(Hypertension. 2001;37:e15.)
© 2001 American Heart Association, Inc.
Hypertension Electronic Pages |
A Comparison Between Systolic and Diastolic Pulse Contour Analysis in the Evaluation of Arterial Stiffness
From the Department of Cardiology and Angiology, Ghent University Hospital, Belgium.
Correspondence to Dr E. Rietzschel, Department of Cardiology and Angiology, Ghent University Hospital, De Pintelaan 185, B-9000 Ghent, Belgium. E-mail daniel.duprez{at}rug.ac.be
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAppendix 1References |
|---|
Abstract—Several methodologically independent measures of arterial stiffness derived from either the systolic or diastolic segments of the arterial pulse have been proposed. The exact nature of the large and small artery elasticity indices (C1 and C2, respectively) derived from diastolic pulse contour analysis remains largely unexplored, although C2 has controversially been termed to be "oscillatory" and "reflective." We investigated the relation between C2 and, respectively, a prototype of arterial reflectivity (ie, the augmentation index, AIx) and a covariate of arterial reflectivity (body height). A validated transfer function is used to transform a tonometrically obtained radial pressure wave into an ascending aortic pressure wave, from which AIx is derived using systolic pulse contour analysis. Diastolic pulse contour analysis using a modified Windkessel model is used to derive C1 and C2. One hundred subjects, who were free from atherothrombotic disease and 19 to 77 years of age, with a wide pressure range (97 to 186/52 to 104 mm Hg) were studied. Mean values of C1, C2, AIx, and body height were, respectively, 13.8±4.3 mL/mm Hgx10, 5.9±3.1 mL/mm Hgx100, 128.5±24.9%, and 169±9 cm. Coefficients of variation were 32.8% for C1, 33.3% for C2, and 6.7% for AIx. C2 was significantly and inversely correlated to AIx (r=-0.707, P<0.001). Both AIx and C2 were correlated to body height (r=-0.487, P<0.001, and r=0.514, P<0.001). In conclusion, the results of this study provide the first clinical evidence that validates a probable biophysical equivalent of the C2 element of a third-order, 4-element modified Windkessel model. We suggest that C2 is, at least in part, a measure of arterial wave reflectance. However, although short-term reproducibility of AIx is excellent, C2 showed markedly increased variability with the devices used.
Key Words: arteries • pulsatile flow • wave reflections • Windkessel model • hypertension, diagnosis • reproducibility of results
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
Hemodynamics research has shifted away from a steady-flow approach toward a pulsatile flow approach, because the former was less predictive in relation to cardiovascular morbidity and mortality.1 2 3 4 5 The growing importance of pulsatile pressure indices (systolic blood pressure [SBP] and pulse pressure) paralleled the notion that not only increases in systemic vascular resistance (SVR) but also increases in arterial stiffness are important in the pathophysiology of hypertension.6 7 This interest in the arterial cushioning function of pulsatile flow has given us a myriad of arterial stiffness indices generated by a wide array of methodologically varied noninvasive measurement techniques.8 9 10 Several working definitions of arterial stiffness have been introduced: parameters of compliance or distensibility, which essentially express changes in mono-dimensional, bidimensional, or tridimensional space for a given pressure difference; parameters quantifying speed of wave propagation along an arterial segment (pulse-wave velocity); and parameters of global vascular elastic behavior derived from the arterial pressure pulse waveform.
Given the complexity of the latter approach, simplified mathematical models have been developed.9 The arterial pressure pulse waveform can be regarded as the result of incident (anterograde) and reflected (retrograde) pressure waves.11 12 13 Wave reflection occurs at sites of discontinuity in calibre (bifurcations, branching points, and arterioles) or discontinuity in elastic properties (atherosclerosis) along the arterial tree.14 Changes in amplitude and timing of wave reflections play a key role in aortic hemodynamics and, on a broader level, give us information on how the arterial vasculature affects the heart.15 16 17 18 19 The SphygmoCor BPAS-1/A device (model SPT-301, PWV Medical Pty Ltd) calculates a number of parameters of ventriculo-arterial coupling and wave reflectance, of which the augmentation index (AIx) is preeminent.11 20 21 Another approach in analyzing the arterial pulse is to regard the waveform as a basic pattern of exponential decay on which damped oscillations are superimposed, building on the "Grundform" - "Grundschwingung" concept pioneered by Otto Frank.22 The Windkessel models of the vasculature9 23 24 25 typify this last approach, and one of these is used by the recently introduced HDI/Pulsewave CR-2000 (Hypertension Diagnostics Inc) to quantify the circulation in terms of SVR, large artery elasticity index (C1), small artery elasticity index (C2), and inductance (L).24 25 26 27
Recently, in the field of arterial stiffness, several authors have underscored the need, in addition to uniformity of terminology, for comparisons of techniques to ascertain which methodology (or methodologies) best suits the different research questions facing us.28 29 Indeed, comparative data between different stiffness indices or devices is very scarce, and, more specifically, data assessing the parameters of Windkessel modelling with reflectivity indices are currently lacking.
