The relationship between atherogenic index of plasma and epicardial adipose tissue in hemodialysis and peritoneal dialysis patients

Abstract

Introduction: Epicardial adipose tissue (EAT) is the true visceral fat depot of the heart. The relationship between coronary artery disease and EAT was shown in end-stage renal disease (ESRD) patients. One of the established risk factor in this population is dyslipidemia. We aimed to determine the relationship between atherogenic index of plasma (AIP) and EAT in ESRD patients. Methods: This was a cross-sectional study involving 76 ESRD patients receiving PD or HD for ≥6 months and 42 healthy subjects. EAT was measured by using an electrocardiogram-gated 64-multidetector computed tomography (MDCT). Atherogenic index of plasma was calculated as the logarithmically transformed ratio of the serum trigliseride to HDL-cholesterol. Results: The etiology of ESRD patients was diabetic nephropathy (n = 16), chronic glomerulonephritis (n = 10), hypertensive nephropathy (n = 23), polycystic kidney disease (n = 7), nephrolithiasis (n = 5) and unknown (n = 15). There were no differences with respect to the following variables between ESRD patients and healthy subjects: age; sex; BMI; predialysis levels of DBP; serum levels of albumin, HDL-cholesterol and hemoglobin. However, ESRD patients had higher serum levels of trigliseride, hs-CRP and AIP when compared to healthy subjects. There was a statistically significant relationship between EAT, BMI and AIP in ESRD patients (r = 0.42, p < 0.001 and r = 0.25, p = 0.028, respectively). The stepwise linear regression analysis revealed that age, as well as BMI were independent predictors of EAT. Conclusion: We found a relationship between EAT as defined by MDCT and AIP in ESRD patients. Further clinical and experimental studies are needed.

• Atherogenic index of plasma
• end-stage renal disease
• epicardial adipose tissue
• high density lipoprotein
• triglyceride

Introduction

Dyslipidemia has been established as a well-known traditional risk factor for cardiovascular disease (CVD) in the general population1 and chronic kidney disease (CKD) patients receiving hemodialysis and peritoneal dialysis.2 Large-scale observational studies have confirmed that dyslipidemia may actively participate in the pathogenesis of atherosclerosis in general population.3 Additionally, it has been well known that CKD patients exhibit significant alterations in lipoprotein metabolism, which may result in the development of severe dyslipidemia in this population.4 Previous studies demonstrated that hypertriglyceridemia might be the earliest laboratory finding among the other lipid abnormalities even in patients who have slightly elevated creatinine levels.4,5 In contrast, high density lipoprotein (HDL) cholesterol was found to inversely related to the cardiovascular risk in non-CKD population.6 End-stage renal disease (ESRD) patients usually have increased concentrations of triglyceride-rich lipoproteins and reduced serum levels of HDL-cholesterol. However, low-density-lipoprotein (LDL) cholesterol values were found to be within normal limits or reduced in this population.7

Epicardial adipose tissue (EAT) originates from the splanchnopleuric mesoderm.8 Mazurek et al.9 concluded that, like abdominal visceral adipose tissue, EAT is also metabolically active because it can secrete proinflammatory cytokines and utilize free fatty acids (FFAs). In a recent study, the authors demonstrated that EAT acts an extremely active organ that produces several bioactive adipokines, as well as proinflammatory and proatherogenic cytokines including tumor necrosis factor (TNF)-α, interleukin (IL)-6, resistin, visfatin, omentin, leptin, plasminogen activator inhibitor-1 (PAI-1) and angiotensinogen.9–13 We recently showed that EAT is increased in HD and PD patients14,15 and this active visceral fat tissue is closely related with malnutrition-inflammation-atherosclerosis/calcification syndrome (MIAC) in this population.16

Logarithmic ratio of triglycerides to HDL was defined as atherogenic index of plasma (AIP) and this index was found to be closely associated with atherosclerosis in general population.17 However, in the literature, data regarding the AIP in ESRD patients are scant. Hence, we aimed to investigate the relationship between AIP and EAT in ESRD patients receiving hemodialysis (HD) and peritoneal dialysis (PD) and to compare these results with the others that obtained from healthy subjects.

