- Email: [email protected]
Treatment with pioglitazone is associated with decreased preprandial ghrelin levels: A randomized clinical trial
Peptides 40 (2013) 89–92
Contents lists available at SciVerse ScienceDirect
Peptides journal homepage: www.elsevier.com/locate/peptides
Short communication
Treatment with pioglitazone is associated with decreased preprandial ghrelin levels: A randomized clinical trial夽 Shervin Taslimi a , Alireza Esteghamati a,∗ , Armin Rashidi b , Hosein Moin Tavakkoli a , Manouchehr Nakhjavani a , Abbas Kebriaee-Zadeh c a
Endocrinology and Metabolism Research Center (EMRC), Vali-Asr Hospital, Tehran University of Medical Sciences, Tehran, Iran Department of Internal Medicine, Eastern Virginia Medical School, Norfolk, VA, United States c Department of Toxicology and Pharmacology, Tehran University of Medical Sciences, Tehran, Iran b
a r t i c l e
i n f o
Article history: Received 4 December 2012 Received in revised form 20 December 2012 Accepted 20 December 2012 Available online 28 December 2012 Keywords: Glucose tolerance Metformin Leptin Ghrelin
a b s t r a c t The effects of metformin and pioglitazone on ghrelin, a physiologic regulator of appetite and food intake, have not been clearly established. In a randomized clinical trial, we randomly assigned 60 type 2 diabetic patients to either metformin (Group A; n = 30) or pioglitazone (Group B; n = 30) treatment groups. The groups were similar in their baseline characteristics. A standard fasting 75 g oral glucose tolerance test was performed at time zero before starting metformin or pioglitazone, and 3 months later. After 3 months of treatment, pioglitazone, but not metformin, was significantly associated with weight gain. Both groups experienced a significant reduction in fasting plasma glucose (p < 0.01), hemoglobin A1c (p < 0.01 in Group A and p < 0.05 in Group B), and insulin resistance (p < 0.01). The effect of metformin on preprandial ghrelin and its response to glucose challenge was not significant, while the pioglitazone group had a significant reduction in preprandial ghrelin levels after treatment (p < 0.05). The effect of pioglitazone on ghrelin was independent of changes in body weight, body mass index, glucose control, insulin resistance, and plasma insulin. In conclusion, treatment with pioglitazone is associated with a decrease in preprandial ghrelin levels and therefore, the weight gain and increased food intake related to pioglitazone use cannot be explained by its effects on ghrelin. The effect of pioglitazone on ghrelin is independent of changes in body weight, body mass index, plasma insulin, insulin resistance, or glucose control. © 2012 Elsevier Inc. All rights reserved.
1. Introduction
2. Materials and methods
Ghrelin and leptin are two intrinsic regulators of food intake and metabolism. Ghrelin, produced mainly by endocrine cells in the stomach [2], enhances appetite and increases food intake [16]. It displays an oscillatory pattern with a preprandial peak and a postprandial trough [5]. The purpose of the present study was to investigate the effects of pioglitazone and metformin on active ghrelin levels during a standard oral glucose tolerance test (OGTT) in a randomized clinical trial on patients with type 2 diabetes.
After obtaining approval from the ethics committee of Tehran University of Medical Sciences (Tehran, Iran), we conducted this prospective, randomized control trial (IRCT201102275917N1) between March 2011 and June 2011 in the Endocrinology and Metabolism Research Center (EMRC), Vali-Asr Hospital, Tehran, Iran. All patients gave written informed consent and agreed to follow the American Diabetes Association dietary advice [1]. We included a total of 60 patients with an established diagnosis of type 2 diabetes who had a hemoglobin A1c (HbA1c) of greater than 7%. All patients were drug-naïve and had not been treated with metformin or pioglitazone before. Exclusion criteria were pregnancy, diabetes diagnosed before the age of 30, known liver disease, creatinine clearance of less than 30 ml/min, and congestive heart failure. Patients were randomly assigned to either of the two groups: Group A received metformin 500 mg twice daily, while Group B received pioglitazone 30 mg daily for 3 months. Age, height, weight and blood pressure were measured at baseline and at 3 months. A standard 75 g fasting OGTT was performed at time zero before starting metformin or pioglitazone, and 3 months later. Plasma glucose, insulin and active ghrelin levels were measured
