- Email: [email protected]
Ghrelin Ratio in Patients with Previous Myocardial Infarction
Archives of Medical Research 43 (2012) 548e554
ORIGINAL ARTICLE
Serum Ghrelin Level and TNF-a/Ghrelin Ratio in Patients with Previous Myocardial Infarction Krisztian V€ or€ os,a Zoltan Prohaszka,b Edit Kaszas,c Anna Alliquander,c Andras Terebesy,d d Sima,d Judit Forrai,d Karoly Cseh,c,d and Laszl Ferenc Horvath, Leonard Janik,d Agnes o Kalabaya a
Department of Family Medicine, b3rd Department of Internal Medicine, Semmelweis University, Budapest, Hungary c Karolyi Sandor Municipality Hospital, Budapest, Hungary d Institute of Public Health, Semmelweis University, Budapest, Hungary Received for publication February 7, 2012; accepted September 6, 2012 (ARCMED-D-12-00067).
Background and Aims. Studies investigating serum ghrelin level in atherosclerosis yielded contradictory results. Interaction of ghrelin with adipocytokines is obscure in cardiovascular diseases. We undertook this study to determine which molecules influence ghrelin level and to see whether post-myocardial infarction (MI) patients have decreased ghrelin levels. Methods. In this cross-sectional study, acyl-ghrelin concentration was determined by radioimmunoassay in sera of 171 patients (ages 62 6 years, mean SD) with previous MI and 81 age-matched referent subjects. We evaluated the associations of ghrelin with insulin, adiponectin, leptin, resistin, fetuin-A and tumor necrosis factor-alpha (TNF-a). Results. Patients had lower ghrelin levels compared to referent subjects (240.55 59.33 vs. 337.96 30.75 pg/mL, p !0.001) even after excluding diabetic and obese patients (240.63 54.08 vs. 337.96 30.75, p !0.001). In multivariate analysis, insulin (b 5 0.327, p !0.001) and adiponectin (b 5 0.301, p !0.001) determined ghrelin level (R2 5 0.199, p !0.001). There was no association between ghrelin and TNF-a levels. In discriminant analysis using ghrelin, adiponectin, leptin, fetuin-A, resistin and TNF-a, the structure matrix revealed ghrelin and TNF-a as strongest predictors for belonging to the patient group (0.760 and 0.569, respectively). Using these two parameters, 89.7% of cases were correctly classified. Subjects with high TNF-a/ghrelin ratio had 11.25 times higher chance for belonging to the patient group (95% CI 5.80e21.80; c2 (1) 5 215.6, p !0.001) Conclusions. Acylated ghrelin levels are decreased in patients with coronary atherosclerosis, independently of body weight and the presence of type 2 diabetes mellitus. Ghrelin level is determined by elevated insulin and decreased adiponectin levels. Ghrelin alone or in combination with TNF-a may prove to be a novel indicator of coronary atherosclerosis. Ó 2012 IMSS. Published by Elsevier Inc. Key Words: Ghrelin, Insulin, Adiponectin, Tumor necrosis factor-a, Myocardial infarction.
Introduction Ghrelin is an orexigenic peptide and an endogenous ligand of the growth hormone secretagogue receptor. In addition to regulating energy balance, it is involved in insulin and
Address reprint requests to: Krisztian V€or€os, Semmelweis University, Department of Family Medicine, Kutv€olgyi ut 4, Budapest H-1125, Hungary; Phone: (þ36)-1-355-8530; FAX: (þ36)-1-214-0841; E-mail: [email protected]
carbohydrate metabolism (1). Numerous in vitro and animal studies proved the protective cardiovascular effects of the protein (2). Ghrelin dilates arteries, thus lowering blood pressure, reducing cardiac afterload and improving cardiac function. Specific heart effects include attenuating interstitial fibrosis, decreasing the frequency of ventricular arrhythmias and protecting cardiomyocytes against apoptosis. Ghrelin decreases the sympathetic activity of the autonomous nervous system, leading to lower heart rate, decreased vasal tone and suppressed plasma norepinephrine level (3). Human
0188-4409/$ - see front matter. Copyright Ó 2012 IMSS. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.arcmed.2012.09.002
Ghrelin and TNF-a/Ghrelin Ratio in Atherosclerosis
studies on the relationship between atherosclerosis and ghrelin levels, however, reported contradictory results, finding both lower (4e7) and higher (8,9) serum concentrations. Interaction between ghrelin and molecules influencing its levels has not been investigated in atherosclerosis. There is a strong connection between diabetes and cardiovascular disease (10). Acylated ghrelin was shown to inhibit insulin secretion; thus, elevated levels of the protein could deteriorate normal insulin response (11). The decreased level found in diabetic patients, however, contradicts the importance of this idea (12). There are limited data on the relationship between ghrelin and adipocytokines, although several studies shed light on the possible relationship between ghrelin, adiponectin and leptin (9,13e16). Subclinical inflammatory process is involved in the pathogenesis of atherosclerosis (17). The role of ghrelin in inflammation has also been suggested. Elevated levels were found in inflammatory diseases and interaction between ghrelin and TNF-a have been reported in animal studies and in vitro (18,19). The relationship between ghrelin and TNF-a has not been extensively investigated in the clinical setting. In our cross-sectional study we determined serum acylghrelin levels in patients surviving MI. To avoid inconsistencies from diagnostic procedures of atherosclerosis, we chose patients with a history of MI, a hard endpoint of coronary atherosclerosis. We planned to see whether serum ghrelin was increased or decreased in patients with established coronary atherosclerosis. To identify the possible determinants of acyl-ghrelin, we also intended to investigate its associations with molecules characteristic of different domains of metabolism and subclinical inflammation: insulin (carbohydrate metabolism), leptin (energy balance), adiponectin and fetuin-A (fatty liver disease, glucose and lipid metabolism) and finally TNF-a and resistin (inflammatory domain). We included patients with type 2 diabetes mellitus (T2DM) to see the effect of ghrelin on insulin and glucose metabolism in coronary atherosclerosis. We also evaluated the possible role of the protein in the subclinical inflammatory process of atherosclerosis.
549
mmol/l or 2 h oral glucose tolerance test $11.1 mmol/l). Sixty-five per cent of patients with previous MI took statins and 70% of them were on aspirin therapy. Only patients with T2DM treated with diet and oral antidiabetic drugs were enrolled. Eighty-one age-matched healthy hospital staff members free from clinical, laboratory and ECG signs of cardiovascular disorders served as the referent group. All patients gave informed consent. The study was approved by the local ethics committee of the Karolyi Sandor Municipality Hospital, Budapest, Hungary. Laboratory Tests Serum ghrelin (Linco Research, St. Charles, MO, Ghrelin (active) RIA kit 125 tubes, cat. no. GHRA-88HK, specific for octanoilated (acylated) on serine 3 variant, sensitivity: 10 pg/mL, specificity: human ghrelin 100%, intra-assay precision [IACV%]: 7.425%, interassay precision [IECV %]: 13.45%, according to manufacturer’s instructions) and adiponectin levels (Linco Research, human adiponectin RIA kit 125 tubes (Cat. # HADP-61HK), sensitivity: 1 ng/ mL, IACV: 3.86%, IECV: 8.47%) were measured by radioimmunoassay. Serum resistin (Linco Research, Resistin ELISA kit, Cat. No. EZHR-95K, sensitivity: 0.16 ng/mL, specificity: human resistin 100%, IACV: 4.0%, IECV: 7.0%), TNF-a (Human TNF-a high sensitivity ELISA kit, Bender MedSystems GMBH, Vienna, Austria, cat. no. BMS223HS, sensitivity: 0.13 pg/mL, IACV: 8.5%, IECV: 9.8%), leptin (DRG International, Mountainside, NJ, cat. no. EIA-2395 sensitivity: 1 ng/mL IACV: 6.91%, IECV: 11.55 %) and insulin (BioSource, Nivelles, Belgium, ELISA kit, cat. no. KAP 1251, sensitivity: 0.17 mIU/mL, IACV: 4.8%, IECV: 8.1%) were measured by ELISA. We used radial immunodiffusion (Litex agarose gel [Sigma], antiAHSG [IgG fraction, Incstar, cat. no. 81931, 13.7 mg/mL, at a final concentration of 84 ml/11.5 mL]) for the determination of serum fetuin-A levels as described previously (20). The intra- and inter-assay coefficient variations of the test were 4.11 and 4.85%, respectively. Other measurements were performed as routine laboratory tests. Statistical Analysis
Materials and Methods We included patients with a history of MI 6e24 months prior to the initiation of the study to avoid the confounding effect of inflammation due to tissue damage. Only STEMI cases were included, diagnosed by electrocardiogram (ECG) and troponin elevation. Exclusion criteria were as follows: clinical or laboratory signs of acute infection, malignant tumor, hepatic disease, renal failure, immune suppression, severe medical or surgical conditions, i.e., MI within 6 months, stroke (at any time), trauma, or surgical procedure. Diabetes was diagnosed according to current EASD criteria (fasting plasma glucose $7.1
Statistical analysis was performed with the SPSS v.19 software (SPSS, Chicago, IL); p values !0.05 were considered significant.