Study population and methods

This was a cross-sectional study involving 76 ESRD patients (30 females, 46 males; mean age, 49.5 ± 14 years) receiving PD or HD for ≥6 months in the Dialysis Unit of Selcuk University and 42 healthy control subjects (19 females, 23 males; mean age, 51 ± 12 years) between February and June 2009.

Patients aged 18–70 years willing to participate in the assessment of EAT by multidetector computed tomography (MDCT) were screened. A review of medical records (including information on age; sex; weight; duration of renal replacement treatment; medications; primary disease of ESRD) was undertaken. Exclusion criteria were: (i) congestive heart failure; (ii) active infection; (iii) autoimmune disease; (iv) secondary hyperparathyroidism; (v) nephrotic-range proteinuria. Ninety patients were evaluated and 14 patients were excluded from the study. Of these 14 patients, 4 patients had congestive heart failure (New York Heart Association (NYHA) class III–IV); 4 patients had active infection; 3 patients had secondary hyperparathyroidism; and 3 patients had autoimmune disease (including systemic lupus erythematosus and microscopic polyangiitis). None of the patients included in the study had nephrotic-range proteinuria and arrhythmia based on electrocardiography (ECG). The remaining 76 ESRD patients fulfilled the above criteria and were enrolled in the study. Forty-two age-matched and sex-matched healthy individuals referred from outpatient clinics of the Internal Medicine Department of Selcuk University were also enrolled as control subjects. They were subject to the same inclusion and exclusion criteria as the patients. HD patients were receiving thrice-weekly dialysis for a 4-h period with a standard bicarbonate-containing dialysate bath using a biocompatible HD membrane (Polysulfone, FX-80 series, Fresenius, Germany). Dialysate flow rates were 500 mL/min and blood-flow rates were 250–300 mL/min. The systolic blood pressure (SBP) and diastolic blood pressure (DBP) of patients and healthy subjects was measured in the upright sitting position after ≥5 min of rest using an Erka sphygmomanometer (PMS Instruments Limited, Berkshire, UK) with an appropriate cuff size. Two readings were recorded for each individual. The mean value of two readings was defined as the blood pressure. Patients with SBP and DBP >140 mmHg and 90 mmHg, respectively, or who were already on antihypertensive treatment were assumed to be hypertensive.

Twenty-three patients were taking antihypertensive drugs (14 of them on angiotensin-converting enzyme (ACE) inhibitors; 8 receiving angiotensin-II receptor blockers (ARB); and 1 receiving a calcium-channel blocker and an ACE inhibitor). Thirty-two patients were taking calcium-containing phosphate binders.

The study protocol was approved by the Medical Ethics Committee of Selcuk University (Meram School of Medicine, Konya, Turkey). Written informed consent was obtained from all subjects included in the study.

Biochemical analyses

Venous blood samples for biochemical analyses were drawn after an overnight fast before first exchange in PD patients and before the midweek session in patients receiving HD. All biochemical analyses including those for glucose, creatinine, total cholesterol (TC), low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol and plasma triglyceride (TG) concentrations were undertaken using an oxidase-based technique by the Roche/Hitachi Modular System (Mannheim, Germany) in the Central Biochemistry Laboratory of the Meram School of Medicine.

Definition of atherogenic index of plasma

Atherogenic index of plasma was calculated as the logarithmically transformed ratio of the serum triglyceride to HDL-cholesterol.18

Evaluation of EAT

Unenhanced CT was quantified on retrospectively ECG-gated cardiac computed tomography (CT) using 64-slice multidetector computed tomography (MDCT; Sensation 64 Siemens Medical Solutions, Erlangen, Germany). The CT protocol was: slice collimation, 64 × 0.6 mm; gantry rotation time, 0.33 s; pitch, 0.2; tube voltage, 120 kV; and tube current, 600 mAs. If the heart rate (HR) was >65 beats per minute (bpm), HR control was achieved with a beta-blocker.

All images analysis was performed on a dedicated workstation (Leonardo Workstation Siemens Medical Solutions Erlangen, Germany) as previously described.19 Adipose tissue quantification was performed in a semi-automated method that required manual definition of borders. CT attenuation thresholds were used to identify fax pixels (window width −195 to −45 HU; window center −120 HU) to calculate adipose volumes.

