Correlation between Fasting Plasma Ghrelin Levels and Age, Body Mass Index (BMI), BMI Percentiles, and 24-Hour Plasma Ghrelin Profiles in Prader-Willi Syndrome

Ghrelin is a GH-releasing acylated peptide found in the stomach and a centrally acting food intake stimulator. Prader-Willi syndrome (PWS) is a genetic disorder characterized by a voracious appetite and increased fasting ghrelin levels. In this report we describe 24-h ghrelin profiles in PWS children (n = 5) and compare these with age, sex, and body mass index (BMI)-matched controls (n = 5). A 3- to 4-fold increase in ghrelin levels was found in PWS over a 24-h period, compared with controls (P < 0.001). Interestingly, there was a greater tendency for the up-regulation of ghrelin level in lean PWS than in obese PWS. To confirm this finding, we measured fasting ghrelin levels in 39 patients with PWS. Inverse correlations were found between plasma ghrelin levels and the following: age (r = −0.408, P = 0.005), BMI (r = −0.341, P = 0.017), percentage of the ideal weight for age (r = −0.382, P = 0.008), and BMI percentile (r = −0.311, P = 0.027). Our data show that there may be a suppressive (or up-regulating) controlling mechanism of ghrelin secretion in obese (or lean) PWS children. We hope that our data may further explain the mechanisms underlying the insatiable appetite and obesity characteristic of PWS.

GHRELIN IS AN endogenous ligand for the GH secretagogue receptor and is synthesized principally in the stomach (1). In addition to having a powerful effect on the secretion of GH, ghrelin signals directly to the hypothalamic regulatory nuclei that control energy homeostasis (2, 3). In mice, peripheral injection with ghrelin for 2 wk resulted in significant body weight increases, which were attributed to an increase in fat mass (4). These mice also displayed an increased respiratory quotient, reflecting a reduced utilization of fat. However, neither energy expenditure nor locomotor activity was found to be significantly altered in these mice (4).

Prader-Willi syndrome (PWS) is a contiguous gene syndrome resulting from the deletion of the parental copies of the imprinted SNRPN gene, the necdin gene, and possibly other genes (5, 6). The syndrome is characterized by diminished fetal activity, obesity, muscular hypotonia, mental retardation, short stature, hypogonadotropic hypogonadism, and small hands and feet (7). It is caused by the deletion or disruption of a gene or several genes on the proximal long arm of the paternal chromosome 15 or maternal uniparental disomy 15 because the genes on the maternal chromosome 15 are virtually inactive through imprinting (8). PWS neonates usually present with feeding difficulties; however, between 12 and 18 months, uncontrollable hyperphagia causes major somatic as well as psychologic problems. Although most PWS children become obese, one study reported on the body compositions and leptin levels in 13 young, underweight children and those in 10 older overweight children with PWS. Both groups showed elevated skinfold sd scores for body mass indexes (BMIs) and elevated BMIs, suggesting a relative increased body fat even in underweight children (9).

Leptin, a protein secreted by adipocytes in proportion to body fat, is known to be a crucial element of the body weight regulatory system, and a mutation in the leptin receptor gene can result in a severe obesity. It was reported that median leptin serum concentrations are similar in PWS and children with nonsyndrome obesity (10). However, circulating plasma leptin levels in nonobese PWS males were nearly 5 times higher than in nonobese control males with similar BMIs (11). To date three reports have correlated high circulating ghrelin with increased appetite and obesity in PWS (12–14). However, these reports showed fasting ghrelin levels only in PWS. Here we present 24-h ghrelin profiles in PWS children (n = 5) and compared these with age-, sex-, and BMI-matched controls (n = 5). Also, we found a correlation of fasting ghrelin with age, BMI, and BMI percentile, which is observed in normal controls, also occurs in PWS.

Subjects and Methods

Subjects

Twenty-four-hour ghrelin measurements.

Six PWS children with BMIs between 19 and 49 kg/m² and five controls with BMIs between 20 and 51 kg/m² were enrolled for 24-h ghrelin monitoring. We explained the purpose and methods of the study, and informed consent had been obtained from parents or guardians in PWS children. Table 1 shows the characteristics of the study subjects. We could match the age, sex, and BMI of the five patients with those of five controls. For the recruitment of the controls, we visited several middle and high schools located in the southern part of Seoul, Korea. We explained the purpose of the study to the teachers, and written study protocol was sent to the parents of the pupils. Among 23 volunteers, we selected the five controls based on the age, sex, and BMI. For subject 6, we failed to match the control because of younger age. All the study subjects were males.

