| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Paediatrics (S.K.), Royal Alexandra Hospital, Brighton BN1 3JN, United Kingdom; Department of Complex Biochemistry (R.G.), Department of Endocrinology and Diabetes (C.C.P., G.L.W., M.R.Z.), and Clinical Epidemiology and Biostatistics Unit (S.D.), Murdoch Children’s Research Institute, The Royal Children’s Hospital, Melbourne 3052, Australia; and Department of Endocrinology (M.H.), Mater Children’s Hospital, Brisbane 4101, Australia
Address all correspondence and requests for reprints to: Dr. Mark Harris, Mater Children’s Hospital, Raymond Terrace, South Brisbane, 4101, Queensland, Australia. E-mail: mark.harris{at}mater.org.au.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Fasting total ghrelin levels were compared in three groups: normal-weight controls (n = 16), obese controls (n = 16), and patients with hypothalamic obesity (n = 16). Obese children had lower fasting total ghrelin levels than normal controls, but there was no difference between the fasting total ghrelin level in obese controls and children with hypothalamic obesity (P = 0.88). These data suggest that it is unlikely that an elevation in fasting total ghrelin is responsible for the obesity that occurs after hypothalamic damage. Therapeutic interventions aimed at reducing fasting total ghrelin may prove ineffective in controlling weight gain in this group.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Prader-Willi syndrome (PWS) is one of the most common genetic obesity syndromes (5). Hypothalamic dysfunction is thought to underlie many of the features of PWS, including the increased appetite and severe obesity typically described in these patients (6). Fasting total ghrelin concentrations in adults and children with PWS are significantly elevated (3- to 4-fold) when compared with obese controls (7, 8, 9). It has been postulated that the elevated fasting total ghrelin contributes to the insatiable appetite and obesity found in patients with PWS. More recently, elevated fasting total ghrelin levels have also been described in patients with pituitary stalk interruption (10).
Given that hypothalamic dysfunction is central to many of the clinical features of PWS and increased weight gain often occurs after hypothalamic damage during childhood, we postulated that fasting total ghrelin levels may be elevated in hypothalamic obesity. The primary aim of this study was to determine whether fasting total ghrelin is increased in children with hypothalamic obesity when compared with body mass index (BMI)-matched obese controls.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Obesity was defined as a BMI greater than the International Obesity Task Force-proposed age- and sex-specific cut-off points, corresponding to a BMI above 30 kg/m2 for an adult (11). Hypothalamic obesity was defined as obesity developing after the diagnosis of a hypothalamic lesion or damage to the hypothalamus (secondary to surgery or radiotherapy) and associated with objective evidence of at least one pituitary hormone deficiency. Normal obese controls were defined as obese children without any previous or current medical problems and not receiving any medications.
Nonobese controls were defined as children with a BMI less than the International Obesity Task Force-proposed age- and sex-specific cut-off points for overweight, corresponding to a BMI of less than 25 kg/m2 as an adult (11) and without any medical problems or receiving any medications.
The study patients and controls were recruited from The Royal Children’s Hospital, Melbourne, Australia. Children with hypothalamic obesity were identified using the patient database of the Endocrinology Department and the hospital medical records. The final diagnosis, presence of pituitary hormone deficiencies, and treatments were recorded for all patients with hypothalamic obesity (Table 1
). Hypothalamic obese children with ACTH deficiency were receiving a mean hydrocortisone replacement dose of 8.1 mg/m2·d. GH was commenced in all hypothalamic obese subjects with proven GH deficiency. The dose was titrated to achieve normal linear growth. GH therapy was ceased when girls reached a bone age of 13.5 yr and boys reached a bone age of 15.5 yr. Obese controls were recruited through hospital outpatient clinics (Endocrine, Weight Management, Dietician and Adolescent clinics). Nonobese controls were recruited from the diabetes clinic (siblings of children with type 1 diabetes). The children were age and sex matched across all three groups and clinically evaluated by a single researcher (S.K.). None of the children with hypothalamic obesity or the obese controls had diabetes mellitus.
|
Measurements
Weight and height were measured using electronic self-zeroing digital sitting scales and a wall-mounted Harpenden’s stadiometer, respectively. Waist (midway between the 10th rib and the iliac crest) and hip (greater femoral trochanter) circumference were measured using a nonstretchable tape measure in a standing position (12).
Fasting morning blood samples were centrifuged, and the samples were separated and stored as follows: the glucose sample was analyzed on the day of collection; the insulin sample was stored at –20 C and analyzed within 1 wk of collection; and the leptin and total ghrelin samples were stored at –70 C until analysis.