Abbreviations: OGTT, oral glucose tolerance test; EMRC, Endocrinology and Metabolism Research Center; HOMA-IR, homeostatic model assessment of insulin resistance; CV, coefficient of variation; BMI, body mass index; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; SEM, standard error of the mean; SD, standard deviation. 夽 Clinical Trial Registration Number: IRCT201102275917N1. ∗ Corresponding author at: Endocrinology and Metabolism Research Center (EMRC), Vali-Asr Hospital, Tehran University of Medical Sciences, P.O. Box 13145784, Tehran, Iran. Tel.: +98 21 88417918; fax: +98 21 64432466. E-mail address: [email protected] (A. Esteghamati). 0196-9781/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.peptides.2012.12.020
90
S. Taslimi et al. / Peptides 40 (2013) 89–92
at 0, 60 and 120 min. Leptin was measured at 0 min. An enzyme linked immunosorbent assay was used to measure plasma levels of active ghrelin (DRG Diagnostics, Marburg, Germany; inter- and intra-assay CV, 3.55% and 3.63% respectively) and leptin (DRG Diagnostics, Marburg, Germany; inter- and intra-assay CVs, 8.66% and 5.95% respectively). Sample collection for the measurement of acylghrelin was performed according to the method recommended by Blatnik and Soderstrom [4]. Briefly, blood was collected in K2 EDTA tubes and 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride (AEBSF) was added immediately after collection. The final concentration was 2 g/l. Plasma insulin was measured using an immunoradiometric assay (Immunotech, Prague, Czech Republic; inter- and intra-assay CV, 3.4% and 4.3% respectively) and its corresponding area under the curve (AUC) was calculated for the three time points as mentioned above. HOMA-IR, our measure of insulin resistance, was calculated using the formula: HOMA-IR = fasting insulin (mU/l) × fasting plasma glucose (FPG; mg/dl)/405 [12]. In this article, we use, for simplicity, the term “ghrelin” instead of active ghrelin. Differences were analyzed using a Student’s t-test and Fisher’s exact method for normal distributions, and Mann–Whitney U-test for skewed distributions. Correlations were determined using the Spearman’s rank method for variables with skewed distributions. Since ghrelin changed significantly after treatment only in the pioglitazone group, multiple regression analysis was applied to this group with preprandial ghrelin change (due to treatment) as the dependent variable and the change in variables that changed significantly with treatment as independent predictors. Data are shown as mean ± standard deviation unless indicated otherwise. p < 0.05 was considered statistically significant.
3. Results Table 1 summarizes the characteristics of the groups before and after treatment. There was no significant difference between the two groups in any of the pretreatment variables. Importantly, there was no significant difference between the groups in their pretreatment ghrelin levels. After 3 months of treatment, there was no significant weight change in Group A, while Group B experienced a statistically significant weight gain (p < 0.01), an increase in BMI (p < 0.01), and a decrease in insulin AUC (p < 0.01). In both groups, there was a significant reduction in FPG (p < 0.01), HbA1c (p < 0.01 in Group A and p < 0.05 in Group B), and insulin resistance as measured by HOMA-IR (p < 0.01). Treatment was associated with a significant decrease (p < 0.05) in leptin levels only in Group A. Pioglitazone treatment was associated with a significant reduction in preprandial ghrelin levels (p < 0.05). Ghrelin was significantly suppressed 1 h post-prandial both before (p < 0.01) and after (p < 0.01) treatment with pioglitazone. The effect of metformin on preprandial ghrelin and its response to glucose challenge was not significant (Fig. 1). Treatment in either group was not associated with any other significant change in the measured variables (Table 1). We then applied multiple regression analysis to Group B, which showed a significant change in preprandial ghrelin with treatment. The purpose of this analysis was to evaluate whether the change in preprandial ghrelin was due to a change in any other variable studied. The dependent variable in the regression model was preprandial ghrelin change with treatment. The set of independent predictors were those variables that showed a significant change with treatment. This set included changes in the following variables: HbA1c, weight, BMI, HOMA-IR, and insulin AUC. To avoid co-linearity, we included weight and BMI in separate models, and also HOMA-IR and insulin AUC in separate models. Importantly, there was no significant correlation between pretreatment ghrelin and either plasma insulin or HOMA-IR. Neither of the
Fig. 1. Ghrelin levels in a standard 75 g oral glucose tolerance test. Ghrelin levels are significantly suppressed 1 h post-challenge in both groups, both before and after treatment. Only pioglitazone decreased preprandial ghrelin levels significantly. p values in both diagrams indicate the significance of difference from preprandial levels. All values are median ± SEM.
