|Home | About | Journals | Submit | Contact Us | Français|
Patients with chronic kidney disease (CKD) have a disproportionate risk of cardiovascular disease. This study was designed to assess the association between two noninvasive measures of cardiovascular risk, pulse wave analysis (PWA), and carotid intima-media thickness (IMT), in a cohort of CKD patients enrolled in the Chronic Renal Insufficiency Cohort (CRIC) study.
Three hundred and sixty-seven subjects with CKD enrolled in the CRIC study at the University of Pennsylvania site (mean age 59.9 years, blood pressure 129/74 mm Hg, estimated glomerular filtration rate 48 ml/min/1.73 m2, IMT 0.8 mm) had both carotid IMT and PWA measurements. Carotid ultrasound was also used to determine the presence of plaque. PWA was used to determine augmentation index (AI), amplification ratio (AMPR), aortic pulse pressure (C_PP), and central aortic systolic pressure (C_SP).
IMT was significantly associated with all PWA-derived measures. However, on multivariable linear regression analysis, only AMPR (regression coefficient −0.072, P = 0.006), C_PP (regression coefficient 0.0025, P < 0.001), and C_SP (regression coefficient 0.0017, P < 0.001) remained significantly associated with IMT. The prevalence of carotid plaque in the cohort was 59%. Of the PWA-derived measures, only C_PP was significantly associated with the presence of carotid plaque (P < 0.001).
PWA-derived measures are associated with carotid IMT and plaque in the CKD. Of these measures, C_PP was most associated with carotid IMT and plaque.
There is a substantial incidence and prevalence of chronic kidney disease (CKD) in the United States, with poor outcomes and profound economic implications.1 The burden of morbidity and mortality from CKD derives from the progression to end-stage renal disease, and the disproportionate risk of cardiovascular disease (CVD).2 Carotid intima-media thickness (IMT) has emerged as a noninvasive means to assess cardiovascular (CV) risk. An IMT measurement above the 95th percentile is predictive of coronary heart disease risk.3 Change in IMT in participants with CKD over 2 years has been associated with cardiovascular events.5
Measures of large artery stiffness are increasingly recognized to provide information about CV risk in certain populations. Pulse wave analysis (PWA) is a useful tool to study arterial wave reflection in the central circulation. Augmentation index (AI) measures the magnitude and timing of wave reflection. In very young, healthy individuals, the reflected wave returns to the central circulation after a longer time interval and arrives during diastole that augments coronary perfusion pressure. In health, the amplitude of the returning wave is relatively low because healthy elastic arteries dampen its magnitude. In disease states, such as hypertension, the arteries become stiffer resulting in a large magnitude of the reflected wave and an increased pulse wave velocity. Thus, the reflected wave will return to central circulation earlier during systole hereby increasing the cardiac workload. This contributes to left ventricular hypertrophy and reduces coronary perfusion pressure.6 Central aortic systolic pressure (C_SP) is an estimate of aortic systolic pressure derived from the radial artery waveform. Devices used in PWA often employ applanation tonometry of a peripheral artery and then, by means of a transfer algorithm, estimate C_SP and the aortic pulse pressure (C_PP).7
PWA provides prognostic information above and beyond that from traditional CV risk factors in the general population,8 but its significance in patients with CKD who are not on dialysis is limited.9 PWA measurements are noninvasive, easily made and well-suited to large epidemiological trials. PWA is a highly reproducible measurement in CKD.10 Central pressures are better predictors of atherosclerosis, as measured by carotid IMT and plaque, compared to traditional brachial blood pressures.11 Central pulse pressures (PPs) predict CVD outcomes in normal, hypertensive, and end-stage renal disease populations.12,13 Increased AI is associated with angiographic evidence of coronary artery disease in CKD.14
Although PWA and IMT measurements provide information about CV risk, little is known about their association among CKD patients. The radial artery tonometry procedures used in PWA are more easily and quickly performed, and seem better suited for epidemiology usage compared with carotid ultrasonography. Because carotid IMT and carotid plaque assessments, although technically difficult to obtain, are considered better measures of CV risk compared to PWA, we sought to determine how well common radial PWA parameters, and which ones in particular, correlated best with these more generally accepted parameters. We hypothesized that PWA would correlate with IMT and carotid plaque, which are established predictors of CVD. Further, we anticipated that all PWA-derived measures of large artery stiffness would not be equivalent in their prediction of carotid IMT. Thus, we sought to examine which of the PWA parameters, AI, C_SP or C_PP, best correlates with IMT and carotid plaque among patients with CKD.
