Determinants of Elevated Pulse Pressure in Middle-Aged and
Older Subjects With Uncomplicated Systolic Hypertension
The Role of Proximal Aortic Diameter and the Aortic Pressure-Flow Relationship
Gary F. Mitchell, MD; Yves Lacourcière, MD; Jean-Pascal Ouellet, MD; Joseph L. Izzo, Jr, MD;
Joel Neutel, MD; Linda J. Kerwin, MS; Alan J. Block, PhD; Marc A. Pfeffer, MD, PhD
Background—Elevated pulse pressure (PP) is associated with increased cardiovascular risk and is thought to be secondary
to elastin fragmentation with secondary collagen deposition and stiffening of the aortic wall, leading to a dilated,
noncompliant vasculature.
Methods and Results—By use of calibrated tonometry and pulsed Doppler, arterial stiffness and pulsatile hemodynamics
were assessed in 128 subjects with uncomplicated systolic hypertension (supine systolic pressure ⱖ140 mm Hg off
medication) and 30 normotensive control subjects of comparable age and gender. Pulse-wave velocity was assessed
from tonometry and body surface measurements. Characteristic impedance (Zc) was calculated from the ratio of change
in carotid pressure and aortic flow in early systole. Effective aortic diameter was assessed by use of the water hammer
equation. Hypertensives were heavier (P⬍0.001) and had higher PP (P⬍0.001), which was attributable primarily to
higher Zc (P⬍0.001), especially in women. Pulse-wave velocity was higher in hypertensives (P⫽0.001); however, this
difference was not significant after adjustment for differences in mean arterial pressure (MAP) (P⬎0.153), whereas
increased Zc remained highly significant (P⬍0.001). Increased Zc in women and in hypertensive men was attributable
to decreased effective aortic diameter, with no difference in wall stiffness at comparable MAP and body weight.
Conclusions—Elevated PP in systolic hypertension was independent of MAP and was attributable primarily to elevated
Zc and reduced effective diameter of the proximal aorta. These findings are not consistent with the hypothesis of
secondary aortic degeneration, dilation, and wall stiffening but rather suggest that aortic function may play an active role
in the pathophysiology of systolic hypertension. (Circulation. 2003;108:1592-1598.)
Downloaded from http://ahajournals.org by on June 14, 2020
Key Words: hypertension 䡲 aorta 䡲 impedance 䡲 pressure 䡲 stiffness
P
ulse pressure (PP), an indirect indicator of arterial stiffness, has recently emerged as a strong independent
predictor of cardiovascular events in patients with hypertension.1– 4 Increased PP and premature timing of the reflected
pressure wave add to load on the left ventricle and arteries
and may trigger ventricular and vascular hypertrophy and
fibrosis.5,6 Vascular stiffening is associated with abnormalities in central aortic flow that can activate endothelium.7
Activation of the endothelium and increased pulsatile stress
on the arterial wall may promote atherogenesis.8 These
adverse consequences of abnormal ventricular–vascular interaction may culminate in myocardial or cerebral infarction
or congestive heart failure and may therefore explain the
observed association between higher PP and increased clinical events.
Abnormal aortic function in hypertension is generally
attributed to accelerated breakdown of elastin in the aorta,
leading to dilatation of the lumen and stiffening of the wall as
elastin is replaced with stiffer collagen.9 However, the status
of “arterial stiffness” and the genesis of increased PP in
hypertension remain controversial. The primary hemodynamic measures of aortic stiffness are pulse-wave velocity (PWV)
and local or characteristic impedance (Zc). PWV affects the
timing of wave reflection and has been shown to predict
adverse events in hypertensives10 and the elderly.11 Both
central (carotid–femoral) and peripheral (carotid– brachial,
carotid–radial) vessels can be assessed by use of this approach. Zc, which has been less well studied in hypertension,
is an indicator of the pressure generated by a given flow
waveform in the proximal aorta during early systole, before
Received December 11, 2002; de novo received May 9, 2003; revision received June 12, 2003; accepted June 13, 2003.
From Cardiovascular Engineering, Inc, Holliston, Mass (G.F.M.); Centre hospitalier de l’Universite Laval, Ste Foy, Quebec, Canada (Y.L.); Q&T
Research, Inc, Sherbrooke, Quebec, Canada (J.-P.O.); State University of New York at Buffalo, Buffalo, NY (J.L.I.); Orange County Research Center,
Orange, Calif (J.N.); Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ (L.J.K., A.J.B.); and Brigham and Women’s Hospital,
Boston, Mass (M.A.P.).