models revealed a statistically significant predictor for preprandial ghrelin change. Therefore, the change in preprandial ghrelin with pioglitazone treatment was not due to changes in any of the other variables studied in this work. 4. Discussion Consistent with previous studies, pioglitazone was associated with weight gain in our study, while metformin did not significantly affect weight [17]. Also, treatment with metformin caused a significant decrease in leptin levels. This result is in line with previous studies showing that insulin sensitization is associated with decreased leptin levels [7]. Moreover, we showed that while metformin did not have a significant effect on preprandial ghrelin levels and post-challenge ghrelin suppression, pioglitazone significantly decreased pre-prandial ghrelin levels. The randomized design of this study eliminated any significant difference between the two groups in any of the pretreatment variables, thus making the difference in treatments the most likely explanation for a significant change in ghrelin levels in the pioglitazone as opposed to the metformin group. Consistent with previous studies, our results suggest that pioglitazone effects are not exerted via effects on leptin [14]. Our regression analysis also suggests that the effects of pioglitazone on ghrelin are not due to changes in insulin resistance, body weight, or long-term glucose control. Pioglitazone treatment in rats is associated with decreased fasting ghrelin levels [13]. A previous study on 16 Japanese type 2 diabetic patients who underwent 4 months of treatment with pioglitazone, treatment was not associated with a significant change in fasting ghrelin levels [10]. A similar study on 36 patients
S. Taslimi et al. / Peptides 40 (2013) 89–92
91
Table 1 Characteristics of the groups before and after treatment. Baseline characteristics
Age (years) Male:female Diabetes duration (months)a Weight (kg) BMI (kg/m2 ) WC (cm) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) HbA1c (%) Fasting plasma glucose (g/dl) Plasma creatinine (mg/dl) LDL (mg/dl) HDL (mg/dl) Total cholesterol (mg/dl) Triglycerides (mg/dl)a HOMA-IRa Leptin (ng/ml)a Ghrelin at time zero (pg/ml)a AUC for insulin (mU/l)a
Group A (metformin; n = 30)
Group B (pioglitazone; n = 30)
Pretreatment
Posttreatment
Pretreatment
Posttreatment
51 (10) 15:15 24 (18) 76.8 (12.5) 28.7 (5.6) 97.8 (11.2) 117 (15) 77 (9) 8.4 (1.5) 186 (67) 0.97 (0.21) 124 (33) 48 (12) 212 (40) 190 (22) 3.6 (0.8) 9.8 (1.5) 41.0 (2.5) 59.9 (12.8)
51 (10) 15:15 24 (18) 76.5 (12.3) 28.7 (5.5) 97.6 (10.9) 121 (18) 77 (10) 7.0 (1.4)†† 135 (48)†† 0.96 (0.23) 101 (26)†† 48 (12) 191 (43)† 145 (10)† 2.3 (0.4)† 7.0 (1.3)† 43.5 (2.0) 46.0 (11.1)†
56 (11) 13:17 36 (10) 73.8 (13.1) 28.4 (4) 96.3 (10.5) 123 (16) 77 (9) 8.2 (1.8) 178 (65) 1.0 (0.17) 116 (36) 49 (11) 197 (41) 149 (16) 3.5 (0.4) 9.8 (2.2) 48.5 (3.1) 52.6 (7.8)
56 (11) 13:17 36 (10) 75.5 (13.7)†† 29.0 (4.1)†† 96.9 (11.2) 123 (19) 77 (11) 7.3 (1.6)† 142 (53)†† 1.0 (0.21) 108 (25) 52 (13) 199 (29) 127 (10) 2.1 (0.2)†† 10 (1.8) 40.5 (2.6)† 44.9 (4.2)††
The two groups were not significantly different in any of the pretreatment variables. AUC, area under the curve at three time points 0, 1 h, and 2 h. a Median ± standard error of the mean; all other results are shown as mean ± SD. † p < 0.05 for comparisons between pretreatment and posttreatment in a given group. †† p < 0.01 for comparisons between pretreatment and posttreatment in a given group.