The goal of this study is to compare 2 methodologically different techniques, predominantly analyzing either the diastolic decay or the systolic segment of the arterial waveform, to describe global vascular elastic behavior. Specifically, C2 as a putative biophysical equivalent of wave reflectance will be assessed. Short-term reproducibility of the 2 techniques used is also analyzed within a single population with a wide pressure and age range.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
Subjects
We enrolled 100 healthy white subjects (41 men, 59 women) with wide age and blood pressure ranges (mean age 46.0±15.1 years and range 19 to 77 years; mean SBP 129±19 mm Hg and range 97 to 186 mm Hg; mean diastolic blood pressure [DBP] 73±11 mm Hg and range 52 to 104 mm Hg). Mean body height was 169±9 cm with a mean weight of 71.6±14.4 kg and a resulting mean body mass index (BMI) of 25.5±4.9 kg/m2. Uncomplicated hypertension was present in 32 subjects, and of these, 27 were currently taking antihypertensive medication. None of the subjects had taken any vasoactive drugs in the past 24 hours before the examinations. Any history of cardiac or vascular disease (including cerebrovascular, renovascular, and peripheral vascular disease) was an exclusion criterion. None of the subjects suffered from known endocrine, hepatic, or renal disease. All subjects gave informed consent to be enrolled in the study, which was approved by the local medical ethics committee.
Study Protocol
The experiments were performed between 8:00 and 10:00
AM, after an overnight fast
and abstinence from tobacco, alcohol, tea, or coffee. The subjects were
allowed 30 minutes of supine rest after which the measurements were
performed.
SphygmoCor/PWV Blood
Pressure Analysis System BPAS-1/A
Pressure pulse waveforms at the level of the radial
artery were
recorded31 38
using a high-fidelity, transcutaneous, single-unit, hand-held
applanation tonometer with an external coplanar
micromanometer tip (Millar Instruments). Under
optimal conditions for applanation (ie, when the flat tonometer end
with coplanar sensor flattens the wall of an artery at the operational
part of the sensor, thus eliminating tangential forces) pressure waves
measured noninvasively are virtually identical to those recorded
with a high-fidelity intra-arterial
transducer.30 31 32 33
To overcome the problem of differences in hold-down pressure of the
applanation tonometer, the peripheral pressure waveforms
are calibrated with a pressure value determined by a mercury
sphygmomanometer at the brachial
artery.31 An averaged radial
pressure waveform derived from an 8-second recording is
subsequently used to derive a corresponding ascending aortic pressure
waveform using a validated general transfer function
(GTF).11 34 35 36 37
From this ascending aortic waveform the AIx is calculated as the height
of the second systolic peak above the wave foot divided by the
height of the first systolic peak above the wave foot expressed
as a
percentage.38 31
HDI/Pulsewave Research
CardioVascular Profiling Instrument
Radial artery waveforms were obtained with the use of
a calibrated proprietary tonometer (model CR-2000, Hypertension
Diagnostics Inc) used according to the
manufacturers’
specifications.39 The
tonometer consists of a 1.27-cm-diameter stainless-steel canister with
a 0.15-mm-thick stainless-steel diaphragm internally connected to a
double-plated ceramic piezoelectric element used to amplify the
waveform signal. The subject’s arm is supported by an angulated
"wrist stabilizer," ensuring a constant wrist position. The
tonometer is housed inside a holding and positioning device that is
wrapped around the supported arm, achieving complete stability of
position after the tonometer has been applied. The obtained waveforms
are calibrated to the systolic and diastolic cuff
pressure values of an integrated oscillometric device (cuff placed on
the contralateral arm with respect to the tonometer). A computer-based
third-order, 4-element modified Windkessel model of the circulation is
used to match the diastolic pressure decay of the
tonometrically obtained waveforms and to quantify changes in
arterial waveform morphology in terms of SVR, C1, C2, and
L. The values of C1, C2, and L given in the report are weighted
averages of the values obtained on individual waveforms during a
30-second recording
period.26 27 39
SVR is calculated as mean arterial pressure (MAP) divided by cardiac output (CO). MAP is derived from the waveform analysis (after calibration with the oscillometrically obtained blood pressure) by integrating the area under each beat and then calculating the average of all beats included in the analysis. CO is calculated from a multivariate algorithm, using cardiac ejection time (CET) derived from the radial artery (see Appendix). The multivariate algorithm includes age, HR, body surface area (BSA), and CET and was derived from and tested against measurements of CO using indocyanine green dye dilution obtained invasively. The noninvasive algorithm tended to overestimate low (invasive) CO and underestimate high (invasive) CO and generally showed fair agreement with 92% of algorithm-derived values, deviating within a margin of 25% from indocyanine green dye dilution measured CO.26 40
The shape of the diastolic decay curve can be represented as the solution to a third-order differential equation with 6 unknown "A" parameters. The "A" values are obtained through a nonlinear curve-fitting routine that matches the shape of the third-order equation to the diastolic decay of the waveform. From the different "A" values and from the previously obtained value of SVR the different elements of the modified Windkessel model (C1, C2, and L) can be calculated.
