Ghrelin and the Hyposomatotropism of Obesity - Wiley Online Library

Ghrelin and the Hyposomatotropism of Obesity Jan H.N. Lindeman,*§ Hanno Pijl,† Franc¸ois M.H. van Dielen,¶ Eef G.W.M. Lentjes,‡ Cees van Leuven,§ and Teake Kooistra§

Abstract LINDEMAN, JAN H.N., HANNO PIJL, FRANC ¸ OIS M.H. VAN DIELEN, EEF G.W.M. LENTJES, CEES VAN LEUVEN, AND TEAKE KOOISTRA. Ghrelin and the hyposomatotropism of obesity. Obes Res. 2002;10: 1161-1166. Objective: Human obesity is characterized by growth hormone (GH) deficiency, which appears primarily related to a central pattern of obesity and is reverted on weight loss. As yet, the metabolic basis of the GH deficiency remains to be elucidated. The recently discovered endogenous ligand for the GH secretagogue receptor, ghrelin, stimulates GH secretion when administered to rodents or healthy humans. It may thus be hypothesized that low ghrelin levels underlie the hyposomatropism in obesity. Research Methods and Procedures: We have tested this hypothesis in individuals with widely varying body mass and fat distribution and evaluated whether the improved GH concentrations on weight loss are associated with enhanced ghrelin levels. Results: Both plasma GH and ghrelin levels were reciprocally related with body mass index (r ⫽ ⫺0.67, p ⬍ 0.001). However, whereas 24-hour GH secretion was negatively related to the visceral fat area (r ⫽ ⫺0.72, p ⬍ 0.01), ghrelin levels showed a positive relationship with the visceral fat area (r ⫽ 0.49, p ⬍ 0.02). Weight loss resulted in increased GH secretion (median 24-hour GH area under the curve: 1983 vs. 4024 mU/day before and after weight loss, respectively; p ⬍ 0.01) but did not affect ghrelin levels. No relationship could be found between GH and ghrelin plasma levels in obese subjects when comparing diurnal concentration profiles. Discussion: We showed that plasma ghrelin and GH levels

Received for review March 11, 2002. Accepted for publication in final form August 19, 2002. *Department of Vascular Surgery, †Department of General Internal Medicine, and ‡Department of Clinical Chemistry, Leiden University Medical Center, Leiden, The Netherlands; §Gaubius Laboratory, TNO Prevention and Health, Leiden, The Netherlands; and ¶Department of Surgery, University Hospital Maastricht, Maastricht, The Netherlands. Address correspondence to J.H.N. Lindeman, M.D., Ph.D., Dept. Vascular Surgery, Leiden University Medical Center, P.O. Box 9600, 2300 RC, Leiden, The Netherlands. E-mail: [email protected] Copyright © 2002 NAASO

are both reciprocally related with body mass index, but no causative relationship could be demonstrated between low ghrelin levels and the hyposomatropism in human obesity. Key words: ghrelin, growth hormone, body-fat distribution, weight loss

Introduction Obesity is associated with numerous metabolic abnormalities such as insulin resistance, type 2 diabetes, dyslipidemia, and elevated levels of the plasminogen activator inhibitor type 1 (PAI-1). These metabolic disturbances contribute to increased cardiovascular risk in obesity, but despite a vast amount of research, the molecular mechanisms underlying these disturbances remain to be defined. Human obesity is characterized by growth hormone (GH) deficiency. The hyposomatropism of obesity and most of the obesity-associated metabolic disturbances seem primarily associated with a central (abdominal) pattern of fat distribution (1,2). GH is the primary regulator of body growth, but it is becoming increasingly clear that GH exerts physiological effects that extend far beyond the stimulation of linear growth (3). Children and adults with GH deficiency demonstrate a cluster of cardiovascular risk factors, including central adiposity characterized by increased visceral fat (4), insulin resistance (5), dyslipidemia (6), and increased PAI-1 (7). Replacement of GH attenuates most of these abnormalities (8), and it has been hypothesized that the GH deficiency underlies (part of) the metabolic abnormalities in obesity. In view of its important action on regulation of energy homeostasis and the potential role of GH in pathological processes associated with obesity, insight into factors that regulate its synthesis and functioning is relevant for understanding the pathogenesis of human obesity and its associated metabolic disturbances. The recent discovery of the peptide hormone ghrelin (9), an endogenous ligand of the GH secretagogue receptor, has provided some important new insights into nutritional homeostasis and suggests an important interplay between ghrelin on one hand and GH on the other hand. Ghrelin is produced by stomach, intestine, placenta, pituitary, and possibly in the hypothalamus (9). OBESITY RESEARCH Vol. 10 No. 11 November 2002

