normal children were similar to those in GHNSD patients, whereas nighttime hormone secretion in the latter group was greater than that in short normal children due to a higher baseline, mean hormone concentration, and number of peaks. GH showed a secretory pattern similar to that of ghrelin, with a nocturnal increase in the hormone concentration in both groups examined. However, in contrast to ghrelin and as expected, the GH increase at night was significantly lower in GHNSD patients than in short normal children.



5493

FIG. 3. Collective graphs depicting the cross-correlation analyses of mean coefficients of correlation over the 24-h period between serum GH and ghrelin raw (A) and smoothed (B) concentrations in GHNSD patients. The area between the dotted lines includes 0 ⫾ 2 SEM calculated from the individual values of rk for all children at lag time k and indicates the limits of significance (P ⬍ 0.05). Arrows indicate the significant correlations.

TABLE 4. GH and ghrelin ApEn and cross-ApEn values in short normal children and GHNSD patients

Short normal children GHNSD patients

ApEn ghrelin

ApEn GH

Cross-ApEn ghrelin-GH

1.15 ⫾ 0.08 1.18 ⫾ 0.08

0.81 ⫾ 0.09a 1.01 ⫾ 0.07

1.49 ⫾ 0.13a 1.71 ⫾ 0.13

Values are the mean ⫾ SD. Length vector (m) ⫽ 1 and tolerance factor (r) ⫽ 20%. a P ⬍ 0.05 vs. GHNSD patients.

A pulsatile ghrelin secretion with an ultradian rhythmicity was previously shown in rats (10). In humans, a pulsatile pattern of ghrelin secretion was asserted, although not statistically substantiated (8, 27, 28). In the present study the evidence of ghrelin pulses was assessed objectively and was confirmed in both groups of children studied together with a nocturnal rise in hormone levels. In accordance with previous studies (29, 30), no gender difference in spontaneous ghrelin secretion was detected in either short normal children or GHNSD patients. The absence of sexual dimorphism in plasma ghrelin concentrations was recently reported in a population of adult male and female subjects of different

ages (31). In that study, ghrelin levels were not different among premenopausal and postmenopausal women taking hormone replacement therapy, indicating that ghrelin levels are not physiologically regulated by estrogen in women (31). However, in another study, ghrelin levels were shown to be higher in women in the late follicular stage than in men, suggesting a primary role of gonadal hormones in determining the sexual dimorphism in serum ghrelin concentrations (8). Recently, changes in ghrelin from prepuberty to puberty were reported, with hormone levels falling with increasing age and with no differences between males and females (30), indicating that in childhood and adolescence, gonadal steroids do not exert a positive effect on ghrelin secretion. The parallelism of ghrelin and GH dynamics hints at a potential relevance of endogenous ghrelin as a promoter of GH release. Studies in patients with genetic GHRH hormone resistance (32, 33) indicate that factors other than GHRH contribute to the pulsatile and diurnal patterns of GH secretion in men, and ghrelin was proposed as a candidate for driving the enhanced nocturnal GH secretion. Although the results in short normal children seem to agree with the latter hypothesis, those in GHNSD are against it. In the GHNSD

