Serum concentration of visfatin is decreased in patients with chronic heart failure

Background. There is an increasing interest in the role of adipocytokines in cardiovascular pathophysiology. Aim. The aim of the study was to compare visfatin levels, a novel adipokine, in patients with heart failure (HF) due to the left ventricular systolic dysfunction with those in ageand body mass index (BMI) — matched healthy controls in relation to the parameters of glucose metabolism and high sensitivity C-reactive protein (hsCRP) levels. Material/ Subjects and Methods. The study population consisted of 28 males with systolic HF referred for cardiopulmonary exercise testing, divided into two subgroups based on their NYHA class (HF patients NYHAI+II, n=17, and HF patients NYHAIII+IV, n=11), and 23 controls. The following indices were measured in a serum samples: visfatin, hsCRP, glucose and lipid metabolism parameters, and the insulin resistance index HOMAIR (homeostasis model assessment insulin resistance) was calculated. Results. Concentrations of visfatin and high-density lipoprotein cholesterol (HDLcholesterol) in the HF subjects were significantly lower (p≤0.01) than in controls. The Kruskal-Wallis test showed significant differences between three groups (controls and both subgroups of heart failure patients) in mean levels of visfatin, hsCRP, glucose, HOMAIR and HDL-cholesterol. Conclusion. Serum visfatin concentrations in patients with systolic HF, particularly with more advanced NYHA classes, are significantly lower in comparison to healthy controls and are independent of age or anthropometric and metabolic parameters.


INTRODUCTION
Heart failure (HF) becomes a serious problem in contemporary medicine because of an increasing number of patients developing this pathology and its relation with high morbidity and mortality.According to the "muscle hypothesis", the main symptom, i.e., exercise intolerance, results from complex peripheral abnormalities, in which low-grade systemic inflammation and metabolic abnormalities play an important role (Clark et al., 1996).
Visfatin is a novel adipocytokine predominantly secreted by visceral adipocytes and is presumed to have proinflammatory properties (Moschen et al., 2007).There is an increasing interest in the role of adipocytokines in cardiovascular pathophysiology (Dahl et al., 2007;Liu et al., 2009;Kadoglou et al., 2011).There are some reports on the relationship between levels of leptin (Wolk & Somers, 2006) and apelin -another new adipokine - (Chong et al., 2006;Ho et al., 2009) and heart failure severity.Their role in pathophysiology of heart failure is suggested.We have found only one report on visfatin in the HF (Ho et al., 2009).
Our aim was to compare the serum visfatin concentrations in patients with heart failure due to the left ventricular systolic dysfunction with those in age-and BMImatched healthy controls in relation to the parameters of glucose metabolism and C-reactive protein levels.

MATERIALS AND METHODS
The study population consisted of 28 males with systolic heart failure referred for cardiopulmonary exercise testing and 23 healthy subjects.Heart failure was diagnosed according to the European Society of Cardiology guidelines.All patients had left ventricular ejection fraction (LVEF) <45% as measured by echocardiography.Coronary artery disease (CAD) was diagnosed in 12 patients (43%) and non-ischemic dilated cardiomyopathy (DCM) in 16 (57%).The patients with heart failure were divided into two subgroups based on their NYHA class: heart failure patients NYHA I+II (n=17) and heart failure patients NYHA III+IV (n=11).

E. Straburzyńska-Migaj and others
All enrolled subjects underwent maximal cardiopulmonary exercise treadmill test performed according to the modified Bruce protocol (adding stage 0: 3 min, 1.7 km/h, 5% grading).The peak oxygen consumption (peak VO 2 ), carbon dioxide production, and minute ventilation were measured using a breath-by-breath technique (Sensor Medics, model V max 29).The equipment was calibrated before each test.There was continuous ECG monitoring and blood pressure measurement at each stage of exercise.Peak VO 2 was defined as the highest 20-second average during the last 60 seconds of exercise.
The control group consisted of male volunteers declaring good health status.The inclusion criteria were age, body mass and BMI as similar as possible to the patients' characteristics.
All subjects gave their informed consent to participate in the study.The study was conducted in compliance with the Helsinki Declaration and was approved by the local Ethics Committee.
The values are given as mean, standard deviation (SD), median (Me), standard error (SE) and interquartile ranges.The normal distribution of the data was verified with the Shapiro-Wilk test.Comparisons between the heart failure patients and controls for age, anthropometric measurements, and biochemical parameters were made using the Student t-test or Mann-Whitney test.
Comparisons for maximal oxygen uptake (VO 2 max) between the heart failure subgroups (by NYHA class) and for visfatin levels between the patients with ischemic heart disease and dilated cardiomyopathy or between patients treated and not treated with statins were made using the Mann-Whitney test.Comparisons between three groups of subjects (heart failure subgroups divided according to NYHA class and the controls) for biochemical parameters were made using the Kruskal-Wallis oneway analysis of variance by ranks with post-hoc test.The Spearman's rank analysis was used to calculate correlation coefficients.A p<0.05 was taken to be statistically significant.Statistical analyses were performed with the Statistica 8.0 software package.

