Arterial Compliance in Adolescents and Young Adults Receiving Chronic Hemodialysis

Abstract

Increased vascular stiffness is an established risk marker of cardiovascular diseases (CVD) in adults with end-stage renal disease, but its role in pediatric patients remains to be defined. We prospectively examined arterial compliances of adolescents and young adults on hemodialysis (HD) using diastolic pulse wave analysis (DPWA). Each of the ten HD patients (age 17.3 ± 3.9 years; mean ± SD) had two DPWA tests within a three-week time period. DPWA measurement was performed before and hourly until the end of three-hour HD. Pre-HD large artery elasticity index (LAEI) was reduced in one patient and small artery elasticity index (SAEI) was reduced in another. Neither patient was hypertensive. Eight other patients had a reduction in both LAEI and SAEI. Among them, six patients had systolic and/or diastolic hypertension, and the other two were normotensive. Serum phosphorus correlated positively with stroke volume and cardiac output indices and negatively with SAEI. The reduction in BP during HD correlated with the amount of fluid removal. LAEI and SAEI were unchanged during HD. In conclusion, the reduction in LAEI and/or SAEI was observed in four normotensive patients, suggesting hypertension was not the only contributing factor for the reduced arterial compliances in our patients. The association between SAEI and serum phosphorus suggests that SAEI derived from DPWA can potentially be an early non-invasive, operator-independent, and volume-independent marker of CVD in adolescents and young adults receiving HD. Longitudinal studies with a larger sample size are needed to confirm our observation and speculation.

Keywords:

• arterial compliance
• hemodialysis
• adolescents

INTRODUCTION

Increased vascular stiffness is an established risk marker of cardiovascular diseases (CVD) in adults with end-stage renal disease (ESRD),[1] but its role in pediatric patients remains to be defined. It has been shown that 43% of young adults who started dialysis before 14 years old had left ventricular hypertrophy (LVH). Among them, 17% of patients had aortic valve calcification.[2] The stiffening of carotid arteries was reported in young adults who had dialysis since childhood.[3] These findings suggest that pediatric dialysis patients are at risk of developing vascular changes when they reach adulthood, which raises the possibility that arterial stiffness may be a surrogate marker of CVD risk in pediatric dialysis patients. In this regard, diastolic pulse wave analysis (DPWA) of the radial artery using a HDI/PulseWave CR-2000 Research CadioVascular Profiling System® (Hypertension Diagnostics Inc., Eagan, Minnesota, USA; abbreviated as HDI/CR-2000®) is a noninvasive, operator-independent technique that can be utilized to assess vascular change.[4] In this study, we prospectively examined the use of HDI/CR-2000® in young adults and adolescents on hemodialysis (HD) therapy.

PATIENTS AND METHODS

This study was approved by the Institutional Board Review of the University of Michigan Health System. Informed consents and assents were obtained from the guardians and participants, respectively, before the study. As recommended by the manufacturer, only patients older than 12 years were recruited. Between September 2004 and January 2005, ten patients receiving chronic HD therapy in pediatric dialysis unit at C.S. Mott Children Hospital, University of Michigan Health System, were recruited. A total of 10 patients were enrolled in the study, seven males and three females. Four patients were Caucasians, four were African Americans, one was of Hispanic origin, and one was of Asian descent. Mean age was 17.3 ± 3.9 years (range 13–23.9) and the duration of HD was 12.7 ± 11.1 months (range 2–38). Causes of end-stage renal diseases were renal dysplasia (n = 1), chronic glomerulonephritis (n = 3), focal segmental glomerulosclerosis (n = 4), obstructive uropathy (n = 1), and bone marrow transplant nephropathy (n = 1). These patients were receiving three three-hour HD sessions weekly using a CT190G cellulose triacetate dialyzer (Baxter Healthcare Corp., Deerfield, Illinois, USA) and a dialysate containing 2.5 mEq/L of calcium and 40 mEq/L of bicarbonate. After consents were obtained, each subject underwent DPWA tests randomly at two HD sessions within a three-week time period. At each HD session, DPWA test was performed immediately before and hourly until the end of HD therapy. Blood pressure (BP) of the brachial artery was obtained on the non-angioaccess arm right before DPWA test using oscillometric method. BP index was the ratio of BP level to the 95^th percentile BP for age, sex, and height in subjects who were younger than 18 years old.[5] The 95^th percentile BPs for patients older than 18 years were 140 and 90 mmHg for systolic BP and diastolic BP, respectively. Hypertension was defined by a BP index of ≥ 1.0. Laboratory data corresponding to each DPWA test were those closest to the date when the test was performed (within 10 days). The Ca × P product was the product of serum calcium and phosphorus levels [(mg/dL)²].