with type 2 diabetes who underwent at least 6 months of treatment with diet alone (n = 11), metformin (n = 12) or pioglitazone (n = 12), the change in fasting ghrelin levels was not significantly different between the groups [9]. The results of our study, with 30 patients in the pioglitazone group, suggest that the Japanese study was probably underpowered and fasting ghrelin levels decrease after treatment with pioglitazone. This does not mean, however, that pioglitazone decreases food intake. In fact, pioglitazone treatment results in increased food consumption in both rats [13] and humans [11]. One potential explanation for the somewhat conflicting outcomes of previous studies on the effects of pioglitazone on ghrelin may be related to the different methods used to stabilize ghrelin in collected samples. Acylghrelin is known to be rapidly converted to des-acylghrelin if blood samples are not immediately stabilized with protease inhibitors. In this study, we used the method recommended by Blatnik and Soderstrom [4], which has been shown to minimize acylghrelin loss. In conclusion, we showed in our randomized clinical trial that treatment with pioglitazone is associated with a reduction in preprandial ghrelin levels, and that this effect is not due to changes in body weight, insulin resistance or long-term glucose control. Importantly, although previous studies have established an inverse relationship between ghrelin levels and BMI [8], our results from regression analysis suggest that the effects of pioglitazone on ghrelin are not due to increased BMI. Similarly, although a negative effect on ghrelin secretion has been demonstrated for insulin in some previous studies [6], our results from regression analysis suggest that the effects of pioglitazone on ghrelin are not due to decreased plasma insulin and improved insulin sensitivity. Considering the appetite-stimulating action of ghrelin, our results suggest that increased food intake subsequent to pioglitazone treatment is not due to the effects of the drug on ghrelin. Similarly, considering the appetite-reducing action of adiponectin [3], which increases after pioglitazone treatment [15], the increased food intake subsequent to pioglitazone treatment is not due to the effects of the drug on adiponectin. This result is consistent with findings of previous studies [9,13]. Hormonal regulation of appetite and body weight is still a complicated subject with many unknown aspects that require further research.
Conflict of interest None.
References [1] American Diabetes Association, Bantle JP, Wylie-Rosett J, Albright AL, Apovian CM, Clark NG, et al. Nutrition recommendations and interventions for diabetes: a position statement of the American Diabetes Association. Diabetes Care 2008;31:S61–78. [2] Ariyasu H, Takaya K, Tagami T, Ogawa Y, Hosoda K, Akamizu T, et al. Stomach is a major source of circulating ghrelin, and feeding state determines plasma ghrelin-like immunoreactivity levels in humans. J Clin Endocrinol Metab 2001;86:4753–8. [3] Bassi M, do Carmo JM, Hall JE, da Silva AA. Chronic effects of centrally administered adiponectin on appetite, metabolism and blood pressure regulation in normotensive and hypertensive rats. Peptides 2012;37:1–5. [4] Blatnik M, Soderstrom CI. A practical guide for the stabilization of acylghrelin in human blood collections. Clin Endocrinol (Oxf) 2011;74:325–31. [5] 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. [6] Flanagan DE, Evans ML, Monsod TP, Rife F, Heptulla RA, Tamborlane WV, et al. The influence of insulin on circulating ghrelin. Am J Physiol Endocrinol Metab 2003;284:313–6. [7] Gautron L, Elmquist JK. Sixteen years and counting: an update on leptin in energy balance. J Clin Invest 2011;121:2087–93. [8] Katsuki A, Urakawa H, Gabazza EC, Murashima S, Nakatani K, Togashi K, et al. Circulating levels of active ghrelin is associated with abdominal adiposity, hyperinsulinemia and insulin resistance in patients with type 2 diabetes mellitus. Eur J Endocrinol 2004;151:573–7. [9] Kiyici S, Ersoy C, Oz Gul O, Sarandol E, Demirci M, Tuncel E, et al. Total and acylated ghrelin levels in type 2 diabetic patients: similar levels observed after treatment with metformin, pioglitazone or diet therapy. Exp Clin Endocrinol Diabetes 2009;117:386–90. [10] Kusaka I, Nagasaka S, Horie H, Ishibashi S. Metformin, but not pioglitazone, decreases postchallenge plasma ghrelin levels in type 2 diabetic patients: a possible role in weight stability. Diabetes Obes Metab 2008;10:1039–46. [11] Martin CK, Gupta AK, Smith SR, Greenway FL, Han H, Bray GA. Effect of pioglitazone on energy intake and ghrelin in diabetic patients. Diabetes Care 2010;33:742–4. [12] Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and b-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412–9. [13] Saitoh Y, Liu R, Ueno H, Mizuta M, Nakazato M. Oral pioglitazone administration increases food intake through ghrelin-independent pathway in Zucker fatty rat. Diabetes Res Clin Pract 2007;77:351–6.
92
S. Taslimi et al. / Peptides 40 (2013) 89–92
[14] Sharma PK, Bhansali A, Sialy R, Malhotra S, Pandhi P. Effects of pioglitazone and metformin on plasma adiponectin in newly detected type 2 diabetes mellitus. Clin Endocrinol (Oxf) 2006;65:722–8. [15] Tsuchiya K, Akaza I, Yoshimoto T, Hirata Y. Pioglitazone improves endothelial function with increased adiponectin and high-density lipoprotein cholesterol levels in type 2 diabetes. Endocr J 2009;56:691–8.
[16] Wren AM, Seal LJ, Cohen MA, Batterham RL, Taheri S, Stanley SA, et al. Ghrelin enhances appetite and increases food intake in humans. J Clin Endocrinol Metab 2001;86:5992–5. [17] Yki-Jarvinen H. Thiazolidinediones. N Engl J Med 2004;351:1106–18.
