The impact of dialysis modality on arterial stiffness in patients with end-stage renal disease

Abstract

Background: Arterial stiffness determined by brachial–ankle pulse wave velocity (baPWV) has been established as a powerful predictor of cardiovascular mortality in hemodialysis (HD) patients. There are, however, few studies regarding the comparative impact of different renal replacement therapies (RRTs) on PWV. Therefore, we conducted a cross-sectional study to compare arterial wall properties and cardiac function between patients treated with continuous ambulatory peritoneal dialysis (CAPD) and those with HD who were matched for age, dialysis duration, and blood pressure. Methods: baPWV and transthoracic echocardiography (TTE) were performed in HD patients (n = 23) after 1 h of midweek dialysis session and CAPD patients (n = 26) with empty abdomen after drainage of dialysate. The baseline data were retrospectively reviewed. Results: baPWV was significantly higher in HD patients than in CAPD patients (18.1 ± 2.8 vs. 16.1 ± 2.7 m/s, p = 0.015). TTE revealed significantly increased E/E′, left atrial volume index (LAVI), and inferior vena cava (IVC) diameter index in HD patients compared with CAPD patients (p < 0.05). In a multivariate regression analysis adjusted for dialysis modality, age, systolic BP, residual glomerular filtration rate, diabetes, and echocardiographic parameters, HD was independently associated with increased baPWV. Conclusion: This study showed that HD patients had significantly increased arterial stiffness and severe diastolic dysfunction compared with CAPD patients.

Keywords:

• arterial stiffness
• continuous ambulatory peritoneal dialysis
• dialysis modality
• hemodialysis
• pulse wave velocity

INTRODUCTION

Cardiovascular disease (CVD) is the leading cause of morbidity and mortality in patients with end-stage renal disease (ESRD). Cardiovascular mortality is responsible for more than 50% of overall morality in the ESRD patients.¹ The velocity of pulse wave travel in the aorta reflects vascular stiffness and predicts coronary heart disease, stroke, and death.²

Pulse wave velocity (PWV) is a measure of arterial stiffness and has proved useful in predicting cardiovascular morbidity and mortality in patients with ESRD.^3,4 With regard to those on dialysis, a population with exceptionally high CVD morbidity and mortality, determination of PWV predicted death from CVD even when blood pressure (BP) is well controlled.³ Advancing age, increasing BP, and the presence of diabetes are typically associated with increased PWV.⁵ This increase in arterial stiffness is certainly of multifactorial origin, including vasoactive humoral factors causing change in arterial elasticity as well as structural changes,^6,7 but the exact mechanisms are not clear.

Comparative studies about cardiac function between hemodialysis (HD) and continuous ambulatory peritoneal dialysis (CAPD) patients suggest that hemodynamic and structural changes in heart are similar in both populations, provided that BP control and fluid status are comparable.^6–9 Cardiac and vascular alterations occur in parallel because the heart and vessels are a coupled interactive physiological system and their changes are induced to a great extent by the same hemodynamic abnormalities.

There are, however, few studies regarding the comparative impact of different renal replacement therapies (RRTs) on PWV and cardiac function. Therefore, we conducted a cross-sectional study to compare arterial compliance properties and cardiac function between patients treated with CAPD and those with HD.

PATIENTS AND METHODS

Study population

Before the study entry, we selected 61 patients undergoing HD (n = 30) or PD (n = 31) who were matched for age, dialysis duration, and BP. All patients had been maintained on RRT for more than 3 months and had baseline transthoracic echocardiography (TTE) data with initiation of dialysis therapy. Excluding 12 patients with systolic dysfunction (ejection fraction (EF) < 45%), valvular heart disease, and coronary artery disease (CAD) 6 months prior to the study entry, a total of 49 patients were included in the final analysis. Of the patients, 23 patients received HD at three sessions per week, lasting 4 h per session and 26 were treated with CAPD at three or four 2 L exchanges per day, using standard dialysates containing glucose. The residual glomerular filtration rate (rGFR) was calculated as the mean of creatinine and urea clearance and corrected for body surface area from a 24-h urine collection.¹⁰ All patients provided written informed consent, and the study protocol was approved by the local ethics Institutional Review Board of NHIC Ilsan Hospital.