Results Clinical data of patients are summarized in Table 1. Patients had higher BMI. Forty-nine of them were obese. T2DM was present in 53 patients. None of the referent subjects was obese or diabetic. The modestly elevated fasting glucose levels in patients were accompanied by pronounced elevation of insulin levels. The lipid profile of controls was
550
V€or€os et al./ Archives of Medical Research 43 (2012) 548e554
Table 1. Clinical data of referent subjects and patients with previous MI Referent subjects Number (male/female) Age (years) BMI (kg/m2) Diabetes Ghrelin (pg/mL) Serum total cholesterol (mmol/l) HDL-cholesterol (mmol/l) Serum triglycerides (mmol/l) Fasting glucose (mmol/l) Insulin (mU/mL) Adiponectin (mg/mL) Fetuin-A (mg/l) Resistin (ng/mL) TNF-a (pg/mL)
81 (46/35) 60 7 22.4 2.4 0/81 337.96 30.75 5.0 1.1 1.44 0.46 1.40 0.6 4.5 5.21 12.77 619 6.40 4.07
0.4 1.12 3.15 96 2.66 0.24
Patients
pa
171 (120/51) 0.028 62 6 NS 28.0 4.1 !0.001 53/171 !0.001 240.55 59.33 !0.001 5.29 1.2 0.030 1.21 0.28 1.85 0.98
0.037 !0.001
!0.001 !0.001 !0.001 !0.001 !0.001 !0.001
5.7 23.73 9.04 673 9.70 6.13
1.6 15.53 4.31 103 6.49 1.77
MI, myocardial infarction; BMI, body mass index; TNF, tumor necrosis factor. Values expressed as mean SD. a Mann-Whitney U test.
superior compared to that of patients, despite lipid-lowering treatment in the latter group.
Serum Ghrelin Concentration in Patients with Previous MI Patients had significantly lower ghrelin levels compared to referent subjects (Figure 1). As both diabetes and obesity are known to alter serum ghrelin levels, we compared the ghrelin levels of nonobese, nondiabetic patients and referent subjects. Concentrations of ghrelin in patients remained significantly lower after the exclusion of obese and diabetic subjects (240.63 54.08, n 5 88 vs. 337.96 30.75, n 5 81, p !0.001). Among diabetic patients, serum insulin levels were elevated compared to
Ghrelin (mean + SD, pg/mL )
500
p <0.001
400
nondiabetic patients (28.30 16.30, n 5 53 vs. 21.68 14.80, n 5 118, p 5 0.008) but there was no significant difference in serum ghrelin levels (231.06 52.9 vs. 244.81 61.7, p 5 0.154). Association of Ghrelin with Fasting Glucose, Insulin, Adiponectin, Resistin, Leptin, Fetuin-A and TNF-a We found no correlation between serum ghrelin levels and fasting glucose concentrations. In univariate analysis, ghrelin had no effect on fasting glucose levels (R2 5 0.003, b 5 0.057, p 5 0.459). The correlation between ghrelin and insulin levels was the strongest one, even after adjusting for BMI and gender (Table 2). However, there was no correlation between ghrelin and insulin levels among diabetic patients. The correlation between serum ghrelin and adiponectin concentrations weakened, whereas that between ghrelin and resistin levels lost significance after correcting for BMI and gender. There was no significant correlation between serum ghrelin levels and concentrations of fetuin-A, leptin, TNF-a or serum lipid levels. In multivariate analysis we evaluated factors determining serum ghrelin level (Table 3). Insulin and adiponectin level proved to be independent predictors of ghrelin level in a model containing age, gender, BMI, fasting glucose, insulin, triglyceride, HDL-cholesterol, adiponectin, resistin, leptin, TNF-a and fetuin-A. We conducted discriminant analysis to evaluate the predictive value of ghrelin, adiponectin, leptin, fetuin-A, resistin and TNF-a for previous MI. These proteins accounted for 57.5% of between-group variability (canonical correlation: 0.758; Wilks’ lambda 0.426, p !0.001). The crossvalidated classification showed that, overall, 90.1% (100% of referent subjects and 85.4% of patients) were correctly classified. Analysis of the structure matrix revealed ghrelin as the strongest predictor for belonging to the patient group (0.760) followed by TNF-a (0.569). Using only these two parameters, 89.7% of cases were correctly classified, whereas classification using only ghrelin discriminated Table 2. Correlations of serum ghrelin with laboratory parameters in patients with previous MI (n 5 171)
300
200
100
0 Patients n = 171
Referent subjects n = 81
Figure 1. Ghrelin concentrations in patients with previous myocardial infarction (n 5 171) and referent subjects (n 5 81) (Mann-Whitney U test).
BMI Fasting glucose Insulin Adiponectin Resistin Fetuin-A TNF-a
ra
pa
Part. corr.b
pb
0.049 0.118 0.369 0.307 0.178 0.111 0.040
0.523 0.125 !0.001 !0.001 0.020 0.150 0.605
e 0.002 0.270 0.280 0.085 0.035 0.006
e 0.978 0.001 0.001 0.317 0.679 0.941
MI, myocardial infarction; BMI, body mass index; TNF, tumor necrosis factor. a Spearman’s rank correlation, significance level. b Partial correlation adjusted for BMI, significance level.
Ghrelin and TNF-a/Ghrelin Ratio in Atherosclerosis Table 3. Determinants of serum ghrelin levels in multivariate regression model (n 5 171) Covariates Age (years) Gender (male, female) BMI Fasting glucose Insulin Triglycerides HDL-cholesterol Adiponectin Resistin Leptin TNF-a Fetuin-A
b (standardized coefficients)
p
0.065 0.002 0.115 0.025 0.327 0.038 0.066 0.301 0.108 0.063 0.071 0.028
0.375 0.975 0.147 0.743 0.001 0.609 0.379 0.001 0.188 0.601 0.388 0.697
BMI, body mass index; HDL, high-density lipoprotein; TNF, tumor necrosis factor. R2 5 0.199, p !0.001.
between the groups less effectively (overall 87.0% of cases were classified correctly). In discriminant analysis, ghrelin and TNF-a were the best predictors and their correlation was the lowest among all proteins investigated (r 5 0.017). Thus, we further analyzed discriminative value of the TNF-a/ghrelin ratio. Patients had significantly higher TNF-a/ghrelin ratio (Mann-Whitney U test, p !0.001, Figure 2). We used the highest TNF-a/ghrelin ratio among controls as the cut-off point (0.014286), thus classifying all controls to the low TNF-a/ghrelin group. One-hundred and sixty-five of the 171 post-MI patients (96.5%) had TNF-a/ghrelin ratio above the cut-off point, whereas only six of these patients belonged to the group of controls with low TNF-a/ghrelin ratio. c2 test showed a significant relationship between TNFa/ghrelin ratio and whether or not subjects belonged to the patient group (c2 [1] 5 215.6, p !0.001). The effect size was 0.932. Risk estimation revealed that subjects
TNF- α / ghrelin ratio (mean + SD, ratio)
p <0.001 0.04
0.03
0.02
0.01
0
Patients n = 171
Referent subjects n = 81
Figure 2. TNF-a/ghrelin ratio in patients with previous myocardial infarction (n 5 171) and referent subjects (n 5 81), (Mann-Whitney U test).
551
having high TNF-a/ghrelin ratio had 11.25 times (95% CI 5.80e21.80) higher chance for belonging to the patient group.