Diastolic Pulse Wave Analysis of the Radial Artery

This test was performed using a piezoelectric tonometer sensor array of HDI/CR-2000®[4] according to manufacturer's recommendation. It analyzes the diastolic pulse waveforms and separately calculates the capacitative compliance of large conduit arteries (large artery elasticity index, LAEI, mL/mmHg × 10) and the oscillatory or reflective compliance of small microcirculatory arteries (small artery elasticity index, SAEI, mL/mmHg × 100).[6] Both LAEI and SAEI correlated with the aortic distensibility as measured by MRI in normotensive or hypertensive adults[6] and the arterial compliances derived from intra-arterial brachial artery waveform analysis in healthy adults.[7] Cardiac output index from HDI/CR-1000® correlated with those measured by cardiac catheterization.[8] Variations of intra-visit and inter-visit repeated measurements of LAEI and SAEI were less than 4% in a multi-center study that involved seven study sites and 308 subjects.[9] Neither arterial compliance was significantly different when measured 1-hr or 52 days apart in healthy subjects.[8]

After a 15-minute rest, the tonometer was placed on the non-angioaccess wrist overlying the radial bony prominence. The tonometer was adjusted carefully to assure good contact with the skin while exerting minimal pressure on the artery. Pulse waves were recorded over 30 seconds, and the wave forms were calibrated to the brachial BP using a calibration system internal to the HDI/CR-2000®. A computer-based third-order four-element Windkessel model of the circulation was used to match a diastolic pressure decay of the waveforms and quantity changes of LAEI and SAEI. Cardiac output was calculated from a multivariate algorithm using cardiac ejection time derived from the radial artery waveform[4] and normalized to the body surface area (cardiac output index, COI, L/min/m²). The mean BP was derived from integrating the area under the pulse curve and calculating the mean area during the 30-second recording. Systemic vascular resistance (SVR, dyne⋅sec⋅cm⁻⁵) was the ratio of mean arterial pressure to cardiac output. Stroke volume was calculated as cardiac output divided by heart rate and indexed to the body surface area (stroke volume index, SVI, mL/beat/m²). The percent changes in DPWA variables during HD therapy were calculated as:

Results of the two DPWA measurements in each patient were averaged before analysis.

Age-related normal values (defined as more than 5^th percentile for a given age and gender group) of LAEI and SAEI are available in a published multi-center study[9]:

• for subjects who were 15–19 years old, LAEI > 16.1 and SAEI > 7.7;

• for those who were 20–29 years old, LAEI > 15.5 and SAEI > 7.1.

The preliminary study by Bennett et al.[10] in 49 normotensive children with a mean age of 12.2 years old reported a LAEI of 14.7 ± 4.9 mL/mmHg × 10 and a SAEI of 8.8 ± 2.7 mL/mmHg × 100 (both mean ± SD) without gender difference. Assuming the data followed a normal distribution and by defining “normal” as those above 5^th percentile, a LAEI > 5.1 and a SAEI > 3.5 were considered normal for this age group in our study.

Hemodynamic Data during Hemodialysis

Critline® (Hema Metrics Corp., Salt Lake City, Utah, USA) was routinely used in our unit to guide patient's fluid removal. It monitors the real-time pre-dialyzer hematocrit and therefore the changes in blood volume. In this study, the ultrafiltration volume was the amount of fluid removed per kg body weight during the entire HD treatment.