TABLE 1.

Characteristics of male study subjects (P1–6, PWS; C1–5, control)

Subject	Age (yr)	% Ideal weight	BMI (kg/m2)	BMI percentile	Genotype
P1	11	131	24	>95	Deletion
P2	16	248	49	>95	Deletion
P3	16	255	48	>95	Deletion
P4	11	204	35	>95	Deletion
P5	12	104	19	50–75	UPD
	Mean± se
	13.2 ± 1.2	188 ± 30.6	35.0 ± 6.1
C1	11	137	25	>95
C2	17	250	51	>95
C3	17	250	45	>95
C4	12	195	35	>95
C5	12	107	20	50–75
	Mean± se
	13.8 ± 1.3	187 ± 29.0	35.2 ± 5.8
P6	9	140	23	>95	UPD

UPD, Uniparental disomy.

Fasting ghrelin measurement.

Thirty-nine PWS children [male (M)/female (F), 26:13; age, 6.23 ± 0.74 yr; BMI, 22.30 ± 1.50 kg/m²], 42 normal controls [age, 8.05 ± 0.60 yr; BMI, 20.42 ± 0.94 kg/m²], which included 16 obese children [obese control, M:F, 7:9; age, 10.52 ± 0.72 yr; BMI, 26.8 ± 1.20 kg/m²] with BMIs over the 95th percentile for age, and the other 26 normal-weight children [lean control, M:F, 12:14; age, 6.53 ± 0.73 yr; BMI, 16.49 ± 0.42 kg/m²] were enrolled for the fasting plasma ghrelin study after obtaining informed consent. All cases of PWS were genetically confirmed by the standard methylation test. Of the PWS patients, 32 patients had deletion of the paternally transmitted chromosome 15q11, five patients had uniparental disomy, and two patients imprinting defects. The study design was reviewed and approved by the Samsung Medical Center Institutional Review Board.

Experimental design

For the 24-h sampling study, all subjects were admitted to the pediatric ward at the Samsung Medical Center, fed a standard hospital diet [meals served at 0800, 1200, and 1730 h, total calories 2100 calories consisting of 60% carbohydrate (310 g), 20% protein (100 g), and 20% fat (50 g)] except patient 6 [meals served at the same time, total calories 2000 calories consisting of 56% carbohydrate (270 g), 19% protein (90 g), 25% fat (55 g)], and allowed to sleep (lights-off) between 2300 and 0700 h. During the rest of the day, the lights were on, and sleeping and snacking were not allowed. Blood samples were obtained through an indwelling venous cannula. For the PWS, blood was withdrawn for measurement of ghrelin and was collected in EDTA tubes every 30 min between 0900 and 2100 h and then hourly between 2100 and 0900 h without adding aprotinin. Thus, 36 samples were collected for each patient. Samples were stored at 4 C during the collecting period, centrifuged for plasma within 2 h, and stored at −70 C until the time of the assay. For the controls, hourly samples (24 samples) were collected instead of 36 samples. For the statistical analysis, samples from patients 1–5 were included and hourly ghrelin values were compared with those of controls 1–5.

For fasting ghrelin determinations, plasma samples were collected after an overnight fast between 0800 and 1000 h into EDTA tubes. Each sample was then centrifuged, the plasma obtained, and stored at −70 C until the time of assay.

Hormonal assay

Plasma immunoreactive ghrelin was measured in duplicate using a commercial ELISA kit (Phoenix Pharmaceuticals, Belmont, CA). Inter- and intraassay coefficients of variance were less than 10%. The lower and upper limits of detection for this assay were 0.75 and 100 ng/ml.

Statistical analysis

Fasting plasma ghrelin levels were compared among groups using the Kruskal-Wallis test. The correlations between plasma ghrelin concentrations and age, BMI, percentage of the ideal weight for age, and BMI percentile were determined using Spearman’s correlation analysis and Pearson’s correlation analysis. Twenty-four-hour plasma ghrelin levels were compared between groups using repeated-measures ANOVA. Data were shown as mean ± se.P < 0.05 were regarded as indicating statistical significance. There was one missing value (P3, 1130 h) in the 24-h ghrelin study, and it was excluded from analysis. All statistical analyses were performed using SAS (version 8.2, SAS Institute, Cary, NC).