Statistical analyses

Statistical analyses were carried out using the Statistical Package for Social Sciences for Windows version 17.0 (SPSS, Chicago, IL). Parametric data are expressed as the mean ± SD. Dichotomous variables were compared using the chi-square test. Statistical differences between parametric data of two groups were analyzed using the Student’s t-test. The Mann–Whitney U test was used to determine differences between non-parametric data. Linear associations between continuous variables were assessed using the Spearman correlation test.

Multivariate linear regression analyses were undertaken to determine independent associations among EAT and other variables. Age, body mass index, AIP, serum LDL, hs-CRP, albumin and dialysis vintage were entered into the regression model as independent variables, and EAT was entered as a dependent variable. The backward elimination method was preferred in the stepwise regression analysis and p > 0.1 used as a criterion for elimination in this model. At the end of the sixth step, two variables were remained statistically significant in the model. p < 0.05 was considered significant for all tests.

Results

Baseline characteristics of patients

The baseline characteristics of 76 ESRD patients and 42 healthy subjects are shown in Table 1. The etiology of ESRD patients was diabetic nephropathy (n = 16), chronic glomerulonephritis (n = 10), hypertensive nephropathy (n = 23), polycystic kidney disease (n = 7), nephrolithiasis (n = 5) and unknown (n = 15). There were no differences with respect to the following variables between ESRD patients and healthy subjects: age; sex; BMI; predialysis levels of DBP; serum levels of albumin, HDL-cholesterol and hemoglobin. However, ESRD patients had higher serum levels of triglyceride, hs-CRP and AIP when compared to healthy subjects (Table 1).

Table 1. Demograhic and biochemical parameters of ESRD patients and healthy subjects.

Parameters	Healthy subjects (n = 42) (Mean ± SD)	ESRD patient (n = 76) (Mean ± SD)	p Value
Age (years)	51 ± 12	49.5 ± 14	0.82
Male/female	23/19	46/30	0.67
BMI (kg/m^2)	26.1 ± 5.4	26.7 ± 5.3	0.53
SBP (mmHg)	130 ± 8	138 ± 28	0.18
DBP (mmHg)	82 ± 7	86 ± 16	0.09
Hemoglobin (g/dL)	12.7 ± 1.9	11.5 ± 1.6	0.001
Albumin (g/dL)	3.8 ± 0.50	3.8 ± 0.49	0.67
Calcium (mg/dL)	–	9.1 ± 0.9	–
Phosphorus (mg/dL)	–	4.6 ± 1.3	–
Parathormone (pg/mL)	–	386 ± 403	–
LDL cholesterol (mg/dL)	109 ± 25	121 ± 35	0.28
HDL cholesterol (mg/dL)	38 ± 11	36 ± 13	0.39
Triglyceride (mg/dL)	110 ± 55	167 ± 91	<0.001
hs-CRP (mg/dL)	9.6 ± 2.9	14.5 ± 16	0.99
AIP	0.43 ± 0.25	0.62 ± 0.26	0.001
Epicardial adipose tissue (cm^3)	104 ± 48	160 ± 76	<0.001

Evaluation of epicardial adipose tissue

Total EAT volume was significantly higher in ESRD patients compared with healthy subjects (160 ± 76 cm³ vs. 122 ±38 cm³, p = 0.04, respectively). When patients were classified according to their median value of EAT as Group1, EAT <136 cm3 and Group 2, EAT ≥ 136 cm³, only AIP levels were found to be statistically significant between two groups (0.65 ±0.25 vs. 0.49 ± 0.27, p = 0.048, respectively) (Table 2).

Table 2. Demographic features and biochemical values of ESRD patients according to EAT group.

