| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Endocrinology, Diabetes, and Metabolism, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School (J.L.C., J.B., J.H.L., C.S.M.), Boston, Massachusetts 02215; and Department of Home Economics and Ecology, Harokopio University (N.Y.), Athens, Greece
Address all correspondence and requests for reprints to: Christos S. Mantzoros, M.D., Division of Endocrinology, Diabetes, and Metabolism, Beth Israel Deaconess Medical Center, Harvard Medical School, 330 Brookline Avenue, ST816, Boston, Massachusetts 02215. E-mail: cmantzor{at}bidmc.harvard.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
As ghrelin and leptin appear to be regulated in opposite directions by acute and chronic alterations in energy homeostasis, it would be reasonable to hypothesize that one may regulate the other or that a common factor regulates both differentially. However, animal studies evaluating the regulation of ghrelin by leptin have not demonstrated a clear relationship between leptin levels and ghrelin levels and/or expression (13, 14, 15, 16, 17, 18). In addition, the regulation of ghrelin by either physiological or pharmacological levels of leptin in humans remains unknown. Although ghrelin has been reported to increase during diet-induced weight loss (19, 20), the effect on ghrelin levels of acute negative energy balance without a significant change in fat mass, such as a prolonged, complete fast that results in a low leptin state, has not yet been reported in humans. Such studies would have direct relevance in clarifying the postulated role of ghrelin as a meal initiator and elucidating whether fasting per se or changes in leptin levels have a direct effect in altering ghrelin levels independently from the potential confounding effect of major changes in weight or fat mass.
We performed cross-sectional and interventional studies in healthy subjects to further elucidate ghrelin physiology in humans. In 120 healthy young subjects, we measured hormone levels (including leptin) and performed body composition, total energy, and macronutrient analyses to evaluate predictors of serum ghrelin levels. To determine whether fasting or leptin directly regulates ghrelin in humans, we performed interventional studies involving fasting and administration of recombinant methionyl human leptin (r-metHuLeptin) to healthy subjects. Six lean men were studied under 3 separate conditions (fed state, 72-h fasting alone, and 72-h fasting with administration of r-metHuLeptin at physiological replacement doses), with measurement of ghrelin and leptin every 30 min over 24 h on the third day of each condition. In a separate group of lean men, we administered r-metHuLeptin at doses ranging from low physiological to high pharmacological on 3 separate days to determine whether leptin has an acute effect on ghrelin levels and, if so, whether this occurs in a dose-dependent fashion.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
One hundred and twenty postpubertal Greek students (age, 17.7 ± 1.7 yr; range, 14–26) were consecutively enrolled in a study approved by the ethics committee at Harokopio University (Athens, Greece) and the institutional review board of Beth Israel Deaconess Medical Center (BIDMC). Subjects and their parents gave written informed consent to participate in the study. Blood samples were collected from all except 2 subjects after an overnight fast (
10 h), resulting in 118 evaluable subjects. Subjects recorded the type and volume of food or beverage consumed (using standard household measures) over 2 consecutive weekdays and 1 weekend day, as previously described (21). Six subjects did not return their food records. The Nutritionist V Diet Analysis software (FirstDataBank, Inc., San Bruno, CA), as modified for the Greek population, was used to calculate total energy and nutrient intake for 61 female and 53 male subjects (21).
Anthropometric and body composition measurements were performed before breakfast in subjects wearing light clothing without shoes. The same observer measured waist (WC) and hip circumference (to a precision of 0.1 cm) and triceps, biceps, subscapular, and suprailiac skinfolds twice on the right side of the body (to a precision of 0.2 mm) using a Lange skinfold caliper (Cambridge Scientific Instruments, Cambridge, MA) with the average of two measurements reported. Bioelectrical impedance analysis (BIA), using a single frequency bioimpedance analyzer (model 101, RJL Systems, Mt. Clemens, MI), was used to calculate fat-free mass and percent fat mass (%FM) (22). The BIA measurements were previously validated (23) against a Lunar DPX dual energy x-ray absorptiometry (DEXA) densitometer (Lunar Corp., Madison, WI; software version 4.7e), which measures total body and trunk fat mass. Estimates of FM and %FM obtained by BIA and DEXA correlated strongly with each other (FM, r = 0.91; %FM, r = 0.88) and with body mass index (BMI; FM and %FM by BIA, r = 0.94; FM and %FM by DEXA, r = 0.84). Trunk fat (a measure of central adiposity) obtained by DEXA correlated highly with WC (r = 0.90) and waist to hip ratio (WHR; r = 0.54). Thus, in our analysis, we used BMI and FM calculated by BIA as markers of overall adiposity, and WC and WHR as markers of central obesity.
