Serum endothelin-1 level positively correlates with waist and hip circumferences in stable coronary artery disease patients

to be higher in subjects with central obesity as compared to those without. Serum endothelin-1 level was significantly correlated with age, hemoglobin level, waist circumference (coefficient of 0.311, p value = 0.023) and hip circumference (coefficient of 0.359, p value=0.010). Multivariableanalysisindicatedthatage(co-efficientof–0.353, p value=0.007)andhipcircumference(coefficient of 0.335, p value = 0.011) were independently correlated with serum endothelin-1. For conclusion, in patients with stable CAD, serum endothelin-1waspositivelycorrelatedwithbothwaistcircumference and hip circumference. Hip circumference independently and posi-tivelycorrelatedwithserumendothelin-1level.


Introduction
Cardiovascular disease (CVD) is one of the leading causes of death in many countries [1,2]. Hypertension, diabetes mellitus, dyslipidemia, smoking behaviors and sedentary lifestyles are the established CVD risk factors for which prevalence are high and increasing in low-and middleincome countries [3]. In Indonesia, besides these risk factors, obesity, especially central obesity, continues to increase significantly and contributes meaningfully to CVD [4]. Individual with central obesity has an excess of visceral adiposity which is associated with sedentary lifestyle and metabolic diseases such as diabetes mellitus, hypertension, and dyslipidemia [5]. It adds multiple burdens of CVD risk factors. The increase in central obesity-related morbidity in Indonesia challenge CVD risk factor controls in a country which still struggles to implement CVD prevention program such as tobacco control policy, promotion of healthy diet and endorsement of healthy lifestyle [5].
Endothelin-1 is a main vasoconstrictor peptide in the circulation which operates in the peripheral, pulmonary, and coronary vascular beds through vascular smooth muscle cells contraction [6]. In a steady-state condition, it sustains the balanced vascular tones by stimulating the release of a vasodilator, nitric-oxide [7]. In vascular dysfunction, the balance between endothelin-1 and nitric-oxide is disrupted which lead to profound increased of endothelin-1, malfunctioning vasodilation, and increased vasoconstrictor tone [8]. Vascular dysfunction, which is indicated by insulin resistance and hyperinsulinemia, frequently accompanies individual with central obesity [9,10]. Excessive visceral adipocytes express endothelin-1 which inhibit their insulin-stimulated glucose uptake, stimulate their lipolysis, release their free fatty acids and induce their production of pro-inflammatory cytokines [11]. Previous study demonstrated the enhanced endothelin-1 activity and endothelin-1-mediated adiposityrelated disruption ofvasodilationin overweight and obese individuals without CVD [12,13].
Individual with excessive fat accumulation in the visceral adipose tissue has higher rate of obesity-related CVD events [14]. Visceral adiposity is associated with increased waist circumference, which is a parameter to determine central obesity [14]. Central obesity is considered as metabolically harmful obesity [8]. Hip circumference indicates the adiposity of lower-body gluteofemoral region, which possess opposite effect to visceral adipose tissue [8]. Increasedserum endothelin-1contributes to metabolic syndrome and higher CVD events in individual with central obesity without coronary artery disease (CAD) [8,13]. In patients with stable CAD, endothelin-1 level in blood circulation is increasing [15,16]. Our previous study indicated that patients survived from acute coronary syndrome (ACS) and who later developed stable CAD, the majority of them were overweight and obese individuals [17]. The association between serumendothelin-1 level and central obesity in individual with stable CAD has not been investigated.
To assess central obesity, waist circumference is a superior indicator and more strongly correlated with intraabdominal visceral adipose content, whereas hip circumference is associated with gluteofemoral peripheral adipose mass [18]. Both measurements had independent and opposite correlation with atherogenic risk factors, glucose intolerance and lipid metabolism disturbances [18]. The relation between serum endothelin-1 in patients with angiographycallyproven stable CAD and parameters of central obesity, namely waist circumference, hip circumference and ratio derived from them needs to be investigated. This study aims to examine the correlation between serum endothelin-1 level and parameters of central obesity in Indonesian patients with stable CAD.