Short-Term Reproducibility
Dual measurements were performed in random order with
each device. Only measurements in which the tonometer was completely
removed and subsequently reapplied (on the ipsilateral side) within a
time frame of 5 minutes were counted as repeat measurements and
withheld for analysis. These criteria were met in all
subjects.
Statistical Analysis
Statistical analysis was performed using SPSS
for Windows release 9.0. All parameters are given as
mean±SD. Relevant parameters were tested for normality
using a Kolmogorov-Smirnov test. Parameters were compared
between groups using ANOVA. Relations between stiffness
parameters and body height were assessed using partial
Spearman 
correlations. Possible confounders were identified first
via univariate correlates (continuous variables) or
ANOVA (categorical variables) and subsequently assessed using
multivariate regression analysis. Standardized
residuals of z scores of linear regression analysis between AIx
and C2 were withheld for further analysis. Coefficients of
variation were calculated as the SD of the difference between 2
measurements divided by the mean value of the mean of both
measurements. Reproducibility data are given as Bland-Altman plots that
depict percentual differences between 2 measurements plotted
against the mean of 2
measurements.41 The
proportion of biological variability in the overall variability was
estimated using multivariate linear regression. A value
of P<0.05 was considered
statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
The Table summarizes the arterial stiffness indices (C1, C2, and AIx), estimates of CET, and stroke volume (SV) obtained by the HDI/Pulsewave CR-2000, as well as coefficients of variation of these parameters. Univariate correlates of C2 comprised (P<0.01, except when indicated otherwise) MAP (r=-0.622), SBP (r=-0.587), age (r=-0.566), body height (r=0.514), DBP (r=-0.516), pulse pressure (r=-0.442), body surface area (r=-0.266), and BMI (r=-0.207, P<0.05). Univariate correlates of AIx comprised (P<0.01) age (r=0.659), MAP (r=0.555), body height (r=-0.487), SBP (r=0.483), DBP (r=0.469), pulse pressure (r=0.320), BMI (r=0.304), and body surface area (r=0.197).
|
In addition to confounders withheld from the univariate analysis (MAP, age, and body height), gender and a history of hypertension/habitual antihypertensive drug use were identified: Women had significantly a lower C2 (5.0±2.6 versus 7.2±3.5 mL/mm Hgx100, P<0.01) and significantly higher a AIx (135±24 versus 119±24%, P<0.01) than men. Subjects who habitually took antihypertensive drugs had a significantly lower C2 (4.2±2.8 versus 6.3±3.1 mL/mm Hgx100 for drug-free subjects, P=0.011) and a nonsignificantly higher AIx (137±18 versus 126±26% for drug-free subjects, P=0.09). However, in a multivariate regression analysis neither gender nor habitual antihypertensive drug use added to a model that included MAP, age, and body height. Colinearities could be inferred because women, despite similar age and blood pressure compared with men, were significantly shorter (P<0.01) and subjects who habitually took antihypertensive drugs had significantly higher blood pressures (P<0.01), were older (P<0.01), and were shorter (P<0.01) than their drug-free counterparts.
The regression plots between AIx, C2, and body height are shown in Figure 1. The correlations between body height and AIx and C2 remained significant after correction for blood pressure and age (r=-0.345, P<0.01, and r=0.314, P<0.01, respectively, for AIx and C2 versus height).
|
AIx was significantly and inversely correlated with C1 (r=-0.424, P<0.001) and C2 (r=-0.707, P<0.001, Figure 2). After controlling for age, height, and blood pressure (MAP), AIx and C2 were still significantly and inversely correlated (r=-0.312, P<0.001). The correlation between C1 and AIx was lost after correction for age (r=-0.161, P>0.05).
|
Coefficients of variation were calculated and averaged 6.7% for AIx, 32.8% for C1, and 33.3% for C2 (Table). Bland-Altman plots were constructed for AIx, C1, and C2 and demonstrate that the 95% confidence interval of the percentual variation varied from -12.4% to +13.8% for AIx, from -63.3% to +69.1% for C1, and from -57.9% to +72.8% for C2. Figure 3 depicts Spearman correlations between first and second measurement of each parameter on the left and corresponding Bland-Altman plots on the right. Differences (absolute and percent) between the dual measurements of C1, C2, and AIx were calculated; no correlations were found between the differences of C1, C2, and AIx and age, height, weight, blood pressure, or heart rate (HR). Variability in measurement of C1 was not a predictor of variability in measurement of C2 and vice versa.
|
To identify the reasons for divergence and the apparent relative overestimation of C2 in a subgroup of subjects, further analysis was performed on the standardized residuals’ z scores of the regression between C2 and AIx. Distance from the regression line between C2 and AIx, expressed as z scores, correlated significantly (P<0.01) with HR (r=-0.380), MAP (r=-0.310), and body height (r=0.267). Mean z scores of standardized residuals for MAP and HR quartiles are shown in Figure 4 (between group differences analyzed using 1-way ANOVA; P<0.01 and P=0.014 for MAP and HR, respectively). A stepwise gradient of the z scores in both HR and MAP quartiles can be clearly seen.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
This study is the first to compare 2 methodologically different techniques, analyzing either the diastolic decay or the systolic segment of the arterial waveform, to describe global vascular elastic behavior. Our results demonstrate that C2 and AIx are significantly and inversely correlated, which raises the possibility that in addition to AIx, C2 could also be a descriptor of reflective phenomena.