Interventional studies
Both interventional study protocols were approved by the BIDMC institutional review board, and subjects gave written informed consent to participate in the study. Clinical quality r-metHuLeptin was supplied by Amgen, Inc. (Thousand Oaks, CA) and administered under an investigator-initiated investigational new drug application approved by the Food and Drug Administration.
Administration of physiological leptin doses in the fasting state. Six normal weight men (age, 23.5 ± 1.3 yr; BMI, 24.0 ± 0.4 kg/m2) were screened for any medical problems and admitted to the BIDMC General Clinical Research Center (GCRC) under three different conditions: baseline fed state, fasting with placebo administration, and fasting with replacement dose r-metHuLeptin administration designed to achieve physiological serum leptin levels similar to those in the fed state. During each fed or fasting study, subjects were admitted to the GCRC the evening before study d 1 and acclimated to a research bed for 2 d before the frequent sampling day (d 3).
During the baseline fed state, subjects received a standardized isocaloric diet: 20% of calories from breakfast (0800 h), 35% from lunch (1300 h), 35% from dinner (1800 h), 10% from a snack (2200 h). During the fasting studies, subjects received only caffeine- and calorie-free liquids for 3 d and NaCl (500 mg), KCl (40 mEq), and a standard multivitamin with minerals daily. After a snack on the evening before study d 1, they fasted until 1000 h on d 4 (light standardized breakfast) with ad libitum feeding starting at 1300 h. Lights were on from 0700–2300 h and were off from 2300–0700 h, during which time the subjects slept. Subjects were weighed in light clothing without shoes, and body composition was assessed by DEXA using a Hologic 2000 scanner [Hologic, Waltham, MA; coefficient of variation (CV), 1.5% for body fat measurements] at the beginning (d 1) and end (d 4) of each fasting state.
Starting at 0830 h on d 3, blood samples were drawn every 15 min for 24 h through an iv catheter placed in the forearm. Pairs of consecutive samples, drawn 15 min apart, were pooled, and hormone measurements were performed on the pooled samples to decrease variability. During the night, blood samples were drawn outside the subject’s room to avoid disturbing sleep. For the fed state, ghrelin levels were not available for two subjects due to insufficient serum. During the two fasting studies, either placebo or r-metHuLeptin was administered as four sc injections per day at a dose of 0.04 mg/kg·d on the first day and 0.1 mg/kg·d on the second and third days (total daily dose each day divided into four equal doses given every 6 h).
Administration of physiological to pharmacological leptin doses in the fed state. Five lean healthy men (age, 22.2 ± 0.9; BMI, 22.0 ± 0.5 kg/m2), screened for any medical problems, were admitted to the GCRC the night before the first study day and placed on an isocaloric diet (same caloric distribution as described above, with breakfast at 0715 h, lunch at 1400 h, dinner at 1800 h, and snack at 2200 h). At 0800 h on d 1, subjects received a single dose of r-metHuLeptin (0.01 mg/kg), with blood samples for ghrelin and leptin drawn at time zero before the dose and at 1, 2, 3, 4, 5, 6, 8, 10, 12, 18, and 24 h after the dose. At 0800 h on the next day, subjects received a second, higher dose of r-metHuLeptin (0.1 mg/kg), with blood drawn at the same time points. (Two subjects received the 0.01 and 0.1 mg/kg doses during two separate 1-d admissions.) The same subjects were admitted to the GCRC again 1–2 months later for a separate 1-d stay (following the same protocol for blood draws and meals), with r-metHuLeptin administered at a dose of 0.3 mg/kg (30 and 3 times higher than the first and second doses, respectively).
Hormone assays
Hormones were measured using commercially available kits: leptin, RIA (Linco Research, Inc., St. Charles, MO; sensitivity, 0.5 ng/ml; intraassay CV, 8.3%); ghrelin, RIA [Linco Research, Inc.; sensitivity, 100 pg/ml; intraassay CV, 3.3–10.0%; spike and recovery experiments, 91–96% recovery; serial dilutions yielded 99–146% of expected values; 100% specificity for human ghrelin, ghrelin-(14–28), and des-octonylghrelin and no cross-reactivity with human leptin or insulin (Linco package insert 17)]; cortisol, RIA (Diagnostic Systems Laboratories, Webster, TX; sensitivity, 13.8 nmol/liter; intraassay CV, 5.3–8.4%); insulin, RIA (Diagnostic Systems Laboratories; sensitivity, 7.8 pmol/liter; intraassay CV, 8.3%); estradiol, RIA (Diagnostic Products Corporation; sensitivity, 29.4 pmol/liter; intraassay CV, 4.3–7.0%); free testosterone, RIA (Diagnostic Systems Laboratories; sensitivity, 0.52 pmol/liter; intraassay CV, 8.0%); and IGF-I, immunoradiometric assay (Diagnostic Systems Laboratories; sensitivity, 0.27 nmol/liter; intraassay CV, 3.9–7.0%). To minimize variability, hormone levels were measured in one assay for all subjects participating in the cross-sectional study and in one assay for each subject in the interventional studies.
