GHRP-2 is a synthetic agonist of ghrelin, the newly-discovered gut peptide which binds to the growth hormone (GH) secretagogue receptor. Ghrelin has been shown to have two major effects, stimulating both GH secretion and appetite/meal initiation. GHRP-2 has been extensively studied for its utility as a growth hormone secretagogue (GHS). Animal studies have shown its effect on food intake. However, whether GHRP-2 can also stimulate appetite in humans when administered acutely is not known. We subcutaneously infused 7 lean, healthy males with GHRP-2 (1μg/kg/h) or saline for 270 minutes and then measured their intake of an ad libitum, buffet-style meal. Similar to what has been reported for ghrelin administration, our subjects ate 35.9±10.9 % more when infused with GHRP-2 vs. saline, with every subject increasing their intake even when calculated per kg body weight (136.0±13.0 kJ/kg vs 101.3±10.5 kJ/kg, p=0.008). The macronutrient composition of consumed food was not different between conditions. As expected, serum GH levels rose significantly during GHRP-2 infusion (AUC 5550±1090 μg/L/240 min vs. 412±161 μg/L/240 min, p=0.003). These data are the first to demonstrate that GHRP-2, like ghrelin, increases food intake, suggesting that GHRP-2 is a valuable tool for investigating ghrelin effects on eating behaviour in humans.
Ghrelin, the recently identified peptide secreted by gastric endocrine cells (1), has attracted much interest for its dual effects. This endogenous ligand for the growth hormone secretagogue receptor (GHS-R) which was cloned in 1996 (2) regulates growth hormone (GH) release (3). Ghrelin also appears to play a role in the regulation of food intake and energy balance. When administered either centrally or peripherally to rodents, ghrelin increases food intake and body weight (4,5). Interestingly, its effects on food intake are independent of GH secretion (4,6–8) and appear to be mediated via the NPY/Agouti gene-related protein (AGRP) neurons in the hypothalamic arcuate nucleus (9,10). Peripheral ghrelin administration has recently been shown to stimulate food intake in lean, healthy men and women (11) and in cancer patients (12).
Data suggest that circulating ghrelin is also implicated in a meal to meal regulation. Ghrelin levels increase in anticipation of a meal (13) and are suppressed by food ingestion (13, 14), but the underlying mechanisms are not known. The meal-related suppression of ghrelin is proportional to the carbohydrate (CHO) content of the meal but does not appear to be directly related to glucose or insulin (14,15), although insulin administration decreases ghrelin (16).
Serum ghrelin levels vary as a function of energy balance. Ghrelin levels are increased in anorexia (17) and decreased in obesity (18). Thus, it is possible that ghrelin may be an important player in food intake behaviour and perhaps in chronic over- and undernutrition as well (19). Because of its dual effects, ghrelin may be a critical hormonal signal of nutritional status to the somatotropic axis, playing a role in integrating energy balance with the growth process (20).
Ghrelin is not available for long-term studies in human subjects in the United States, however, the synthetic GHS-R agonist GHRP-2 (DalaDßNalAlaTrpDPheLysNH2) is available for clinical studies. GHRP-2 belongs to a family of GHS(s) discovered in the 1980’s and extensively studied for their effect on GH release (21). Despite the very different chemistry of natural ghrelin and synthetic GHS, evidence strongly and increasingly supports that they have the same biological actions. Like ghrelin, GHS(s) increase food intake and body weight in rodents (22). GHRP-2 has been shown to increase appetite ratings in children with idiopathic GH treated chronically with oral GHRP-2 (23).
The aim of this study was to investigate whether GHRP-2 stimulates food intake in healthy human subjects. Similar to the study by Wren et al. (11) which used ghrelin, we tested the effect of a short-term subcutaneous (sc) infusion of GHRP-2 on food intake during a buffet meal.
Seven lean men, (age 26.0 ± 1.6 years, BMI 21.5 ± 1.7 kg/m2, fat mass 14.4 ± 1.1 % by anthropometrics, weight stable, healthy, on no medications, non-smoking, non-dieting, without any eating disorder, participated in the study. The Institutional Review Board of St. Luke’s/Roosevelt Hospital approved the study protocol and informed written consent was obtained prior to inclusion in the study. A screening session included a full physical exam, routine laboratory tests, a taste test and a test meal. Subjects had to rate their liking of the food a ‘5’. or more on average (on a scale 0-7) and had to eat at least 2090 kJ at the screening test meal.