Drs Mitchell, Lacourcière, Ouellet, Izzo, Neutel, and Pfeffer have received grants from Bristol-Myers Squibb; Drs Kerwin and Block are employees
of Bristol-Myers Squibb. Dr Mitchell is owner of Cardiovascular Engineering, Inc, a company that makes devices that measure vessel stiffness.
Correspondence to Gary F. Mitchell, MD, Cardiovascular Engineering, Inc, 327 Fiske St, Holliston, MA 01746. E-mail garyfmitchell@mindspring.com
© 2003 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.org
DOI: 10.1161/01.CIR.0000093435.04334.1F
1592
Mitchell et al
Downloaded from http://ahajournals.org by on June 14, 2020
return of the reflected pressure wave. These related measures
of arterial stiffness have important differences. Each has a
square root dependence on the product of wall elastance and
thickness. However, PWV is relatively insensitive to changes
in lumen diameter (inverse square root), whereas Zc has an
amplified sensitivity to diameter.12 This fundamentally important role of geometry is apparent from the water hammer
equation, Zc⫽4⫻PWV⫻/D2, where is density of blood
and D is vessel diameter. If the paradigm of accelerated aortic
elastin breakdown, wall stiffening, and lumen dilatation is
correct, one would expect a far greater abnormality in PWV
than Zc in hypertensives compared with normotensives, because the increase in diameter should attenuate the effects of
increased wall stiffness on Zc. As a result, conclusions
regarding aortic stiffness may differ depending on whether
PWV or Zc was assessed.
Several studies have demonstrated increased stiffness of
central arteries in hypertensives,13–15 whereas others have
demonstrated normal or even reduced stiffness in peripheral
arteries.16,17 One possible explanation for this divergence of
findings lies in the structural and functional diversity of
central elastic and peripheral muscular vessels. For example,
with normal aging, stiffness of the central arteries increases,18
whereas stiffness of peripheral arteries remains unchanged or
decreases.19 This divergent change in peripheral vessel properties may moderate the abnormalities in total arterial compliance (TAC) and pulse-wave transmission that would otherwise result from deterioration in aortic properties. A similar
pattern of differential alterations in peripheral and central
vessel stiffness was recently reported in patients with congestive heart failure.20 Thus, conclusions regarding vessel
stiffness in hypertension may also differ depending on the
vascular territory studied.
In addition to these technical issues, the possibility remains
that abnormalities in vascular stiffness may be a consequence
of associated conditions, such as diabetes, smoking-related
disease, or atherosclerosis. Therefore, we designed this study
to evaluate changes in central and peripheral PWV and
central Zc in otherwise healthy adults with uncomplicated
systolic hypertension.
Methods
Study Subjects
Male or female subjects 40 to 75 years of age with a history of
hypertension were included if they were in sinus rhythm and had
systolic or mixed systolic– diastolic hypertension. Patients were
eligible for inclusion if their qualifying seated systolic blood pressure
(SBP) was ⱖ160 mm Hg and ⬍200 mm Hg and diastolic BP (DBP)
ⱕ110 mm Hg and if their supine SBP was ⱖ140 mm Hg at the time
of the hemodynamic study after withdrawal of all antihypertensive
medications for at least 1 week. Patients with known or suspected
secondary forms of hypertension were excluded. Patients with a
history of diabetes, coronary artery disease, or peripheral vascular
disease and current smokers were excluded. Control subjects had no
known history of cardiovascular disease, diabetes, lipid disorder that
was treated or that warranted treatment, malignancy (aside from skin
cancer requiring only local therapy), or other medical condition that
required pharmacotherapy, other than hormone replacement therapy.
They were in sinus rhythm and had a supine SBP ⬍140 mm Hg and
a DBP ⬍90 mm Hg at the time of the hemodynamic study. They
were recruited by each clinical center in proportion to the number of
hypertensive patients studied at that center, and their recruitment was
Hypertension and Aortic Properties
1593
stratified by median age and gender on the basis of demographics of
the overall hypertensive cohort. The study protocol was approved by
an institutional review board at each clinical center, and each patient
gave written informed consent before enrollment.