5494


group, in fact, despite the greater nighttime elevation of ghrelin levels compared with that in short normal children, GH levels were lower, suggesting that ghrelin is not a compelling candidate for generating the nighttime increase in GH secretion, but it might be regulated, in turn, by GH. The low nocturnal GH secretion in the presence of high ghrelin levels in GHNSD indicates that GH might exert feedback on ghrelin secretion, as recently described in rats, in which the presence of a stomach-ghrelin-pituitary axis was hypothesized based on the responsiveness of ghrelin to changes in systemic GH levels (34). Correlation analysis demonstrated that GH and ghrelin correlated to each other over time in both a positive and a negative fashion. When cross-correlation analysis was performed between 24-h GH and ghrelin raw and smoothed values, the highest positive correlation was observed when GH values preceded ghrelin values with 1.5 and 0 h lag times for raw and smoothed values, respectively, in short normal children and 7 h in GHNSD patients, with GH leading ghrelin. This positive correlation between GH and ghrelin might reflect the effect of SMS on GH and ghrelin secretion. The pulsatile pattern of GH secretion is due to the interplay of GHRH-induced GH pulses modulated by prevailing SMS tone and rapid SMS oscillations (35). Thus, the occurrence of a GH pulse is preceded by low SMS levels. Because SMS was shown to suppress ghrelin levels (12, 36), the low SMS concentrations preceding GH pulses might explain the high ghrelin levels coinciding with a GH pulse and vice versa, as indicated by the 0 h lag time between smoothed values of GH and ghrelin in short normal children. The longer lag time observed in GHNSD patients compared with short normal children denotes a weaker significance of the mathematical correlation between the spontaneous secretion of the two hormones, which is presumably explained by the alterations of the GH interplay with the different neurotransmitters and/or neurohormones (i.e. SMS) in GHNSD patients. There was also a negative correlation between GH and ghrelin raw and smoothed values at lag times of 4.5 and 7 h for short normal children and GHNSD patients, respectively, with ghrelin leading GH. This means that in both groups a burst of ghrelin is followed by a nadir in GH levels at these time intervals and vice versa. The inverse relationship between ghrelin and GH might again reflect the effects of SMS on ghrelin and vice versa. Continuous ghrelin treatment was reported to increase SMS mRNA expression in the periventricular nuclei of moderately GH-deficient rats (37). In humans, acute ghrelin administration showed a stimulating effect of the hormone on peripheral SMS levels (38). Conversely, in healthy subjects, the administration of either SMS (36) or octreotide (8) was shown to suppress the plasma ghrelin concentration. Thus, a complex interactive loop may be active under physiological conditions underlying the inverse correlation between ghrelin and GH detected in both short normal children and GHNSD patients. Again, the longer lag time observed in GHNSD patients compared with short normal children denotes a distinct shape of the line of correlation between these two hormones, suggesting once more that in GHNSD patients the interplay of GH with the different neurotransmitters is altered. This is in agreement with the recently shown reduction in GHNSD children of the


GH releasable pool evaluated by the provocative and potent combined GHRH and arginine test (14). These findings together with those reported herein suggest that altered control of the hypothalamic GHRH-secreting neurons by neurotransmitters and neuropeptides may be a plausible cause of impaired GH secretion in GHNSD patients. The observed differences in GH ApEn between patients with GHNSD and short normal children reflect a more disorderly secretion of GH in the former compared with the latter. The absence of a significant difference in ghrelin ApEn between the two groups of children indicates that ghrelin is not directly involved with GH secretion. Therefore, the differences in ghrelin-GH cross-ApEn between the two groups may presumably be attributed to the more disorderly GH ApEn secretion detected in GHNSD children. Although the results obtained from short normal children are likely to reflect normal physiology, an as yet unknown alteration of hormones in familial short stature might also contribute to the described hormone relationships. In conclusion, the results of the present study indicate that ghrelin is secreted in a pulsatile fashion with diurnal variation in both prepubertal short normal children and patients with GHNSD. Ghrelin levels were similar in males and females and were not correlated with the auxological parameters in either group examined. The greater nighttime elevation of ghrelin levels detected in GHNSD patients compared with short normal children accompanied by lower GH levels suggest that ghrelin is not a compelling candidate for generating the nighttime increase in GH secretion. Although the parallelism of ghrelin and GH dynamics hints at a potential relevance of endogenous ghrelin as a promoter of GH release, our data do not support this hypothesis. We suggest that the interactions of ghrelin and GH are the result of the SMS action. SMS inhibits GH secretion not only by a direct effect on the pituitary and by inhibiting hypothalamic GHRH, but also through the suppression of ghrelin release. Acknowledgments Received February 13, 2004. Accepted August 12, 2004. Address all correspondence and requests for reprints to: Dr. Lucia Ghizzoni, Department of Pediatrics, University of Parma, Via Gramsci 14, 43100 Parma, Italy. E-mail: [email protected].

References 1. Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K 1999 Ghrelin is a growth hormone-releasing acylated peptide from stomach. Nature 402:656 – 660 2. Seoane LM, Tovar S, Baldelli R, Arvat E, Ghigo E, Casanueva F, Dieguez C 2000 Ghrelin elicits a marked stimulatory effect on GH secretion in freelymoving rats. Eur J Endocrinol 143:R7–R9 3. Tolle V, Zizzari P, Tomasetto C, Rio M-C, Epelbaum J, Bluet-Pajot M-T2001 In vivo and in vitro effects of ghrelin/motilin-related peptide on growth hormone secretion in the rat. Neuroendocrinology 73:54 – 61 4. Takaya K, Ariyasu H, Kanamoto N, Iwakura H, Yoshimoto A, Harada M, Mori K, Komatsu Y, Usui T, Shimatsu A, Ogawa Y, Hosoda K, Akamizu T, Kojima M, Kangawa K, Nakao K 2000 Ghrelin strongly stimulates growth hormone (GH) release in humans. J Clin Endocrinol Metab 85:4908 – 4911 5. Arvat E, Di Vito L, Broglio F, Capotti M, Muccioli G, Dieguez C, Casanueva F, Deghenghi R, Capanni F, Ghigo E 2000 Preliminary evidence that ghrelin, the natural GH secretagogue (GHS)-receptor ligand, strongly stimulates GH secretion in humans. J Endocrinol Invest 23:493– 495 6. Korbonits M, Bustin SA, Kojima M, Jordan S, Adams EF, Lowe DG, Kangawa K, Grossman AB 2001 The expression of the growth hormone secretagogue receptor ligand ghrelin in normal and abnormal human pituitary and other neuroendocrine tumors. J Clin Endocrinol Metab 86:881– 887