Contents lists available at SciVerse ScienceDirect
Peptides journal homepage: www.elsevier.com/locate/peptides
Short communication
Treatment with pioglitazone is associated with decreased preprandial ghrelin levels: A randomized clinical trial夽 Shervin Taslimi a , Alireza Esteghamati a,∗ , Armin Rashidi b , Hosein Moin Tavakkoli a , Manouchehr Nakhjavani a , Abbas Kebriaee-Zadeh c a
Endocrinology and Metabolism Research Center (EMRC), Vali-Asr Hospital, Tehran University of Medical Sciences, Tehran, Iran Department of Internal Medicine, Eastern Virginia Medical School, Norfolk, VA, United States c Department of Toxicology and Pharmacology, Tehran University of Medical Sciences, Tehran, Iran b
a r t i c l e
i n f o
Article history: Received 4 December 2012 Received in revised form 20 December 2012 Accepted 20 December 2012 Available online 28 December 2012 Keywords: Glucose tolerance Metformin Leptin Ghrelin
a b s t r a c t The effects of metformin and pioglitazone on ghrelin, a physiologic regulator of appetite and food intake, have not been clearly established. In a randomized clinical trial, we randomly assigned 60 type 2 diabetic patients to either metformin (Group A; n = 30) or pioglitazone (Group B; n = 30) treatment groups. The groups were similar in their baseline characteristics. A standard fasting 75 g oral glucose tolerance test was performed at time zero before starting metformin or pioglitazone, and 3 months later. After 3 months of treatment, pioglitazone, but not metformin, was significantly associated with weight gain. Both groups experienced a significant reduction in fasting plasma glucose (p < 0.01), hemoglobin A1c (p < 0.01 in Group A and p < 0.05 in Group B), and insulin resistance (p < 0.01). The effect of metformin on preprandial ghrelin and its response to glucose challenge was not significant, while the pioglitazone group had a significant reduction in preprandial ghrelin levels after treatment (p < 0.05). The effect of pioglitazone on ghrelin was independent of changes in body weight, body mass index, glucose control, insulin resistance, and plasma insulin. In conclusion, treatment with pioglitazone is associated with a decrease in preprandial ghrelin levels and therefore, the weight gain and increased food intake related to pioglitazone use cannot be explained by its effects on ghrelin. The effect of pioglitazone on ghrelin is independent of changes in body weight, body mass index, plasma insulin, insulin resistance, or glucose control. © 2012 Elsevier Inc. All rights reserved.
1. Introduction
2. Materials and methods
Ghrelin and leptin are two intrinsic regulators of food intake and metabolism. Ghrelin, produced mainly by endocrine cells in the stomach [2], enhances appetite and increases food intake [16]. It displays an oscillatory pattern with a preprandial peak and a postprandial trough [5]. The purpose of the present study was to investigate the effects of pioglitazone and metformin on active ghrelin levels during a standard oral glucose tolerance test (OGTT) in a randomized clinical trial on patients with type 2 diabetes.
After obtaining approval from the ethics committee of Tehran University of Medical Sciences (Tehran, Iran), we conducted this prospective, randomized control trial (IRCT201102275917N1) between March 2011 and June 2011 in the Endocrinology and Metabolism Research Center (EMRC), Vali-Asr Hospital, Tehran, Iran. All patients gave written informed consent and agreed to follow the American Diabetes Association dietary advice [1]. We included a total of 60 patients with an established diagnosis of type 2 diabetes who had a hemoglobin A1c (HbA1c) of greater than 7%. All patients were drug-naïve and had not been treated with metformin or pioglitazone before. Exclusion criteria were pregnancy, diabetes diagnosed before the age of 30, known liver disease, creatinine clearance of less than 30 ml/min, and congestive heart failure. Patients were randomly assigned to either of the two groups: Group A received metformin 500 mg twice daily, while Group B received pioglitazone 30 mg daily for 3 months. Age, height, weight and blood pressure were measured at baseline and at 3 months. A standard 75 g fasting OGTT was performed at time zero before starting metformin or pioglitazone, and 3 months later. Plasma glucose, insulin and active ghrelin levels were measured
Abbreviations: OGTT, oral glucose tolerance test; EMRC, Endocrinology and Metabolism Research Center; HOMA-IR, homeostatic model assessment of insulin resistance; CV, coefficient of variation; BMI, body mass index; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; SEM, standard error of the mean; SD, standard deviation. 夽 Clinical Trial Registration Number: IRCT201102275917N1. ∗ Corresponding author at: Endocrinology and Metabolism Research Center (EMRC), Vali-Asr Hospital, Tehran University of Medical Sciences, P.O. Box 13145784, Tehran, Iran. Tel.: +98 21 88417918; fax: +98 21 64432466. E-mail address: [email protected] (A. Esteghamati). 0196-9781/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.peptides.2012.12.020