Statistics

Paired Student's t-test was used to compare the pre- and post-HD DPWA variables. Pearson's correlation product was used to examine correlations between two continuous variables. All data are mean ± SD. A p ≤ 0.05 is considered statistically significant.

RESULTS

Eight patients in our study were greater than 15 years old. Using the age-related reference values from Zimlichman et al.,[9] one patient had a low SAEI, and the other seven patients had reduced LAEI and SAEI. Two other patients were younger than 15 years. Using the normal values of LAEI > 5.1 and SAEI > 3.5 by Bennett et al.,[10] one patient had a lower SAEI and the other one had a lower LAEI and SAEI. All together, there were eight patients in our study who had reduced LAEI and SAEI. Among them, two patients had both systolic and diastolic hypertension, four patients had systolic hypertension, and two patients did not have hypertension (see Table 1). Five of the six patients with systolic and/or diastolic hypertension were taking calcium channel blockers (n = 4), angiotensin-converting enzyme inhibitors (n = 4), metoprolol (n = 1), and labetalol (n = 1), while the sixth one was not taking antihypertensive medications. Two normotensive patients with reduced LAEI and SAEI were taking angiotensin-converting enzyme inhibitors. The two patients with either a reduced LAEI or a reduced SAEI were normotensive and they did not take antihypertensive medications.

Table 1 Numbers of subjects with hypertension according to their arterial compliances (n = 10)

	Systolic hypertension	Diastolichypertension
Normal LAEI and low SAEI, n = 1	0 (0%)	0 (0%)
Low LAEI and small SAEI, n = 1	0 (0%)	0 (0%)
Low LAEI and low SAEI, n = 8	6 (75%)	2 (25%)

Data are number (%).

Abbreviations: LAEI = large artery elasticity index, mL/mmHg × 10; SAEI = small artery elasticity index, mL/mmHg × 100.

Correlation coefficients between the pre-hemodialysis DPWA variables and demographic and laboratory data are shown in Table 2. Height correlated positively with LAEI. Serum intact PTH level correlated positively with the pulse pressure. Serum phosphorus correlated negatively with SAEI and positively with SVI and COI.

Table 2 Correlation coefficients between the pre-hemodialysis DPWA variables and demographic and laboratory data (n = 10)

	SBP index	DBP index	PP	LAEI	SAEI	SVI	COI	SVR
HD month, 12.7 ± 11.1	NS	NS	NS	NS	NS	NS	NS	NS
Height, 156 ± 12 cm	NS	NS	NS	0.67	NS	NS	NS	NS
Hemoglobin, 11.0 ±1.7 g/L	NS	NS	NS	NS	NS	NS	NS	NS
Hematocrit, 34.2 ± 5.6%	NS	NS	NS	NS	NS	NS	NS	NS
Serum protein, 6.47 ± 1.20 g/dL	NS	NS	NS	NS	NS	NS	NS	NS
Serum albumin, 3.36 ± 0.87 g/dL	NS	NS	NS	NS	NS	NS	NS	NS
Serum intact PTH, 441 ± 242 pg/mL	NS	NS	0.65	NS	NS	NS	NS	NS
Serum calcium, 8.77 ± 1.59 mg/dL	NS	NS	NS	NS	NS	NS	NS	NS
Serum phosphorus, 6.91 ± 2.24 mg/dL	NS	NS	NS	NS	−0.72	0.84	0.66	NS
Ca×P product, 60.6 ± 22.9 (mg/dL)^2	NS	NS	NS	NS	NS	NS	NS	NS

*p < 0.05

Abbreviations: SBP index = systolic blood pressure/95^th percentile systolic blood pressure, 1.09 ± 0.16; DBP index = diastolic blood pressure/95^th percentile diastolic blood pressure, 0.91 ± 0.17; PP = pulse pressure, 65.1 ± 7.3 mmHg; LAEI = large artery elasticity index, 8.73 ± 2.98 mL/mmHg × 10; SAEI = small artery elasticity index, 4.78 ± 2.01 mL/mmHg × 100; SVI = stroke volume index, 36.6 ± 7.5 mL/beat/m²; COI = cardiac output index, 3.13 ± 0.33 L/min/m²; SVR = systemic vascular resistance, 1664 ± 348 dyne⋅sec⋅cm⁻⁵; NS, not statistically significant.