Statistical analysis
SPSS 8.0 software (SPSS, Inc., Chicago, IL) was used for statistical analysis. Demographic, anthropometric, and hormonal parameters are expressed as the mean ± SD for the cross-sectional study, with comparison of means (between males and females) performed using t tests. Pearson’s correlation coefficients were calculated. Using bivariate analysis, we evaluated associations between ghrelin levels and gender, age, anthropometric and body compositions (BMI, WHR, FM, %FM, and sum of skinfolds), and hormonal (leptin, cortisol, estradiol, free testosterone, insulin, and IGF-I) parameters, all expressed as continuous variables. Multivariate linear regression analysis was used to adjust for potential confounders, as shown in the relevant tables. Nonnormally distributed variables were logarithmically transformed to obtain a normal distribution. We used bivariate and two separate multivariate linear regression analysis models to evaluate the correlation between total energy or energy from specific macronutrients (protein, carbohydrates, and fat) with ghrelin levels (21, 24). Using an energy partition model, ghrelin levels regressed on caloric intake from each of the three macronutrients were adjusted for gender, FM, and calories from sources other than the variable of interest, thus providing analysis of the specific nutrient’s effect on ghrelin. To allow an isocaloric evaluation of each macronutrient’s effect on ghrelin controlling for total energy intake, a multivariate nutrient density model was used to evaluate the correlation between ghrelin and caloric intake from each macronutrient adjusted for 1) total energy intake only, or 2) gender, FM, and total energy intake. The size of the cross-sectional study provided more than 80% power to detect clinically significant associations (r 0.22) at the conventional 
= 0.05 level. P < 0.05 (two-tailed) was considered statistically significant for these analyses.
For the interventional studies, the area under the curve (AUC) for hormone measurements (ghrelin, leptin, or cortisol) over the study time period (24 h for the fasting study and 18 h for the fed study) was calculated using the standard trapezoid method. Values are expressed as the mean ± SE. Nonparametric ANOVA was used to evaluate significant differences in AUC results between the fed vs. fasting state with placebo vs. fasting with r-metHuLeptin states for the first interventional study and among the three doses of r-metHuLeptin (0.01, 0.1, and 0.3 mg/kg) for the second interventional study. P < 0.05 was considered statistically significant for ANOVA. For the first interventional study, the percent changes in ghrelin, leptin, and cortisol values for the fed, fasting, and fasting with replacement r-metHuLeptin admissions are calculated based on the 0830 h point for the fed stay. For the two subjects who did not have ghrelin levels measured in the fed stay, the average of the 0830 h ghrelin levels of the other subjects was used for calculation. Wilcoxon signed-ranks test was used to evaluate significant changes in weight or body composition between d 1 and 4 of each admission, with Bonferroni correction used to adjust for multiple comparisons. P < 0.0167 was considered statistically significant, and 0.0167 
P < 0.05 was of borderline significance.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the cross-sectional study of 120 subjects, the 56 men and 62 women were of normal weight (BMI, 22.3 ± 3.6 kg/m2), with a mean ghrelin level of 856.1 ± 433.3 pg/ml (females, 943.0 ± 503.2 pg/ml; males, 759.3 ± 316.6 pg/ml; P = 0.02) and a mean leptin level of 6.8 ± 5.9 ng/ml (females, 9.9 ± 6.2 ng/ml; males, 3.3 ± 2.6 ng/ml; P < 0.001; Table 1
). Females reported lower total energy intake than males (1863 ± 794 vs. 2796 ± 1009 kcal/d; P < 0.001), but there was no difference in relative macronutrient contribution to total energy intake between genders (data not shown).