Subjects
This is a cross-sectional study. Subjects were patients diagnosed with stable CAD. The subjects were enrolled consecutively during the performance of coronary angiography (CAG) with/without stenting in Integrated Heart Center (Pusat Jantung Terpadu) Dr. Sardjito Hospital, Yogyakarta, Indonesia. The inclusion criteria were: (1) subjects underwent elective CAG, (2) subjects with age 30-75 years, (3) subjects with significant CAD (namely stenosis ≥50% in left main coronary artery or stenosis ≥70% in other main branches), and (4) subjects agreed to participate in the research by signing an informed consent form. The exclusion criteria were: (1) subjects with chronic heart failure reduced ejection fraction, (2) subjects with chronic kidney disease (MDRD <30 mL/min/1.73 m 2 ), (3) subjects with hepatic cirrhosis, (4) subjects with on-going treatment of malignancy, (5) subjects with history of previous coronary revascularisation (percutaneous coronary intervention (PCI) or coronary-artery bypass graft (CABG), and (6) subjects with coronary anatomy anomalies.

Procedures
Subjects were enrolled during hospitalization for CAG with/without PCI. The demographic and medical histories were collected by anamnesis. Clinical data were collected during day-1 hospitalization (before CAG procedure). The bodyweight (kg) and body height (m) were measured by standard body scale and height ruler respectively at day-1 hospitalization (before CAG procedure). The body mass index (BMI) was calculated as bodyweight/(body height) 2 . Obesity based on BMI was determined as obese II (BMI ≥30), obese I (BMI 25-29.9), overweight (BMI 23-24.9) and normal (BMI <23) [19]. The waist circumference (cm) was measured by standard body tape to the closest 0.1 cm in the horizontal plane at middle-point between lowest rib and the iliac-crest. The hip circumference (cm) was measured at the maximum protrusion of hip by standard body tape to the closest 0.1 cm. Both measurements were performed during day-2 or day-3 hospitalization (after CAG procedure) by trained investigators in the morning (after 8 hours fasting). The waist-to-hip ratio was calculated from waist circumference (cm)/hip circumference (cm). The waist-to-height ratio was calculated from waist circumference (cm)/body height (cm). The central obesity was determined by criteria of male waist circumference ≥90 cm and female waist circumference ≥80 cm [19].
The elective CAG with/without PCI procedures and subsequent treatments for subjects were done at the discretion of attending cardiologists. The research procedure was approved by The Medical and Health Research Ethics Committee Faculty of Medicine, Public Health and Nursing Universitas Gadjah Mada and Dr. Sardjito Hospital Yogyakarta, Indonesia.

Laboratory tests
Peripheral blood samples were obtained from subjects while reclining in a supine position from antecubital venous access on day-1 (before CAG procedure). The blood samples were taken after at least 15 min of supine resting on vacutainer tubes; in each subject a first sample was prepared for routine and blood chemistry, whereas the second sample was prepared for endothelin-1 measurement. The blood samples were transferred into Vacutainer tubes (BD, USA) and left at a room temperature to form clotting at 20-30 min. The tubes were sent to a hospital central laboratory for routine hematology and blood chemistry examinations. For endothelin-1 measurement, the tubes containing clotted blood samples were centrifuged at 200 g for 20 minutes and the supernatant was stored at -80 • C freezer until analysed. The procedures of measurement and quantification of endothelin-1 followed the manufacturer's instructions (Endothelin-1 immunoassay Quantikine® ELISA kit (R&D Systems, Minneapolis, MN, USA) without replication [20].