During HD therapy, there were significant decreases in systolic BP, diastolic BP, and SVI, with a concomitant increase in heart rate. LAEI, SAEI, COI, and SVR remained unchanged (see Table 3). The changes in DPWA variables during dialysis did not correlate with ultrafiltration volume or the reduction in blood volume (data not shown). However, the ultrafiltration volume (29.5 ± 11.2 mL/kg) correlated significantly (r = −0.64, p = 0.045) with the percent reduction in diastolic BP (−9.23 ± 8.88%) and marginally (r = −0.60, p = 0.066) with the percent reduction in systolic BP (−8.14 ± 3.56%).

Table 3 Changes in DPWA variables during hemodialysis therapy (n = 10)

	Pre-hemodialysis	Post-dialysis
Systolic BP, mmHg	142 ± 17	129 ± 13
Diastolic BP, mmHg	76.7 ± 12.7	68.2 ± 8.9
Heart rate, beats/min	86.7 ± 14.0	95.5 ± 11.8
Pulse pressure, mmHg	65.1 ± 7.3	61.6 ± 11.5
LAEI, mL/mmHg × 10	8.73 ± 2.98	10.0 ± 5.0
SAEI, mL/mmHg × 100	4.78 ± 2.01	4.67 ± 2.00
Stroke volume index, mL/beat/m^2	36.6 ± 7.5	30.1 ± 5.6
Cardiac output index, L/min/m^2	3.13 ± 0.33	2.80 ± 0.27
Systemic vascular resistance,dyne.sec.cm^−5	1664 ± 348	1763 ± 516

*p < 0.05.

Abbreviations: SBP index = systolic blood pressure/95^th percentile systolic blood pressure, 1.09 ± 0.16; DBP index = diastolic blood pressure/95^th percentile diastolic blood pressure, 0.91 ± 0.17; PP = pulse pressure, 65.1 ± 7.3 mmHg; LAEI = large artery elasticity index, 8.73 ± 2.98 mL/mmHg × 10; SAEI = small artery elasticity index, 4.78 ± 2.01 mL/mmHg × 100; SVI = stroke volume index, 36.6 ± 7.5 mL/beat/m²; COI = cardiac output index, 3.13 ± 0.33 L/min/m²; SVR = systemic vascular resistance, 1664 ± 348 dyne⋅sec⋅cm⁻⁵; NS, not statistically significant.

DISCUSSION

The significant findings from our study are that LAEI and/or SAEI were reduced in all patients, regardless of whether they were hypertensive or not. Serum phosphorus level correlated negatively with SAEI and positively with SVI and COI. During HD, the reduction in blood pressure was related to the amount of fluid removal. There was no significant change in either LAEI or SAEI during HD treatment.

Consistent with findings from adult studies employing different methods to measure arterial stiffness,[11] we observed a reduction in both LAEI and SAEI in eight patients and either a reduced LAEI or a reduced SAEI in the other two patients. Because the reduced LAEI and/or SAEI were present in four normotensive patients, it is likely that hypertension was not the only contributing factor. Although the reduced LAEI in our patients might be related to their height, it was not observed in SAEI. SAEI was selectively decreased by inhibition of nitric oxide synthase (NOS),[12] suggesting its association with the endothelial function. In vivo gene transfer of endogenous NOS to the wall of rat carotid arteries inhibited cell growth and neointimal formation following balloon injury.[13] SAEI was a diagnostic marker for the extent of coronary artery stenosis[14] as well as a predictor of CVD in normal and hypertensive individuals.[15] Therefore, SAEI may be a candidate marker of cardiovascular remodeling in adolescents and young adults receiving HD. This speculation needs to be tested on longitudinal studies.