However, before reaching this conclusion, possible confounders of the relation between C2 and AIx should be addressed. In addition to the covariates of C2 and AIx previously identified in literature (age and blood pressure), we investigated the effect of body height, because arterial wave reflections are not only dependent on arterial stiffness but are also related to the length of the arterial path and a fortiori to body height.11 42 43 44 We found significant correlations between body height and C2 and AIx. The correlation between C2 and body height has not been described previously and is approximately of the same magnitude as the inverse correlation we confirmed between AIx and body height, which has been described previously. In these reports, shorter stature was associated with an increase in magnitude and earlier return of reflected waves due to a shorter arterial tree. In contrast, body height had only marginal effects on nonwave reflection–related stiffness indices. Because of a theoretical relation between body size and HR, on the basis of data from comparative physiology (ie, increasing HR with decreasing body size of the species), as well as on scarce human data, we took care to exclude an effect of HR.45 46 47 After adjustment, the correlation between C1 and AIx was lost; however, the correlation between AIx and C2 did not change significantly after correction for age, height, and blood pressure.
The concept of wave reflectance is preeminent in a controversy in literature concerning the nature of C2, which has (arbitrarily) been termed to be "oscillatory" and "reflective."8 48 The controversy endures because first it has not been possible to convert the C1 and C2 capacitors of the electrical analogue to a distinct anatomical or physiological substrate. With regard to this point, the authors themselves stated: "Each model element may not explicitly describe a single vascular property but should be viewed as contributing to the resulting pressure waveform."25 Second, wave reflection of any kind in a Windkessel model is, from a theoretical point of view, impossible, because pulse-wave velocity is assumed to be infinite. However, the correlation between C2 and AIx combined with the individual relations to body height, which were inverse but of similar magnitude, does suggest that C2 is, at least in part, an expression of wave reflectance and thus supports the use of the term "reflective." These results are the first clinical data supporting the "reflective" nature of C2, a potentially clinically relevant finding, since indices of wave reflectivity have been shown to be independently associated with left ventricular hypertrophy.16 17 18 19
Previous attempts have been made to provide a biological significance for C2, on the basis of which C2 was claimed to represent distal ("small arterial") compliance. The evidence for this stems largely from vasodilatory experiments using nitroprusside, nitro-glycerine, and hydralazine.49 50 It was observed that these drugs induced increases of C2 several-fold larger than increases in C1, which were in the case of the nitrovasodilators (nitroprusside and nitroglycerin) largely independent of effects on SVR. On the basis of the results, C2 was taken to represent a measure of "small arterial" and hence "distal" compliance. It should be noted that the effects of hydralazine on C2 were less than the effects of the nitrovasodilators. Nitroglycerin and nitroprusside have their effect predominantly as dilators of small arteries with much less effect on an arteriolar level. In the study of reflectance phenomena, arterial dilators are characterized by a prominent reduction in wave reflectance with minimal effect on SVR. Hydralazine is more representative for arteriolar dilators, which are characterized by prominent reductions of SVR with minimal effect on wave reflections. Hence, with hindsight, these earlier published concepts could equally have been explained by modulation of wave reflectivity. Thus, for a parameter that has been termed reflective, distal, and small artery, the data that support the latter 2 claims may actually be more supportive of a reflective nature of C2.
The position of C2 relative to the regression line with AIx is influenced by blood pressure and HR. Lower blood pressures and/or slower HRs give rise to a gradient that favors relative overestimation of C2. The inverse holds for higher blood pressures or faster HRs (Figure 4). The pressure-dependent gradient can be explained on the basis of the calculation of SVR. Lower MAP values yield lower SVR values, which independently from the waveform analysis (the "A" indices) will increase values of C1, C2, and L (see Appendix). The overestimating gradient produced by decreasing HR cannot be explained by the effect on SVR. A lower HR would (because of a lower CO and thus higher SVR) be expected to cause a relative underestimation of C2. These findings are contrary to the observed data, and we can only surmise a potent, opposite effect of HR on the derivation of the "A" values.
Short-Term Reproducibility
A short-term reproducibility analysis was
performed and the results indicate a wide gap between the coefficients
of variation of C1 and C2 (33% and 34%, respectively) and of AIx
(7%). The level of variability for AIx is very similar to values
previously
reported.51 52
Causes of variability could occur at 3 separate stages (signal
acquisition, signal calibration and waveform analysis), which
will therefore be individually addressed.