Subjects were eligible to participate if they had enrolled in the Chronic Renal Insufficiency Cohort Study (CRIC) at the University of Pennsylvania site and also participated in the CRIC Carotid IMT Ancillary Study. The study design and methods of the CRIC study have been described elsewhere.15 Briefly, adults between the ages of 21 and 74 years with an estimated glomerular filtration rate between 20 and 70 ml/min/1.73 m2 according to the simplified Modification of Diet in Renal Disease equation and not on dialysis were recruited for inclusion in the study. Informed consent was obtained from all participants. This study was approved by the University’s Institutional Review Board.
Blood pressure measurements were obtained using an aneroid sphygmomanometer (Welch Allyn Tycos classic handheld aneroid). A similar Tyco aneroid device was recently shown to be an adequate substitute for standard mercury sphygmomanometry.16 All blood pressures taken in this study were performed by research coordinators trained and certified to American Heart Association’s standards. Measurements were done in triplicate in the seated position, and the mean value was used for analysis. Diabetes was determined by patient self-report, or currently taking insulin or hypoglycemic medications. Hypertension was defined as use of antihypertensive medications or a systolic blood pressure >140 mm Hg or diastolic blood pressure >90 mm Hg. Serum creatinine was measured at the University of Pennsylvania laboratory and calibrated based on standard measurements made from the Cleveland Clinic Foundation laboratory in Cleveland, OH.15
PWA provides measures of large artery stiffness and function, such as AI, C_PP, and C_SP. PWA was performed noninvasively using the SphygmoCor system (AtCor Medical, West Ryde, Australia). Subjects typically had their blood pressure performed, assumed a supine position and then underwent the PWA study. Coordinators obtaining measurements were all trained by one investigator (RRT) and certified to perform measurements with yearly retraining. Reproducibility of measurements made by this coordinator group has been published.14 After obtaining a brachial blood pressure, as described above, applanation tonometry of the radial artery was performed. This was performed with the patient supine after 5 min of rest. Tonometry was performed using the right radial artery for a period long enough to obtain an acceptable pressure waveform (10 s of recorded data), as determined by visual inspection and the software’s built-in quality control parameters. The SphygmoCor software uses a validated mathematical algorithm to convert the radial waveform into a corresponding central waveform.17 The data obtained from the radial arterial waveform are used to estimate C_SP, C_PP, and AI. AI is defined as the difference between the first and second peaks of the reflected waveform and is expressed as a percentage of the PP. As a consequence of aging and arterial stiffness, the amplification of PP from aorta to brachial arteries is reduced. To examine this further, the amplification ratio (AMPR), or brachial PP/C_PP, was determined. Mean arterial pressure was calculated by systolic blood pressure/3 + diastolic blood pressure × 2/3.
ECG-gated B-mode common carotid artery images for IMT measurement, performed by trained sonographer, were acquired with an ultrasound imager using a linear array 7.5 MHz probe following established guidelines.18 The images were then transferred from the ultrasound machine hard drive to a magneto-optical disk for off-line analysis. The proximal portion of the carotid bulb was included in all images as an anatomical reference point for standardization of IMT measurements. The far wall was used for this study because measurement of near-wall thickness is more subject to artifacts of ultrasound imaging.19 The presence of plaque, defined as the presence of focal thickening at least 50% greater than that of the surrounding vessel wall, was noted.20
An automated US Food and Drug Administration–approved commercially available IMT analysis system (Medical Imaging Applications, Iowa City, IA) was used to locate the lumen-intima and media-adventitia echo boundaries at sub-pixel resolution.19,21,22 The calculation is done horizontally. The analysis of the images was performed by an experienced technician blinded to the study participants. Intra-and interoperator intraclass correlations for our vascular laboratory are 0.97 with coefficient of variation within operators of 0.04%.
Continuous data are described as mean ± s.d., and categorical data as proportions. The relationships between PWA parameters and IMT were examined using Pearson correlation and linear regression. We regressed IMT on PWA parameters and other variables; adjusted regression coefficients and P values for PWA parameters are reported from multivariable linear regression model that adjusts for age, race, and sex. Specifically, the final multivariable model was derived using a backwards selection method, where all of the variables other than the PWA parameters are included in the model and then effects removed one at a time until only significant effects remain. In the final multivariable model, PWA parameters were adjusted for variables found to be significant covariates in univariable analysis that included age, sex, and race. PWA parameters were considered both as continuous and binary predictors; for dichotomization, we used clinically relevant cut-points as determined from mean values of within-cohort. Means and standard deviations for IMT were calculated for these dichotomous groups. We also examined unadjusted and adjusted odds ratios for the association of PWA parameters with the presence of carotid plaque using logistic regression analysis. All analyses were done using SAS 9.1 (SAS, Cary, NC).