Hemodynamic Data Acquisition
Subjects were studied in the supine position after ⬇10 minutes of
rest. With a semiautomated computer-controlled device, auscultatory
BP was obtained 3 to 5 times at 2-minute intervals with a goal of
obtaining 3 sequential readings that agreed to within 5 mm Hg for
SBP and DBP. Arterial tonometry and ECG were obtained from the
brachial, radial, femoral, and carotid arteries with a custom transducer. Next, subjects were placed in the left lateral decubitus position
to image the left ventricular outflow tract in a parasternal long-axis
view. This was followed by duplicate acquisitions of simultaneous
tonometry of the carotid artery and pulsed Doppler of the left
ventricular outflow tract from an apical 5-chamber view. Finally,
body surface measurements were assessed from suprasternal notch to
brachial, radial, femoral, and carotid recording sites. All data were
digitized during the primary acquisition (ECG and tonometry pressures at 1000 Hz, audio at 12 kHz, and video at 30 frames/s),
transferred to CD-ROM, and shipped to the Core Laboratory at
Cardiovascular Engineering, Inc, for analysis.
Data Analysis
Tonometry waveforms were signal-averaged with the ECG used as a
fiducial point.21 BPs were overread by 2 reviewers in the core
laboratory. Average systolic and diastolic cuff pressures were used to
calibrate the peak and trough of the signal-averaged brachial pressure
waveform. Diastolic and integrated mean brachial pressures were
then used to calibrate carotid, radial, and femoral pressure tracings.22
Carotid-brachial, carotid-radial, and carotid-femoral PWVs were
calculated from tonometry waveforms, and body surface measurements were corrected for parallel transmission as described previously.23 Systolic ejection period was measured from the foot of the
carotid pressure waveform to the dicrotic notch. Augmentation index
was calculated as described previously.24 True peripheral amplification was calculated by taking the difference in amplitude of the
central primary wave and peripheral PP (PPP) and expressing this
difference as a percentage of the central primary wave. Apparent
amplification, which is reduced by the presence of pressure augmentation in the central aorta, was calculated by taking the difference in
PPP and central PP (CPP) and expressing this difference as a
percentage of CPP.25 The central primary wave was defined as the
difference between DBP and pressure at the first systolic inflection
point or peak. Zc was estimated in the time domain as described
previously.20,26,27 Pressure waveforms were decomposed into forward (Pf) and backward (Pb) waves in the time domain28; the ratio of
their amplitudes was taken as an index of global reflection, and the
extent of their temporal overlap, expressed as a percentage of
systolic ejection period, was taken as an index of abnormal reflected
wave timing. Proximal aortic compliance per unit length was
calculated as described previously.12
Statistical Analysis
Baseline characteristics were tabulated and compared by use of a 2
statistic for dichotomous variables and ANOVA for continuous
variables. By use of a general linear model, comparisons were
adjusted for differences in relevant covariates, which were identified
by correlation analysis. The hemodynamic correlates of PP were
assessed by forward stepwise linear regression. Variance inflation
factors were evaluated at each step in these models and were
confirmed to be well below 10. The contribution of each predictor
variable to the observed difference in PP between control and
hypertensive subjects was estimated by multiplying the raw regression coefficient for predictor variables that entered the model by the
observed difference in predictor variable between groups. All models
included a gender term or were evaluated separately by gender.
Reproducibility for impedance variables, such as Zc, obtained by use
of a noninvasive approach is high, as we20 and others22 have shown
1594
Circulation
September 30, 2003
TABLE 1.
Clinical Characteristics
Normotensive
Hypertensive
P
Variable
Female
(n⫽11)
Male
(n⫽19)
Female
(n⫽50)
Male
(n⫽78)
HTN
Gender
H⫻G
Age, y
56⫾10
60⫾10
61⫾8
60⫾8
0.126
0.320
0.093
160⫾6
175⫾8
0.524
⬍0.001
0.939
Height, cm
160⫾8
176⫾7
Weight, kg
62⫾10
80⫾14
79⫾20
91⫾15
⬍0.001
⬍0.001
0.371
Body mass index, kg/m2
24⫾3
26⫾3
31⫾7
30⫾4
⬍0.001
0.994
0.201
HTN indicates hypertension; H⫻G, interaction term between hypertension and gender.
previously. Values are presented as the mean⫾SD except as noted.
A 2-sided value of P⬍0.05 was considered significant.
Results
Downloaded from http://ahajournals.org by on June 14, 2020
Characteristics of the study participants are presented in Table 1.