Signal Acquisition
Both devices use a tonometer for recording the
radial arterial pulse wave. The
SphygmoCor BPAS-1/A device uses a
Millar hand-held tonometer with an external
micromanometer tip. There is general agreement that
applanation tonometry can accurately reproduce
intra-arterial waveforms (when conditions of correct
applanation are achieved, which seems to be the case in the majority of
subjects).11 30 32 33 53
The study by Sato et al33
showed an excellent relation between invasive and tonometric signals
with a flat gain up to 7Hz, which covered most of the frequency ranges
of interest. However, AIx retains some dependency on the high-frequency
content of the signal (8 to 10
Hz).37 Noise reduction
accomplished by introducing a low-pass filter in the system in the
range of 9 to 12Hz could therefore be responsible for underestimation
of AIx obtained noninvasively compared with invasively obtained AIx.
The HDI/Pulsewave CR-2000 device uses a
proprietary tonometer (see Methods) in combination with a "holding
and positioning device" and an angulated "wrist stabiliser,"
which ensures a constant wrist position and a complete stability of
position after the tonometer has been applied, in contrast with the
SphygmoCor BPAS-1/A pencil probe, which has to
be manually held in position. Consecutive measurements with the
tonometer kept in fixed position might be very reproducible but do not
comply with daily reality, neither in cross-sectional population
studies (or as proposed in risk stratification) nor in longitudinal
interventional or observational studies. Therefore, we opted to
completely remove the tonometers from the skin between measurements. In
these conditions, the only theoretical advantage of the stabilized HDI
tonometer is a more constant position within a single measurement. This
theoretical advantage was, however, not translated into improved
reproducibility in our population. Because variability in measurement
of C1 was not a predictor of variability in measurement of C2 and vice
versa, acquisition of the signal or bad-signal quality does not seem to
be the prime suspect in the greater variability of the
HDI/Pulsewave CR-2000. There is no published
information on the HDI proprietary tonometer regarding the accuracy of
the tonometrically obtained radial artery waveform compared with
intra-arterial recordings.
Signal Calibration
In both devices, the peripheral pressure
waveforms obtained at the radial artery are calibrated with a blood
pressure value determined at the brachial artery. This potentially
introduces a systematical error because it assumes equivalence between
radial and brachial
pressures.11 37
Although there is a degree of systolic amplification from
brachial artery to radial artery, the differences in pressure are
likely to be very small and probably insignificant in comparison to the
error introduced by noninvasive blood pressure measurement per se.
Furthermore, this systematical error occurs only in the
SphygmoCor BPAS-1/A device with regard to the
central aortic pressure data; because AIx is an expression of the
relative height of 2 parts of the same waveform, this systematical
error is not present in the calculation of AIx. In contrast, all
"A"-dependent components of the modified Windkessel model (C1, C2,
and L) are pressure dependent, because SVR (calculated as MAP/CO) is
used in the calculation of all 3. Inaccuracies in measurement of MAP
will therefore introduce inaccuracies in SVR and a fortiori in C1, C2,
and L.
Waveform Analysis
In the SphygmoCor BPAS-1/A, an
averaged radial pressure waveform is derived from an 8-second
recording, which is used to derive a corresponding ascending
aortic pressure waveform using an integral GTF. Because AIx is
calculated on the ascending aortic waveform, the accuracy of AIx hinges
on the accuracy of these GTFs. Since initial publication, several
authors have independently produced similar GTFs and provided
validating
evidence.34 36 37
GTFs were obtained in individuals in steady state and during
hemodynamic transients (after vasodilatation with
nitroglycerine and during handgrip maneuvers) and used
to synthesize ascending aortic pressure waves, which were compared with
invasively recorded pressure waves. Induction of
hemodynamic transients, despite producing marked
variations in hemodynamic status, had only a marginal
effect on the calculated GTFs, which remained practically identical to
steady-state GTFs. Karamanoglu et
al34 constructed
patient-specific transfer functions, but these proved to be only
marginally better than a GTF in matching ascending aortic pressure
data. The authors concluded that clinically acceptable predictions of
central aortic pressure and waveform could be obtained by mathematical
manipulation of radial pressure waves using a single GTF and that this
seems to hold true in a wide variety of hemodynamic
states. These conclusions are, however, not universally accepted, and
the main criticisms are the small number of subjects in the validating
studies, the variations in methodology in the cited validating studies,
and the inherent problems related to the noninvasive calibration of the
radial artery waveform as discussed
above.54 55 56 57
In the HDI/Pulsewave CR-2000 device, a 30-second collection of radial artery waveform data are analyzed using a third-order, 4-element modified Windkessel model of the circulation, matching the diastolic pressure decay using a proprietary nonlinear curve-fitting routine. The advantages and shortcomings of Windkessel models are well known and described in literature.9 25 Three main problems with the methodology should be addressed. First, the model assumes a measurement-site independence, a assumption that has been shown to be unfounded by the study of Fogliardi et al.58 Second, the model seems to be critically dependent on which segment of the diastolic decay curve is used, and wide variations of "A" values can be produced by small differences in portions of the diastolic decay used for parameter estimation.58 Third, induction of hemodynamic transients seems to exacerbate the problems created by the assumption of measurement-site independence.58
Causes for the increased variability of the HDI/Pulsewave CR-2000 device could theoretically be found at several stages. However, taking into account that variability in C1 was not a predictor of variability in C2 and vice versa and that the device has reproducible measurements of SV (and CO) reflecting good reproducibility of CET, which is also derived from the same radial waveform with minimal variability, it must be assumed that signal acquisition or bad signal quality does not seem to be the prime suspect in the greater variability of the HDI/Pulsewave CR-2000 device. Furthermore, only a minimum of variability can be attributed to biological variations in blood pressure or HR between measurements in supine subjects at rest during a 5-minute time frame (the measurements were, as expected for subjects in supine rest, small: in the order of 3% to 4% blood pressure or HR variability, accounting for at most 8% to 9% of the variability of C1 or C2 in a stepwise multiple regression analysis). It therefore seems likely that the proprietary parameter estimating algorithm of the HDI/Pulsewave CR-2000 is primarily responsible for the variability of C1 and C2. Although reproducibility does not imply validity, in future clinical or validating studies with stiffness devices, power calculations on the number of subjects needed for inclusion are likely to be several times larger for the HDI/Pulsewave CR-2000 device.