RESULTS
Basic characteristics of the examined groups of subjects are presented in Table 1.
The heart failure group and controls did not differ significantly regarding age, body weight, BMI and waist circumference.The heart failure subjects had significantly lower resting systolic blood pressure (p≤0.01) and diastolic blood pressure (p<0.05), and a higher heart rate (p≤0.01)than the controls.
Results of the laboratory assessments are shown in Table 2. Concentrations of visfatin and HDL-cholesterol in the HF subjects were significantly lower (p≤0.01)than in the controls.There were significantly higher levels of triglycerides (p<0.05),glucose (p≤0.01) and HOMA IR (p<0.05) in the HF patients in comparison to the control group.In 53% and 26% of subjects from the heart failure and control groups, respectively, the values of HO-MA IR represent insulin resistance (HOMA ≥2.5).
In the HF group there were no significant differences in visfatin concentrations between the patients with ischemic heart disease (n=12) and dilated cardiomyopathy (n=16), (Me±SE respectively: 0.74±0.81ng/ ml vs 0.51±0.54ng/ml; p=0.5155) and between the patients treated (n=18) and not treated with statins (n=10), The Kruskal-Wallis test showed significant differences between three groups (the controls and both subgroups of HF patients divided according to NYHA I+II vs III+IV class) in levels of visfatin, hsCRP (figure 1, a and b, respectively), HOMA IR , HDL-cholesterol (Fig. 2, a and b, respectively) and glucose (p=0.0038).The NYHA III+IV subjects had significantly lower HDL-cholesterol concentrations (p≤0.01) and only the tendency to decreased visfatin concentrations (p=0.061).Moreover, they had higher hsCRP, HOMA IR and glucose levels (p<0.05)than the controls.In the NYHA I+II subjects, the levels of all these biochemical parameters did not differ from the controls.There was a significant difference between the two heart failure subgroups with respect to hsCRP concentrations (p=0.0255).No significant differences were found between three investigated groups with respect to age, body mass, BMI and waist circumference.