Discussion Serum ghrelin levels of patients with previous MI were significantly decreased compared to age-matched referent subjects. Most studies focusing on the association between ghrelin and atherosclerosis used a surrogate marker to determine the presence and degree of vascular sclerosis and had contradictory results. Early studies reported a positive correlation between carotid intima media thickness and serum ghrelin level (8,9). Recently, Kadoglou and coworkers reported decreased ghrelin levels in patients with coronary artery disease (CAD) (7) and several studies found a negative correlation between ghrelin and carotid intima media thickness in patients with diabetes (4), hypertension (5) and metabolic syndrome (6). Our results in patients with established CAD support the idea that ghrelin level is decreased in coronary atherosclerosis. In line with the clinical situation, our study population was heterogeneous, including numerous patients with obesity and diabetes. Lower ghrelin concentrations observed in these pro-atherosclerotic states could explain our finding (12,21). However, even after excluding diabetic and obese patients from the analysis, we found lower ghrelin levels in patients with CAD. Thus, it is possible that ghrelin levels are decreased in a wide range of CAD patients and low concentration of the protein may be an early sign of atherosclerosis. We investigated the relationship between ghrelin, insulin and adipocytokines to see which of these molecules may determine ghrelin level and contribute to its lower concentration in CAD patients. Insulin level was the strongest and independent determinant of ghrelin level. The effect of ghrelin on carbohydrate metabolism has widely been studied but results are somewhat contradictory. It has been shown that acyl-ghrelin can inhibit insulin secretion, which may explain the negative correlation that we found (11). The biological role of ghrelin as inhibitor of insulin secretion has been questioned because of the extremely high concentrations needed to blunt insulin response (22). Moreover, if ghrelin deteriorates insulin secretion and is thus a causative agent in impaired insulin and glucose metabolism, one would expect higher ghrelin concentrations in patients with diabetes, obesity or cardiovascular disease. Our finding of lower ghrelin concentration in CAD patients and studies showing decreased levels in obese and diabetic patients makes it unlikely that the inhibition of insulin secretion by ghrelin lies behind this association (12,21,23). An alternative explanation for the negative correlation we observed may be that the insulin resistant state and the accompanying elevated blood sugar
552
V€or€os et al./ Archives of Medical Research 43 (2012) 548e554
and insulin levels decrease serum ghrelin level. It has been shown that both insulin (24) and glucose load (25) can lower serum ghrelin level. Moreover, the ghrelin O-acyltransferase enzyme responsible for the activation of ghrelin is also inhibited by insulin (26), which could also be implicated in our finding of decreased acyl-ghrelin level. Insulin levels were elevated in diabetic compared to nondiabetic patients. However, this difference was not accompanied by lower ghrelin levels in the former group, and there was also no correlation between ghrelin and insulin in diabetic patients. In two studies it has been shown that insulin decreases ghrelin levels, but this association is modified by obesity and diabetes (27,28). In subjects with initially high ghrelin level, Sondegaard et al. found a marked decrease in ghrelin level during hyperinsulinemic-euglycemic clamp, whereas in obese subjects with initially lower ghrelin level, the decrease was smaller (27). Eventually, both groups reached similar ghrelin levels. In diabetic patients the decrease of ghrelin level to insulin infusion was also smaller compared to healthy controls (28). We hypothesize that insulin can decrease ghrelin levels only to a certain extent, until the individually lowest ghrelin level is reached. When this lowest ghrelin level is reached, further elevation of insulin level is not accompanied by an additional drop in ghrelin level. We assume that the already elevated insulin level in normal weight, nondiabetic patients is sufficient to maximally suppress ghrelin level, and the further elevated insulin level in diabetic patients could not lead to even lower ghrelin levels. We found a positive association between adiponectin and ghrelin even after adjusting for body mass index (BMI) and gender. Little is known about the interaction of ghrelin and adiponectin. A positive correlation has been reported in postmenopausal women (29) and in obese women without insulin resistance (13). Intravenous administration of unacylated and acylated ghrelin had no effect (14) or decreased adiponectin levels (30). According to our knowledge, the effect of adiponectin on ghrelin level has not yet been investigated. Cardiomyocytes and epicardial adipose tissue is likely to have a direct impact on atherosclerosis by producing certain adipokines and cytokines, e.g., adiponectin. Studies on cultures of these cells could evaluate the role of adipokines of this origin. Based on previous findings and our findings, it is possible that in addition to elevated insulin, decreased adiponectin level also contributes to low ghrelin concentration in CAD patients. We found no association between ghrelin and leptin levels. Although Ukkola et al. found a negative correlation between ghrelin and leptin (9), others did not (15,31). Erdman and coworkers found a negative correlation in normoinsulinemic patients but found no correlation in hyperinsulinemic subjects with obesity or T2DM. These results are in line with our observation that, in our patients, insulin but not leptin determined ghrelin levels.
Fetuin-A has been shown to be associated with adiponectin level in patients with atherosclerosis (32,33). We found no correlation between serum ghrelin and fetuin-A levels. Serum lipids were also not associated with ghrelin level. Several studies described a positive correlation with HDLcholesterol levels (34) and a negative correlation with triglycerides (12,34,35). In vitro, blocking lipolysis (36) and promoting triglyceride uptake by adipocytes (36,37) were suggested mechanisms. The correlation with triglycerides, however, was not independent and no association with cholesterol levels was reported. Our finding of no correlation has to be interpreted cautiously because 65% of our patients were on lipid-lowering treatment. Subclinical inflammation is known to be an important factor in the progression of atherosclerosis (17). The acute inflammatory phase of MI usually subsides within 2e3 weeks (38). The healing phase initiates after 4e6 weeks (39) and after 8 weeks tissue regeneration and remodeling is dominant, with fibroblast activity (40). We enrolled patients 6e24 months after MI so we did not expect considerable inflammation due to tissue damage at this point. However, subclinical inflammation was present and we found elevated TNF-a levels in our patients. Acylated ghrelin was also implicated in inflammation regulation. Elevated levels were found in humans with chronic inflammatory diseases (41,42). Animal and in vitro studies provided evidence that acylated ghrelin can blunt the inflammation induced by TNF-a or other pro-inflammatory substances and decrease the expression of TNF-a and some other interleukins (18,19). Taken together, the elevated levels of ghrelin in inflammatory diseases and its anti-inflammatory effects may represent a defense mechanism against chronic inflammation. However, we could not detect any significant relationship between ghrelin and TNF-a level. Considering this independency, we analyzed whether the TNF-a/ghrelin ratio was a helpful parameter to discriminate between patients with atherosclerosis and referent subjects. Patients having low ghrelin and elevated TNF-a level (high TNF-a/ghrelin ratio) had more than 11 times higher chance for belonging to the patient group. These two proteins and their ratio may be useful novel indicators of coronary atherosclerosis in the clinical setting. It seems promising to determine the predictive power of these indicators. Our patients were on preventive medications according to guidelines. Lipid-lowering and antidiabetic drugs were reported not to alter or to elevate serum ghrelin levels (43,44). The cross-sectional design and relatively small number of patients are the major limitations of our study. Cardiac remodeling may have altered ghrelin level, although we found no data in the literature regarding this issue. In conclusion, our data support the concept that acylated ghrelin levels are decreased in patients with coronary
Ghrelin and TNF-a/Ghrelin Ratio in Atherosclerosis
atherosclerosis, independent of body weight and the presence of T2DM. Ghrelin levels are determined by elevated insulin and decreased adiponectin levels in patients with previous MI. Whether ghrelin alone or in combination with TNF-a may prove to be a novel indicator of coronary atherosclerosis warrants further studies.
Acknowledgments This work was supported by grants OTKA KON T046427 and ETT368/2009 of the Hungarian Ministry of Health. The authors declare that they have no conflicts of interest.
