Gastric bypass does not normalize obesity-related changes in ghrelin profile and leads to higher acylated ghrelin fraction




Gastric bypass (GBP) lowers food intake, body weight, and insulin resistance in severe obesity (SO). Ghrelin is a gastric orexigenic and adipogenic hormone contributing to modulate energy balance and insulin action. Total plasma ghrelin (T-Ghr) level is low and inversely related to body weight and insulin resistance in moderately obese patients, but these observations may not extend to the orexigenic acylated form (A-Ghr) whose plasma concentration increase in moderate obesity.

Design and Methods:

We investigated the impact of GBP on plasma T-, A-, and A/T-Ghr in SO patients (n = 28, 20 women), with measurements at baseline and 1, 3, 6, and 12 months after surgery. Additional cross-sectional comparison was performed between nonobese, moderately obese, and SO individuals before GBP and at the end of the follow-up period.


Before GBP, SO had lowest T-Ghr and highest A/T-Ghr profile compared with both nonobese and moderately obese individuals. Lack of early (0-3 months from GBP) T-Ghr changes masked a sharp increase in A-Ghr and A/T-Ghr profile (P < 0.05) that remained elevated following later increments (6-12 months) of both T- and A-Ghr (P < 0.05). Levels of A-Ghr and A/T-Ghr at 12 months of follow-up remained higher than in matched moderately obese individuals not treated with surgery (P < 0.05).


The data show that following GBP, early T-Ghr stability masks elevation of A/T-Ghr, that is stabilized after later increments of both T- and A-hormones. GBP does not normalize the obesity-associated elevated A/T-Ghr ratio, instead resulting in enhanced A-Ghr excess. Excess A-Ghr is unlikely to contribute to, and might limit, the common GBP-induced declines of appetite, body weight, and insulin resistance.


Ghrelin is a gastric orexigenic and adipogenic hormone (1) but its total plasma concentration is low in obesity (2) which is likely due to feedback mechanisms involving elevated circulating insulin and leptin (3, 4), and total plasma ghrelin is accordingly associated with low-insulin resistance in both lean and obese individuals (5). Ghrelin acylation appears to be required for its orexigenic effects (1) and more recent observations indicate that circulating acylated ghrelin may be preserved in moderately obese patients (5). In addition, acylated hormone has not been associated with favorable metabolic profile and higher insulin sensitivity in moderately obese humans (5), likely resulting from its complex tissue-specific effects that may lead to a potentially beneficial impact in skeletal muscle with negative effects in the liver (6-9).

Gastric bypass (GBP) surgery is a most effective treatment for severely obese patients, leading to reduced appetite, sustained loss of body mass and fat, and improvement of obesity-associated metabolic abnormalities (10-13). The positive impact of this procedure on energy balance and intermediate metabolism has been hypothesized to extend beyond its combined restrictive and malabsorptive effects (11, 12). GBP-induced changes in gut hormone release and plasma concentrations have been indeed proposed to contribute to its beneficial metabolic effects and might potentially contribute to reduce appetite (11, 12). Gut hormone profiles associated with GBP-induced reductions of appetite and insulin resistance in vivo remain, however, incompletely understood. In particular, the impact of GBP on plasma ghrelin profile remains incompletely defined (13). A negative impact of surgically-induced anatomical changes on total ghrelin secretion by the excluded stomach has been initially proposed (14) but not confirmed by later studies (13). No information is available on GBP-induced absolute and relative changes in total and acylated hormone concentrations, and whether they normalize potential alterations in ghrelin profile of severe obesity (SO).

In this study, we therefore investigated the time course of changes in plasma total and acylated ghrelin in patients undergoing laparoscopic GBP. Lean and moderately obese control subjects were also studied to determine whether GBP-induced changes normalize potential alterations in plasma ghrelin profile as compared to non-GBP individuals. The 1-year duration of follow-up was chosen as BMI at 12 months after surgery is reported to be stable in longer follow-up periods (10), and 12-month values are therefore likely to reflect long-term changes for anthropometric, metabolic, and hormonal parameters.