1161

Ghrelin, Growth Hormone, and Obesity, Lindeman et al.

Ghrelin administration was found to strongly stimulate GH secretion in both rodents (10) and humans (11,12), and rat serum ghrelin concentrations were increased by fasting and reduced by re-feeding (13). Regarding its potent effects on GH release, we hypothesized that low ghrelin levels underlie the observed hyposomatropism in central obesity, and that the improved GH concentrations on weight loss are related to increases in circulating ghrelin. To test this hypothesis, ghrelin levels were assessed in individuals widely varying in body mass and fat distribution and compared with GH and leptin concentrations. Weight loss increases GH levels, and we studied whether the improved GH concentrations on weight loss are paralleled by enhanced ghrelin levels. The putative causative relationship between GH and ghrelin concentrations was further evaluated by comparing diurnal plasma ghrelin and GH concentration profiles, because in the case of a direct relationship, circulating ghrelin levels should resemble the pulsatile GH secretion pattern.

Research Methods and Procedures Study Population Normal weight and moderately obese subjects (N ⫽ 80, 43 women and 37 men, ages 28 to 48 years) were recruited through advertisements in local newspapers. Severely obese individuals [body mass index (BMI) ⬎ 40 kg/m2] were all scheduled to undergo gastric banding for morbid obesity. Except for their obesity, all patients were healthy according to medical history, clinical examination, and routine laboratory findings. The study was approved by the Medical Ethical Committee of the Leiden University Medical Center. Blood Sampling and Analysis Blood sampling was performed between 8:30 AM and 9:30 AM after an overnight fast of at least 8 hours. Blood samples were immediately put on melting ice and centrifuged within 1 hour of sampling. Plasma was stored at ⫺70 °C until analysis. Plasma ghrelin and leptin were determined by radioimmunoassay (Phoenix Peptide, Mountain View, CA and Linco Research, St. Charles, MO, respectively). To minimize fluctuations in the ghrelin assay, all assays were performed in kits from a single batch. Plasma GH concentrations were assessed by an ultra sensitive 22 kDa–specific immunofluorometric assay (Delfia hGH kit; Wallac Oy, Turku, Finland). PAI-1 antigen was measured by an inhouse enzyme-linked immunosorbent assay (ELISA), which recognizes both free and complexed forms of PAI-1. The intra- and within coefficient of variations of this assay are 6.3% and 6.0%, respectively. Anthropometrics, Fat Distribution, and Weight Loss All anthropometric measurements were made with the subjects wearing underwear. Body weight was rounded to 1162