References 1. Pulkkinen L, Ukkola O, Kolehmainen M, et al. Ghrelin in diabetes and metabolic syndrome. Int J Pept 2010;2010. http://dx.doi.org/10.1155/ 2010/248948. 2. Zhang G, Yin X, Qi Y, et al. Ghrelin and cardiovascular diseases. Curr Cardiol Rev 2010;6:62e70. 3. Kishimoto I, Tokudome T, Hosoda H, et al. Ghrelin and cardiovascular diseases. J Cardiol 2012;59:8e13. 4. Kadoglou NP, Sailer N, Moumtzouoglou A, et al. Visfatin (nampt) and ghrelin as novel markers of carotid atherosclerosis in patients with type 2 diabetes. Exp Clin Endocrinol Diabetes 2010;118:75e80. 5. Yano Y, Toshinai K, Inokuchi T, et al. Plasma des-acyl ghrelin, but not plasma HMW adiponectin, is a useful cardiometabolic marker for predicting atherosclerosis in elderly hypertensive patients. Atherosclerosis 2009;204:590e594. 6. Kotani K, Sakane N, Saiga K, et al. Serum ghrelin and carotid atherosclerosis in older Japanese people with metabolic syndrome. Arch Med Res 2006;37:903e906. 7. Kadoglou NP, Lampropoulos S, Kapelouzou A, et al. Serum levels of apelin and ghrelin in patients with acute coronary syndromes and established coronary artery diseaseeKOZANI STUDY. Transl Res 2010; 155:238e246. 8. Poykko SM, Kellokoski E, Ukkola O, et al. Plasma ghrelin concentrations are positively associated with carotid artery atherosclerosis in males. J Intern Med 2006;260:43e52. 9. Ukkola O, Poykko S, Paivansalo M, et al. Interactions between ghrelin, leptin and IGF-I affect metabolic syndrome and early atherosclerosis. Ann Med 2008;40:465e473. 10. Haffner SM, Lehto S, Ronnemaa T, et al. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med 1998;339:229e234. 11. Tong J, Prigeon RL, Davis HW, et al. Ghrelin suppresses glucosestimulated insulin secretion and deteriorates glucose tolerance in healthy humans. Diabetes 2010;59:2145e2151. 12. Poykko SM, Kellokoski E, Horkko S, et al. Low plasma ghrelin is associated with insulin resistance, hypertension, and the prevalence of type 2 diabetes. Diabetes 2003;52:2546e2553. 13. Krzyzanowska-Swiniarska B, Kempa A, Miazgowski T, et al. Serum acylated ghrelin, adiponectin and leptin levels in normal-weight and obese premenopausal women. Horm Metab Res 2007;39:835e839. 14. Kiewiet RM, Hazell MJ, Van Aken MO, et al. Acute effects of acylated and unacylated ghrelin on total and high molecular weight adiponectin in morbidly obese subjects. J Endocrinol Invest 2011;34: 434e438. 15. Caliskan Y, Gorgulu N, Yelken B, et al. Plasma ghrelin levels are associated with coronary microvascular and endothelial dysfunction in peritoneal dialysis patients. Ren Fail 2009;31:807e813.
553
16. Erdmann J, Lippl F, Wagenpfeil S, et al. Differential association of basal and postprandial plasma ghrelin with leptin, insulin, and type 2 diabetes. Diabetes 2005;54:1371e1378. 17. Corrado E, Rizzo M, Coppola G, et al. An update on the role of markers of inflammation in atherosclerosis. J Atheroscler Thromb 2010;17:1e11. 18. Hattori N. Expression, regulation and biological actions of growth hormone (GH) and ghrelin in the immune system. Growth Horm IGF Res 2009;19:187e197. 19. Baatar D, Patel K, Taub DD. The effects of ghrelin on inflammation and the immune system. Mol Cell Endocrinol 2011;340:44e58. 20. Kalabay L, Jakab L, Prohaszka Z, et al. Human fetuin/alpha2HSglycoprotein level as a novel indicator of liver cell function and short-term mortality in patients with liver cirrhosis and liver cancer. Eur J Gastroenterol Hepatol 2002;14:389e394. 21. Tschop M, Weyer C, Tataranni PA, et al. Circulating ghrelin levels are decreased in human obesity. Diabetes 2001;50:707e709. 22. Meyer C. Final answer: ghrelin can suppress insulin secretion in humans, but is it clinically relevant? Diabetes 2010;59:2726e2728. 23. Palik E, Baranyi E, Melczer Z, et al. Elevated serum acylated (biologically active) ghrelin and resistin levels associate with pregnancyinduced weight gain and insulin resistance. Diabetes Res Clin Pract 2007;76:351e357. 24. Saad MF, Bernaba B, Hwu CM, et al. Insulin regulates plasma ghrelin concentration. J Clin Endocrinol Metab 2002;87:3997e4000. 25. Broglio F, Gottero C, Prodam F, et al. Ghrelin secretion is inhibited by glucose load and insulin-induced hypoglycaemia but unaffected by glucagon and arginine in humans. Clin Endocrinol 2004;61:503e509. 26. An W, Li Y, Xu G, et al. Modulation of ghrelin O-acyltransferase expression in pancreatic islets. Cell Physiol Biochem 2010;26: 707e716. 27. Sondergaard E, Gormsen LC, Nellemann B, et al. Visceral fat mass is a strong predictor of circulating ghrelin levels in premenopausal women. Eur J Endocrinol 2009;160:375e379. 28. Anderwald C, Brabant G, Bernroider E, et al. Insulin-dependent modulation of plasma ghrelin and leptin concentrations is less pronounced in type 2 diabetic patients. Diabetes 2003;52:1792e1798. 29. Soni AC, Conroy MB, Mackey RH, et al. Ghrelin, leptin, adiponectin, and insulin levels and concurrent and future weight change in overweight, postmenopausal women. Menopause 2011;18:296e301. 30. Ott V, Fasshauer M, Dalski A, et al. Direct peripheral effects of ghrelin include suppression of adiponectin expression. Horm Metab Res 2002; 34:640e645. 31. Jiang ZT, Zou YP, Huang H, et al. Mechanism of laparoscopic adjustable gastric banding in the treatment of obesity with type 2 diabetes mellitus. Zhonghua Wei Chang Wai Ke Za Zhi 2010;13:520e523. 32. Ix JH, Sharma K. Mechanisms linking obesity, chronic kidney disease, and fatty liver disease: the roles of fetuin-A, adiponectin, and AMPK. J Am Soc Nephrol 2010;21:406e412. 33. Voros K, Graf L Jr, Prohaszka Z, et al. Serum fetuin-A in metabolic and inflammatory pathways in patients with myocardial infarction. Eur J Clin Invest 2011;41:703e709. 34. Nanjo Y, Adachi H, Hirai Y, et al. Factors associated with plasma ghrelin level in Japanese general population. Clin Endocrinol 2011; 74:453e458. 35. Palik E, Birkas KD, Faludi G, et al. Correlation of serum ghrelin levels with body mass index and carbohydrate metabolism in patients treated with atypical antipsychotics. Diabetes Res Clin Pract 2005;68(suppl 1):S60eS64. 36. Miegueu P, St Pierre D, Broglio F, et al. Effect of desacyl ghrelin, obestatin and related peptides on triglyceride storage, metabolism and GHSR signaling in 3T3-L1 adipocytes. J Cell Biochem 2011;112:704e714. 37. Theander-Carrillo C, Wiedmer P, Cettour-Rose P, et al. Ghrelin action in the brain controls adipocyte metabolism. J Clin Invest 2006;116: 1983e1993.
554
V€or€os et al./ Archives of Medical Research 43 (2012) 548e554
38. Bonvini Rf HT, Camenzind E. Inflammatory response post-myocardial infarction and reperfusion: a new therapeutic target? Eur Heart J Suppl 2005;7(suppl I):127e136. 39. Liehn EA, Postea O, Curaj A, et al. Repair after myocardial infarction, between fantasy and reality: the role of chemokines. J Am Coll Cardiol 2011;58:2357e2362. 40. Frangogiannis NG, Smith CW, Entman ML. The inflammatory response in myocardial infarction. Cardiovasc Res 2002;53:31e47. 41. Karmiris K, Koutroubakis IE, Xidakis C, et al. Circulating levels of leptin, adiponectin, resistin, and ghrelin in inflammatory bowel disease. Inflamm Bowel Dis 2006;12:100e105.
42. Toussirot E, Streit G, Nguyen NU, et al. Adipose tissue, serum adipokines, and ghrelin in patients with ankylosing spondylitis. Metabolism 2007;56:1383e1399. 43. Kadoglou NP, Tsanikidis H, Kapelouzou A, et al. Effects of rosiglitazone and metformin treatment on apelin, visfatin, and ghrelin levels in patients with type 2 diabetes mellitus. Metabolism 2010; 59:373e379. 44. Otto C, Otto B, Frost RJ, et al. Short-term therapy with atorvastatin or fenofibrate does not affect plasma ghrelin, resistin or adiponectin levels in type 2 diabetic patients with mixed hyperlipoproteinaemia. Acta Diabetol 2007;44:65e68.