Methods and Procedures

Subjects and experimental protocol

The study protocol was approved by the Ethics Committee at Trieste University Hospital. All participants were given detailed oral and written information on the study aims and risks, and they gave written consent before entering the study. A total of 28 patients (age 41 ± 2 years, eight men) were recruited from the Unit for Bariatric Surgery at the Division of General Surgery at Trieste University Hospital and followed up in the Internal Medicine Outpatient Metabolic Syndrome Clinic. In all the participants, clinical history was collected along with complete physical examination. BMI was calculated as weight (in kilograms) divided by height (in meters) squared. Waist circumference was measured on bare skin during mid-respiration at the natural indentation between the 10th rib and the iliac crest to the nearest 0.5 cm. Inclusion criteria were those internationally accepted for GBP surgery, with key role for SO and BMI >40 kg/m2. Exclusion criteria were those for GBP and also included renal failure (plasma creatinine above 1.5 mg/dl); females taking hormonal estrogen therapy and patients with thyroid disease were also excluded from the study, based on measurement of thyroid hormones and thyroid-stimulating hormone as part of routine screening before GBP surgery. The latter criterion led to the exclusion of two subjects. Prevalence of type 2 diabetes mellitus based on clinical history with previous diagnosis from plasma glucose concentration ≥ 126 mg/dl or hemoglobin A1c ≥ 6.5% was 5 out of 28 patients. For patients undergoing GBP, one blood sample was collected 1-2 weeks before surgery under 10-h fasted conditions; plasma was separated and stored at −80 °C till biochemical and hormonal measurements were done. Additional measurements and sampling were performed under identical conditions at 1, 3, 6 and 12 months after surgery. The GBP procedure was performed laparoscopically in all subjects. Briefly, dissection was begun between the first and second vascular arcades on the lesser curvature using an ultrasonic coagulation device, and a 50-70 ml gastric pouch was created by means of laparoscopic linear stapler. Omentum and transverse mesocolon were lifted upward and Treitz flexure was identified. The biliary limb was measured 100 cm distal to the Treitz ligament and placed in the left ipocondrium via antecolic transposition (first loop). A suspended stitch was fixed between the bowel and gastric pouch; starting from this level, the alimentary limb was measured up to 120 cm and fixed distally by stitch to the biliary limb in the left upper quadrant (second loop). Manual single layer gastrojejunostomy was performed between the gastric pouch and alimentary limb, and manual side to side single layer jejunojejunostomy was performed between alimentary and biliary limbs.

Nonobese and moderately obese individuals who did not undergo GBP (n = 18 each) were also recruited for the study, as two groups of weight-stable healthy volunteers. In addition to sampling for hormone measurements for cross-sectional comparisons, they underwent a routine examination and biochemical work-up to assess exclusion criteria that were similar to those for GBP patients. Moderately obese individuals had BMI comparable with GBP patients at 12-month follow-up, but were weight-stable and had not undergone bariatric surgery procedures. Ghrelin profile in GBP patients at the end of the 12-month follow-up period was therefore also compared with that of the moderately obese control individuals alone, to assess the potential impact of GBP on total, acylated, ghrelin independently of major anthropometric and metabolic variables.

Plasma analyses

Plasma glucose, total and high-density lipoprotein cholesterol, and plasma triglycerides were measured using standard methods. Plasma insulin was measured by enzyme-linked immunosorbent assay (Insulin Human Ultrasensitive ELISA; DRG Instruments, Marburg, Germany). Insulin sensitivity was assessed by the homeostatic model assessment (HOMA) index (5) using the following formula: HOMA = (FPG × FPI)/22.5, where FPG and FPI are fasting plasma glucose (mmol) and fasting plasma insulin (µU/ml), respectively. Plasma total (intra-assay coefficient of variation: 4.4%; interassay coefficient of variation: 9.2%) and acylated (intra-assay coefficient of variation: 4.9%; interassay coefficient of variation: 8.9%) ghrelin were measured using radioimmunoassay (Linco, St Charles, MO).