|
). Although there was only a weak correlation between age and ghrelin levels by bivariate analysis (ß = 0.14; P = 0.13), adjustment for gender indicated an association of borderline significance (ß = 0.17; P = 0.08) between age and ghrelin levels. Further adjustment for markers of adiposity (BMI, WHR, FM, and %FM) resulted in a significant positive correlation between age and ghrelin levels (ß = 0.20–0.22; P = 0.02–0.03), indicating that ghrelin correlates positively with age in individuals of the same gender and level of adiposity.
|
). Similarly, ghrelin was negatively correlated with FM (ß = -0.20; P = 0.03), independent of gender and age, and even more strongly with %FM after sequentially adjusting for gender (ß = -0.30; P = 0.01) and age (ß = -0.34; P = 0.005). Similar results were obtained for the sum of skinfold thickness, another marker of adiposity (data not shown). Finally, ghrelin showed a highly significant negative correlation with WC (ß = -0.37, P < 0.001), a validated marker of central adiposity, which persisted after adjusting for gender, age, BMI, FM, and %FM, indicating that the correlation of ghrelin with central obesity is independent of overall obesity. Similar results were obtained for ghrelin and WHR (Table 2).
Although ghrelin and leptin were not correlated by bivariate analysis (ß = -0.11; P = 0.27), adjusting for gender resulted in a significant negative correlation (ß = -0.44; P < 0.001), which persisted even after accounting for total and central adiposity (Table 2). This indicates that leptin and ghrelin are negatively associated within a given gender independent of overall and central adiposity. In contrast, insulin had only a weak negative association with ghrelin (ß = -0.17; P = 0.07) independent of gender and age that became nonsignificant after adjustment for adiposity. IGF-I correlated negatively with ghrelin (ß = -0.19; P = 0.04), but the association was of borderline significance after controlling for gender (ß = -0.16; P = 0.09) and was not significant after adjusting for age or adiposity. Similar results were seen for free testosterone (ß = -0.20; P = 0.03), with no association after adjustment for gender. There was no significant correlation between ghrelin and either cortisol or estradiol levels.
We then used energy partition and multivariate nutrient density models to evaluate the relationship between total energy and macronutrient energy intake with serum ghrelin levels. With the exception of a borderline association between serum ghrelin and amount of energy derived from protein (ß = -0.18; P = 0.07) by bivariate analysis, which became nonsignificant after adjusting for gender, FM, and caloric intake from carbohydrates and fat, ghrelin was not associated with total energy or macronutrient intake (data not shown).
Interventional studies
Given the negative association between ghrelin and leptin in the cross-sectional study, we evaluated the potential regulation of ghrelin by leptin in an interventional study involving 72-h fasting with and without leptin replacement in healthy lean men. After 3 d of fasting, body weight tended to decrease by approximately 2 kg [fasting, 75.7 ± 2.9 to 73.7 ± 2.9 kg (P = 0.028); fasting and r-metHuLeptin, 77.3 ± 3.3 to 75.0 ± 3.3 kg (P = 0.028)] with a small decrease in fat mass [fasting, 13.0 ± 2.8 to 12.4 ± 2.9 kg (P = 0.07); fasting and r-metHuLeptin, 14.3 ± 2.9 to 12.9 ± 2.9 kg (P = 0.028)] that was of borderline significance. In both fasting studies, the %FM was not significantly different compared with baseline [fasting, 17.0 ± 3.3 to 16.8 ± 3.6 kg (P = 0.50); fasting and r-metHuLeptin, 18.3 ± 3.4 to 18.3 ± 3.6 kg (P = 0.89)]. In the baseline fed state, leptin levels demonstrate the typical diurnal pattern with an increase of approximately 50% during the overnight hours compared with daytime levels (25) (Fig. 1A). After subjects fasted for 3 d, leptin levels were suppressed to approximately 10% of baseline, with complete loss of the diurnal rhythm (i.e. out of proportion to changes in weight and body fat). When r-metHuLeptin was administered to subjects during the 72-h fast, serum leptin levels were fully restored to baseline levels. The resultant leptin levels were higher than those in the fed state, but were within the physiological range for lean men (Table 3A).
|
|
). Fasting for 3 d abolished the meal-related pattern of ghrelin levels, but the diurnal variation remained, with higher levels overnight, followed by a decline between 0200–0400 h. When replacement dose r-metHuLeptin was administered during fasting, the lack of a meal-related pattern in ghrelin levels was also observed, with minimal variation in ghrelin levels during the day and evening before declining from 0200–0500 h. The AUC and 24-h mean levels for ghrelin in the three different states were not significantly different by ANOVA (Table 3A).