90
S. Taslimi et al. / Peptides 40 (2013) 89–92
at 0, 60 and 120 min. Leptin was measured at 0 min. An enzyme linked immunosorbent assay was used to measure plasma levels of active ghrelin (DRG Diagnostics, Marburg, Germany; inter- and intra-assay CV, 3.55% and 3.63% respectively) and leptin (DRG Diagnostics, Marburg, Germany; inter- and intra-assay CVs, 8.66% and 5.95% respectively). Sample collection for the measurement of acylghrelin was performed according to the method recommended by Blatnik and Soderstrom [4]. Briefly, blood was collected in K2 EDTA tubes and 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride (AEBSF) was added immediately after collection. The final concentration was 2 g/l. Plasma insulin was measured using an immunoradiometric assay (Immunotech, Prague, Czech Republic; inter- and intra-assay CV, 3.4% and 4.3% respectively) and its corresponding area under the curve (AUC) was calculated for the three time points as mentioned above. HOMA-IR, our measure of insulin resistance, was calculated using the formula: HOMA-IR = fasting insulin (mU/l) × fasting plasma glucose (FPG; mg/dl)/405 [12]. In this article, we use, for simplicity, the term “ghrelin” instead of active ghrelin. Differences were analyzed using a Student’s t-test and Fisher’s exact method for normal distributions, and Mann–Whitney U-test for skewed distributions. Correlations were determined using the Spearman’s rank method for variables with skewed distributions. Since ghrelin changed significantly after treatment only in the pioglitazone group, multiple regression analysis was applied to this group with preprandial ghrelin change (due to treatment) as the dependent variable and the change in variables that changed significantly with treatment as independent predictors. Data are shown as mean ± standard deviation unless indicated otherwise. p < 0.05 was considered statistically significant.
3. Results Table 1 summarizes the characteristics of the groups before and after treatment. There was no significant difference between the two groups in any of the pretreatment variables. Importantly, there was no significant difference between the groups in their pretreatment ghrelin levels. After 3 months of treatment, there was no significant weight change in Group A, while Group B experienced a statistically significant weight gain (p < 0.01), an increase in BMI (p < 0.01), and a decrease in insulin AUC (p < 0.01). In both groups, there was a significant reduction in FPG (p < 0.01), HbA1c (p < 0.01 in Group A and p < 0.05 in Group B), and insulin resistance as measured by HOMA-IR (p < 0.01). Treatment was associated with a significant decrease (p < 0.05) in leptin levels only in Group A. Pioglitazone treatment was associated with a significant reduction in preprandial ghrelin levels (p < 0.05). Ghrelin was significantly suppressed 1 h post-prandial both before (p < 0.01) and after (p < 0.01) treatment with pioglitazone. The effect of metformin on preprandial ghrelin and its response to glucose challenge was not significant (Fig. 1). Treatment in either group was not associated with any other significant change in the measured variables (Table 1). We then applied multiple regression analysis to Group B, which showed a significant change in preprandial ghrelin with treatment. The purpose of this analysis was to evaluate whether the change in preprandial ghrelin was due to a change in any other variable studied. The dependent variable in the regression model was preprandial ghrelin change with treatment. The set of independent predictors were those variables that showed a significant change with treatment. This set included changes in the following variables: HbA1c, weight, BMI, HOMA-IR, and insulin AUC. To avoid co-linearity, we included weight and BMI in separate models, and also HOMA-IR and insulin AUC in separate models. Importantly, there was no significant correlation between pretreatment ghrelin and either plasma insulin or HOMA-IR. Neither of the
Fig. 1. Ghrelin levels in a standard 75 g oral glucose tolerance test. Ghrelin levels are significantly suppressed 1 h post-challenge in both groups, both before and after treatment. Only pioglitazone decreased preprandial ghrelin levels significantly. p values in both diagrams indicate the significance of difference from preprandial levels. All values are median ± SEM.