Statistical analysis

The StatView software was used for all statistical analyses. Time effects for each variable in the GBP group were analyzed using ANOVA and paired t-test. Differences between non-GBP and GBP individuals were analyzed using ANOVA and unpaired t-test. Due to unequal variances log-transformed values for HOMA, total and acylated ghrelin profiles were used for analyses. P values of ≤ 0.05 were considered statistically significant.


BMI, plasma insulin, glucose, HOMA index

Before GBP, severely obese patients had highest BMI and waist circumference by design (Table 1). These alterations were associated with highest plasma insulin, glucose, and HOMA index as compared to nonobese and moderately obese individuals, with an additional trend for higher plasma free fatty acids (Table 1). GBP expectedly led to a substantial, progressive reduction in BMI and waist circumference, associated with lower plasma glucose and plasma insulin, resulting in substantial reduction in HOMA index (Table 2). At variance with BMI and glucose metabolism indexes, concentration of free fatty acids was higher at 1 month after GBP than before GBP, and it progressively returned at baseline levels at subsequent time points. No early reductions were also observed for plasma total cholesterol and triglycerides whereas high-density lipoprotein cholesterol was lower at 1 month after GBP than before GBP. Lowering of plasma triglycerides with higher high-density lipoprotein cholesterol was in turn observed at subsequent time points (Table 2). Patients at 12 months after GBP surgery had BMI, waist circumference, plasma, free fatty acids, and HOMA comparable to those in moderately obese non-GBP individuals (all P = not significant).

Table 1. Anthropometric characteristics and metabolic profile in control groups and in severely obese individuals before gastric bypass surgery
inline image
Table 2. Anthropometric characteristics and metabolic profile in severely obese individuals before GBP surgery and during a 12-month follow-up period
inline image

Plasma ghrelin profile in GBP

Consistent with previous studies, total plasma ghrelin (T-Ghr) was lower and orexigenic acylated form (A-Ghr) was preserved with higher A/T-Ghr ratio in moderately obese compared with nonobese individuals (Figure 1). Before GBP, severely obese patients had further increments in A/T-Ghr ratio with lowest plasma T-Ghr (Figure 1). Following GBP, plasma T-Ghr was unchanged at early follow-up stages (0-3 months, Figure 2a) but lack of early T-Ghr changes masked marked early increments in acylated hormone concentration (Figure 2b). Increments in T-Ghr were observed with further reduction in BMI at 6 and 12 months after surgery, associated with further parallel increase of A-Ghr and persistent stable elevation of circulating A-Ghr fraction (Figure 2a-c). Compared with moderately obese individuals matched for BMI and metabolic profile who did not undergo GBP, GBP patients had marked alterations of 12-month plasma ghrelin profile with persistently lower T-Ghr and substantially higher A- and A/T-Ghr (Figure 2).

Figure 1.

(a) Plasma total ghrelin (T-Ghr), (b) orexigenic acylated form (A-Ghr) and (c) A/T-Ghr ratio in nonobese (lean), moderately obese (obese), and severely obese individuals. For each variable in each panel, different letters denote statistically significant differences (P < 0.05) among groups by ANOVA and post hoc tests; the presence of the same letter conversely indicates lack of statistically significant differences.

Figure 2.

Changes in (a) plasma total ghrelin (T-Ghr), (b) orexigenic acylated form (A-Ghr), and (c) A/T-Ghr ratio over 12 months of observation in severely obese patients undergoing gastric bypass (GBP) surgery, compared to moderately obese patients not treated with GBP (Ob-Non GBP). For each variable in each panel, different letters denote statistically significant differences (P < 0.05) among groups by ANOVA and post hoc tests; the presence of the same letter conversely indicates lack of statistically significant differences.


This study demonstrated that GBP surgery for SO is associated with marked time-dependent changes in plasma ghrelin profile. In particular: (i) in the early observation period for up to 3 months, total ghrelin does not change but a marked increase in acylated hormone level occurs; (ii) in the later period of 6-12 months after surgery, a parallel increase occurs for both total and acylated ghrelin; (iii) acylated ghrelin fraction as expressed by the acylated-to-total ghrelin ratio sharply increases and remains elevated up to a stage of follow-up that was reported to reach BMI stabilization. GBP-induced changes therefore do not normalize alterations of circulating ghrelin profile observed in pre-GBP severely obese patients, and GBP-induced weight loss results in moderately obese individuals with marked acylated hormone excess.