Parameters	Group 1 (EAT < 136 cm^3) (Mean ± SD) (n = 37)	Group 2 (EAT ≥ 136 cm^3) (Mean ± SD) (n = 39)	p Value
Age (years)	43.4 ± 13.5	55.2 ± 12.3	<0.001
DM n (%)	3 (7.1%)	3 (10.7%)	0.46
Dialysis vintage (months)	52 ± 27	61 ± 48	0.50
Kt/V	1.85 ± 0.53	1.82 ± 0.68	0.32
BMI (kg/m^2)	25.1 ± 4.5	28.3 ± 5.6	0.005
SBP (mmHg)	143 ± 33	134 ± 24	0.22
DBP (mmHg)	81 ± 11	89 ± 14	0.34
Uric acid (mg/dL)	4.8 ± 1.1	5.3 ± 1.2	0.09
Calcium (mg/dL)	9.2 ± 1.0	9.0 ± 0.9	0.25
Phosphorus (mg/dL)	4.8 ± 1.3	4.4 ± 1.3	0.21
Parathormone (pg/mL)	396 ± 356	378 ± 449	0.29
Albumin (g/dL)	3.8 ± 0.4	3.8 ± 0.5	0.90
LDL (mg/dL)	122 ± 39	120 ± 32	0.98
Triglyceride (mg/dL)	126 ± 69	176 ± 97	0.09
HDL (mg/dL)	47 ± 11	50 ± 28	0.37
Total cholesterol (mg/dL)	178 ± 51	190 ± 45	0.71
hs-CRP (mg/L)	13 ± 16	13 ± 11	0.34
AIP	0.49 ± 0.27	0.65 ± 0.25	0.05

SBP, systolic blood pressure; DBP, diastolic blood pressure; DM, diabetes mellitus; LDL, low-density lipoprotein; HDL, high-density lipoprotein; hs-CRP, high sensitive C-reactive protein.

Correlation analysis between EAT and other variables

There was a statistically significant relationship between EAT, BMI and AIP in ESRD patients (r = 0.42, p < 0.001 and r = 0.25, p = 0.028, respectively). The correlation between EAT and AIP was shown in Figure 1. However, EAT was not found to be correlated with AIP in healthy subjects (r = 0.19, p = 0.21) (Figure 2).

Figure 1. The correlation between EAT and AIP in ESRD patients.

Figure 2. The correlation between EAT and AIP in healthy controls.

Correlation analysis between AIP and other variables

AIP was found to be positively correlated with BMI, serum albumin and total cholesterol as well as EAT (Table 3).

Table 3. Univariate correlation analysis of parameters of ESRD patients related to AIP.

Parameters	r	p Value
Albumin (g/dL)	0.245	0.033
BMI (kg/m^2)	0.229	0.047
Total cholesterol (mg/dL)	0.310	0.006
LDL cholesterol (mg/dL)	0.06	0.67
EAT (cm3)	0.252	0.0028

BMI, body mass index; EAT, epicardial adipose tissue; r, correlation coefficient.

Predictors of EAT

The stepwise linear regression analysis revealed that age, as well as BMI were independent predictors of EAT. Regression results are shown in Table 4.

Table 4. The independent predictors of EAT in linear regression analysis.

Parameters	Standardized Beta‡	t	p Value	95% CI
STEP 1
Age (years)	0.501	3.680	0.001	1.24–4.31
BMI (kg/m^2)	0.326	2.384	0.023	0.785–9.858
AIP	0.167	1.036	0.307	−44.56–137.31
LDL (mg/dL)	0.101	0.778	0.440	−0.385–0.857
Albumin (d/dL)	0.044	0.268	0.790	−49.55–64.59
hs-CRP (mg/dL)	0.004	0.023	0.982	−1.31–1.348
Dialysis vintage (months)	−0.120	−0.873	0.389	−9.045–3.61
r^2 = 0.509
Adjusted r^2 = 0.407
p = 0.001
STEP 6
Age (years)	2.509	3.660	0.45	1.12–3.89
BMI (kg/m^2)	6.271	3.102	0.38	2.18–10.36
r^2 = 0.509
Adjusted r^2 = 0.440
p < 0.001

Discussion

There were five main findings of this study. First, AIP and EAT measurements were increased in ESRD patients when compared with healthy subjects. Second, EAT was significantly correlated with BMI and AIP in ESRD patients. Third, there was a statistically significant relationship between AIP and BMI, serum albumin and total cholesterol in ESRD patients. Fourth, AIP was significantly increased in ESRD patients who have higher EAT volumes. Lastly, age as well as BMI was found to be independent predictor of EAT.