ORIGINAL ARTICLE
Serum Ghrelin Level and TNF-a/Ghrelin Ratio in Patients with Previous Myocardial Infarction Krisztian V€ or€ os,a Zoltan Prohaszka,b Edit Kaszas,c Anna Alliquander,c Andras Terebesy,d d Sima,d Judit Forrai,d Karoly Cseh,c,d and Laszl Ferenc Horvath, Leonard Janik,d Agnes o Kalabaya a
Department of Family Medicine, b3rd Department of Internal Medicine, Semmelweis University, Budapest, Hungary c Karolyi Sandor Municipality Hospital, Budapest, Hungary d Institute of Public Health, Semmelweis University, Budapest, Hungary Received for publication February 7, 2012; accepted September 6, 2012 (ARCMED-D-12-00067).
Background and Aims. Studies investigating serum ghrelin level in atherosclerosis yielded contradictory results. Interaction of ghrelin with adipocytokines is obscure in cardiovascular diseases. We undertook this study to determine which molecules influence ghrelin level and to see whether post-myocardial infarction (MI) patients have decreased ghrelin levels. Methods. In this cross-sectional study, acyl-ghrelin concentration was determined by radioimmunoassay in sera of 171 patients (ages 62 6 years, mean SD) with previous MI and 81 age-matched referent subjects. We evaluated the associations of ghrelin with insulin, adiponectin, leptin, resistin, fetuin-A and tumor necrosis factor-alpha (TNF-a). Results. Patients had lower ghrelin levels compared to referent subjects (240.55 59.33 vs. 337.96 30.75 pg/mL, p !0.001) even after excluding diabetic and obese patients (240.63 54.08 vs. 337.96 30.75, p !0.001). In multivariate analysis, insulin (b 5 0.327, p !0.001) and adiponectin (b 5 0.301, p !0.001) determined ghrelin level (R2 5 0.199, p !0.001). There was no association between ghrelin and TNF-a levels. In discriminant analysis using ghrelin, adiponectin, leptin, fetuin-A, resistin and TNF-a, the structure matrix revealed ghrelin and TNF-a as strongest predictors for belonging to the patient group (0.760 and 0.569, respectively). Using these two parameters, 89.7% of cases were correctly classified. Subjects with high TNF-a/ghrelin ratio had 11.25 times higher chance for belonging to the patient group (95% CI 5.80e21.80; c2 (1) 5 215.6, p !0.001) Conclusions. Acylated ghrelin levels are decreased in patients with coronary atherosclerosis, independently of body weight and the presence of type 2 diabetes mellitus. Ghrelin level is determined by elevated insulin and decreased adiponectin levels. Ghrelin alone or in combination with TNF-a may prove to be a novel indicator of coronary atherosclerosis. Ó 2012 IMSS. Published by Elsevier Inc. Key Words: Ghrelin, Insulin, Adiponectin, Tumor necrosis factor-a, Myocardial infarction.
Introduction Ghrelin is an orexigenic peptide and an endogenous ligand of the growth hormone secretagogue receptor. In addition to regulating energy balance, it is involved in insulin and
Address reprint requests to: Krisztian V€or€os, Semmelweis University, Department of Family Medicine, Kutv€olgyi ut 4, Budapest H-1125, Hungary; Phone: (þ36)-1-355-8530; FAX: (þ36)-1-214-0841; E-mail: [email protected]
carbohydrate metabolism (1). Numerous in vitro and animal studies proved the protective cardiovascular effects of the protein (2). Ghrelin dilates arteries, thus lowering blood pressure, reducing cardiac afterload and improving cardiac function. Specific heart effects include attenuating interstitial fibrosis, decreasing the frequency of ventricular arrhythmias and protecting cardiomyocytes against apoptosis. Ghrelin decreases the sympathetic activity of the autonomous nervous system, leading to lower heart rate, decreased vasal tone and suppressed plasma norepinephrine level (3). Human
0188-4409/$ - see front matter. Copyright Ó 2012 IMSS. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.arcmed.2012.09.002
Ghrelin and TNF-a/Ghrelin Ratio in Atherosclerosis
studies on the relationship between atherosclerosis and ghrelin levels, however, reported contradictory results, finding both lower (4e7) and higher (8,9) serum concentrations. Interaction between ghrelin and molecules influencing its levels has not been investigated in atherosclerosis. There is a strong connection between diabetes and cardiovascular disease (10). Acylated ghrelin was shown to inhibit insulin secretion; thus, elevated levels of the protein could deteriorate normal insulin response (11). The decreased level found in diabetic patients, however, contradicts the importance of this idea (12). There are limited data on the relationship between ghrelin and adipocytokines, although several studies shed light on the possible relationship between ghrelin, adiponectin and leptin (9,13e16). Subclinical inflammatory process is involved in the pathogenesis of atherosclerosis (17). The role of ghrelin in inflammation has also been suggested. Elevated levels were found in inflammatory diseases and interaction between ghrelin and TNF-a have been reported in animal studies and in vitro (18,19). The relationship between ghrelin and TNF-a has not been extensively investigated in the clinical setting. In our cross-sectional study we determined serum acylghrelin levels in patients surviving MI. To avoid inconsistencies from diagnostic procedures of atherosclerosis, we chose patients with a history of MI, a hard endpoint of coronary atherosclerosis. We planned to see whether serum ghrelin was increased or decreased in patients with established coronary atherosclerosis. To identify the possible determinants of acyl-ghrelin, we also intended to investigate its associations with molecules characteristic of different domains of metabolism and subclinical inflammation: insulin (carbohydrate metabolism), leptin (energy balance), adiponectin and fetuin-A (fatty liver disease, glucose and lipid metabolism) and finally TNF-a and resistin (inflammatory domain). We included patients with type 2 diabetes mellitus (T2DM) to see the effect of ghrelin on insulin and glucose metabolism in coronary atherosclerosis. We also evaluated the possible role of the protein in the subclinical inflammatory process of atherosclerosis.
549
mmol/l or 2 h oral glucose tolerance test $11.1 mmol/l). Sixty-five per cent of patients with previous MI took statins and 70% of them were on aspirin therapy. Only patients with T2DM treated with diet and oral antidiabetic drugs were enrolled. Eighty-one age-matched healthy hospital staff members free from clinical, laboratory and ECG signs of cardiovascular disorders served as the referent group. All patients gave informed consent. The study was approved by the local ethics committee of the Karolyi Sandor Municipality Hospital, Budapest, Hungary. Laboratory Tests Serum ghrelin (Linco Research, St. Charles, MO, Ghrelin (active) RIA kit 125 tubes, cat. no. GHRA-88HK, specific for octanoilated (acylated) on serine 3 variant, sensitivity: 10 pg/mL, specificity: human ghrelin 100%, intra-assay precision [IACV%]: 7.425%, interassay precision [IECV %]: 13.45%, according to manufacturer’s instructions) and adiponectin levels (Linco Research, human adiponectin RIA kit 125 tubes (Cat. # HADP-61HK), sensitivity: 1 ng/ mL, IACV: 3.86%, IECV: 8.47%) were measured by radioimmunoassay. Serum resistin (Linco Research, Resistin ELISA kit, Cat. No. EZHR-95K, sensitivity: 0.16 ng/mL, specificity: human resistin 100%, IACV: 4.0%, IECV: 7.0%), TNF-a (Human TNF-a high sensitivity ELISA kit, Bender MedSystems GMBH, Vienna, Austria, cat. no. BMS223HS, sensitivity: 0.13 pg/mL, IACV: 8.5%, IECV: 9.8%), leptin (DRG International, Mountainside, NJ, cat. no. EIA-2395 sensitivity: 1 ng/mL IACV: 6.91%, IECV: 11.55 %) and insulin (BioSource, Nivelles, Belgium, ELISA kit, cat. no. KAP 1251, sensitivity: 0.17 mIU/mL, IACV: 4.8%, IECV: 8.1%) were measured by ELISA. We used radial immunodiffusion (Litex agarose gel [Sigma], antiAHSG [IgG fraction, Incstar, cat. no. 81931, 13.7 mg/mL, at a final concentration of 84 ml/11.5 mL]) for the determination of serum fetuin-A levels as described previously (20). The intra- and inter-assay coefficient variations of the test were 4.11 and 4.85%, respectively. Other measurements were performed as routine laboratory tests. Statistical Analysis
Materials and Methods We included patients with a history of MI 6e24 months prior to the initiation of the study to avoid the confounding effect of inflammation due to tissue damage. Only STEMI cases were included, diagnosed by electrocardiogram (ECG) and troponin elevation. Exclusion criteria were as follows: clinical or laboratory signs of acute infection, malignant tumor, hepatic disease, renal failure, immune suppression, severe medical or surgical conditions, i.e., MI within 6 months, stroke (at any time), trauma, or surgical procedure. Diabetes was diagnosed according to current EASD criteria (fasting plasma glucose $7.1
Statistical analysis was performed with the SPSS v.19 software (SPSS, Chicago, IL); p values !0.05 were considered significant.