In conclusion, the results of this study provide the first clinical validating evidence for a probable biophysical equivalent of the C2 element of a third-order, 4-element modified Windkessel model. We suggest that C2 is, at least in part, a measure of arterial wave reflectance. However, although short-term reproducibility of AIx with the SphygmoCor BPAS-1/A is good, measurement of C1 and C2 with the HDI/Pulsewave CR-2000 showed markedly increased variability.
|
Appendix 1 |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
Formulas
 |
 |
 |
 |
 |
 |
 |
 |
|
Acknowledgments |
|---|
Dr E. Rietzschel is currently a research fellow of the Fund for Scientific Research - Flanders (Fonds voor Wetenschappelijk Onderzoek - Vlaanderen No 60012.98). The authors would like to thank Frida Brusselmans, An Maas, and Marnix Bobelyn for their generous help, technical assistance, and all-round support and Marc Gillis for preparing the illustrations.
Received December 7, 2000; first decision December 22, 2000; accepted January 15, 2001.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1References |
|---|
1. Kannel WB, Wolf PA, McGee DL, Dawber TR, McNamara P, Castelli WP. Systolic blood pressure, arterial rigidity, and risk of stroke: the Framingham study. JAMA. 1981;245:1225–1229.
2.
Kannel WB. Clinical
misconceptions dispelled by epidemiological research.
Circulation. 1995;92:3350–3360.
3.
Benetos A, Safar M,
Rudnichi A, Smulyan H, Richard JL, Ducimetieere P, Guize L. Pulse
pressure: a predictor of long-term cardiovascular
mortality in a French male population.
Hypertension. 1997;30:1410–1415.
4.
Domanski MJ, Davis
BR, Pfeffer MA, Kastantin M, Mitchell GF. Isolated systolic
hypertension: prognostic information provided by pulse pressure.
Hypertension. 1999;34:375–380.
5.
O’Rourke MF,
Avolio AP. Pulsatile flow and pressure in human systemic arteries:
studies in man and in a multibranched model of the human systemic
arterial tree. Circ
Res. 1980;46:363–372.
6. O’Rourke MF. Isolated systolic hypertension, pulse pressure, and arterial stiffness as risk factors for cardiovascular disease. Curr Hypertens Rep. 1999;3:204–211.
7. Cohn JN. Arteries, myocardium, blood pressure and cardiovascular risk: towards a revised definition of hypertension. J Hypertens. 1998;16:2117–2124.[Medline] [Order article via Infotrieve]
8. O’Rourke MF, Mancia G. Arterial stiffness. J Hypertens. 1999;17:1–4.[Medline] [Order article via Infotrieve]
9. Burattini R, Pagani M, Zannoli R. Mechanical properties of the arterial system. High Blood Press. 1996;5:228–240.
10. Cohn JN. Pathophysiologic and prognostic implications of measuring arterial compliance in hypertensive disease. Prog Cardiovasc Dis. 1999;41:441–450.[Medline] [Order article via Infotrieve]
11. Nichols WW, O’Rourke MF. Sphygmocardiography. In: Nichols WW, O’Rourke MF, eds. McDonald’s Blood Flow in Arteries. London, England: Arnold; 1998:450–477.
12.
O’Rourke MF,
Blazek JV, Morreels CLJ, Krovetz LJ. Pressure wave transmission along
the human aorta: changes with age and in arterial
degenerative disease. Circ Res. 1968;23:567–579.
13. O’Rourke MF. The arterial pulse in health and disease. Am Heart J. 1971;82:687–702.[Medline] [Order article via Infotrieve]
14. Nichols WW, O’Rourke MF. Wave reflections. In: Nichols WW, O’Rourke MF, eds. McDonald’s Blood Flow in Arteries. London, England: Arnold; 1998:201–222.
15.