7. Wright JC, Bartke A, Ghrelin administration decreases insulin secretion in Ames dwarf mice. Proc of the 83rd Annual Meeting of The Endocrine Society, Denver, CO, 2001, pp 170 –171 (Abstract P1-98) 8. Barkan AL, Dimaraki EV, Jessup SK, Symons KV, Ermolenko M, Jaffe CA 2003 Ghrelin secretion in humans is sexually dimorphic, suppressed by somatostatin, and not affected by the ambient growth hormone levels. J Clin Endocrinol Metab 88:2180 –2184 9. Janssen JAMJL, van der Toorn FM, Hofland LJ, van Koetsveld P, Broglio F, Ghigo E, Lamberts SWJ, van der Lely AJ 2001 Systemic ghrelin levels in subjects with growth hormone deficiency are not modified by one year of growth hormone replacement therapy. Eur J Endocrinol 145:711–716 10. Bagnasco M, Kalra PS, Kalra SA 2002 Ghrelin and leptin pulse discharge in fed and fasted rats. Endocrinology 143:726 –729 11. Tolle V, Bassant M-H, Zizzari P, Poindessous-Jazat F, Tomasetto C, Epelbaum J, Bluet-Pajot M-T 2002 Ultradian rhythmicity of ghrelin secretion in relation with GH, feeling behavior, and sleep-wake patterns in rats. Endocrinology 143:1353–1361 12. Tannenbaum GS, Epelbaum J, Bowers CY 2003 Interrelationship between the novel peptide ghrelin and somatostatin/growth hormone-releasing hormone in regulation of pulsatile growth hormone secretion. Endocrinology 144:967–974 13. Spiliotis BE, August GP, Hung W, Sonis W, Mendelson W, Bercu B 1984 Growth hormone neurosecretory dysfunction: a treatable cause of short stature. JAMA 251:2223–2230 14. Aimaretti G, Bellone S, Bellone J, Chiabotto P, Baffoni C, Corneli G, Origlia C, de Sanctis C, Capanni F, Ghigo E 2000 Reduction of the pituitary GH releasable pool in short children with neurosecretory dysfunction. Clin Endocrinol (Oxf) 52:287–293 15. Jorgensen EV, Schwartz ID, Hvizdala E, Barbosa J, Phuphanich S, Shulman DI, Root AW, Estrada J, Hu CS, Bercu BB 1993 Neurotransmitter control of growth hormone secretion in children after cranial radiation therapy. J Pediatr Endocrinol 6:131–142 16. Johnson ML, Straume M 1999 Innovative quantitative neuroendocrine techniques. In: Giustina A, ed. Sex-steroid interactions with growth hormone. New York: Springer-Verlag; 318 –326 17. Pincus SM, Hartman ML, Roelfsema F, Thorner MO, Veldhuis JD 1999 Hormone pulsatility discrimination via coarse and short time sampling. Am J Physiol 277:E948 –E957 18. Veldhuis JD, Pincus SM, Garcia-Rudaz MC, Ropelato MG, Escobar ME, Barontini M 2001 Disruption of the synchronous secretion of leptin, LH, and ovarian androgens in nonobese adolescents with the polycystic ovarian syndrome. J Clin Endocrinol Metab 86:3772–3778 19. Kowarski A, Thompson RG, Migeon CJ, Blizzard RM 1971 Determination of integrated concentrations and true secretion rates of human growth hormone. J Clin Endocrinol Metab 32:356 –360 20. Greulich WW, Pyle SJ 1959 Radiographic atlas of skeletal development of the hand and wrist. 2nd ed. Stanford, CA: Stanford University Press 21. Merriam GR, Wachter KW 1982 Algorhythms for the study of episodic hormone secretion. Am J Physiol 243:E310 –E318 22. Ghizzoni L, Mastorakos G, Street ME, Vottero A, Mazzardo G, Vanelli M, Chrousos GP, Bernasconi S 1997 Spontaneous thyrotropin and cortisol secretion interactions in patients with nonclassical 21-hydroxylase deficiency and control children. J Clin Endocrinol Metab 82:1–7