Results

Twenty-four-hour ghrelin study

Plasma ghrelin levels in the PWS group were higher (mean 8.63 ± 0.38 ng/ml), compared with the controls (mean 2.43 ± 0.21 ng/ml) (P < 0.001) throughout the 24-h ghrelin monitoring period (Fig. 1). However, the 24-h ghrelin profile for subject P5 was much higher than all others. When this patient was excluded from the analysis, plasma ghrelin levels in the PWS group were still significantly higher than the controls (mean 5.61 ± 0.50 ng/ml) (P < 0.001). Although there seemed to be a nocturnal increase and a postprandial decrease in ghrelin, the data were inconsistent and showed wide individual variations (Figs. 2 and 3³). The young and lean control with BMI 20 kg/m² (control 5, C5) showed more increased ghrelin levels than other subjects (Fig. 2). Also, it was noted that the PWS subjects with BMI of 24.2 kg/m² (patient 1, P1) and 18.6 kg/m² (P5) showed higher ghrelin levels than PWS subjects with higher BMIs [49.1, 47.6, 35.3 kg/m² (P2, P3, P4)] (Fig. 3). These data suggest that higher BMI is correlated with lower ghrelin levels, even among the PWS patients.

Fig. 1.

Twenty-four-hour mean plasma ghrelin profiles (mean ± se) in PWS (n = 5) and sex- and age-matched controls (n = 5). Hourly ghrelin values were compared with those of controls 1–5. Plasma ghrelin levels in the PWS group were higher (mean 8.63 ± 0.38 ng/ml), compared with the controls (mean 2.43 ± 0.21 ng/ml) (P < 0.001) throughout the 24-h ghrelin monitoring period.

Fig. 2.

Twenty-four-hour plasma ghrelin profiles in five controls. The young and lean subject (C5) with BMI 20 kg/m² showed more increased ghrelin levels than the other subjects.

Fig. 3.

Twenty-four-hour plasma ghrelin profiles in six PWS patients. The young PWS subjects with BMI of 18.6 kg/m² (P5) and 24.2 kg/m² (P1) showed more increased ghrelin levels than PWS subjects with higher BMIs [49.1, 47.6, 35.3, kg/m² (P2, P3, P4)].

The higher ghrelin levels in the lean PWS patients prompted us to investigate the fasting ghrelin levels in sizable PWS groups. We were able to sample another 39 PWS plasma specimens. In children with PWS, fasting ghrelin concentrations (mean 5.88 ± 0.95 ng/ml) were significantly higher than in obese normal controls (P = 0.026). In PWS, inverse correlation was found between plasma ghrelin levels and the following characteristics: age (r = −0.408, P = 0.005), BMI (r = −0.341, P = 0.017), percentage of the ideal weight for age and height (r = −0.382, P = 0.008), and BMI percentile (r = −0.311, P = 0.027) (Fig. 4), as was observed for ghrelin levels in normal (lean plus obese) controls, i.e. age (r = −0.516, P < 0.001), BMI (r = −0.549, P < 0.001), percentage of the ideal weight for age (r = −0.535, P < 0.001), and BMI percentile (r = −0.514, P < 0.001). No significant gender difference was found in fasting ghrelin levels in PWS, obese control, or lean control (P > 0.05). No significant difference was seen in fasting ghrelin levels in PWS, depending on the genotype of the individual (deletion or uniparental disomy).

Fig. 4.

In PWS inverse correlations were found between plasma ghrelin levels and the following: age (r = −0.408, P = 0.005) (A), BMI (r = −0.341, P = 0.017) (B), percentage of the ideal weight for age and height (r = −0.382, P = 0.008) (C), and BMI percentile (r = −0.311, P = 0.027) (D).

Discussion

Our data show a 3- to 4-fold increase in the mean ghrelin levels in PWS over all 24-h periods, compared with lean control or obese control in the first time. Several reports have emphasized increased fasting ghrelin level in PWS, but these reports did not demonstrate that both minimal and maximal ghrelin levels were increased (12–14). It should also be noted that marked individual variations were observed according to individual body composition. Patients 1 and 5 were rather slim children in the PWS group (BMI 24.2 and 18.6 kg/m²), and their peak ghrelin levels and 24-h mean ghrelin levels were markedly higher than those in other PWS subjects with higher BMIs [49.1, 47.6, and 35.3 kg/m² (P2, P3, and P4)]. These data strongly suggest that ghrelin levels are correlated with body composition, even among the PWS groups. Serum ghrelin levels in normal children were previously reported to be inversely correlated with BMI and age (12). Our data confirmed these findings in normal children (P < 0.001), but here we demonstrate for the first that the same trend is present in PWS. Low r value might be due to small patient number.