Murgo JP,
Westerhof N, Giolma JP, Altobelli SA. Aortic input impedance in normal
man: relationship to pressure wave forms.
Circulation. 1980;62:105–116.
16. Roman MJ, Pini R, Pickering TG, Devereux RB. Non-invasive measurements of arterial compliance in hypertensive compared with normotensive adults. J Hypertens Suppl. 1992;10:S115–S118.[Medline] [Order article via Infotrieve]
17.
Roman MJ, Ganau
A, Saba PS, Pini R, Pickering TG, Devereux RB. Impact of
arterial stiffening on left ventricular
structure. Hypertension. 2000;36:489–494.
18. Saba PS, Roman MJ, Pini R, Spitzer M, Ganau A, Devereux RB. Relation of arterial pressure waveform to left ventricular and carotid anatomy in normotensive subjects. J Am Coll Cardiol. 1993;22:1873–1880.[Abstract]
19.
Chen CH, Nakayama
M, Nevo E, Fetics BJ, Maughan WL, Kass DA. Coupled
systolic-ventricular and vascular stiffening with
age: implications for pressure regulation and cardiac reserve in the
elderly. J Am Coll
Cardiol. 1998;32:1221–1227.
20.
O’Rourke MF,
Yaginuma T. Wave reflections and the arterial pulse.
Arch Intern Med. 1984;144:366–371.
21. O’Rourke MF, Gallagher DE. Pulse wave analysis. J Hypertens Suppl.. 1996;14:S147–S157.[Medline] [Order article via Infotrieve]
22. Frank O. Die Grundform der Arteriellen Pulses. Acta Biol.. 1899;37:426–483.
23. Goldwyn RM, Watt TB. Arterial pressure pulse contour analysis via a mathematical model for qualification of human vascular properties. IEEE Trans Biomed Eng. 1967;BME-14:11–17.
24.
Watt TB, Burrus
CS. Arterial pressure contour analysis for
estimating human vascular properties.
J Appl Physiol. 1976;40:171–176.
25. Finkelstein SM, Cohn JN. First- and third-order models for determining arterial compliance. J Hypertens Suppl.. 1992;10:S11–S14.[Medline] [Order article via Infotrieve]
26.
Cohn JN,
Finkelstein S, McVeigh G, Morgan D, LeMay L, Robinson J, Mock J.
Noninvasive pulse wave analysis for the early detection of
vascular disease. Hypertension. 1995;26:503–508.
27.
McVeigh GE,
Bratteli CW, Morgan DJ, Alinder CM, Glasser SP, Finkelstein SM, Cohn
JN. Age-related abnormalities in arterial compliance
identified by pressure pulse contour analysis: aging and
arterial compliance.
Hypertension. 1999;33:1392–1398.
28. Carretta R. Guidelines regarding nomenclature of the mechanics of large artery walls: recommendations of the study group of the Italian Society of Hypertension. High Blood Press. 1998;1997:7:48–54.
29. Lehmann ED, Hopkins KD, Gosling RG. Multiple definitions of ‘compliance.’ Clin Sci (Colch ). 1996;90:433–434.
30. Drzewiecki G, Melbin J, Noordegraaf AV. Arterial tonometry: review and analysis. J Biomech. 1983;16:141–152.[Medline] [Order article via Infotrieve]
31.
Kelly R, Hayward
C, Avolio A, O’Rourke M. Noninvasive determination of age-related
changes in the human arterial pulse.
Circulation. 1989;80:1652–1659.
32. Pressman GL, Newgard PM. A transducer for the continuous measurement of arterial blood pressure. IEEE Trans Biomed Eng. 1963;10:73–81.[Medline] [Order article via Infotrieve]
33.
Sato T, Nishinaga
M, Kawamoto A, Ozawa T, Takatsuji H. Accuracy of a continuous blood
pressure monitor based on arterial tonometry.
Hypertension. 1993;21:866–874.
34.
Karamanoglu M,
O’Rourke MF, Avolio AP, Kelly RP. An analysis of the
relationship between central aortic and peripheral upper
limb pressure waves in man. Eur Heart
J. 1993;14:160–167.
35.
Cameron JD,
McGrath BP, Dart AM. Use of radial artery applanation tonometry and a
generalized transfer function to determine aortic pressure augmentation
in subjects with treated hypertension.
J Am Coll Cardiol. 1998;32:1214–1220.
36.
Chen CH, Ting CT,
Nussbacher A, Nevo E, Kass DA, Pak P, Wang SP, Chang MS, Yin FC.
Validation of carotid artery tonometry as a means of estimating
augmentation index of ascending aortic pressure.
Hypertension. 1996;27:168–175.
37.
Chen CH, Nevo E,
Fetics B, Pak PH, Yin FC, Maughan WL, Kass DA. Estimation of central
aortic pressure waveform by mathematical transformation of radial
tonometry pressure: validation of generalized transfer function.
Circulation. 1997;95:1827–1836.
38.
Kelly RP, Gibbs
HH, O’Rourke MF, Daley JE, Mang K, Morgan JJ, Avolio AP.