Statistical analysis
The normal distribution was tested with Kolmogorov-Smirnov test. The comparison between normally distributed continuous data was performed with Student's t-test, while Mann-Whitney test was used for not normally distributed continuous data. The bivariate correlation analysis was performed with Pearson correlation test or Spearman correlation test where applicable. A multivariate regression analysis was done to determine the strength of correlation among different independent co-variables. Co-variables were selected from bivariate analysis which had p value < 0.05. A p value ≤ 0.05 was considered statistically significant. ± 0.7 pg/mL (n = 16), overweight: 2.2 ± 0.8 pg/mL (n = 16) and normal: 2.2 ± 0.5 pg/mL (n = 13), p value = 0.503). There was a trend that highest mean level of endothelin-1 was in subjects with obese II. Table 1 shows the characteristics of research subjects. Fifty-consecutive subjects were enrolled in this study. Mean endothelin-1 level in all subjects was 2.2 ± 0.7 pg/mL. Majority of subjects were males (80.0%) with mean age of 58.8 ± 8.5 years old. Most of subjects were with hypertension (86.0%), dyslipidemia (68%) and smokers (52%). Only minority of patients had diabetes mellitus (28.0%). Additionally, most subjects had history of ACS (64%) within previous 1 year. Anthropometric parameters indicated that mean bodyweight was 66.3 ± 11.5 kg, mean body height was 1.62 ± 0.07 m, and mean BMI was 25.1 ± 3.5. Based on different obesity groups by BMI parameters, there was a trend that the highest level of endothelin-1 was in obese II subjects (as shown in Fig. 1).

Results
The circumference measurements indicated that mean waist circumference was 87.6 ± 11.4 cm, hip circumference was 95.3 ± 10.5 cm, mean waist-to-hip ratio was 0.92 ± 0.06, and mean waist-to-height ratio was 0.54 ± 0.07. Based on waist circumference, central obesity occurred in 32% of subjects, as shown in Table 1. Serum endothelin-1 level tended to be higher in subjects with central obesity as compared to those without central obesity (as shown in Fig. 2). Serum endothelin-1 did not significantly differ based on demographic and CVD risk factors variables ( Table 2).

Fig. 2. Comparison of endothelin-1 level between subjects with central obesity (n = 16) and no central obesity (n = 34).
There was no signicant difference in endothelin-1 level between two groups (mean ± SD: 2.4 ± 0.8 pg/mL vs. 2.2 ± 0.7 pg/mL, p value = 0.425). There was a trend that higher mean level of endothelin-1 was in subjects with central obesity. Table 3 shows result of correlation analysis between serum endothelin-1 level and other continuous variables. Among clinical, anthropometric and laboratory variables, serum endothelin-1 level was significantly correlated with age in years (correlation coefficient of -0.376, p value = 0.007) and hemoglobin level (correlation coefficient of 0.316, p value 0.026). Serum endothelin-1 was significantly correlated with parameters of central obesity, namely waist circumference (correlation coefficient of 0.311, p value = 0.023) and hip circumference (correlation coefficient of 0.359, p value = 0.010). Other variables, i.e., age, body mass index, systolic blood pressure, diastolic blood pressure, heart rate, hemoglobin level, leucocytes count, platelet count, glucose level and cre-  atinine level, were not significantly correlated with serum endothelin-1 level. Table 4 shows the result of multivariate regression analysis. Multivariable analysis were included age, hemoglobin, waist circumference and hip circumference as covariates and endothelin-1 as the dependent variable. Correlation were observed inage (coefficient of -0.353, p value = 0.007) and hip circumference (coefficient of 0.335, p value = 0.011) with endothelin-1 level. Waist circumference did not significantly correlate independently with serum endothelin-1 level.