OBESITY RESEARCH Vol. 10 No. 11 November 2002

the nearest 0.1 kg and body height was measured to the nearest 0.1 cm using a wall-mounted stadiometer. BMI was calculated as weight (kilograms) divided by height (square meters). Age did not influence plasma ghrelin levels, and therefore no adjustments were made. The influence of fat distribution and weight loss was assessed in 11 moderately obese, premenopausal women [BMI, 33.9 ⫾ 3.6 kg/m2, ages 35.1 ⫾ 6.0 years (SD)]. Women were carefully selected to cover a wide range of waist-to-hip ratios. The visceral adipose tissue area was assessed by magnetic resonance imaging: transverse abdominal scans were made at three levels, with each scan 10 mm thick and a gap of 2 mm between levels. The highest scan was at the level of the cranial intervertebral facies of the forth lumbar vertebra. Images were analyzed on a SUN workstation, and the visceral adipose tissue area was calculated. All measurements were repeated after the subjects had lost 50% of their prestudy weight excess (average weight loss, 15.3 ⫾ 4.4% of the initial body weight) through a liquid hypocaloric diet (2 MJ/day, Modifast; Novartis, Veenendaal, The Netherlands; 1 MJ ⫽ 238.95 kcal). Twenty-Four Hour Ghrelin and Growth Hormone Profiles Subjects were admitted to the research center at 8:00 AM after an overnight fast. From 3 days before the procedure to during the procedure, participants consumed a standard isocaloric (8.3 MJ/day) liquid diet (Modifast; Novartis and Nutridrink; Numico, Zoetermeer, The Netherlands) and refrained from other snacks. Standard meals were consumed in equal portions at 9:30 AM, 1:00 PM, and 6:30 PM. During the test, patients were allowed to walk inside the research center. Blood collection was established through a nontrombogenic catheter, which was connected to a constant withdrawal pump (Conflo; Carmeda AB, Taeby, Sweden). The reservoir tubes were replaced every 10 minutes for a 24hour period. The 24-hour GH area under the concentration curve was determined using the linear trapezoidal rule. Statistical Methods Results are expressed as mean ⫾ SD. The significance within the groups was tested by Student’s paired t test or Wilcoxon rank-order test for paired observations when applicable. Base log10 transformation was used to normalize ghrelin and 24-hour GH concentrations, and correlations with BMI were sought after Pearson’s least squares method. A p value ⱕ0.05 was considered significant. All statistical calculations were performed using SPSS for Windows V10.0 (SPSS, Chicago, IL).

Results Plasma ghrelin levels exponentially declined with BMI (N ⫽ 80; r ⫽ ⫺0.67, p ⬍ 0.001) falling from maximal


Figure 1: Plasma ghrelin levels are exponentially related to body mass index (BMI; N ⫽ 80; r ⫽ 0.67, p ⬍ 0.001 after log10 transformation).

values of ⬃400 pg/mL at a BMI of 20 kg/m2 to values ⬍80 pg/mL at indices over 35 kg/m2 (Figure 1). In agreement with the literature (1), there was also an exponential decline in the integrated 24-hour plasma GH concentrations with BMI (n ⫽ 22; r ⫽ ⫺0.67, p ⬍ 0.001). In contrast, plasma levels of leptin, a hormone that reduces food intake and increases metabolic rate, showed a significant positive linear relationship with BMI (N ⫽ 80; r ⫽ 0.85, p ⬍ 0.001), which is believed to reflect the leptin resistance associated with human obesity (14). Superimposed on BMI and total-body fat, GH secretion is strongly influenced by fat distribution: accumulation of visceral adipose tissue has been found to be negatively associated with plasma GH values. (1,15) The influence of visceral fat mass on plasma GH and ghrelin levels was studied in 11 moderately obese, weight-matched women (BMI, 33.0 ⫾ 3.4 kg/m2; percentage total body fat, 42.2 ⫾ 3.9%). Women were selected to cover a wide range of body fat distribution, and the visceral fat area ranged from 200 to

863 cm2 (median, 376 cm2). The 24-hour GH area under the curve was inversely related to the visceral fat area (r ⫽ ⫺0.72, p ⬍ 0.01 after base log10 transformation), whereas a linear relationship was observed between the visceral fat area and ghrelin (r ⫽ 0.49, p ⬍ 0.02). The association between plasma ghrelin and the log 24-hour GH area under the curve (r ⫽ 0.84, p ⬍ 0.01) disappeared after correction for visceral fat area. Loss of 50% overweight in these women significantly decreased visceral fat mass, leptin levels (p ⬍ 0.01), and PAI-1 concentrations (p ⬍ 0.02), and increased GH secretion, but did not influence plasma ghrelin levels (Table 1). Results from the above studies are not consistent with a role of ghrelin in the diminished integrated GH secretion in obesity but do not exclude a role of ghrelin in the acute regulation of plasma GH levels (i.e., peak initiation, peak amplitude). We further sought evidence for a direct relationship between ghrelin and plasma GH concentrations by comparing the daily patterns of GH and ghrelin plasma levels in normal and overweight subjects. All subjects showed a pulsatile GH secretion pattern characterized by 7 to 9 peaks per 24 hours with major peaks during sleep, albeit peak amplitudes were significantly lower in the obese individuals. Three examples, typical for normal-weight and obese individuals, are shown in Figure 2. In the normal-weight individuals, diurnal ghrelin patterns are most adequately described by a cosine with a minimum early in the morning, followed by a 2 to 3 hour hump and a second minimum before lunch. After lunch, levels reached a plateau and remained essentially stable. The diurnal ghrelin pattern in the obese individuals seemed much more smooth, without an apparent influence of food intake. For the range of BMIs tested (19.6 to 34.8 kg/m2), no evident direct relationship between GH pulses and ghrelin was observed, indicating that the acute changes in plasma GH levels are not regulated by ghrelin either.