Based on our findings, we speculate that the overall increase in ghrelin secretion may be related to the oversecretion of ghrelin due to hyperplasia of the ghrelin-secreting cells in the stomach or to the relative insensitivity of the hypothalamus to ghrelin levels, which results in an attenuated negative feedback mechanism that remains to be identified. Thus, we propose that an investigation of the distribution of ghrelin-secreting cells and ghrelin contents in the stomach would prove worthwhile.

The increased concentrations of ghrelin found in children with PWS may lead to the stimulation of appetite through the hypothalamic neuropeptide Y/agouti gene-related peptide signaling pathways. Because the human ghrelin gene on chromosome 3p25–26 is not located within 15q11.2-q12, which causes PWS, it is unclear why this elevation in circulating ghrelin occurs in the syndrome.

It should be stressed that our determination of ghrelin has a limitation. It is known that the acyl form of ghrelin is the functionally active form. Purified ghrelin is a peptide of 28 amino acids, in which the serine 3 residue is n-octanoylated (1). However, the antibody used in ELISA detects both the acyl and des-acyl forms of ghrelin. Thus, a more specific antibody is needed to clarify this inherent problem.

Our 24-h ghrelin monitoring provides some explanation for the inconsistencies of previous findings. The first issue is related to the nocturnal rise and marked periprandial dynamics of ghrelin, typified by preprandial elevations and rapid postprandial declines. There are two 24-h ghrelin monitoring reports in the literature. Cummings et al. (14) determined 24-h plasma ghrelin profiles in 13 obese subjects before and after a 6-month dietary program for weight loss. Plasma ghrelin levels sharply increased before meals and fell shortly afterward. However, Barkan et al. (15) studied ghrelin and GH concentrations over a 24-h period in young healthy men and women and acromegaly patients. They questioned whether the nocturnal rise and the marked periprandial dynamics of ghrelin exist and emphasized that ghrelin secretion showed sexual dimorphism. Although a tendency of a nocturnal rise and an increase and decline of ghrelin around the mealtime were observed, ghrelin levels showed wide fluctuations and individual variations in our study. However, we must comment that our study looks at a much younger population than that used in previous studies for the periprandial dynamics of ghrelin (14, 15). No significant sexual difference in fasting ghrelin levels was observed in PWS patients and normal prepubertal children. It may be due to much younger ages of our subjects.

Previous studies have shown that GH secretion is reduced in PWS children, despite high ghrelin levels, although this is difficult to evaluate due to the concomitant obesity that suppresses GH secretion (16). The low GH levels might be due to impaired hypothalamopituitary secretion of GH rather than artifact related to obesity (17). Recently GH secretagogue receptor(Ghsr)-null mice was reported (18). In contrast to wild-type mice, acute treatment of Ghsr-null mice with ghrelin stimulated neither GH release nor food intake, showing that the GHSR is a biologically relevant ghrelin receptor. Serum IGF-I levels and body weights of mature Ghsr-null mice are modestly reduced, compared with wild-type littermates, which is consistent with ghrelin’s property as an amplifier of GH pulsatility and its speculated role in establishing an IGF-I set point for maintaining anabolic metabolism (18). Thus, we speculate that ghrelin’s regulation of GH is impaired like the Ghsr knockout state in PWS, but ghrelin’s control of appetite is working. But this speculation needs to be elucidated by further investigations.

In conclusion, we found a 3- to 4-fold increase in mean 24-h ghrelin levels in PWS vs. age-, sex-, and BMI-matched controls. In PWS, inverse correlations were found between plasma ghrelin levels and age, BMI, percentage of the ideal weight for age, and height and BMI percentile. Further study will be needed to address the significance of individual ghrelin vs. the production/secretion of ghrelin in the stomach and sensitivity/negative feedback mechanisms of ghrelin in the hypothalamus to elucidate the mechanisms underlying the voracious appetite in PWS.

Acknowledgements

This work was supported by a grant from the Korean Health 21 R&D Project, Ministry of Health and Welfare (01-PJ10-01GN15-0001).

Abbreviations:

• BMI,

  Body mass index;

• F,

  female;

• Ghsr,

  GH secretagogue receptor;

• M,

  male;

• PWS,

  Prader-Willi syndrome.