5495

23. 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 81:1–5 24. Ghizzoni L, Mastorakos G, Ziveri M, Furlini M, Solazzi A, Vottero A, Bernasconi S 2001 Interactions of leptin and thyrotropin 24-hour secretory profiles in short normal children. J Clin Endocrinol Metab 86:2065–2072 25. Ghizzoni L, Mastorakos G, Street ME, Mazzardo G, Vottero A, Vanelli M, Bernasconi S 2001 Leptin, cortisol, and GH secretion interactions in short normal prepubertal children. J Clin Endocrinol Metab 86:3729 –3734 26. StatSoft 1994 Statistica software for the Windows operating system 27. Cummings DE, Purnell JQ, Frayo RS, Schmidova K, Wisse BE, Weigle DS 2001 A preprandial rise in plasma ghrelin levels suggests a role in meal initiation in humans. Diabetes 50:1714 –1719 28. Cummings DE, Weigle DSW, Frayo RS, Breen PA, Ma MK, Dellinger EP, Purnell JQ 2002 Plasma ghrelin levels after diet-induced weight loss or gastric bypass surgery. N Engl J Med 346:1623–1630 29. Bellone S, Rapa A, Vivenza D, Castellino N, Petri A, Bellone J, Me E, Broglio F, Prodam F, Ghigo E, Bona G 2002 Circulating ghrelin levels as function of gender, pubertal status and adiposity in childhood. J Endocrinol Invest 25:13–15 30. Whatmore AJ, Hall CM, Jones J, Westwood M, Calyton PE 2003 Ghrelin concentrations in healthy children and adolescents. Clin Endocrinol (Oxf) 59:649 – 654 31. Purnell JQ, Weigle DS, Breen P, Cummings DE 2003 Ghrelin levels correlate with insulin levels, insulin resistance, and high-density lipoprotein cholesterol, but not with gender, menopausal status, or cortisol levels in humans. J Clin Endocrinol Metab 88:5747–5752 32. Roelfsema F, Biermasz NR, Veldman RG, Veldhuis JD, Fro¨lich M, StokvisBrantsma WH, Wit J-A 2001 Growth hormone (GH) secretion in patients with an inactivating defect of the GH-releasing hormone (GHRH) receptor is pulsatile: evidence for a role for non-GHRH inputs into the generation of GH pulses. J Clin Endocrinol Metab 86:2459 –2464 33. Maheshwari HG, Pezzoli SS, Rahim A, Shalet SM, Thorner MO, Baumann G 2002 Pulsatile growth hormone secretion persists in genetic growth hormone-releasing hormone resistance. Am J Physiol 282:E943–E951 34. Qi X, Reed J, Englander EW, Chandrashekar V, Bartke A, Greeley Jr GH 2003 Evidence that growth hormone exerts a feedback effect on stomach ghrelin production and secretion. Exp Biol Med 228:1028 –1032 35. Tannenbaum GS, Ling N 1984 The interrelationship of growth hormone (GH)-releasing factor and somatostatin in generation of the ultradian rhythm of GH secretion. Endocrinology 115:1952–1957 36. Nørrelund H, Hansen TK, Ørskov H, Hosoda H, Kojima M, Kangawa K, Weeke J, Møller N, Christiansen JS, Jørgensen JOL 2002 Ghrelin immunoreactivity in human plasma is suppressed by somatostatin. Clin Endocrinol (Oxf) 57:539 –546 37. Thompson NM, Jeffrey SD, Mode A, Houston PA, Wells T 2003 Patterndependent suppression of growth hormone (GH) pulsatility by ghrelin and GH-releasing peptide-6 in moderately GH-deficient rats. Endocrinology 144: 4859 – 4867 38. Arosio M, Ronchi CL, Gebbia C, Cappiello V, Beck-Peccoz P, Peracchi M 2003 Stimulatory effects of ghrelin on circulating somatostatin and pancreatic polypeptide levels. J Clin Endocrinol Metab 88:701–704

JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.