Cortisol levels were measured in the same samples to obtain a marker of circadian variability that could serve as a benchmark for comparison with ghrelin levels. In the baseline state, cortisol levels peaked in the early morning hours, with a nadir overnight as expected. Fasting resulted in mild activation of the adrenal axis, as manifested by an increase in 24-h AUC and mean levels, with similar results obtained during the fasting study with r-metHuLeptin administration (Table 3A and Fig. 1C). During both fasting studies, an increase in ghrelin levels during the early morning hours occurred at approximately the same time as the early morning rise in cortisol levels.
Given the lack of a significant effect of a low leptin state or replacement dose r-metHuLeptin on ghrelin levels, we then evaluated whether high pharmacological doses of r-metHuLeptin would acutely regulate ghrelin levels. In response to a single dose of r-metHuLeptin administration, leptin levels increased from a baseline of 2.32 ± 0.22 ng/ml to a peak of 7.19 ± 0.66 ng/ml at 4 h after 0.01 mg/kg r-metHuLeptin (low physiological), from 2.11 ± 0.22 ng/ml to a peak of 68.45 ± 3.88 ng/ml at 4 h after 0.1 mg/kg r-metHuLeptin (pharmacological), and from 2.97 ± 0.30 ng/ml to a peak of 147.54 ± 12.71 ng/ml at 3 h after 0.3 mg/kg r-metHuLeptin (high pharmacological). The AUC of leptin levels increased significantly with increasing dose of r-metHuLeptin (Table 3B), but over the time course of high leptin levels induced by pharmacological doses of r-metHuLeptin, the AUC of serum ghrelin levels did not change significantly (Table 3B). Small changes in ghrelin levels corresponded to the timing of meals, but ghrelin levels overall were clearly not affected by the marked change in leptin levels (Fig. 2).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Ghrelin has generated significant scientific interest over the past few years, as available observational and experimental evidence indicates that it is an important peripheral orexigen that increases appetite and food intake. The decrease in this stomach-derived hormone immediately after a meal (8, 11, 26) suggests that ghrelin may be involved in the short-term control of energy intake and is regulated by circulating (i.e. glucose and/or insulin) and/or local (i.e. stomach distention) factors. However, although one study found that oral and iv glucose decrease ghrelin levels (8), other studies have reported no regulation of ghrelin by hypoglycemia (27), hyperglycemia (27, 28), or the combination of sc insulin and iv glucose designed to simulate a meal (26). Although supraphysiological insulin levels during clamp studies can suppress ghrelin levels, physiological insulin levels do not appear to regulate ghrelin (27, 28, 29). Finally, stomach distention with water alone had no effect on ghrelin levels in one study (8), but ghrelin levels did decrease after consumption of noncaloric fiber to a similar degree as after a standard meal (30). Thus, although available data suggest regulation of ghrelin by a putative meal or nutrient-related factor, the exact mechanism remains unknown. It is possible that other factors, such as stomach pH, may explain the discrepant findings and require further study.
If ghrelin secretion is altered in response to meals, one might expect that lack of meals would conversely result in a loss of this meal-related pattern. Consistent with this idea, we report for the first time that ghrelin levels show no significant fluctuations related to meals during the third day of complete fasting. We then hypothesized that fasting for 3 d (resulting in increased appetite) may be associated with elevated ghrelin levels compared with the fed state. It has been suggested that ghrelin may have a role in meal initiation (11). In rodents, fasting for 2 d increased pulsatile ghrelin secretion (31) and ghrelin gene expression in the stomach (13, 14), but ghrelin peptide content was paradoxically decreased (14). In humans, although a decrease in ghrelin levels overnight to a nadir at 0600 h has been noted (11), the effect of a more prolonged fast on ghrelin levels has not been well characterized. Surprisingly, we found that prolonged fasting for 72 h, which resulted in a very low leptin state, did not significantly increase ghrelin levels compared with those during the baseline fed state in the same subjects, raising the possibility that the observed meal-related pattern of ghrelin secretion may be due primarily to postprandial decreases rather than preprandial increases in ghrelin levels. Alternatively, it is also possible that although ghrelin levels may increase in response to short-term fasting, tachyphylaxis occurs after more prolonged fasting. This needs to be explored further in future studies.
More long-term hypocaloric diets have also provided relevant insights into the role of ghrelin in appetite regulation. Although ghrelin has been reported to increase during diet-induced weight loss (19, 20), a low fat diet, causing moderate weight loss, was not associated with increased ghrelin levels (32). In obese subjects who lost weight by undergoing gastric bypass surgery, 24-h ghrelin levels were significantly lower compared with those in obese controls dieting for weight loss (20). Although of shorter duration, the paradigm in our study involved a complete fast, and the lack of a change in 24-h ghrelin levels during fasting suggests that increasing ghrelin levels are not necessary for appetite regulation and/or meal initiation and indicate that ghrelin is not regulated by a low leptin state independently of significant changes in body weight and/or FM.