Results Clinical data of patients are summarized in Table 1. Patients had higher BMI. Forty-nine of them were obese. T2DM was present in 53 patients. None of the referent subjects was obese or diabetic. The modestly elevated fasting glucose levels in patients were accompanied by pronounced elevation of insulin levels. The lipid profile of controls was
550
V€or€os et al./ Archives of Medical Research 43 (2012) 548e554
Table 1. Clinical data of referent subjects and patients with previous MI Referent subjects Number (male/female) Age (years) BMI (kg/m2) Diabetes Ghrelin (pg/mL) Serum total cholesterol (mmol/l) HDL-cholesterol (mmol/l) Serum triglycerides (mmol/l) Fasting glucose (mmol/l) Insulin (mU/mL) Adiponectin (mg/mL) Fetuin-A (mg/l) Resistin (ng/mL) TNF-a (pg/mL)
81 (46/35) 60 7 22.4 2.4 0/81 337.96 30.75 5.0 1.1 1.44 0.46 1.40 0.6 4.5 5.21 12.77 619 6.40 4.07
0.4 1.12 3.15 96 2.66 0.24
Patients
pa
171 (120/51) 0.028 62 6 NS 28.0 4.1 !0.001 53/171 !0.001 240.55 59.33 !0.001 5.29 1.2 0.030 1.21 0.28 1.85 0.98
0.037 !0.001
!0.001 !0.001 !0.001 !0.001 !0.001 !0.001
5.7 23.73 9.04 673 9.70 6.13
1.6 15.53 4.31 103 6.49 1.77
MI, myocardial infarction; BMI, body mass index; TNF, tumor necrosis factor. Values expressed as mean SD. a Mann-Whitney U test.
superior compared to that of patients, despite lipid-lowering treatment in the latter group.
Serum Ghrelin Concentration in Patients with Previous MI Patients had significantly lower ghrelin levels compared to referent subjects (Figure 1). As both diabetes and obesity are known to alter serum ghrelin levels, we compared the ghrelin levels of nonobese, nondiabetic patients and referent subjects. Concentrations of ghrelin in patients remained significantly lower after the exclusion of obese and diabetic subjects (240.63 54.08, n 5 88 vs. 337.96 30.75, n 5 81, p !0.001). Among diabetic patients, serum insulin levels were elevated compared to
Ghrelin (mean + SD, pg/mL )
500
p <0.001
400
nondiabetic patients (28.30 16.30, n 5 53 vs. 21.68 14.80, n 5 118, p 5 0.008) but there was no significant difference in serum ghrelin levels (231.06 52.9 vs. 244.81 61.7, p 5 0.154). Association of Ghrelin with Fasting Glucose, Insulin, Adiponectin, Resistin, Leptin, Fetuin-A and TNF-a We found no correlation between serum ghrelin levels and fasting glucose concentrations. In univariate analysis, ghrelin had no effect on fasting glucose levels (R2 5 0.003, b 5 0.057, p 5 0.459). The correlation between ghrelin and insulin levels was the strongest one, even after adjusting for BMI and gender (Table 2). However, there was no correlation between ghrelin and insulin levels among diabetic patients. The correlation between serum ghrelin and adiponectin concentrations weakened, whereas that between ghrelin and resistin levels lost significance after correcting for BMI and gender. There was no significant correlation between serum ghrelin levels and concentrations of fetuin-A, leptin, TNF-a or serum lipid levels. In multivariate analysis we evaluated factors determining serum ghrelin level (Table 3). Insulin and adiponectin level proved to be independent predictors of ghrelin level in a model containing age, gender, BMI, fasting glucose, insulin, triglyceride, HDL-cholesterol, adiponectin, resistin, leptin, TNF-a and fetuin-A. We conducted discriminant analysis to evaluate the predictive value of ghrelin, adiponectin, leptin, fetuin-A, resistin and TNF-a for previous MI. These proteins accounted for 57.5% of between-group variability (canonical correlation: 0.758; Wilks’ lambda 0.426, p !0.001). The crossvalidated classification showed that, overall, 90.1% (100% of referent subjects and 85.4% of patients) were correctly classified. Analysis of the structure matrix revealed ghrelin as the strongest predictor for belonging to the patient group (0.760) followed by TNF-a (0.569). Using only these two parameters, 89.7% of cases were correctly classified, whereas classification using only ghrelin discriminated Table 2. Correlations of serum ghrelin with laboratory parameters in patients with previous MI (n 5 171)
300
200
100
0 Patients n = 171
Referent subjects n = 81
Figure 1. Ghrelin concentrations in patients with previous myocardial infarction (n 5 171) and referent subjects (n 5 81) (Mann-Whitney U test).
BMI Fasting glucose Insulin Adiponectin Resistin Fetuin-A TNF-a
ra
pa
Part. corr.b
pb
0.049 0.118 0.369 0.307 0.178 0.111 0.040
0.523 0.125 !0.001 !0.001 0.020 0.150 0.605
e 0.002 0.270 0.280 0.085 0.035 0.006
e 0.978 0.001 0.001 0.317 0.679 0.941
MI, myocardial infarction; BMI, body mass index; TNF, tumor necrosis factor. a Spearman’s rank correlation, significance level. b Partial correlation adjusted for BMI, significance level.
Ghrelin and TNF-a/Ghrelin Ratio in Atherosclerosis Table 3. Determinants of serum ghrelin levels in multivariate regression model (n 5 171) Covariates Age (years) Gender (male, female) BMI Fasting glucose Insulin Triglycerides HDL-cholesterol Adiponectin Resistin Leptin TNF-a Fetuin-A
b (standardized coefficients)
p
0.065 0.002 0.115 0.025 0.327 0.038 0.066 0.301 0.108 0.063 0.071 0.028
0.375 0.975 0.147 0.743 0.001 0.609 0.379 0.001 0.188 0.601 0.388 0.697
BMI, body mass index; HDL, high-density lipoprotein; TNF, tumor necrosis factor. R2 5 0.199, p !0.001.
between the groups less effectively (overall 87.0% of cases were classified correctly). In discriminant analysis, ghrelin and TNF-a were the best predictors and their correlation was the lowest among all proteins investigated (r 5 0.017). Thus, we further analyzed discriminative value of the TNF-a/ghrelin ratio. Patients had significantly higher TNF-a/ghrelin ratio (Mann-Whitney U test, p !0.001, Figure 2). We used the highest TNF-a/ghrelin ratio among controls as the cut-off point (0.014286), thus classifying all controls to the low TNF-a/ghrelin group. One-hundred and sixty-five of the 171 post-MI patients (96.5%) had TNF-a/ghrelin ratio above the cut-off point, whereas only six of these patients belonged to the group of controls with low TNF-a/ghrelin ratio. c2 test showed a significant relationship between TNFa/ghrelin ratio and whether or not subjects belonged to the patient group (c2 [1] 5 215.6, p !0.001). The effect size was 0.932. Risk estimation revealed that subjects
TNF- α / ghrelin ratio (mean + SD, ratio)
p <0.001 0.04
0.03
0.02
0.01
0
Patients n = 171
Referent subjects n = 81
Figure 2. TNF-a/ghrelin ratio in patients with previous myocardial infarction (n 5 171) and referent subjects (n 5 81), (Mann-Whitney U test).
551
having high TNF-a/ghrelin ratio had 11.25 times (95% CI 5.80e21.80) higher chance for belonging to the patient group.