Although both total and acylated ghrelin concentrations were higher at 12 months after GBP, the increase in acylated hormone occurred earlier and was more pronounced, leading to higher acylated ghrelin fraction throughout the observation period. The ghrelin acylating enzyme ghrelin O-acyl transferase (GOAT) has been recently characterized and its expression was described to be enhanced in the stomach by prolonged caloric restriction (15, 16). In addition, fatty acid substrate availability per se may enhance GOAT activity and ghrelin acylation (15). Low-food intake therefore likely contributed to early and sustained increments in acylated ghrelin following bypass, and transient fatty acid elevation could have contributed to enhance its early rise.

Earlier studies indicated variable changes in total circulating ghrelin following GBP, reporting suppressed, unchanged, or higher total hormone concentrations at variable times after surgery (13). Although some discrepancies and an early study reporting substantial suppression of total ghrelin (13, 14) have been difficult to interpret, most reports suggest that changes in plasma total ghrelin are related to weight loss, with highest increments in patients with largest declines in BMI (13). This relationship could be mediated by the attenuation of negative feedback mechanisms that lower ghrelin secretion during positive-energy balance, including plasma insulin as indeed observed in the current model (3, 4). In this study, total circulating ghrelin only increased after substantial decline in BMI at later follow-up stages, consistent with a role of sustained weight loss in GBP-induced total ghrelin changes. Most importantly, however, this study extends the above findings to acylated ghrelin and it identifies acylated hormone as a major contributor to ghrelin changes after GBP. On the basis of these observations, existing literature discrepancies could be due at least in part to undetected variability in acylated ghrelin concentration, and acylated hormone should therefore be measured when investigating GBP-induced changes in ghrelin profile. It should also be pointed out that tight control and inclusion criteria in this study population likely contributed to study results, and larger studies in diverse populations will be needed to extend the validation of the current findings.

Ghrelin has a major role in the regulation of appetite and energy balance, with emerging complex contributions to the modulation of whole body and tissue insulin action. Acylated ghrelin is responsible for orexigenic ghrelin effects and higher acylated ghrelin fraction is associated with less-favorable metabolic profile and insulin action in obese patients (5), likely resulting from combined potentially positive muscle effects and negative impact in the liver (6-9). This study provides further evidence for positive interaction between obesity and acylated ghrelin, by demonstrating a substantial increase of its circulating fraction from moderately to severely obese individuals. Also importantly, the data demonstrate that GBP does not correct obesity-associated alterations in ghrelin profile. Instead, GBP leads to further increments in acylated hormone as compared to similarly obese individuals who have not undergone bariatric surgery, thereby identifying a novel alteration of plasma ghrelin profile specifically associated with GBP treatment. On the basis of current knowledge and the earlier observations, excess acylated ghrelin in GBP patients is unlikely to contribute to, and it could instead limit, the GBP-induced reductions in appetite, body weight, and insulin resistance. Long-term follow-up studies suggest that BMI values at 12 months after surgery remain stable and predict body weight loss throughout subsequent observation (10). Substantial attenuation of the rate of body weight loss through the last 6 months of observation, as well as normalization of plasma free fatty acids, are in agreement with the earlier observations in suggesting that ghrelin profile at the and of the current follow-up is unlikely to undergo further major changes due to additional weight loss. The potential long-term impact of excess acylated ghrelin on maintenance of body weight, fat mass, and insulin sensitivity in GBP should be determined in long-term investigations.

In conclusion, this study identifies substantial alterations of ghrelin profile in GBP patients. Early total ghrelin stability masks elevation of acylated ghrelin and its acylated fraction that remains high after later further increments of total and acylated hormones. GBP fails to normalize, but rather enhances excess acylated ghrelin levels observed in SO. GBP-related changes in ghrelin profile with marked acylated ghrelin elevation are unlikely to contribute to reduced insulin resistance and to the commonly reported GBP-associated reduction in appetite and body fat.