Discussion
Results of our study indicated that in patients with angiographycally-proven stable CAD, serum endothelin-1 level was positively correlated with both waist circumference and hip circumference. The independent correlation was significantly observed in age (inverse correlation) and hip circumference (positive correlation). Serum endothelin-1 tended to be higher in stable CAD patients with central obesity and patients with BMI-derived obesity compared to their counterparts. The measurement of serum endothelin-1 as a prognostic indicator in stable CAD patients with central obesity needs to be corroborated by larger and more extensive research.
Previous study indicated that obese individuals (BMI ≥30.5 males and BMI ≥27.3) had significantly higher levels of serum endothelin-1 as compared to lean individual (BMI <25 for males and BMI <24.7 for females). Obese subjects with hypertension had the highest endothelin-1 level [21]. This study also indicated that obese individuals had higher waist-to-hip ratio [21]. Activity of endothelin-1-mediated vasoconstriction in vascular tone is elevated in obese individual with waist circumference >100 cm [13]. Its enhanced activity contributes to the adiposity-related impairment in endothelium-dependent vasodilation, which is prevalent in obese individuals [13]. It was postulated that increased endothelin-1 level and enhanced activity in obese patients are due to an insulin resistance mechanism [13]. However, these studies had excluded subjects with established CAD.
An important mechanism by which obesity leads to the development of vascular diseases is the development of insulin resistance and inflammation [8]. Furthermore, in patients with stable CAD, within a spectrum of advanced atherosclerosis, insulin resistance and inflammation may have developed for long time before any stable CAD events. Evidence indicates that endothelin-1 contributes to insulin resistance and inflammation through numerous mechanisms, including impairment of insulin signaling in endothelial cells as a precursor of early atherosclerosis [22,23]. Atherosclerosis itself is associated with excess endothelin-1, which is by several mechanisms known to enhance the atheroma formation and atherosclerotic disease progression [24,25]. Furthermore, excess endothelin-1 inhibits insulin-stimulated glucose uptake in adipocytes and skeletal muscle cells, as well as enhancing lipolysis, release free fatty acids and pro-inflammatory cytokines in adipocytes [11]. In obese patients with stable CAD, all factors that contribute to increased serum endothelin-1 levels were found. Hypertension, which predominates in our study subjects, also contribute to elevated level of serum endothelin-1 [26].
Our study indicated positive correlation between serum endothelin-1 level with both waist and hip circumferences, but not with other obesity and central parameters. As an individual measure, waist circumference is a measure of visceral and subcutaneous adipose tissue in the abdominal region, whereas hip circumference is a measure of adipose tissue and also muscle mass in the lower body part [27]. After adjustment with covariables, unexpectedly, hip circumference independently correlated positively with serum endothelin-1. From this current observation, we speculate that increased hip circumference reflects increased subcutaneous adipose storage and muscle mass of the gluteofemoral region which had specific lipid storage and secretion of adipose tissue-related proteins [27]. Adipose tissue in the gluteofemoral region had active role in the removal of circulating nonesterified fatty acids which could limit the development of insulin resistance [28]. Similarly, thigh subcutaneous adipocytesc correlated with greater insulin sensitivity [28]. Both adipocytes and skeletal muscle cells express endothelin-1 and its receptors [11]. Whether the activity of gluteofemoral adipose tissue is mediated by endothelin-1 needed further research.
This study had several limitations. The major limitation was the small number of subjects enrolled in the analysis. Further study with larger sample size is needed to be performed to confirm our findings. Another limitation was laboratory parameters of various metabolic disturbances were not measured in this research. Further study with measurements of parameters of metabolic disturbances as additional confounding variables needs to be performed in patients with angiographycally-proven stable CAD.
In conclusion, serum endothelin-1 level was significantly positively correlated with waist circumference and hip circumference, as parameters of central obesity among Indonesian patients with angiographycally-proven stable CAD. The independent positive correlation was found between serum endothelin-1 level and hip circumference.

Author contributions
ABH designed the study, performed subject's enrollment, data collection, statistical analysis and manuscript writing. JF performed data collection, laboratory analysis and manuscript writing. IP performed subjects enrollment, data collection, and statistical analysis. FSTD contributed to data collection, data analysis and manuscript writing. All authors gave final approval of the manuscript.

Ethics approval and consent to participate
Ethics approval for this research had been granted by The Medical and Health Research Ethic Commitee of Faculty of Medicine, Public Health and Nursing Universitas Gadjah Mada -Dr. Sardjito Hospital, Yogyakarta, Indonesia.