There were 507 subjects enrolled in the CRIC study at the University of Pennsylvania site. Of these, 436 underwent carotid IMT measurement and 418 underwent PWA. Our cohort consists of 367 subjects who had both IMT and PWA measurements. The median time between IMT and PWA measurements was 90 days. Baseline characteristics of study subjects are presented in Table 1. Male gender (69%), hypertension (85%), and diabetes (53%) were common. The mean age was 59.9 + 10.4 years. The mean estimated glomerular filtration rate was 48.4 + 15.5 ml/min/1.73 m2. Mean systolic and diastolic blood pressures were 129.0 + 20.2 mm Hg and 73.7 + 12.8 mm Hg, respectively. The mean carotid IMT was 0.80 + 0.19 mm. The mean C_SP was 119.1 + 20.1 mm Hg. The mean AI was 25.9 + 13.0%. The mean C_PP was 43.6 + 17.0 mm Hg. Many of these baseline characteristics were associated with C_SP (Table 1). Particularly, Black race (P = 0.002), diabetes (P = 0.001), IMT (P = 0.01), parathyroid hormone (P = 0.002), and β-blocker use (P = 0.001) were associated with C_SP > 125 mm Hg. Although, most subjects were hypertensive, the mean BP was in the normotensive range reflecting the fact that the majority of subjects were receiving two or more antihypertensive agents (Table 2).
Associations between PWA parameters and IMT are presented in Figure 1. IMT was associated with C_SP, AI, and C_PP when examined as dichotomized variables. Elevations in C_SP, AI, and C_PP were associated with greater IMT. Increased AMPR (brachial PP/C_PP) was associated with lower IMT.
All PWA parameters were associated with carotid IMT (Table 3) in multivariable linear regression analysis. The associations between C_PP with IMT and brachial PP with IMT were similar. In an adjusted regression analysis, most of the associations were attenuated, yet still present, except for the association between AI and IMT, which nearly disappeared. A further adjustment for β-blocker use had no effect on the association. The correlation between C_PP and IMT was significant across all strata according to age, gender, race, diabetes status, and estimated glomerular filtration rate.
The presence of plaque in the carotid arteries was also noted. For the cohort, 59% had plaque present. Associations between PWA parameters and the presence of carotid plaque are presented in Table 4. Of the PWA parameters, higher C_PP and higher brachial PP were associated with carotid plaque. These were attenuated but still significant after adjustment for β-blocker use. Increased AI and C_SP, and lower AMPR were associated with the presence of plaque after adjusting for other variables, but these associations were not significant.
We found associations between several measures derived from PWA and carotid IMT and plaque in CKD patients. We show that measurements derived from PWA are not equally associated with carotid IMT. C_PP, C_SP, and AMPR showed significant correlation with IMT, but not AI. Additionally, carotid plaque was associated with C_PP, but not AI, AMPR, or C_SP. Our study is first to describe the relationship between PWA-derived measures and carotid IMT and plaque in CKD. Of the various PWA measures, C_PP had the strongest association with carotid IMT and plaque among this population of CKD patients.
Patients with CKD have an extraordinarily high rate of cardiovascular morbidity and mortality.2 Carotid IMT is a well established cardiovascular risk marker;23–25 thus, our goal was to determine how well PWA-derived measures correlate with carotid IMT.
Our findings are consistent with studies examining associations between IMT and PWA in other disease groups, and extend these observations to the CKD population. One study determined the association between central pressures and IMT26 in a cohort of over 200 diabetics, excluding subjects with CKD. The investigators determined that C_SP was a determinant of IMT independent of age. In their study, IMT did correlate with AI and central augmentation pressure; however, their average IMT values were considerable higher than those in our study (1.05 mm in theirs compared to 0.80 mm in ours), and their population was selected for diabetes. Fukui and colleagues similarly found a correlation between central augmentation pressure and IMT among a cohort of diabetic patients.27 Subjects with “nephropathy” were included in that study. Specifically, 3% of the diabetic subjects had a serum creatinine >2.0 mg/dl and 8% had a urine albumin excretion of >300 mg/g creatinine. In this study, parathyroid hormone was associated with increased C_SP. This was not surprising, as disordered bone mineral metabolism has been linked to vascular disease and arterial stiffness in CKD.