By design, hypertensive and control subjects were evenly
distributed by age and gender. Hypertension was associated with
increased weight and body mass index in both genders. Higher
SBP was associated with higher mean arterial pressure (MAP)
and PP in hypertensives of either gender (Table 2). PP was
disproportionately elevated in hypertensive women, as indicated
by the significant interaction term. Cardiac output, peak aortic
flow, and heart rate did not differ by BP status, although peak
flow was reduced in hypertensives (P⫽0.021) when adjusted for
weight. Gender differences in cardiac output were not significant
(P⫽0.384) when adjusted for weight, whereas differences in
peak flow (P⬍0.001) and stroke volume (P⫽0.004) persisted,
indicating that cardiac output was successfully maintained by
higher heart rate in women. Proximal aortic stiffness (Zc) was
increased and TAC was reduced in women compared with men
and in hypertensives compared with normotensives (Table 2).
TABLE 2.
Carotid-brachial, carotid-radial, and especially carotid-femoral
PWVs were all elevated in hypertensive subjects at ambient
pressures (Table 2).
Hypertension and female gender were associated with earlier
return of the reflected pressure wave to the central aorta and with
longer systolic ejection period, despite the higher heart rate in
women, resulting in increased temporal overlap between forward and reflected pressure waves (Table 3). This excess
overlap, coupled with increased global reflection coefficient
(Pb/Pf) in hypertensives, resulted in increased augmentation
index and delayed timing of peak pressure in hypertensives and
women. True amplification was higher in hypertensives, consistent with the elevated peripheral resistance and global reflection
coefficient, whereas apparent amplification did not differ significantly (Table 3).
Because vessel stiffness may be dependent on MAP and
weight, which differed between BP groups, we performed
correlation analyses adjusted for gender and BP status (Table 4).
On the basis of these results, we evaluated general linear models
that included terms for gender and BP group (all models) and a
Hemodynamic Data
Normotensive
Variable
Female
(n⫽11)
Male
(n⫽19)
Hypertensive
P
Female
(n⫽50)
Male
(n⫽78)
HTN
Gender
H⫻G
SBP, mm Hg
114⫾12
125⫾9
167⫾16
162⫾12
⬍0.001
0.301
0.004
DBP, mm Hg
64⫾6
69⫾8
83⫾9
88⫾10
⬍0.001
0.033
0.813
PP, mm Hg
50⫾8
56⫾13
84⫾18
74⫾13
⬍0.001
0.623
0.012
MAP, mm Hg
85⫾8
91⫾6
118⫾10
117⫾10
⬍0.001
0.295
0.067
Heart rate, bpm
68⫾10
62⫾9
65⫾10
63⫾9
0.687
0.032
0.245
Cardiac output, mL/s
68⫾9
75⫾15
69⫾20
77⫾18
0.602
0.054
0.961
Stroke volume, mL
61⫾9
75⫾16
64⫾13
74⫾15
0.597
⬍0.001
0.487
Peripheral resistance, DSC
1686⫾220
1673⫾353
2415⫾577
2131⫾534
⬍0.001
0.166
0.208
First flow modulus, mL/s
117⫾14
134⫾26
117⫾29
135⫾30
0.919
First impedance modulus, DSC
236⫾48
221⫾63
444⫾147
314⫾105
Peak flow, mL/s
276⫾28
351⫾69
273⫾57
Characteristic impedance, DSC
185⫾35
159⫾43
Proximal aortic compliance, CD
0.75⫾0.29
Total arterial compliance, mL/mm Hg
1.40⫾0.35
Carotid-brachial PWV, m/s
0.870
0.005
⬍0.001
0.003
0.017
335⫾71
0.598
⬍0.001
0.625
268⫾85
208⫾67
⬍0.001
0.003
0.187
0.76⫾0.28
0.37⫾0.19
0.46⫾0.21
⬍0.001
0.197
0.337
1.79⫾0.53
1.03⫾0.40
1.28⫾0.40
⬍0.001
⬍0.001
0.427
8.2⫾1.9
8.2⫾1.4
8.7⫾1.9
10.4⫾2.4
0.004
0.101
0.094
Carotid-radial PWV, m/s
9.4⫾1.7
9.6⫾1.3
10.3⫾1.5
11.3⫾1.9
⬍0.001
0.141
0.274
Carotid-femoral PWV, m/s
8.2⫾2.3
9.4⫾2.8
12.7⫾4.3
12.3⫾2.9
⬍0.001
0.630
0.220
DSC indicates (dyne䡠s)/cm5; CD, 10⫺5 cm4/dyne.
Mitchell et al
TABLE 3.