models revealed a statistically significant predictor for preprandial ghrelin change. Therefore, the change in preprandial ghrelin with pioglitazone treatment was not due to changes in any of the other variables studied in this work. 4. Discussion Consistent with previous studies, pioglitazone was associated with weight gain in our study, while metformin did not significantly affect weight [17]. Also, treatment with metformin caused a significant decrease in leptin levels. This result is in line with previous studies showing that insulin sensitization is associated with decreased leptin levels [7]. Moreover, we showed that while metformin did not have a significant effect on preprandial ghrelin levels and post-challenge ghrelin suppression, pioglitazone significantly decreased pre-prandial ghrelin levels. The randomized design of this study eliminated any significant difference between the two groups in any of the pretreatment variables, thus making the difference in treatments the most likely explanation for a significant change in ghrelin levels in the pioglitazone as opposed to the metformin group. Consistent with previous studies, our results suggest that pioglitazone effects are not exerted via effects on leptin [14]. Our regression analysis also suggests that the effects of pioglitazone on ghrelin are not due to changes in insulin resistance, body weight, or long-term glucose control. Pioglitazone treatment in rats is associated with decreased fasting ghrelin levels [13]. A previous study on 16 Japanese type 2 diabetic patients who underwent 4 months of treatment with pioglitazone, treatment was not associated with a significant change in fasting ghrelin levels [10]. A similar study on 36 patients
S. Taslimi et al. / Peptides 40 (2013) 89–92
91
Table 1 Characteristics of the groups before and after treatment. Baseline characteristics
Age (years) Male:female Diabetes duration (months)a Weight (kg) BMI (kg/m2 ) WC (cm) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) HbA1c (%) Fasting plasma glucose (g/dl) Plasma creatinine (mg/dl) LDL (mg/dl) HDL (mg/dl) Total cholesterol (mg/dl) Triglycerides (mg/dl)a HOMA-IRa Leptin (ng/ml)a Ghrelin at time zero (pg/ml)a AUC for insulin (mU/l)a
Group A (metformin; n = 30)
Group B (pioglitazone; n = 30)
Pretreatment
Posttreatment
Pretreatment
Posttreatment
51 (10) 15:15 24 (18) 76.8 (12.5) 28.7 (5.6) 97.8 (11.2) 117 (15) 77 (9) 8.4 (1.5) 186 (67) 0.97 (0.21) 124 (33) 48 (12) 212 (40) 190 (22) 3.6 (0.8) 9.8 (1.5) 41.0 (2.5) 59.9 (12.8)
51 (10) 15:15 24 (18) 76.5 (12.3) 28.7 (5.5) 97.6 (10.9) 121 (18) 77 (10) 7.0 (1.4)†† 135 (48)†† 0.96 (0.23) 101 (26)†† 48 (12) 191 (43)† 145 (10)† 2.3 (0.4)† 7.0 (1.3)† 43.5 (2.0) 46.0 (11.1)†
56 (11) 13:17 36 (10) 73.8 (13.1) 28.4 (4) 96.3 (10.5) 123 (16) 77 (9) 8.2 (1.8) 178 (65) 1.0 (0.17) 116 (36) 49 (11) 197 (41) 149 (16) 3.5 (0.4) 9.8 (2.2) 48.5 (3.1) 52.6 (7.8)
56 (11) 13:17 36 (10) 75.5 (13.7)†† 29.0 (4.1)†† 96.9 (11.2) 123 (19) 77 (11) 7.3 (1.6)† 142 (53)†† 1.0 (0.21) 108 (25) 52 (13) 199 (29) 127 (10) 2.1 (0.2)†† 10 (1.8) 40.5 (2.6)† 44.9 (4.2)††
The two groups were not significantly different in any of the pretreatment variables. AUC, area under the curve at three time points 0, 1 h, and 2 h. a Median ± standard error of the mean; all other results are shown as mean ± SD. † p < 0.05 for comparisons between pretreatment and posttreatment in a given group. †† p < 0.01 for comparisons between pretreatment and posttreatment in a given group.
with type 2 diabetes who underwent at least 6 months of treatment with diet alone (n = 11), metformin (n = 12) or pioglitazone (n = 12), the change in fasting ghrelin levels was not significantly different between the groups [9]. The results of our study, with 30 patients in the pioglitazone group, suggest that the Japanese study was probably underpowered and fasting ghrelin levels decrease after treatment with pioglitazone. This does not mean, however, that pioglitazone decreases food intake. In fact, pioglitazone treatment results in increased food consumption in both rats [13] and humans [11]. One potential explanation for the somewhat conflicting outcomes of previous studies on the effects of pioglitazone on ghrelin may be related to the different methods used to stabilize ghrelin in collected samples. Acylghrelin is known to be rapidly converted to des-acylghrelin if blood samples are not immediately stabilized with protease inhibitors. In this study, we used the method recommended by Blatnik and Soderstrom [4], which has been shown to minimize acylghrelin loss. In conclusion, we showed in our randomized clinical trial that treatment with pioglitazone is associated with a reduction in preprandial ghrelin levels, and that this effect is not due to changes in body weight, insulin resistance or long-term glucose control. Importantly, although previous studies have established an inverse relationship between ghrelin levels and BMI [8], our results from regression analysis suggest that the effects of pioglitazone on ghrelin are not due to increased BMI. Similarly, although a negative effect on ghrelin secretion has been demonstrated for insulin in some previous studies [6], our results from regression analysis suggest that the effects of pioglitazone on ghrelin are not due to decreased plasma insulin and improved insulin sensitivity. Considering the appetite-stimulating action of ghrelin, our results suggest that increased food intake subsequent to pioglitazone treatment is not due to the effects of the drug on ghrelin. Similarly, considering the appetite-reducing action of adiponectin [3], which increases after pioglitazone treatment [15], the increased food intake subsequent to pioglitazone treatment is not due to the effects of the drug on adiponectin. This result is consistent with findings of previous studies [9,13]. Hormonal regulation of appetite and body weight is still a complicated subject with many unknown aspects that require further research.