Assessment of arterial pulse wave velocity and augmentation index

Arterial PWV was determined using transcutaneous Doppler flow recordings and the foot-to-foot method. A VP-2000 PWV® (Nippon Colin Ltd., Komaki city, Japan) system was used to assess PWV. Briefly, waveforms are obtained transcutaneously over the common carotid artery and the right femoral artery, and the time delay (t) is measured between the feet of the two waveforms. The distance (D) covered by the waves is assimilated to the distance measured between the two recording sites. The brachial–ankle PWV (baPWV) was calculated from the equation: (D1 - D2)/T. D1 is the distance between the heart and ankle, D2 is the distance between the heart and brachium, and T is the transit between the right brachial arterial wave and right tibial arterial wave. The distances between the sampling points are automatically calculated from the patient's height and are divided by the time interval for the waveform from each measuring point.^11–13 Augmentation index (AIx) is measured by the difference between the first and second systolic peaks on the aortic pressure waveform divided by the pulse wave height. The baPWV and AIx measurements were performed at 1 h after HD on midweek dialysis session or in CAPD patients with empty abdomen after drainage of dialysate. Two measurements were performed in each arm, and the average value was used for the analysis.

Assessment of transthoracic echocardiography

Transthoracic echocardiographic studies were performed by standard techniques with a 2.5-MHz transducer interfaced with commercially available ultrasound systems (Sequoia C512, Acuson, Inc., Mountain View, California, USA). Measurements were performed by the same investigator according to the methods of the American Society of Echocardiography.¹⁴ Transthoracic studies were performed with the patients in the left lateral decubitus position and images were obtained from parasternal and apical windows. The Doppler study included color, pulse, and continuous-wave examinations. Two-dimensional echocardiographic images of the left atrium (LA) and left ventricle (LV) were obtained in the standard parasternal long-axis and apical two- and four-chamber views with second harmonic imaging. LVEF was assessed by the modified Quinones method.¹⁵ LA volumes were calculated from the parasternal long-axis and apical four-chamber view using the prolate ellipsoid model. For evaluation of diastolic function, sample volume (size 2 mm) of the pulsed wave Doppler was placed between the tips of the mitral leaflets in the apical four-chamber view. Early (E) and late (A) transmitral flow velocities were obtained. Pulsed wave Doppler tissue imaging (DTI) was performed by activating DTI function in the same machine. Sample volume was located at the septal side of the mitral annulus. Early (E′) and late (A′) diastolic mitral annulus velocities and the ratio of early diastolic transmitral flow velocities (E) to early diastolic mitral annulus velocities (E/E′) were obtained. Studies were stored digitally and analyzed off-line. The diameter of the inferior vena cava (IVC) was measured in the hepatic segment just below the diaphragm. The diameter of IVC was expressed as an index to the body surface area in mm/m². Overhydration was defined as IVC diameter index more than 11.5 mm/m².¹⁶

Statistical analysis

Data are expressed as mean ± standard deviation (SD). SPSS (version 11.0; SPSS Inc., Chicago, Illinois, USA) was used for all statistical analyses. Kolmogorov–Smirnov test was used to analyze the normality of the distribution of the parameters measured. The comparisons between the two groups were made by Student's t-test and χ²-test. Skewed data were analyzed with the Mann–Whitney U-test. Relationships between paired parameters were analyzed by Pearson's correlation coefficient. The independent association between serum baPWV and clinical parameters was analyzed further by multiple linear regression analysis adjusting other factors with p-value less than 0.10 on univariate analysis. A p < 0.05 was taken to indicate statistical significance.

RESULTS

Patient characteristics

The demographic features and baseline echocardiographic data of the two groups are shown in Tables 1 and 2. The mean age of the subjects was 54 years and 55% were male. All patients were on antihypertensive treatment. The most common cause of ESRD was diabetes mellitus (DM) (57%), which was followed by hypertension (19%), glomerulonephritis (14%), polycystic kidney disease (2%), others (4%), and unknown diseases (4%). The average spKt/V in HD patients was 1.54 ± 0.30 per dialysis and the average Kt/V in CAPD patients was 2.01 ± 0.52 per week. The dose of erythropoietin was 5452 ± 3127 U/week in patients receiving HD and 3074 ± 1035 U/week in those treated with CAPD. The vascular access types in patients with HD included arteriovenous fistulas in 14 patients and arteriovenous grafts in 9 patients. There were no significant differences in age, gender, the proportion of DM, dialysis duration, angiotensin receptor blocker (ARB) or angiotensin-converting enzyme (ACE) inhibitor use, statin use, and current smoker between the two groups. In addition, laboratory parameters such as hemoglobin, high-sensitivity C-reactive protein (hsCRP), calcium, phosphate, fasting glucose, hemoglobin A1C and intact parathyroid hormone (iPTH) levels, and baseline echocardiographic data measured with initiation of dialysis therapy were not different. However, patients treated with CAPD had better preserved rGFR than those on HD (3.2 ± 2.6 vs. 0.2 ± 0.7 mL/min/1.73 m², p < 0.001). Total cholesterol, triglyceride, low density lipoprotein (LDL) cholesterol, and ferritin levels were significantly higher in patients on CAPD compared with those on HD (p < 0.001) Serum iron and transferrin saturation were significantly higher in patients on CAPD compared with those with HD (p = 0.004).