DISCUSSION
The presented data show that serum visfatin concentrations are significantly lower in male subjects with heart failure due to systolic dysfunction in comparison to healthy controls.Furthermore, visfatin concentrations are lower in patients with more severe HF (NYHA III+IV ) in comparison to mild HF (NYHA I+II ).Concentrations of visfatin were independent of age, BMI, levels of glucose, insulin, cholesterol, and triglicerydes, HOMA IR , heart rate, and peak VO 2 .
An adipocytokine exerting insulin-mimicking effects, visfatin, attracts particular interest of investigators dealing with diabetes and obesity (Fukuhara et al., 2005)   and subjects with or without visceral fat accumulation (Ersoy et al., 2010).A study conducted in women with rheumatoid arthritis showed lower plasma visfatin concentrations in patients with abdominal obesity in comparison to those without abdominal obesity (Straburzyńska-Lupa et al., 2010).The results regarding relations between anthropometric parameters and visfatin levels are also inconsistent.Some investigators found correlations between visfatin and BMI, others show no correlations (Zahorska-Markiewicz et al., 2007;Alghasham & Bakarat, 2008;Jin et al., 2008;Caixàs et al., 2009) or even negative correlations (Chen et al., 2006).In our study there were no differences in the BMI or waist circumference between the patients and controls.We demonstrated that there was no significant correlations between visfatin levels and BMI in the HF group, while there was significant positive correlation in the control group.
Similarly controversial are results regarding relationships between visfatin and both, glucose and insulin resistance.In obesity and type 2 diabetes there is a well established association between visceral adipose tissue, lipid metabolism and insulin resistance, which in turn are associated with increased cardiovascular risk (Wajchenberg, 2000).Correlations of visfatin and insulin resistance were shown in some studies (Chen et al., 2006;Lewandowski et al., 2007) and no correlations were found in others (Pagano et al., 2006;Alghasham & Bakarat, 2008;de Luis et al., 2008).Heart failure is regarded as an insulin resistance state in which glucose metabolism abnormalities are related to the disease severity and prognosis (Doehner et al., 2005;Straburzyńska-Migaj et al., 2007).We have demonstrated significantly higher glucose and insulin levels, and HOMA IR in our patients compared to the healthy controls, which were further higher in NYHA III+IV vs. NYHA I+II ; however, we failed to demonstrate any correlations between visfatin and serum levels of glucose, insulin or HOMA IR .
Some investigations imply that visfatin may play a role in lipid metabolism (Jin et al., 2008;Esteghamati et al., 2011), while others (Zahorska-Markiewicz et al., 2007;de Luis et al., 2008;Ersoy et al., 2010) have shown no correlations between visfatin and lipid parameters in various study populations.We have demonstrated significant correlation between visfatin and HDL-cholesterol in heart failure male patients, which confirms the results of a few other studies (Smith et al., 2006;Jin et al., 2008).No correlations were found between visfatin and lipids in the healthy controls.Wang et al. (2007), in studies on Caucasian subjects, have proposed that the association of visfatin with HDL-cholesterol may be the effect of en-zymatic function of visfatin in NAD metabolism.Recent studies have documented that statin therapy may suppress visfatin serum levels (Kadoglou et al., 2011).However, in our study the comparative analysis of visfatin levels between heart failure individuals treated and not treated with statins did not show any significant difference, but the small sample size and the large variation in visfatin concentrations could blunt the results.
Recently, a role of visfatin in inflammatory processes such as atherosclerosis or subclinical inflammation with concomitant type 2 diabetes is suggested (Alghasham & Bakarat, 2008;Liu et al., 2009).Liu et al. (2009) have documented that plasma visfatin concentrations are significantly higher in patients with chronic coronary artery disease and acute coronary syndromes compared with control groups.In their study serum visfatin concentrations correlated with inflammatory factors (monocyte chemoattractant protein -MCP-1 and interleukin-6), but not with hsCRP or fasting glucose levels, age, waist circumference, BMI, HOMA IR , or levels of cholesterol and triglycerides.Heart failure is also considered a low grade inflammatory process in which inflammatory parameters are related to disease severity and prognosis (Rauchhaus et al., 2000;Straburzyńska-Migaj et al., 2004).In the presented study we have shown that hsCRP levels are significantly higher in NYHA III+IV patients compared to NYHA I+II ones, but we documented no correlations between hsCRP and visfatin.Furthermore, we found no significant differences in visfatin levels between the patients with ischemic and non-ischemic heart failure.However, the limitation of our study is the small number of participants and the results must be approached with caution.
We have found one paper evaluating the role of visfatin and four other adipocytokines as potential prognostic factors in patients with severe heart failure and high levels of B-type natriuretic peptide (Ho et al., 2009).Visfatin was found to have no clinical importance in those patients.
Visfatin has been reported to exert several different effects with respect to cardiovascular disease (Hausenloy, 2009).On the one hand, these include endothelial dysfunction, angiogenesis, atherosclerotic plaque instability, and on the other -cardioprotection.Interesting and potentially important is a finding that visfatin may directly protect myocardium against the effects of acute ischemia-reperfusion injury at the cardiomyocyte level (Hausenloy, 2009).We are clearly at the beginning of a difficult way to explaining the role of visfatin in cardiovascular disease, including heart failure.

Figure 1 .
Figure 1.Serum visfatin and C-reactive protein (hsCRP) concentrations in controls and heart failure patients divided into NYHA I+II and NYHA III+IV class.*p<0.05,significantly different between two subgroups of heart failure patients, results are expressed as median ±S.E; (a) serum visfatin concentrations in controls and in both subgroups of HF patients; (b) C-reactive protein (hsCRP) concentrations in controls and in both subgroups of HF patients.

Figure 2 .
Figure 2. Homeostasis Model Assessment Insulin Resistance index (HOMA IR ) values and high density lipoprotein (HDL) cholesterol concentrations in controls and heart failure patients divided into NYHA I+II and NYHA III+IV class.**p≤0.01,*p<0.05,significantly different between subgroup of heart failure patients with NYHAIII+IV and controls, results are expressed as median ±S.E.(a) homeostasis model assessment insulin resistance index (HOMA IR ) in controls and in both subgroups of HF patients (b) high density lipoprotein (HDL) cholesterol concentrations in controls and in both subgroups of HF patients.