We then addressed whether total daily caloric intake and specific macronutrients (carbohydrate, protein, and fat) in the diet affect ghrelin levels. In a recent study, a low fat, high carbohydrate diet that induced moderate weight loss did not change ghrelin levels (32), but the effect of varying macronutrient intake in a weight-maintaining setting has not yet been studied. We found no association between normal variations in macronutrient or total energy intake (over 3 d) and ghrelin levels, suggesting either that only the complete absence of caloric intake or the total amount of energy stored in adipose tissue or energy per meal may be more important in ghrelin regulation than dietary composition. Further interventional studies are needed to clarify this issue as well as whether micronutrients (including vitamins) or fatty acids affect ghrelin levels.
An orexigenic molecule that is causally related to the development of obesity would be expected to be higher in obese subjects. However, the converse (lower ghrelin levels in obesity) has been demonstrated in several studies (7, 8, 33), indicating down-regulation of ghrelin and, thus, a physiological adaptation to the positive energy balance associated with obesity (7). Similarly, ghrelin levels are higher in states characterized by chronic negative energy balance, such as anorexia nervosa (8, 34, 35) and bulimia nervosa (36). In contrast, leptin levels correlate positively with adiposity and thus reflect long-term energy stores (6, 10). If ghrelin is important in short-term energy homeostasis and antagonizes leptin action in the hypothalamus, it would be of interest to study whether ghrelin and leptin may interact, as the systems for governing energy intake over the short-term and body weight over the long-term are probably closely linked to tightly defend body weight. Our observational study revealed that ghrelin is negatively associated with leptin (after controlling for gender), and that this association is independent of adiposity. Our observations are consistent with two prior studies (7, 35), but not with a third study that reported no significant association in obese children, although multivariate adjustments were not performed (37).
Although ghrelin has been demonstrated to have an inverse association with both adiposity and leptin levels, whether regulation of ghrelin by leptin in animal models and humans is direct has been less clear. Studies in mice have shown both decreased (13) and increased (14) ghrelin gene expression in the stomach after leptin administration. Both transgenic "skinny" mice overexpressing leptin (16) and central adenoviral-mediated expression of leptin in rats resulting in weight loss and low peripheral leptin levels have increased ghrelin levels (15, 38), but peripheral leptin infusion, causing only moderate weight loss, prevented the rise in ghrelin seen in pair-fed, calorie-restricted rats (17). Obese ob/ob mice with leptin deficiency and db/db mice with leptin resistance and high leptin levels have low ghrelin levels (16), but leptin treatment of ob/ob mice does not change ghrelin levels (16). Similarly in humans, ghrelin levels in genetic forms of obesity (due to mutations in the leptin, leptin receptor, or MC4 receptor genes) are comparable to those in obese controls, with the exception of Prader-Willi children who have several-fold higher ghrelin levels (39, 40). In our study we found that neither fasting-induced low leptin levels nor restoring leptin to fed state levels alter ghrelin levels, and that exogenous administration of r-metHuLeptin to achieve high physiological or high pharmacological leptin levels had no direct effect on ghrelin levels. Taken together, these data suggest that ghrelin may track body weight changes more closely than peripheral leptin levels per se and that ghrelin may sense body fat or energy stores through mechanisms independent of leptin. The lack of regulation of ghrelin by leptin levels is consistent with the idea that ghrelin may have evolved before the development of adipocytes as a gut-derived hormone that could sense adiposity directly through fat stored in the gastrointestinal system (18), making ghrelin one of the few molecules important in energy homeostasis that is not downstream of leptin.
We also found a weak negative association between ghrelin and fasting insulin levels, a surrogate marker of insulin resistance, but this correlation did not persist after adjustment for adiposity. Previous studies have found a negative correlation between ghrelin and insulin resistance (37, 41, 42), and treatment with the insulin sensitizer metformin significantly increases ghrelin levels (41), but whether this is independent from overall vs. central adiposity has not yet been investigated. We found that independent of overall adiposity, ghrelin has a strong negative association with central adiposity, which is known to correlate with insulin resistance. Taken together, these findings suggest that the relationship between ghrelin and insulin may be explained by the link between ghrelin and central adiposity.