TABLE 1. Characteristics of patients

	CAPD patients (n = 26)	HD patients (n = 23)	p-Value
Age (years)	53.3 ± 12.5	56.4 ± 12.8	NS
Sex (male/female)	11/15	13/10	NS
Diabetes mellitus, n (%)	12 (46.2%)	15 (65.2%)	NS
Dialysis duration (months)	27.7 ± 17.0	26.5 ± 15.6	NS
CT ratio	0.48 ± 0.05	0.48 ± 0.06	NS
Body mass index (kg/m^2)	23.9 ± 3.6	23.1 ± 3.1	NS
Systolic BP (mmHg)	151.2 ± 23.1	152.0 ± 15.1	NS
Diastolic BP (mmHg)	90.6 ± 12.4	84.6 ± 9.0	NS
ARB or ACE inhibitor, n (%)	22 (84.6%)	18 (78.3%)	NS
Statin, n (%)	13 (50%)	9 (39.1%)	NS
rGFR (mL/min/1.73 m^2)	3.2 ± 2.5	0.2 ± 0.7	<0.001
Albumin (g/dL)	3.06 ± 0.38	3.14 ± 0.39	NS
hsCRP (mg/dL)	0.32 ± 0.25	0.30 ± 0.59	NS
Total cholesterol (mg/dL)	224.7 ± 53.8	134.8 ± 42.6	<0.001
Triglyceride (mg/dL)	159.8 ± 82.0	80.4 ± 59.4	<0.001
LDL cholesterol (mg/dL)	117.2 ± 39.0	65.3 ± 35.9	<0.001
HDL cholesterol (mg/dL)	43.0 ± 15.5	40.4 ± 14.5	NS
Hemoglobin (g/dL)	10.1 ± 1.3	10.1 ± 1.1	NS
Calcium (mg/dL)	8.8 ± 1.0	8.94 ± 0.94	NS
Phosphorus (mg/dL)	5.0 ± 1.6	5.1 ± 1.4	NS
Iron (μg/dL)	84 ± 29	61 ± 23	0.004
Transferrin saturation (%)	33 ± 12	23 ± 10	0.004
Ferritin (ng/mL)	352 ± 177	176 ± 112	<0.001
Fasting glucose (mg/dL)	170 ± 113	193 ± 139	NS
Hemoglobin A1C (%)	6.3 ± 1.5	7.1 ± 2.0	NS
iPTH (pg/dL)	184.1 ± 151.4	247.7 ± 172.4	NS
Current smoker	0 (0%)	3 (13%)	NS

Notes: CT, cardiothoracic; BP, blood pressure; ARB, angiotensin receptor blocker; ACE, angiotensin-converting enzyme; rGFR, residual glomerular filtration rate; hsCRP, high-sensitivity C-reactive protein; LDL, low density lipoprotein; HDL, high density lipoprotein; iPTH, intact parathyroid hormone.

TABLE 2. Baseline transthoracic echocardiographic data of study patients at the initial stage of dialysis therapy

	CAPD patients (n = 26)	HD patients (n = 23)	p-Value
LVEDD^19	54.0 ± 6.2	57.8 ± 5.6	NS
LVESD^19	35.7 ± 8.2	40.3 ± 8.6	NS
LVMI (g/m^2)	126.8 ± 20.3	128.3 ± 21.4	NS
EF (%)	65.7 ± 11.8	66.1 ± 12.5	NS
LAVI (mL/m^2)	31.0 ± 10.4	36.7 ± 9.9	NS
E/E′ ratio	13.1 ± 4.8	15.7 ± 8.6	NS
RVP (mmHg)	40.9 ± 21.3	43.5 ± 21.9	NS
IVC diameter index (mm/m^2)	12.1 ± 0.4	12.4 ± 0.3	NS

Notes: LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LVMI, left ventricular mass index; EF, ejection fraction; LAVI, left atrial volume index; E, early mitral inflow; E′, early diastolic mitral annulus motion; RVP, right ventricular pressure; IVC, inferior vena cava.