Plasma glucose was measured by the glucose oxidase method on the Vitros 250 Chemistry System (Ortho Clinical Diagnostics, Rochester, NY). This assay demonstrates a between-run coefficient of variation (CV) of 2.8 and 1.3% at 5.2 and 15.9 mmol/liter, respectively, with the manufacturer’s detection limit quoted as 1.0 mmol/liter.
Serum insulin was measured by a noncompetitive chemiluminescence assay on the Immulite 2000 (Diagnostic Products Corp., Los Angeles, CA). This assay demonstrates a between-run CV of 6.5 and 5.0% at 11.2 and 44.3 pmol/liter, respectively, with the manufacturer’s detection limit quoted as 2 pmol/liter. When the measured level was less than 2 pmol/liter, it was assumed for calculation purposes as 2 pmol/liter. Plasma leptin levels were analyzed using a commercially available human leptin RIA kit (Linco Research, Inc., St. Charles, MO). This assay demonstrates a within-run CV of 8.2 and 5.8% at 2.8 and 19.2 ng/ml, respectively, with the detection limit determined by the lowest calibrator at 0.5 ng/ml. Total plasma ghrelin levels were analyzed using a commercially available RIA kit (Linco Research, Inc.). This assay demonstrated a within-run CV of 6.9% at 2192 pg/ml with the detection limit determined by the lowest calibrator at 100 pg/ml. The homeostatic model assessment (HOMA) of insulin resistance was derived from the following formula: fasting insulin (microunits per milliliter) x fasting glucose (millimoles per liter)/22.5 (13). Fasting insulin values were converted from picomoles per liter to microunits per milliliter by multiplying by 0.1394.
Statistical analysis
Statistical analysis was performed using STATA statistics software, version 8 (StataCorp, College Station, TX). Data are presented as medians and interquartile ranges (IQR). Age- and sex-adjusted Z scores based on 1990 British growth data (11) were calculated for BMI. Nonparametric statistical methods were applied for the intergroup comparisons (Kruskall-Wallis or Mann-Whitney U test), because of the relatively small sample size and skewed distribution of the data. P values less than 0.05 were taken as statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Sixteen patients with hypothalamic obesity agreed to participate in the study (Table 1
). The mean age at diagnosis and on entering the study was 5.6 and 12.3 yr, respectively. All of the children were deficient in at least one pituitary hormone, either as a result of the underlying pathology or as a consequence of the treatment received. Ten of the children had undergone surgery alone, one patient had received only radiotherapy, and the rest had been treated with a combination of surgery, radiotherapy, and chemotherapy. The most common lesion was a craniopharyngioma (9 of 16 patients). Other pathologies included hypothalamic astrocytoma (two patients), hypothalamic hamartoma (two patients), hypothalamic glioma (one patient), optic glioma (one patient), and medulloblastoma (one patient).
Clinical comparison of the hypothalamic obese children with the obese and normal-weight controls
The patients with hypothalamic obesity were matched with normal-weight and obese controls in terms of male to female ratio and age (Table 2). The two obese groups were also matched in terms of BMI and waist to hip ratio. Children with hypothalamic obesity were somewhat shorter than the obese controls.
|
There were significant intergroup differences for fasting insulin (P < 0.001), HOMA (P < 0.001), leptin (P < 0.001), fasting total ghrelin (P = 0.01), and fasting glucose (P = 0.01) (Table 3). There was no difference in leptin (P = 0.91), ghrelin (P = 0.88), insulin (P = 0.78), or HOMA (P = 0.96) between children with hypothalamic obesity and obese controls. Children with hypothalamic obesity had lower fasting glucose than obese controls (P = 0.005).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Insulin appears to negatively regulate total ghrelin (14, 15). Fasting insulin was similarly elevated in the two obese groups in the current study and would be consistent with insulin reducing fasting total ghrelin in children with hypothalamic obesity. The two obese groups also had comparable leptin levels; however, leptin by itself is unlikely to have a significant inhibitory effect on ghrelin levels because patients with leptin deficiency or insensitivity have low ghrelin levels (7, 8). Moreover, administration of leptin to healthy fasting subjects does not alter ghrelin levels (16).
The discrepancy between fasting total ghrelin levels in children with hypothalamic obesity and subjects with PWS raises some interesting questions. By definition, children with hypothalamic obesity have suffered some form of hypothalamic insult; however, the current findings indicate that hypothalamic damage by itself does not necessarily result in elevated fasting total ghrelin levels. Perhaps the difference in insulin secretion, given its inhibitory effect on ghrelin, contributes to the divergence between fasting ghrelin levels in PWS subjects and patients with hypothalamic obesity. Hypoinsulinemia has been reported in PWS patients (17), whereas increased insulin levels are often a feature of hypothalamic obesity (18). It is also possible that hyperghrelinemia in PWS patients could be secondary to a processing defect (19).