Subjects were studied after an overnight fast on two non-consecutive days separated by at least a week. In a randomized, double-blind fashion, GHRP-2 (1 μg/kg/h) or placebo was administered for the entire duration of the experiment (including meal time) via sc catheter attached to a pump (Minimed 508). Blood was collected for subsequent hormone measurement through an iv catheter placed in the forearm, sampled every 30 minutes until the beginning of the meal. The buffet meal included several choices of food served in excess to accommodate any appetite. While eating in the laboratory, the subjects were observed with a camera to ensure their safety and that they did not dispose of the food in any other way than eating. A fixed, standard test breakfast was given 120 minutes after the beginning of the infusion and the ad libitum lunch was offered at 240 min. The breakfast consisted of 1.5 English Muffins, 5 g of butter, and 125 ml of apple juice (1254 kJ, 65% CHO, 4% protein, 31% fat). Prior to the buffet lunch, subjects listened to a taped instruction to ‘eat as much as they wanted’. The items served during the buffet lunch were pasta, bread rolls, peas and carrots, chocolate chip cookies, applesauce, chicken nuggets and breaded fillets, and water. Food was weighed before and after the meal and caloric intake calculated. After the meal, the infusion was stopped. Visual analogue scales (VAS, from 0 ‘not at all’ to 150 mm ‘extremely’) to assess hunger and fullness rating were administered prior to starting the infusion, before and after breakfast, and before and after completion of the lunch meal. The positions of the marks on the 150 mm scale were measured in mm by a blinded investigator.
Serum GH concentrations were measured in duplicate by a radioisotopic kit for quantitative determination of HGH from Nichols Institute Diagnostic (San Juan, CA). The assay sensitivity was 0.2 μg/L and the median intra- and interassay coefficients of variation were 2.9% and 7.5% respectively. Serum cortisol levels were measured by radioimmunoassay (kit from DSL, Webster, TX). The assay sensitivity was 0.11 ng/dl. The median intra- and interassay variation was 3.8% and 5.9% respectively. Blood collection was incomplete for one subject, therefore, data on 6 subjects are reported.
Comparison between GHRP-2 and placebo was done by a paired t-test for all variables. A hormonal response was assessed as area under the curve (AUC) as measured by the trapezoidal method. A p-value of 0.05 was used for level of significance. Data are presented as mean ± SEM unless otherwise stated. SPSS version 11.5 was used for statistical analysis.
No side effects from the infusion were reported. Food intake at the buffet meal was increased by 35.9±10.9 % (range 12 to 95%) with GHRP-2 compared to placebo. Each and every subject responded to the GHRP-2 infusion by increasing their food intake, regardless of the amount of food eaten during the saline control experiment (Figure 1). The total amount of calories eaten was greater with GHRP-2 than with placebo (9409±1229 kJ vs. 7118±1078 kJ, p=0.004) as were the kJ ate per kg of body weight (136±13 kJ/kg vs. 101±10 kJ/kg, p=0.008). Subjects ate more CHO (58±14 g, p=0.006), protein (17±4 g, p=0.005) and fat (38±10 g, p=0.008) with GHRP-2 than placebo and the proportion of calories obtained from each nutrient was not different between the 2 conditions (results not shown). The duration of the ad libitum meal was the same with and without the GHRP-2 (30 ± 8 min vs. 32±6 min, p=0.667). Thus the food intake per interval (calories are eaten during the meal divided by the duration of the meal) was greater, but not significantly, with GHRP-2 than with placebo (322±33 kJ/min vs. 276±38 kJ/min, p=0.179). Based on the VAS, baseline hunger and fullness ratings were not different between conditions. However, during the GHRP-2 infusion, subjects felt hungrier (112.6±7.5 mm vs. 89.7±12.2, p=0.013) before lunch and showed a trend to be fuller after lunch (134.1±6.0 mm vs. 118.6±10.3 mm, p=0.054).