Comparison of baPWV and TTE data between CAPD and HD patients

baPWV was significantly higher in HD patients compared with CAPD patients suggesting stiffer arteries in the former (18.1 ± 2.8 vs. 16.1 ± 2.7 m/s, p = 0.015). However, there was no significant difference in AIx between the two groups. TTE revealed significantly elevated E/E′ (17.1 ± 77 vs. 12.7 ± 3.7, p = 0.018), LAVI (41.5 ± 28.0 vs. 28.4 ± 7.0 mL/m², p = 0.025), IVC diameter index (11.1 ± 2.5 vs. 8.7 ± 2.4 mm/m², p = 0.003), and LV end-systolic (35.2 ± 6.8 vs. 31.5 ± 3.3 mm, p = 0.018) and LV end-diastolic diameter (54.6 ± 6.1 vs. 51.0 ± 3.9 mm, p = 0.024) in HD patients compared with CAPD patients (p < 0.05). However, other parameters such as LV mass index (LVMI) and LVEF were similar in both the groups (Table 3).

TABLE 3. Pulse wave velocity and transthoracic echocardiographic data

	CAPD patients (n = 26)	HD patients (n = 23)	p-Value
baPWV (m/s)	16.1 ± 2.7	18.1 ± 2.8	0.015
AIx	1.15 ± 0.09	1.15 ± 0.16	NS
LVEDD^19	51.0 ± 3.9	54.6 ± 6.1	0.024
LVESD^19	31.5 ± 3.3	35.2 ± 6.8	0.018
LVMI (g/m^2)	107.8 ± 28.3	125.3 ± 40.5	NS
EF (%)	67.0 ± 4.6	64.5 ± 7.0	NS
LAVI (mL/m^2)	28.4 ± 7.0	41.5 ± 28.0	0.025
E/E′ ratio	12.7 ± 3.7	17.1 ± 7.7	0.018
RVP (mmHg)	23.5 ± 6.1	29.8 ± 17.8	NS
IVC diameter index (mm/m^2)	8.7 ± 2.4	11.1 ± 2.5	0.003

Notes: baPWV, brachial–ankle pulse wave velocity; AIx, augmentation index; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LVMI, left ventricular mass index; EF, ejection fraction; LAVI, left atrial volume index; E, early mitral inflow; E′, early diastolic mitral annulus motion; RVP, right ventricular pressure; IVC, inferior vena cava.

Factors related to baPWV

Pearson's correlation analysis revealed that baPWV significantly correlated with age (r = 0.380, p = 0.007), dialysis modality (r = 0.348, p = 0.014), systolic BP (r = 0.400, p = 0.004), and IVC diameter index (r = 0.376, p = 0.014) (Table 4).

TABLE 4. Correlation between pulse wave velocity related parameters

	R	p-Value
Age	0.380	0.007
Dialysis modality	0.348	0.014
Dialysis duration	0.055	NS
Body mass index	−0.150	NS
CT ratio	−0.161	NS
Diabetes mellitus	−0.051	NS
Systolic BP	0.400	0.004
Diastolic BP	0.121	NS
rGFR	−0.139	NS
hsCRP	0.056	NS
Albumin	0.047	NS
Total cholesterol	−0.088	NS
LDL cholesterol	−0.100	NS
Triglyceride	−0.175	NS
Calcium	−0.173	NS
Phosphorus	−0.115	NS
iPTH	0.040	NS
Iron	−0.115	NS
Transferrin saturation	−0.041	NS
Ferritin	0.031	NS
Fasting glucose	0.109	NS
Hemoglobin A1C	0.182	NS
Current smoker	−0.022	NS
LVMI	−0.136	NS
EF	−0.185	NS
LAVI	0.133	NS
E/E′	0.277	NS
IVC diameter index	0.376	0.014

Notes: CT, cardiothoracic; BP, blood pressure; rGFR, residual glomerular filtration rate; hsCRP, high-sensitivity C-reactive protein; LDL, low density lipoprotein; iPTH, intact parathyroid hormone; LVMI, left ventricular mass index; EF, ejection fraction; LAVI, left atrial volume index; E, early mitral inflow; E′, early diastolic mitral annulus motion; IVC, inferior vena cava.

In a multiple linear regression analysis for baPWV adjusting dialysis modality, age, systolic BP, rGFR, DM, and echocardiographic parameters, HD was independently associated with increased baPWV (β = 0.525, p = 0.010) (Table 5).