Table 1. Effect of loss of 50% of the prestudy overweight on visceral fat area, leptin and plasma PAI-1, and GH and ghrelin levels (mean ⫾ SD or median [range]) in 11 moderately obese women

Weight (kg) Visceral fat area (cm2) Leptin (␮g/L) PAI-1 (ng/L) GH (mU/day) Ghrelin (pg/L)

Before

After

94.7 ⫾ 12.1 426 ⫾ 207 36.5 ⫾ 21 161 ⫾ 82 1983 (550 to 5999) 74.7 ⫾ 13.7

80.1 ⫾ 10.0 260 ⫾ 137 16.7 ⫾ 9.5 71 ⫾ 38 4024 (1788 to 10,060) 74.3 ⫾ 33.6

p Value p p p p p

⬍ 0.01 ⬍ 0.01 ⬍ 0.01 ⬍ 0.02 ⬍ 0.01 NS

PAI-1, plasminogen activator inhibitor type 1; GH, growth hormone; NS, not significant.


1163


Figure 2: Representative 24-hour profiles of plasma ghrelin: (——), 20-minute sampling and growth hormone; (. . . . . . . . . . ), 10-minute sampling from a normal-weight subject; and two moderately obese subjects with either a large (LVFA) and a small visceral fat area (SVFA). Meals are indicated by the dashed line (breakfast (B), lunch (L), dinner (D).

Discussion Several observations from animal and human studies support the hypothesis that ghrelin is a physiological regulator of GH release. In this paper we explored the putative role of ghrelin in the hyposomatropism associated with obesity. Although plasma ghrelin and GH show a concordant, reciprocal relationship with BMI, no evidence for a role of ghrelin in the regulation plasma GH in obesity was found. First, GH secretion in obesity is negatively related with the visceral fat mass whereas ghrelin levels are actually positively related with the amount of visceral fat. Second, weight loss increases GH levels and decreases leptin and 1164


PAI-1 levels in obese subjects, but these changes were not paralleled by increased ghrelin levels. Third, plasma ghrelin levels increased nearly 2-fold immediately before breakfast and lunch and showed a diurnal rhythm that was in phase with that of leptin, but which did not match the pulsatile GH secretion pattern at all. Ghrelin has originally been described as an endogenous ligand for the GH secretagogue receptor (16), and the putative role of ghrelin as a regulator of GH secretion is based on the ability of exogenous ghrelin to elicit a robust GH response in both animals and humans. In such a concept, the hyposomatropism of obesity can easily be explained by the exponential decline in plasma ghrelin levels with increasing BMIs, as shown in this paper and in the recent paper of Tscho¨ p et al. (17). However, further research indicates that, at least in obese individuals, regulation of GH secretion occurs independently of plasma ghrelin levels. Assuming that weight loss does not induce ghrelin sensitization and altered expression levels of GH secretagogue receptors in the pituitary or brain, our results demonstrate that the potency of exogenous ghrelin as a stimulus of GH secretion cannot be simply extrapolated to endogenous ghrelin. Our findings that plasma ghrelin levels are not influenced by weight loss apparently contrast with recent studies of Krarup Hansen et al. (18) and Cummings et al. (19). This may however relate to differences in caloric intake during the reassessment of ghrelin levels before and after weight loss. In the study of Krarup Hansen et al. (18), ghrelin levels were assessed during active weight loss induced by a negative energy balance and a fitness program. Cummings et al. (19) evaluated plasma ghrelin levels in weight-stable obese individuals who had lost 17% of their initial body weight. These subjects were weight stable for at least 3 months and had adjusted their diet to maintain a stable weight (19). Reportedly, maintenance of body weight at 20% below the baseline in obese subjects is associated with ⬃8-kcal/kg reduction in total energy expenditure (20); therefore, caloric intake in the study by Cummings et al. (19) is likely to differ before and after weight loss. An increase in caloric intake has been found to reduce plasma ghrelin levels, and the results of Karup Hansen et al. and Cummings et al. may relate to a lower caloric intake during the assessment of plasma ghrelin levels after weight loss (21). In the present study, metabolic differences during the assessment of pre- and post-weight loss parameters were minimized: subjects were at least 3 weeks weight stable before retesting and consumed a standardized, isocaloric diet in the days before and during the procedures. Physical exercise was not used as a means of inducing weight loss (18); indeed, subjects were explicitly asked not to change their physical activity level. Although our findings do not provide evidence for a direct role of ghrelin in GH secretion in obese subjects, we cannot exclude that ghrelin may act as a GH secretagogue at