The ability of AI to predict mortality in kidney disease patients is somewhat controversial. In a study of older (majority >50 years of age) hemodialysis patients, AI (obtained from carotid artery tonometry) was associated with all-cause and CV mortality.28 However, similar to our findings, a study of children undergoing hemodialysis found no correlation between AI and IMT.29 Additionally, no association between AI and mortality was found in a cohort of young adult patients undergoing hemodialysis.30
The association between AMPR and IMT has been demonstrated among middle-aged males. Nijdam et al.31 report an inverse association between IMT and AMPR in a cross-sectional analysis of males aged 40–80 years, including hypertensives and diabetics. These findings support an association between lower PP amplification and increased cardiovascular vasculopathy because the effects of reflected waves are not transmitted directly to organs closest to branches of the aorta (i.e., brain, heart, and kidneys).
Increasingly, it is recognized that carotid plaque is an important, and possibly superior, predictor of CVE.32 Among endstage renal disease patients, greater plaque score is associated with increased risk for CVE and all-cause mortality.33 In our study, we found a correlation between carotid plaque, and C_PP and brachial PP. A large epidemiological study of middle- aged adults, including many with diabetes and/or hypertension, similarly found a strong correlation between carotid plaque and C_PP.11 That study also noted a strong correlation between C_PP and IMT.
Brachial PP is known to be a risk factor for CAD,34 heart failure, stroke, and kidney disease.35–37 Although brachial PP has been established as a good surrogate for large artery stiffness, particularly among the elderly, recently, investigators have shown that C_PP more closely correlates with markers of CVD, particularly carotid IMT, compared to brachial PP.38 Increasingly, it is recognized that measures of central pressure are better predictors of CV risk compared to brachial artery measurements, as they are a better estimate of the actual pressures seen by the heart and other major vascular beds. C_PP was more strongly related to the two outcomes than was brachial PP in a non-CKD cohort (r = 0.36 vs. 0.31 for plaque score; P < 0.001 for comparison of Spearman correlation coefficient; r = 0.29 vs. 0.25 for intima-media thickness; P < 0.002).11 Similarly, in our analysis, C_PP was superior to brachial PP in predicting carotid IMT and the presence of carotid plaque.
The PP ratio (brachial PP/C_PP), also known as AMPR, or PP amplification, may provide additional information regarding CV risk. The variability in the ratio is largely driven by large artery stiffness and wave reflection, so that a higher ratio (lower central pressures) indicates a low CV risk profile. McEniery et al. demonstrated that an elevated PP ratio (C_PP/brachial PP) was associated with CV risk, including hypertension, hyperlipidemia, and diabetes in a cohort of subjects enrolled in the Anglo-Cardiff Collaborative Trial.38
We found an association between increased C_SP and β-blocker use. This was also demonstrated in a large study of the impact of various antihypertensive classes on central blood pressures.13 However, β-blocker use did not affect the association between PWA parameters, and presence of carotid plaque or IMT.
This study was performed at a single CRIC site; thus, generalizability of the findings to other populations, including other sites within the CRIC study,15 is unknown. There was likely some subject misclassification, as there was a single measurement of the IMT and PWA. Although, the SphygmoCor device can be used to perform carotid artery tonometry as well, only radial artery tonometry was performed in this study. However, radial artery–derived measures of central pressures are widely accepted as valid.39 The carotid approach (which does not need a transfer algorithm) is technically more challenging to do,40 in part, because of lack of easily accessed bone to support the vessel unlike the radial approach at the wrist.39 Additionally, analysis did not account for all antihypertensive therapies taken by the subjects. Therefore, it is possible that other anti-hypertensive regimens used by the subjects potentially confounded results.
In conclusion, PWA-derived measures are associated with carotid IMT and plaque in CKD. Of the various measures obtained from PWA, C_PP has the strongest association with carotid plaque and IMT among CKD patients; thus, it may be a more useful PWA measure to employ when planning studies in this population. The longitudinal follow-up of the CRIC study will provide further opportunity to determine the utility of both central pressure measurements and carotid IMT and plaque as predictors of CVE in CKD.
S.S.D. was supported by a research supplement to National Institutes of Health (NIH) grant DK-067390. S.E.R. was supported by NIH grant DK-064343 from the National Institute of Diabetes and Digestive and Kidney Diseases, and VA Merit Award IIR05-247 from the Veterans Health Administration. R.R.T. was supported by NIH grants DK-067390 and DK-060984. The project described was also supported by grant no. UL1RR024134 from the National Center for Research Resources. We thank all investigators of the Chronic Renal Insufficiency Cohort Study for their contributions.
Disclosure: The authors declared no conflict of interest.