Conflict of interest None.
References [1] American Diabetes Association, Bantle JP, Wylie-Rosett J, Albright AL, Apovian CM, Clark NG, et al. Nutrition recommendations and interventions for diabetes: a position statement of the American Diabetes Association. Diabetes Care 2008;31:S61–78. [2] Ariyasu H, Takaya K, Tagami T, Ogawa Y, Hosoda K, Akamizu T, et al. Stomach is a major source of circulating ghrelin, and feeding state determines plasma ghrelin-like immunoreactivity levels in humans. J Clin Endocrinol Metab 2001;86:4753–8. [3] Bassi M, do Carmo JM, Hall JE, da Silva AA. Chronic effects of centrally administered adiponectin on appetite, metabolism and blood pressure regulation in normotensive and hypertensive rats. Peptides 2012;37:1–5. [4] Blatnik M, Soderstrom CI. A practical guide for the stabilization of acylghrelin in human blood collections. Clin Endocrinol (Oxf) 2011;74:325–31. [5] 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. [6] Flanagan DE, Evans ML, Monsod TP, Rife F, Heptulla RA, Tamborlane WV, et al. The influence of insulin on circulating ghrelin. Am J Physiol Endocrinol Metab 2003;284:313–6. [7] Gautron L, Elmquist JK. Sixteen years and counting: an update on leptin in energy balance. J Clin Invest 2011;121:2087–93. [8] Katsuki A, Urakawa H, Gabazza EC, Murashima S, Nakatani K, Togashi K, et al. Circulating levels of active ghrelin is associated with abdominal adiposity, hyperinsulinemia and insulin resistance in patients with type 2 diabetes mellitus. Eur J Endocrinol 2004;151:573–7. [9] Kiyici S, Ersoy C, Oz Gul O, Sarandol E, Demirci M, Tuncel E, et al. Total and acylated ghrelin levels in type 2 diabetic patients: similar levels observed after treatment with metformin, pioglitazone or diet therapy. Exp Clin Endocrinol Diabetes 2009;117:386–90. [10] Kusaka I, Nagasaka S, Horie H, Ishibashi S. Metformin, but not pioglitazone, decreases postchallenge plasma ghrelin levels in type 2 diabetic patients: a possible role in weight stability. Diabetes Obes Metab 2008;10:1039–46. [11] Martin CK, Gupta AK, Smith SR, Greenway FL, Han H, Bray GA. Effect of pioglitazone on energy intake and ghrelin in diabetic patients. Diabetes Care 2010;33:742–4. [12] Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and b-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412–9. [13] Saitoh Y, Liu R, Ueno H, Mizuta M, Nakazato M. Oral pioglitazone administration increases food intake through ghrelin-independent pathway in Zucker fatty rat. Diabetes Res Clin Pract 2007;77:351–6.
92
S. Taslimi et al. / Peptides 40 (2013) 89–92
[14] Sharma PK, Bhansali A, Sialy R, Malhotra S, Pandhi P. Effects of pioglitazone and metformin on plasma adiponectin in newly detected type 2 diabetes mellitus. Clin Endocrinol (Oxf) 2006;65:722–8. [15] Tsuchiya K, Akaza I, Yoshimoto T, Hirata Y. Pioglitazone improves endothelial function with increased adiponectin and high-density lipoprotein cholesterol levels in type 2 diabetes. Endocr J 2009;56:691–8.
[16] Wren AM, Seal LJ, Cohen MA, Batterham RL, Taheri S, Stanley SA, et al. Ghrelin enhances appetite and increases food intake in humans. J Clin Endocrinol Metab 2001;86:5992–5. [17] Yki-Jarvinen H. Thiazolidinediones. N Engl J Med 2004;351:1106–18.