Nitroglycerin has more favourable effects on left
ventricular afterload than apparent from measurement of
pressure in a peripheral artery.
Eur Heart J. 1990;11:138–144.
39. Hypertension Diagnostics Inc. HDI/Pulsewave CR-2000 Research CardioVascular Profiling System Operator’s Manual.Logan, Minn: Hypertension Diagnostics; 1999.
40. Finkelstein SM, Cohn JN, inventors. Method and apparatus or measuring cardiac output. US patent 5241966. 1993.
41. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–310.[Medline] [Order article via Infotrieve]
42.
London GM, Guerin
AP, Pannier B, Marchais SJ, Stimpel M. Influence of sex on
arterial hemodynamics and blood pressure:
role of body height.
Hypertension. 1995;26:514–519.
43. London GM, Guerin AP, Pannier BM, Marchais SJ, Metivier F. Body height as a determinant of carotid pulse contour in humans. J Hypertens Suppl. 1992;10:S93–S95.[Medline] [Order article via Infotrieve]
44.
Smulyan H,
Marchais SJ, Pannier B, Guerin AP, Safar ME, London GM. Influence of
body height on pulsatile arterial
hemodynamic data. J
Am Coll Cardiol. 1998;31:1103–1109.
45.
Bonaa KH, Arnesen
E. Association between heart rate and atherogenic blood lipid fractions
in a population: the Tromso study.
Circulation. 1992;86:394–405.
46. Westerhof N. Heart period is related to the time constant of the arterial system and not to minimum of impedance modulus. In: Hosoda S, Yaginuma T, Sugawara M, Taylor MG, Caro CG, eds. Recent Progress in Cardiovascular Mechanisms. Newark, New Jersey: K Harwood Academic Publishers; 1994:115–128.
47. O’Rourke MF. Arterial Function in Health and Disease. Edinburgh, England: Churchill-Livingstone; 1982:170–182.
48. Cohn JN. Arterial Stiffness. J Hypertens. 1999;17:1227.[Medline] [Order article via Infotrieve]
49.
Finkelstein SM,
Collins VR, Cohn JN. Arterial vascular compliance response
to vasodilators by Fourier and pulse contour analysis.
Hypertension. 1988;12:380–387.
50. Zobel LR, Finkelstein SM, Carlyle PF, Cohn JN. Pressure pulse contour analysis in determining the effect of vasodilator drugs on vascular hemodynamic impedance characteristics in dogs. Am Heart J. 1980;100:81–88.[Medline] [Order article via Infotrieve]
51. Siebenhofer A, Kemp C, Sutton A, Williams B. The reproducibility of central aortic blood pressure measurements in healthy subjects using applanation tonometry and sphygmocardiography. J Hum Hypertens. 1999;13:625–629.[Medline] [Order article via Infotrieve]
52. Wilkinson IB, Fuchs SA, Jansen IM, Spratt JC, Murray GD, Cockcroft JR, Webb DJ. Reproducibility of pulse wave velocity and augmentation index measured by pulse wave analysis. J Hypertens. 1998;16:2079–2084.[Medline] [Order article via Infotrieve]
53. Saba PS, Cavallini C, Scorzoni D, Longhini C, Pini R, Ganau A, on behalf of the study group "mechanics of large arteries" IsoH. Arterial tonometry: principles and clinical application in hypertension. High Blood Press. 1996;5:241–250.
54. Lehmann ED. Regarding the accuracy of generalized transfer functions for estimating central aortic blood pressure. J Hypertens. 1999;17:1225–1227.[Medline] [Order article via Infotrieve]
55. Lehmann ED. Regarding the accuracy of generalised transfer functions for estimating central aortic blood pressure when calibrated non-invasively. J Hypertens. 2000;18:347–350.[Medline] [Order article via Infotrieve]
56. Wilkinson IB, Cockcroft JR, Webb DJ. Regarding the accuracy of generalized transfer functions for estimating central aortic blood pressure. J Hypertens. 1999;17:1226–1227.
57. Wilkinson IB, Cockcroft JR, Webb DJ. Regarding the accuracy of generalized transfer functions for estimating central aortic blood pressure. J Hypertens. 2000;18:349–350.
58. Fogliardi R, Burattini R, Shroff SG, Campbell KB. Fit to diastolic arterial pressure by third-order lumped model yields unreliable estimates of arterial compliance. Med Eng Phys. 1996;18:225–233. [Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  F. Compton, M. Wittrock, J.-H. Schaefer, W. Zidek, M. Tepel, and A. Scholze Noninvasive Cardiac Output Determination Using Applanation Tonometry-Derived Radial Artery Pulse Contour Analysis in Critically Ill Patients Anesth. Analg., January 1, 2008; 106(1): 171 - 174. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X J Jiang, M F O'Rourke, W Q L Jin, L S Liu, C W Li, P C Tai, X C Zhang, and S Z Liu Quantification of glyceryl trinitrate effect through analysis of the synthesised ascending aortic pressure waveform Heart, August 1, 2002; 88(2): 143 - 148. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||