Discussion Serum ghrelin levels of patients with previous MI were significantly decreased compared to age-matched referent subjects. Most studies focusing on the association between ghrelin and atherosclerosis used a surrogate marker to determine the presence and degree of vascular sclerosis and had contradictory results. Early studies reported a positive correlation between carotid intima media thickness and serum ghrelin level (8,9). Recently, Kadoglou and coworkers reported decreased ghrelin levels in patients with coronary artery disease (CAD) (7) and several studies found a negative correlation between ghrelin and carotid intima media thickness in patients with diabetes (4), hypertension (5) and metabolic syndrome (6). Our results in patients with established CAD support the idea that ghrelin level is decreased in coronary atherosclerosis. In line with the clinical situation, our study population was heterogeneous, including numerous patients with obesity and diabetes. Lower ghrelin concentrations observed in these pro-atherosclerotic states could explain our finding (12,21). However, even after excluding diabetic and obese patients from the analysis, we found lower ghrelin levels in patients with CAD. Thus, it is possible that ghrelin levels are decreased in a wide range of CAD patients and low concentration of the protein may be an early sign of atherosclerosis. We investigated the relationship between ghrelin, insulin and adipocytokines to see which of these molecules may determine ghrelin level and contribute to its lower concentration in CAD patients. Insulin level was the strongest and independent determinant of ghrelin level. The effect of ghrelin on carbohydrate metabolism has widely been studied but results are somewhat contradictory. It has been shown that acyl-ghrelin can inhibit insulin secretion, which may explain the negative correlation that we found (11). The biological role of ghrelin as inhibitor of insulin secretion has been questioned because of the extremely high concentrations needed to blunt insulin response (22). Moreover, if ghrelin deteriorates insulin secretion and is thus a causative agent in impaired insulin and glucose metabolism, one would expect higher ghrelin concentrations in patients with diabetes, obesity or cardiovascular disease. Our finding of lower ghrelin concentration in CAD patients and studies showing decreased levels in obese and diabetic patients makes it unlikely that the inhibition of insulin secretion by ghrelin lies behind this association (12,21,23). An alternative explanation for the negative correlation we observed may be that the insulin resistant state and the accompanying elevated blood sugar
552
V€or€os et al./ Archives of Medical Research 43 (2012) 548e554
and insulin levels decrease serum ghrelin level. It has been shown that both insulin (24) and glucose load (25) can lower serum ghrelin level. Moreover, the ghrelin O-acyltransferase enzyme responsible for the activation of ghrelin is also inhibited by insulin (26), which could also be implicated in our finding of decreased acyl-ghrelin level. Insulin levels were elevated in diabetic compared to nondiabetic patients. However, this difference was not accompanied by lower ghrelin levels in the former group, and there was also no correlation between ghrelin and insulin in diabetic patients. In two studies it has been shown that insulin decreases ghrelin levels, but this association is modified by obesity and diabetes (27,28). In subjects with initially high ghrelin level, Sondegaard et al. found a marked decrease in ghrelin level during hyperinsulinemic-euglycemic clamp, whereas in obese subjects with initially lower ghrelin level, the decrease was smaller (27). Eventually, both groups reached similar ghrelin levels. In diabetic patients the decrease of ghrelin level to insulin infusion was also smaller compared to healthy controls (28). We hypothesize that insulin can decrease ghrelin levels only to a certain extent, until the individually lowest ghrelin level is reached. When this lowest ghrelin level is reached, further elevation of insulin level is not accompanied by an additional drop in ghrelin level. We assume that the already elevated insulin level in normal weight, nondiabetic patients is sufficient to maximally suppress ghrelin level, and the further elevated insulin level in diabetic patients could not lead to even lower ghrelin levels. We found a positive association between adiponectin and ghrelin even after adjusting for body mass index (BMI) and gender. Little is known about the interaction of ghrelin and adiponectin. A positive correlation has been reported in postmenopausal women (29) and in obese women without insulin resistance (13). Intravenous administration of unacylated and acylated ghrelin had no effect (14) or decreased adiponectin levels (30). According to our knowledge, the effect of adiponectin on ghrelin level has not yet been investigated. Cardiomyocytes and epicardial adipose tissue is likely to have a direct impact on atherosclerosis by producing certain adipokines and cytokines, e.g., adiponectin. Studies on cultures of these cells could evaluate the role of adipokines of this origin. Based on previous findings and our findings, it is possible that in addition to elevated insulin, decreased adiponectin level also contributes to low ghrelin concentration in CAD patients. We found no association between ghrelin and leptin levels. Although Ukkola et al. found a negative correlation between ghrelin and leptin (9), others did not (15,31). Erdman and coworkers found a negative correlation in normoinsulinemic patients but found no correlation in hyperinsulinemic subjects with obesity or T2DM. These results are in line with our observation that, in our patients, insulin but not leptin determined ghrelin levels.
Fetuin-A has been shown to be associated with adiponectin level in patients with atherosclerosis (32,33). We found no correlation between serum ghrelin and fetuin-A levels. Serum lipids were also not associated with ghrelin level. Several studies described a positive correlation with HDLcholesterol levels (34) and a negative correlation with triglycerides (12,34,35). In vitro, blocking lipolysis (36) and promoting triglyceride uptake by adipocytes (36,37) were suggested mechanisms. The correlation with triglycerides, however, was not independent and no association with cholesterol levels was reported. Our finding of no correlation has to be interpreted cautiously because 65% of our patients were on lipid-lowering treatment. Subclinical inflammation is known to be an important factor in the progression of atherosclerosis (17). The acute inflammatory phase of MI usually subsides within 2e3 weeks (38). The healing phase initiates after 4e6 weeks (39) and after 8 weeks tissue regeneration and remodeling is dominant, with fibroblast activity (40). We enrolled patients 6e24 months after MI so we did not expect considerable inflammation due to tissue damage at this point. However, subclinical inflammation was present and we found elevated TNF-a levels in our patients. Acylated ghrelin was also implicated in inflammation regulation. Elevated levels were found in humans with chronic inflammatory diseases (41,42). Animal and in vitro studies provided evidence that acylated ghrelin can blunt the inflammation induced by TNF-a or other pro-inflammatory substances and decrease the expression of TNF-a and some other interleukins (18,19). Taken together, the elevated levels of ghrelin in inflammatory diseases and its anti-inflammatory effects may represent a defense mechanism against chronic inflammation. However, we could not detect any significant relationship between ghrelin and TNF-a level. Considering this independency, we analyzed whether the TNF-a/ghrelin ratio was a helpful parameter to discriminate between patients with atherosclerosis and referent subjects. Patients having low ghrelin and elevated TNF-a level (high TNF-a/ghrelin ratio) had more than 11 times higher chance for belonging to the patient group. These two proteins and their ratio may be useful novel indicators of coronary atherosclerosis in the clinical setting. It seems promising to determine the predictive power of these indicators. Our patients were on preventive medications according to guidelines. Lipid-lowering and antidiabetic drugs were reported not to alter or to elevate serum ghrelin levels (43,44). The cross-sectional design and relatively small number of patients are the major limitations of our study. Cardiac remodeling may have altered ghrelin level, although we found no data in the literature regarding this issue. In conclusion, our data support the concept that acylated ghrelin levels are decreased in patients with coronary
Ghrelin and TNF-a/Ghrelin Ratio in Atherosclerosis
atherosclerosis, independent of body weight and the presence of T2DM. Ghrelin levels are determined by elevated insulin and decreased adiponectin levels in patients with previous MI. Whether ghrelin alone or in combination with TNF-a may prove to be a novel indicator of coronary atherosclerosis warrants further studies.
Acknowledgments This work was supported by grants OTKA KON T046427 and ETT368/2009 of the Hungarian Ministry of Health. The authors declare that they have no conflicts of interest.