Our study appears to be the first to demonstrate a link between serum phosphorus levels and some of the DPWA variables. The observation that pre-dialysis SAEI correlated negatively with the serum phosphorus level is consistent with a prior study in which carotid arterial stiffness correlated with intact PTH level and the doses of vitamin D and phosphorus binders in 16 pediatric dialysis patients.[16] We also observed a correlation between serum phosphorus and stroke volume index and cardiac output index. One possible explanation for the above observations in our study is the association between arterial calcification and vascular remodeling.[17],[18] Further studies examining the association between arterial calcification and DPWA variables are needed to confirm this speculation.

Our study also appears to be the first to examine changes in DPWA variables during HD in young adults and adolescents. We observed correlations between the amount of fluid removal and the reduction in both systolic and diastolic blood pressure during HD. This would explain why stroke volume index was decreased and heart rate was increased. The lack of changes in LAEI and SAEI is consistent with these findings, as arterial compliance is determined by the changes in volume relative to those in pressure (ΔV/ΔP).[19] It could also be partly due to a paradoxical withdrawal of sympathetic vasoconstrictor drive during HD.[20] However, both Cohen et al.[21] and Gadegbeku et al.[22] observed a significant decrease in SAEI and a stable BP, pulse pressure, and heart rate during HD in their adult patients. Reasons for this discrepancy are not clear. Our patients were younger than patients in their studies. The difference in the types of antihypertensive medications between our study and theirs might be another explanation, as calcium channel blockers,[23] angiotensin-converting enzyme inhibitors,[24] and angiotensin receptor antagonists[23] affect LAEI and SAEI. Further studies are needed to clarify these findings.

Our findings were consistent with a previous study in which examination with Pulse Wave Velocity and Augmentation Index indicated an increased arterial stiffness in children receiving chronic hemodialysis.[25] In addition, the measured arterial stiffness was not affected by dialysis but was associated with serum calcium and phosphorus levels, findings that were similar to ours. In this aspect, our method is non-invasive and operator-independent, suitable for practical clinical application. Furthermore, longitudinal studies have demonstrated that changes in arterial compliances derived from DWPA corresponded to the initiation and discontinuation of pharmacotherapy,[26] suggesting its potential use for longitudinal monitoring of treatment.

One limitation of our study is that we used the reference values of LAEI and SAEI by Zimlichman et al.[9] and Bennett et al.[10] as our control values. Although we believe it was reasonable to use the data by Zimlichman et al.[9] for those older than 15 years, as their multicenter study demonstrated an intra-patient variations of less than 4%, we can not explain the discrepancy in reference values in LAEI and SAEI between studies by Zimlichman et al.[9] and Bennett et al.[10] It should be noted that the latter were obtained in a limited number of subjects and may not be representative of all healthy children. Another limitation of our study is the small sample size and the cross-section examination. This might have affected our statistical analysis and conclusion. We did not examine the association between fasting cholesterol or LDL-cholesterol and DPWA variables in our study because they were not uniformly available in all patients. However, this should not have affected our findings that pre-dialysis DPWA variables were closely associated with calcium-phosphorus metabolism. It has been shown that the changes in elastic- and muscular-type arteries in patients with renal failure are independent of the presence of atherosclerotic plaques.[27] It should be noted that the validity of Windkessel model theory has been questioned, as the compliances measured in the legs were greater than those in the wrists.[28] It is possible that this difference in leg and wrist compliances is affected by regional factors that are still unclear at present time.

In conclusion, our study demonstrated that LAEI and SAEI from HDI/CR-2000® are reduced in young adults and adolescents receiving hemodialysis, regardless of whether they were hypertensive or not. We observed a negative association between serum phosphorus level and SAEI, which is consistent with the known association between calcium-phosphorus metabolism and arterial stiffness. In addition, SAEI is not affected by fluid removal during HD. These findings suggest that SAEI derived from HDI/CR-2000® can be a candidate marker for the CVD risk in adolescents and young adults receiving hemodialysis therapy. Further studies with a larger sample size are needed to validate our observation.