TABLE 5. Multiple regression analysis of pulse wave velocity and significantly associated variables

	β	t-Value	p-Value
Dialysis modality (HD vs. CAPD)	0.525	2.711	0.010
Age	0.397	3.068	0.004
Systolic BP	0.415	3.061	0.004
rGFR	0.232	1.385	NS
Diabetes mellitus	0.107	0.838	NS
IVC diameter index	−0.100	−0.644	NS
E/E′	0.072	0.528	NS

Notes: HD, hemodialysis; CAPD, continuous ambulatory peritoneal dialysis; BP, blood pressure; rGFR, residual glomerular filtration rate; IVC, inferior vena cava; E, early mitral inflow; E′, early diastolic mitral annulus motion.

DISCUSSION

In this study, baPWV was significantly increased in HD patients than in CAPD patients. Our finding is robust because we controlled age and BP, which were recognized as dominant factors related to arterial stiffness.^17,18 There are few cross-sectional studies comparing arterial stiffness between two dialysis modalities. Konings et al.¹⁹ showed that the distensibility coefficient of the right common carotid artery in PD patients was significantly less than in the nonrenal population, and better than in HD patients. This finding is consistent with our results, although arterial stiffness was assessed by a different technique. By contrast, Covic et al.²⁰ showed that CAPD is associated with stiffer arteries compared with matched HD subjects. This discrepancy may be explained by different characteristics of the study population. In their study, most HD patients (73%) were not prescribed any antihypertensive agent whereas all CAPD patients had been taking more than one antihypertensive agent to maintain comparable BP between the two groups. Therefore, it is possible that in that study, CAPD patients might have stiffer arteries due to long-standing hypertension. Moreover, they excluded DM and current users of ACE inhibitors or ARBs. This disparity of the characteristics of the study subjects may have led to the different results. It is uncertain why arterial stiffness is worse in HD patients than in CAPD patients.

One retrospective study showed that arterial stiffness is independently correlated with residual renal function (RRF) in PD patients.²¹ Mimura et al. showed that PWV became significantly faster in HD patients when it was serially assessed at the time of starting dialysis and 1 year after dialysis initiation.²² In this study, they suggested that the more preserved RRF conferred by PD might account for the inhibition of increases in PWV in patients treated with this dialysis modality. Patients treated with PD had a 65% lower risk of RRF loss than those on HD.²³ The better preservation of RRF in stable PD patients corresponded with greater cardiovascular stability such as rare hypotensive episodes and minimizing severe fluctuations in fluid status compared with HD patients.²⁴ Therefore, it can be speculated that compared with PD patients, rapid intravascular contraction and abrupt fluctuation in fluid status due to rapid loss of RRF in HD may be responsible for stiffer artery. In our study, CAPD patients had better preserved RRF than HD patients. However, there was no relationship between baPWV and the degree of RRF.

Of note, in this study, echocardiographic findings showed that LAVI and E/E′ were significantly increased in HD patients than in CAPD patients. LA volume, especially LAVI, is known to be associated with atherosclerosis and adverse cardiovascular outcomes in CKD and ESRD patients.^25,26 A recent prospective cohort study reported that change in LAV predicted incident cardiovascular events in dialysis patients independent of LV systolic function and LV mass.²⁷ In addition, Fujiu et al. showed that PWV was a significant independent determinant of E/E′, a potent diastolic marker, and diastolic LV dysfunction is associated with arterial stiffness in ESRD patients.²⁸ Therefore, it is possible that stiffer arteries in HD patients may partly contribute to the more impaired diastolic dysfunction in this study.

We also showed that IVC diameter index was significantly wider in HD patients by echocardiographic studies, despite the strict measurement of optimal dry weight in HD groups. In addition, IVC diameter index significantly associated with LAVI. The mean IVC diameter index was within the normal range in both the groups;¹⁶ however, there was significantly larger IVC diameter index in HD patients compared with CAPD patients suggesting inadequate volume control in the former group. Although RRF and peritoneal ultrafiltration capability are maintained, PD may be superior to HD for fluid and BP control.²⁹ Although the size of IVC alone cannot fully represent the volume status, in this study, it is likely that CAPD provided better fluid control than HD presumably due to more preserved RRF. Volume overload increases arterial distension, thus resulting in arterial stiffness.³⁰ Therefore, this finding may also explain the worse arterial stiffness in HD patients than in PD patients.

Our study has several limitations. A cross-sectional comparative analysis, despite carefully correcting for all known potential confounders, cannot exclude the possibility of baseline differences between the two dialysis populations and small size of patients may have significant limitation. Besides, given the limitations of this study, the results need to be confirmed in larger randomized clinical trials.

In conclusion, HD patients had increased arterial stiffness and LA volume compared to age-, dialysis duration-, and BP-matched CAPD patients and the increased arterial stiffness in patients with HD may be responsible for severe diastolic dysfunction.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this paper.