Multiple mechanisms are likely to influence the increased weight gain associated with hypothalamic damage. This study was not designed to assess the relative contribution of hyperphagia, decreased energy expenditure, or pituitary hormone deficiencies in this condition. Because fasting total ghrelin levels were suppressed, it seems likely that mechanisms other than ghrelin-induced hyperphagia are responsible for the increased weight gain in these patients. Indeed, hyperphagia was not a consistent symptom in our hypothalamic obesity group, in keeping with a previous report (20). None of the patients with ACTH deficiency had clinical features of hypercortisolism consistent with the average hydrocortisone dose (8.1 mg/m2·d). GH insufficiency, however, may have played a role in some GH-deficient subjects. Two patients with GH deficiency who had attained their final adult height were not receiving GH at the time of the study.
Only fasting ghrelin levels were measured in this study; therefore, a defect in postprandial suppression of ghrelin has not been excluded in patients with hypothalamic obesity. There is evidence that obese insulin-resistant subjects have an attenuated postprandial suppression of ghrelin, and it has been suggested that this blunted suppression of ghrelin may inhibit satiety (21).
It is uncertain whether chronically elevated total ghrelin levels cause obesity in humans. Certainly injections of ghrelin in human subjects increase appetite (22), and weight loss after gastric bypass surgery may be associated with reductions in total ghrelin (23). Patients with anorexia nervosa, however, have elevated total ghrelin levels (24), suggesting that these patients are resistant to the orexigenic actions of ghrelin. In addition somatostatin-induced reductions in total ghrelin were not associated with reduced food intake in patients with PWS (25). Due to the pleiotropic actions of somatostatin, however, it is possible that this treatment may have caused a concomitant decrease in an anorexigenic factor.
The total ghrelin measurement may not accurately reflect a specific biological action of ghrelin. Ghrelin circulates in an acylated and unacylated form, with the unacylated form being present in a 2.5-fold higher concentration than the acylated form (26). Until recently, it was felt that acylation at serine 3 was essential for the biological activity of ghrelin (27). However, a recent paper reported that unacylated ghrelin is able to antagonize the metabolic but not the neuroendocrine response elicited by acylated ghrelin (28). Although there is little published data on the relative amounts of acylated and unacylated ghrelin in patients with different forms of obesity, nonhypothalamic obese subjects were reported to have an increased proportion of acylated ghrelin when compared with normal-weight controls (29).
Most total ghrelin levels previously reported in the literature are lower than the values obtained in this study. This finding may result from the use of differing commercially available RIAs. Total plasma ghrelin levels in the current study were analyzed using a RIA kit obtained from Linco Research, Inc., whereas a number of previous studies have used an RIA kit provided by Phoenix Pharmaceuticals (Belmont, CA) (8, 10). A recent comparison of these two RIAs demonstrated that although the r value for the two assays was 0.976, the values obtained for the Linco assay were up to10-fold higher than the Phoenix assay (30). Exchanging the calibrators provided with each RIA kit significantly reduced the difference between the values obtained with the two assays. It is important therefore to be aware of this factor when reviewing published data.
A type II error has to be considered when interpreting any negative result. The current study had an 80% power to detect a 22% difference (i.e. 314 pg/ml) between the two obese groups with a significance level of 0.05 (two-tailed). Given that total ghrelin levels are 3- to 4-fold higher in PWS patients when compared with BMI-matched controls (7, 8, 9), it was felt that the difference in total ghrelin levels between the two obese groups (1345 vs. 1399 pg/ml) was unlikely to be statistically or biologically significant.
BMI was chosen to match the two obese groups in this study. It is possible that the use of BMI may have underestimated the degree of adiposity in the hypothalamic group, given that many of these subjects were either GH deficient or were receiving glucocorticoids. The similar leptin levels and waist to hip ratios in these two groups, however, suggest that both fat mass and fat distribution were not significantly different.
The results of this study suggest that therapeutic interventions aimed at reducing total ghrelin are unlikely to prove effective in controlling weight gain after hypothalamic damage. Body weight control is the result of complex, overlapping systems, many of which involve the hypothalamus and protect against weight loss. The added challenge in treating hypothalamic obesity is that a reduction in plasma ghrelin may not guarantee weight loss, because hypothalamic damage could alter an individual’s response to lowered ghrelin levels.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online March 15, 2005
Abbreviations: BMI, Body mass index; CV, coefficient of variation; HOMA, homeostatic model assessment; IQR, interquartile ranges; PWS, Prader-Willi syndrome.
Received November 4, 2004.
Accepted March 4, 2005.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