Finally, the hormonal regulation of ghrelin remains incompletely understood. Although ghrelin is a potent GH secretagogue (43), ghrelin levels do not correlate with IGF-I or GH levels (44, 45). The effect of GH therapy on ghrelin levels in GH-deficient adults remains controversial (44, 46), although octreotide treatment in acromegaly has a suppressive effect on ghrelin levels (45, 47). We found that ghrelin and IGF-I are not associated after adjusting for gender. The interaction between ghrelin, GH, and IGF-I probably involves a complex system that is regulated by stimulatory and inhibitory factors in the hypothalamus and influenced by effects of gender, age, and adiposity. A negative correlation between ghrelin and androstenedione levels has been reported in women with polycystic ovarian syndrome and obese women, suggesting an interaction between ghrelin and sex steroid synthesis or action (42), but we observed no association between ghrelin and free testosterone or estradiol.
In summary, these cross-sectional and interventional studies of healthy subjects provide novel insights into the physiology of ghrelin. Women have higher ghrelin levels than men, which may be explained by differences in total and/or central adiposity. Ghrelin levels correlate negatively with central adiposity, even after adjustment for total adiposity, and the relationship between ghrelin and insulin levels reflects primarily an effect of central adiposity on insulin resistance. Although ghrelin levels are affected by adiposity, total energy and specific macronutrient intake are not associated with ghrelin levels. In addition, complete fasting for 3 d that does not result in major changes in body fat mass does not significantly increase 24-h ghrelin levels, as might be expected if ghrelin acts as a signal to increase appetite, but does result in loss of the meal-related pattern of secretion. Finally, although ghrelin and leptin levels are inversely correlated, physiological and pharmacological leptin levels do not regulate ghrelin over the course of hours to a few days. Future studies to fully elucidate ghrelin physiology in humans are clearly needed.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: AUC, Area under the curve; BIA, bioelectrical impedance analysis; BMI, body mass index; CV, coefficient of variation; DEXA, dual energy x-ray absorptiometry; %FM, percent fat mass; r-metHuLeptin, recombinant methionyl human leptin; WC, waist circumference; WHR, waist to hip ratio.
Received August 12, 2003.
Accepted October 7, 2003.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  E. Tham, J. Liu, S. Innis, D. Thompson, B. D. Gaylinn, R. Bogarin, A. Haim, M. O. Thorner, and J.-P. Chanoine Acylated ghrelin concentrations are markedly decreased during pregnancy in mothers with and without gestational diabetes: relationship with cholinesterase Am J Physiol Endocrinol Metab, May 1, 2009; 296(5): E1093 - E1100. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Al, L. Ng, P. Tyrrell, J. Bargman, T. Bradley, and E. Silverman Adipokines as novel biomarkers in paediatric systemic lupus erythematosus Rheumatology, May 1, 2009; 48(5): 497 - 501. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Liu, C. E. Prudom, R. Nass, S. S. Pezzoli, M. C. Oliveri, M. L. Johnson, P. Veldhuis, D. A. Gordon, A. D. Howard, D. R. Witcher, et al. Novel Ghrelin Assays Provide Evidence for Independent Regulation of Ghrelin Acylation and Secretion in Healthy Young Men J. Clin. Endocrinol. Metab., May 1, 2008; 93(5): 1980 - 1987. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Hwang, J. L. Chan, G. Ntali, D. Malkova, and C. S. Mantzoros Leptin Does Not Directly Regulate the Pancreatic Hormones Amylin and Pancreatic Polypeptide: Interventional studies in humans Diabetes Care, May 1, 2008; 31(5): 945 - 951. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Pacifico, C. Anania, J. F Osborn, E. Ferrara, E. Schiavo, M. Bonamico, and C. Chiesa Long-term effects of Helicobacter pylori eradication on circulating ghrelin and leptin concentrations and body composition in prepubertal children Eur. J. Endocrinol., March 1, 2008; 158(3): 323 - 332. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Zhao, I. Sakata, Y. Okubo, K. Koike, K. Kangawa, and T. Sakai Gastric leptin, but not estrogen and somatostatin, contributes to the elevation of ghrelin mRNA expression level in fasted rats J. Endocrinol., March 1, 2008; 196(3): 529 - 538. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Marzullo, A. Caumo, G. Savia, B. Verti, G. E. Walker, S. Maestrini, A. Tagliaferri, A. M. Di Blasio, and A. Liuzzi Predictors of Postabsorptive Ghrelin Secretion after Intake of Different Macronutrients J. Clin. Endocrinol. Metab., October 1, 2006; 91(10): 4124 - 4130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Bellone, R. Baldelli, G. Radetti, A. Rapa, D. Vivenza, A. Petri, S. Savastio, M. Zaffaroni, F. Broglio, E. Ghigo, et al. Ghrelin Secretion in Preterm Neonates Progressively Increases and Is Refractory to the Inhibitory Effect of Food Intake J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1929 - 1933. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. A. Tritos and E. G. Kokkotou The Physiology and Potential Clinical Applications of Ghrelin, a Novel Peptide Hormone Mayo Clin. Proc., May 1, 2006; 81(5): 653 - 660. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Bouros, A. Tzouvelekis, S. Anevlavis, M. Doris, S. Tryfon, M. Froudarakis, V. Zournatzi, and A. Kukuvitis Smoking acutely increases plasma ghrelin concentrations. Clin. Chem., April 1, 2006; 52(4): 777 - 778. [Full Text] [PDF]
|
|
|
|
|
|
M. Misra, K. K. Miller, K. Kuo, K. Griffin, V. Stewart, E. Hunter, D. B. Herzog, and A. Klibanski Secretory dynamics of ghrelin in adolescent girls with anorexia nervosa and healthy adolescents Am J Physiol Endocrinol Metab, August 1, 2005; 289(2): E347 - E356. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Chiesa, J. F. Osborn, L. Pacifico, G. Tellan, P. M. Strappini, R. Fazio, and G. Delogu Circulating Ghrelin in Patients Undergoing Elective Cholecystectomy Clin. Chem., July 1, 2005; 51(7): 1258 - 1261. [Full Text] [PDF]
|
|
|
|
|
|
J. Calissendorff, O. Danielsson, K. Brismar, and S. Rojdmark Inhibitory effect of alcohol on ghrelin secretion in normal man Eur. J. Endocrinol., May 1, 2005; 152(5): 743 - 747. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Avram, C. A. Jaffe, K. V. Symons, and A. L. Barkan Endogenous Circulating Ghrelin Does Not Mediate Growth Hormone Rhythmicity or Response to Fasting J. Clin. Endocrinol. Metab., May 1, 2005; 90(5): 2982 - 2987. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kanumakala, R. Greaves, C. C. Pedreira, S. Donath, G. L. Warne, M. R. Zacharin, and M. Harris Fasting Ghrelin Levels Are Not Elevated in Children with Hypothalamic Obesity J. Clin. Endocrinol. Metab., May 1, 2005; 90(5): 2691 - 2695. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Erdmann, F. Lippl, S. Wagenpfeil, and V. Schusdziarra Differential Association of Basal and Postprandial Plasma Ghrelin With Leptin, Insulin, and Type 2 Diabetes Diabetes, May 1, 2005; 54(5): 1371 - 1378. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kojima and K. Kangawa Ghrelin: Structure and Function Physiol Rev, April 1, 2005; 85(2): 495 - 522. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Espelund, T. K. Hansen, K. Hojlund, H. Beck-Nielsen, J. T. Clausen, B. S. Hansen, H. Orskov, J. O. L. Jorgensen, and J. Frystyk Fasting Unmasks a Strong Inverse Association between Ghrelin and Cortisol in Serum: Studies in Obese and Normal-Weight Subjects J. Clin. Endocrinol. Metab., February 1, 2005; 90(2): 741 - 746. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. K Heilbronn, S. R Smith, C. K Martin, S. D Anton, and E. Ravussin Alternate-day fasting in nonobese subjects: effects on body weight, body composition, and energy metabolism Am. J. Clinical Nutrition, January 1, 2005; 81(1): 69 - 73. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. R. Vulliemoz, E. Xiao, L. Xia-Zhang, M. Germond, J. Rivier, and M. Ferin Decrease in Luteinizing Hormone Pulse Frequency during a Five-Hour Peripheral Ghrelin Infusion in the Ovariectomized Rhesus Monkey J. Clin. Endocrinol. Metab., November 1, 2004; 89(11): 5718 - 5723. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Arosio, C. L. Ronchi, P. Beck-Peccoz, C. Gebbia, C. Giavoli, V. Cappiello, D. Conte, and M. Peracchi Effects of Modified Sham Feeding on Ghrelin Levels in Healthy Human Subjects J. Clin. Endocrinol. Metab., October 1, 2004; 89(10): 5101 - 5104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Zimmermann-Belsing and U. Feldt-Rasmussen Obesity: The New Worldwide Epidemic Threat to General Health and Our Complete Lack of Effective Treatment Endocrinology, April 1, 2004; 145(4): 1501 - 1502. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