References 1. Pulkkinen L, Ukkola O, Kolehmainen M, et al. Ghrelin in diabetes and metabolic syndrome. Int J Pept 2010;2010. http://dx.doi.org/10.1155/ 2010/248948. 2. Zhang G, Yin X, Qi Y, et al. Ghrelin and cardiovascular diseases. Curr Cardiol Rev 2010;6:62e70. 3. Kishimoto I, Tokudome T, Hosoda H, et al. Ghrelin and cardiovascular diseases. J Cardiol 2012;59:8e13. 4. Kadoglou NP, Sailer N, Moumtzouoglou A, et al. Visfatin (nampt) and ghrelin as novel markers of carotid atherosclerosis in patients with type 2 diabetes. Exp Clin Endocrinol Diabetes 2010;118:75e80. 5. Yano Y, Toshinai K, Inokuchi T, et al. Plasma des-acyl ghrelin, but not plasma HMW adiponectin, is a useful cardiometabolic marker for predicting atherosclerosis in elderly hypertensive patients. Atherosclerosis 2009;204:590e594. 6. Kotani K, Sakane N, Saiga K, et al. Serum ghrelin and carotid atherosclerosis in older Japanese people with metabolic syndrome. Arch Med Res 2006;37:903e906. 7. Kadoglou NP, Lampropoulos S, Kapelouzou A, et al. Serum levels of apelin and ghrelin in patients with acute coronary syndromes and established coronary artery diseaseeKOZANI STUDY. Transl Res 2010; 155:238e246. 8. Poykko SM, Kellokoski E, Ukkola O, et al. Plasma ghrelin concentrations are positively associated with carotid artery atherosclerosis in males. J Intern Med 2006;260:43e52. 9. Ukkola O, Poykko S, Paivansalo M, et al. Interactions between ghrelin, leptin and IGF-I affect metabolic syndrome and early atherosclerosis. Ann Med 2008;40:465e473. 10. Haffner SM, Lehto S, Ronnemaa T, et al. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med 1998;339:229e234. 11. Tong J, Prigeon RL, Davis HW, et al. Ghrelin suppresses glucosestimulated insulin secretion and deteriorates glucose tolerance in healthy humans. Diabetes 2010;59:2145e2151. 12. Poykko SM, Kellokoski E, Horkko S, et al. Low plasma ghrelin is associated with insulin resistance, hypertension, and the prevalence of type 2 diabetes. Diabetes 2003;52:2546e2553. 13. Krzyzanowska-Swiniarska B, Kempa A, Miazgowski T, et al. Serum acylated ghrelin, adiponectin and leptin levels in normal-weight and obese premenopausal women. Horm Metab Res 2007;39:835e839. 14. Kiewiet RM, Hazell MJ, Van Aken MO, et al. Acute effects of acylated and unacylated ghrelin on total and high molecular weight adiponectin in morbidly obese subjects. J Endocrinol Invest 2011;34: 434e438. 15. Caliskan Y, Gorgulu N, Yelken B, et al. Plasma ghrelin levels are associated with coronary microvascular and endothelial dysfunction in peritoneal dialysis patients. Ren Fail 2009;31:807e813.
553
16. Erdmann J, Lippl F, Wagenpfeil S, et al. Differential association of basal and postprandial plasma ghrelin with leptin, insulin, and type 2 diabetes. Diabetes 2005;54:1371e1378. 17. Corrado E, Rizzo M, Coppola G, et al. An update on the role of markers of inflammation in atherosclerosis. J Atheroscler Thromb 2010;17:1e11. 18. Hattori N. Expression, regulation and biological actions of growth hormone (GH) and ghrelin in the immune system. Growth Horm IGF Res 2009;19:187e197. 19. Baatar D, Patel K, Taub DD. The effects of ghrelin on inflammation and the immune system. Mol Cell Endocrinol 2011;340:44e58. 20. Kalabay L, Jakab L, Prohaszka Z, et al. Human fetuin/alpha2HSglycoprotein level as a novel indicator of liver cell function and short-term mortality in patients with liver cirrhosis and liver cancer. Eur J Gastroenterol Hepatol 2002;14:389e394. 21. Tschop M, Weyer C, Tataranni PA, et al. Circulating ghrelin levels are decreased in human obesity. Diabetes 2001;50:707e709. 22. Meyer C. Final answer: ghrelin can suppress insulin secretion in humans, but is it clinically relevant? Diabetes 2010;59:2726e2728. 23. Palik E, Baranyi E, Melczer Z, et al. Elevated serum acylated (biologically active) ghrelin and resistin levels associate with pregnancyinduced weight gain and insulin resistance. Diabetes Res Clin Pract 2007;76:351e357. 24. Saad MF, Bernaba B, Hwu CM, et al. Insulin regulates plasma ghrelin concentration. J Clin Endocrinol Metab 2002;87:3997e4000. 25. Broglio F, Gottero C, Prodam F, et al. Ghrelin secretion is inhibited by glucose load and insulin-induced hypoglycaemia but unaffected by glucagon and arginine in humans. Clin Endocrinol 2004;61:503e509. 26. An W, Li Y, Xu G, et al. Modulation of ghrelin O-acyltransferase expression in pancreatic islets. Cell Physiol Biochem 2010;26: 707e716. 27. Sondergaard E, Gormsen LC, Nellemann B, et al. Visceral fat mass is a strong predictor of circulating ghrelin levels in premenopausal women. Eur J Endocrinol 2009;160:375e379. 28. Anderwald C, Brabant G, Bernroider E, et al. Insulin-dependent modulation of plasma ghrelin and leptin concentrations is less pronounced in type 2 diabetic patients. Diabetes 2003;52:1792e1798. 29. Soni AC, Conroy MB, Mackey RH, et al. Ghrelin, leptin, adiponectin, and insulin levels and concurrent and future weight change in overweight, postmenopausal women. Menopause 2011;18:296e301. 30. Ott V, Fasshauer M, Dalski A, et al. Direct peripheral effects of ghrelin include suppression of adiponectin expression. Horm Metab Res 2002; 34:640e645. 31. Jiang ZT, Zou YP, Huang H, et al. Mechanism of laparoscopic adjustable gastric banding in the treatment of obesity with type 2 diabetes mellitus. Zhonghua Wei Chang Wai Ke Za Zhi 2010;13:520e523. 32. Ix JH, Sharma K. Mechanisms linking obesity, chronic kidney disease, and fatty liver disease: the roles of fetuin-A, adiponectin, and AMPK. J Am Soc Nephrol 2010;21:406e412. 33. Voros K, Graf L Jr, Prohaszka Z, et al. Serum fetuin-A in metabolic and inflammatory pathways in patients with myocardial infarction. Eur J Clin Invest 2011;41:703e709. 34. Nanjo Y, Adachi H, Hirai Y, et al. Factors associated with plasma ghrelin level in Japanese general population. Clin Endocrinol 2011; 74:453e458. 35. Palik E, Birkas KD, Faludi G, et al. Correlation of serum ghrelin levels with body mass index and carbohydrate metabolism in patients treated with atypical antipsychotics. Diabetes Res Clin Pract 2005;68(suppl 1):S60eS64. 36. Miegueu P, St Pierre D, Broglio F, et al. Effect of desacyl ghrelin, obestatin and related peptides on triglyceride storage, metabolism and GHSR signaling in 3T3-L1 adipocytes. J Cell Biochem 2011;112:704e714. 37. Theander-Carrillo C, Wiedmer P, Cettour-Rose P, et al. Ghrelin action in the brain controls adipocyte metabolism. J Clin Invest 2006;116: 1983e1993.
554
V€or€os et al./ Archives of Medical Research 43 (2012) 548e554
38. Bonvini Rf HT, Camenzind E. Inflammatory response post-myocardial infarction and reperfusion: a new therapeutic target? Eur Heart J Suppl 2005;7(suppl I):127e136. 39. Liehn EA, Postea O, Curaj A, et al. Repair after myocardial infarction, between fantasy and reality: the role of chemokines. J Am Coll Cardiol 2011;58:2357e2362. 40. Frangogiannis NG, Smith CW, Entman ML. The inflammatory response in myocardial infarction. Cardiovasc Res 2002;53:31e47. 41. Karmiris K, Koutroubakis IE, Xidakis C, et al. Circulating levels of leptin, adiponectin, resistin, and ghrelin in inflammatory bowel disease. Inflamm Bowel Dis 2006;12:100e105.
42. Toussirot E, Streit G, Nguyen NU, et al. Adipose tissue, serum adipokines, and ghrelin in patients with ankylosing spondylitis. Metabolism 2007;56:1383e1399. 43. Kadoglou NP, Tsanikidis H, Kapelouzou A, et al. Effects of rosiglitazone and metformin treatment on apelin, visfatin, and ghrelin levels in patients with type 2 diabetes mellitus. Metabolism 2010; 59:373e379. 44. Otto C, Otto B, Frost RJ, et al. Short-term therapy with atorvastatin or fenofibrate does not affect plasma ghrelin, resistin or adiponectin levels in type 2 diabetic patients with mixed hyperlipoproteinaemia. Acta Diabetol 2007;44:65e68.