Despite recent developments in the management of ESRD patients, CVD remains the main cause of morbidity and mortality in this population.20,21 This heightened risk can be attributed to many factors including advanced age, atherosclerosis, hypertension, dyslipidemia, anemia, hyperparathyroidism, chronic inflammation, diabetes (and its macrovascular and microvascular complications), left-ventricular hypertrophy, malnutrition and vascular calcification. Disturbed lipid profile was found to be closely related to atherosclerosis in general population22,23 and in CKD patients.4 Hemodialysis patients usually display increased concentrations of triglyceride-rich lipoproteins including very low density lipoprotein (VLDL), reduced serum levels of HDL-cholesterol and elevated concentrations of lipoprotein a (Lp(a)). Total and LDL-cholesterol values are usually within normal limits or slightly reduced in this population.24

Despite the neutral effect of dialysis on serum lipid profile, certain dialysis-related parameters may significantly affect lipoprotein metabolism and modify the features of dyslipidemia in HD patients. It has also been demonstrated that the use of high-flux membranes is closely associated with a significant reduction in serum triglyceride levels as well as by an increase in HDL-cholesterol levels.25,26 This improvement might be attributed to an increase in the apolipoprotein C-II/C-III ratio which increases the activity of lipoprotein lipase and facilitates the intravascular lipolysis of triglyceride-rich lipoproteins and the type of dialysate that may significantly affect the serum levels of lipoproteins in this population.4 Additionally, chronic heparin usage during HD sessions might be another important factor in which heparin releases lipoprotein lipase from the endothelial surface and this may result in defective catabolism of triglyceride-rich lipoproteins via decreasing the serum levels of lipoprotein lipase.4 Lastly, the new non-calcium-containing phosphate binder sevelamer hydrochloride was found to be effective in decreasing total cholesterol and apolipoprotein B in HD patients.27

Recent studies indicated that PD patients have higher serum triglyceride, Lp (a), small dense LDL-cholesterol levels and reduced HDL-cholesterol concentrations when compared to HD patients.28,29 This heightened risk might be attributed to many factors including lose of proteins via peritoneal membrane, increased absorption of glucose from dialysate and increased serum levels of insulin in this population.30 In accordance to these study results, usage of icodextrin-containing dialysis solutions was found to be superior to glucose-containing dialysis solutions in terms of lowering serum triglycerides and small dense LDL concentrations.31

EAT, like abdominal visceral adipose tissue, is a metabolically active tissue that can secrete proinflammatory cytokines and utilize free fatty acids (FFAs).9 Despite the smaller adipocyte size, EAT has a higher rate of uptake and secretion of fatty acids than other visceral fat depots. In health, EAT may act as a “buffering system” by scavenging excess FFAs that are toxic to the myocardium. However, under ischemic conditions, EAT may serve as a local energy source by providing FFAs for the increased metabolism of the myocardium.32,33 In a recent study, the authors concluded that EAT acts as an extremely active organ that produces several bioactive adipokines, as well as proinflammatory and proatherogenic cytokines.9–13 Levels of most of these proinflammatory cytokines were, in general, increased and these cytokines were found to be associated with atherosclerosis in ESRD patients.34–36

In this study, we demonstrated that AIP and EAT measurements were increased in ESRD patients when compared with healthy subjects. Increased EAT was found to be positively correlated with advanced age, BMI and AIP in ESRD patients. AIP was significantly increased in ESRD patients who have higher EAT volumes. This association might be attributed to increased levels of proinflammatory cytokines secreted by EAT. Studies have shown an important association between obesity and cardiovascular morbidity and mortality in CKD populations.20,37,38 Our findings were consistent with those of other studies. Therefore, defining EAT and AIP could be valuable for further studies such as the determination of cardiovascular risk stratification in ESRD patients.

Our study had two main limitations. First, this was a cross-sectional analysis of ESRD patients focusing on AIP and EAT. Second, the sample size was relatively small. This was not a prospective controlled study, so we cannot draw cause-and-effect relationships from our findings.

In conclusion, we found a relationship between EAT as defined by MDCT and AIP in ESRD patients. Further, clinical and experimental studies are needed to determine the relationship between EAT and AIP.

Declaration of interest

This study was supported by Scientific Investigation and Project Foundation of Selcuk University Meram School of Medicine.