levels found in lean individuals: the pathogenesis of hyposomatropism in obesity is complex and poorly understood (22), the GH surge in response to virtually every GH secretagogue is considerably reduced in obesity (23) and we cannot exclude that human obesity is also associated with some form of ghrelin resistance. In two small clinical trials, ghrelin has been administered to normal-weight human volunteers, but no study in obese individuals is yet available. Moreover, we cannot rule out that our findings are influenced by the fact that we measured total ghrelin rather than its biologically active acylated form. Preliminary evidence from a rat study suggests that ratio acylated/ nonacylated ghrelin is influenced by feeding status (24), and it cannot be excluded that this proportion is also influenced by more chronic conditions such as fat distribution or weight loss. The observation that plasma ghrelin levels surge before breakfast and lunch rather than before the GH peak suggests that ghrelin may be involved in regulation of food intake (25). A similar conclusion was reached by Cummings et al., who found a cosine-like diurnal rhythm with superimposed surges in ghrelin levels before each meal (26), and who accordingly hypothesized that ghrelin acts as meal initiator. Interestingly, in the present study and a study of Cummings et al. (19), such a phenomenon was not observed in obese individuals; the diurnal rhythm appeared rather smoothed. This finding is in line with the proposed role of ghrelin as an anabolic hormone regulating energy stores. In conclusion, the concentrations of circulating ghrelin and GH exponentially decline with increasing BMI, but no causative relationship between the low ghrelin levels and the hyposomatropism of obesity could be demonstrated. The surge of plasma ghrelin levels immediately before a meal in normal-weight (but not in obese) individuals is intriguing and suggests that ghrelin might have a role in the (central) regulation of metabolism and appetite.

Acknowledgments The financial support of the Netherlands Heart Foundation (NHS 97.100) is gratefully acknowledged. References 1. Giustina A, Veldhuis JD. Pathophysiology of the neuroregulation of growth hormone secretion in experimental animals and the human. Endocrine Rev. 1998;19:717–97. 2. Langendonk JG, Meinders E, Burggraaf J, et al. Influence of obesity and body fat distribution on growth hormone kinetics in humans. Am J Physiol. 1999;277:E824 –E9. 3. Salomon F, Cuneo RC, Hesp R, So¨nkson PH. The effects of treatment with recombinant human growth hormone on body composition and metabolism in adults with growth hormone deficiency. N Eng J Med. 1989;321:1797– 803. 4. de Boer H, Blok G-J, Voerman B, Derriks P, van der Veen E. Changes in subcutaneous and visceral fat mass during growth hormone replacement therapy in adult man. Int J Obes Relat Metab Disord. 1996;20:580 –7.

5. Johansson JO, Fowelin J, Landin K, Lager I, Bengtsson BÅ. Growth hormone-deficient adults are insulin-resistant. Metabolism. 1995;44:1126 –9. 6. Sesmilo G, Biller BMK, Llevadot J, et al. Effects of growth hormone administration on inflammatory and other cardiovascular risk markers in men with growth hormone deficiency. Ann Intern Med. 2000;133:111–22. 7. Johansson JO, Landin K, Tengborn L, Rose´n T, Bengtsson BÅ. High fibrinogen and plasminogen activator inhibitor activity in growth hormone-deficient adults. Arterioscler Thromb. 1994;14:434 –7. 8. Borson-Chavot F, Serusclat A, Kalfallah Y, et al. Decrease in carotid intima media thickness after one year growth hormone treatment in adults with GH deficiency. J Clin Endocinol Metab. 1999;84:1329 –33. 9. Horvath TL, Diano S, Sotonyi P, Heiman M, Tscho˜p M. Minireview: ghrelin and the regulation of energy balance-a hypothalamic perspective. Endocrinology. 2001;142: 4163–9. 10. Wren AM, Small CJ, Ward HL, et al. The novel hypothalamic peptide ghrelin stimulates food intake and growth hormone secretion. Endocrinology. 2000;141:4325– 8. 11. Arvat E, Di Vito L, Broglio F, et al. Preliminary evidence that ghrelin, the natural GH secretagogue receptor ligand, strongly stimulates GH secretion in humans. J Endocrinol Invest. 2000;23:493–5. 12. Takaya K, Ariyasu H, Kanamoto N, et al. Ghrelin strongly stimulates growth hormone release in humans. J Clin Endocrinol Metab. 2001;12:4908 –11. 13. Nakazato M, Murakami N, Date Y, et al. A role for ghrelin in the central regulation of feeding. Nature. 2001;409: 194 – 8. 14. Considine RV, Sinha MK, Heiman ML, et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med. 1996;334:292–5. 15. Pijl H, Langendonk JG, Burggraaf J, et al. Altered neuroregulation of GH secretion in viscerally obese premenopausal women. J Clin Endocrinol Metab. 2001;86:5509 –15. 16. Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K. Ghrelin is a growth hormone releasing acylated peptide from the stomach. Nature. 1999;142:4163–9. 17. Tscho¨p M, Weyer C, Tataranni PA, Devanarayan V, Ravussin E, Heiman ML. Circulating ghrelin levels are decreased in human obesity. Diabetes. 2001;50:707–9. 18. Krarup Hansen T, Dall R, Hosoda H, et al. Weight loss increases circulating levels of ghrelin in human obesity. Clin Endocrinol (Oxf). 2002;56:203– 6. 19. Cummings DE, Weigle DS, Frayo RS, et al. Plasma ghrelin levels after diet induced weight loss or gastric bypass surgery. N Engl J Med. 2002;346:1623–30. 20. Mantzoros CS, Flier JS. Leptin as a therapeutic agent-trials and tribulations. J Clin Endocrinol Metab. 2000;85:4000 –1. 21. Shiiya T, Nakazato M, Mizuta M, et al. Plasma ghrelin levels in lean and obese humans and the effect of glucose on ghrelin secretion. J Clin Endocrinol Metab. 2002;87:240 – 4. 22. Scacchi M, Pincelli AI, Cavagnini F. Growth hormone in obesity. Int J Obes Relat Metab Disord. 1999;23:260 –70. OBESITY RESEARCH Vol. 10 No. 11 November 2002

1165


23. Rahim A, O’Neill P, Shalet SM. The effect of body composition on hexarelin-induced growth hormone release in normal elderly subjects. Clin Endocrinol (Oxf). 1998;49:659 – 64. 24. Date Y, Kojima M, Hosoda H, et al. Ghrelin, a novel growth hormone–releasing acylated peptide is synthesized in a distinct endocrine cell type in the gastrointestinal tract of rats and humans. Endocrinology. 2000;141:4255– 61.

1166


25. Wren AM, Cohen MA, Brynes AE, et al. Ghrelin enhances appetite and increases food intake in humans. J Clin Endocrinol Metab. 2001;86:5992–5. 26. Cummings DE, Purnell JQ, Frayo RS, Schmidova K, Wisse BE, Weigle DS. A preprandial rise in plasma ghrelin levels suggests a role in meal initiation in humans. Diabetes. 2001;50:1714 –9.