Hypertension and Aortic Properties
1595
Waveform Morphology
Normotensive
Variable
Hypertensive
P
Female
(n⫽11)
Male
(n⫽19)
Female
(n⫽50)
Male
(n⫽78)
HTN
118⫾21
127⫾23
94⫾20
114⫾24
⬍0.001
Gender
H⫻G
0.002
0.243
Timing from foot, ms
Inflection point, ti
Pressure peak, tmax
188⫾46
194⫾48
221⫾19
213⫾27
⬍0.001
0.904
0.241
Systolic ejection period, tes
313⫾25
310⫾25
339⫾25
315⫾24
0.003
0.008
0.036
Pressure at inflection point
105⫾11
114⫾7
143⫾12
146⫾14
⬍0.001
0.022
0.247
Peak pressure
114⫾11
122⫾11
169⫾17
161⫾14
⬍0.001
0.986
0.006
89⫾8
99⫾10
127⫾11
128⫾11
⬍0.001
0.015
0.055
Pressures, mm Hg
End-systolic pressure
Augmentation index, %
16⫾13
11⫾13
30⫾9
20⫾10
⬍0.001
0.001
0.152
Forward wave, Pf, mm Hg
41⫾7
44⫾10
64⫾16
56⫾11
⬍0.001
0.394
0.050
Reflected wave, Pb, mm Hg
17⫾2
19⫾5
29⫾6
26⫾5
⬍0.001
0.629
0.011
Pb/Pf amplitude ratio, %
42⫾7
44⫾8
47⫾9
47⫾8
0.014
0.632
0.430
Pb/Pf temporal overlap, %
62⫾8
59⫾9
72⫾7
64⫾8
⬍0.001
⬍0.001
0.111
Apparent amplification, %
1.2⫾8.8
6.1⫾11.2
0.108
0.022
0.943
True amplification, %
23.4⫾19.4
23.3⫾14.4
42.5⫾22.3
29.8⫾18.4
0.002
0.119
0.125
Natural log Eh, ln(dyne/cm)
14.3⫾0.6
14.7⫾0.6
15.2⫾0.8
15.3⫾0.5
⬍0.001
0.042
0.226
2.4⫾0.3
2.9⫾0.5
2.6⫾0.5
2.9⫾0.5
0.375
⬍0.001
0.619
Effective aortic diameter, cm
Downloaded from http://ahajournals.org by on June 14, 2020
gender– hypertension interaction (PPP, CPP) and that adjusted
further for MAP (regional PWVs), weight (Zc, TAC), or both
(PPP, CPP). Differences in PPP, CPP, Zc, and TAC remained
(all P⬍0.001), whereas differences in regional PWV were no
longer significant (all P⬎0.191) (Figure). Because this was a
multicenter study, we repeated these models and included a
grouping variable for center number. Differences in PPP, CPP,
Zc, and TAC remained significant. Because the relationship
between MAP and vessel stiffness may be nonlinear, we
repeated those models with a MAP term (CPP, PPP, regional
PWVs) and included a MAP2 term as well. Differences in
regional PWV remained not significant (P⬎0.500), whereas
differences in CPP and PPP remained highly significant
(P⬍0.001).
A predominant increase in Zc with a lesser increase in
carotid-femoral PWV in hypertensives implied abnormal aortic
geometry as a contributing factor, because changes in wall
TABLE 4. Correlations Between MAP, Weight, and
Stiffness Measures
Variable
MAP, mm Hg
Weight, kg
R
R
P
P
PPP, mm Hg
0.192
0.019 ⫺0.167
0.042
CPP, mm Hg
0.191
0.020 ⫺0.161
0.049
Characteristic impedance, DSC
0.017
0.840 ⫺0.262
0.001
Proximal aortic compliance, CD
⫺0.106
0.199
0.114
0.168
Total arterial compliance, mL/mm Hg
⫺0.088
0.287
0.317 ⬍0.001
Carotid-brachial PWV, m/s
0.321 ⬍0.001
0.157
0.055
Carotid-radial PWV, m/s
0.338 ⬍0.001
0.140
0.088
Carotid-femoral PWV, m/s
0.195
0.017 ⫺0.048
0.560
DSC indicates (dyne䡠s)/cm5; CD, 10⫺5 cm4/dyne.
⫺2.5⫾9.4
2.7⫾10.9
stiffness or thickness have equal impact on Zc and PWV.12 To
assess this hypothesis, effective aortic diameter and the aortic
elastance–wall thickness product (Eh) were calculated as described previously.12 Eh was natural log transformed (lnEh) to
normalize variance between groups. At ambient pressure, hypertensive patients had comparable diameter and increased lnEh.
Correlation analysis of diameter and lnEH with MAP and
weight, adjusted for hypertension and gender, demonstrated
significant relations of diameter with MAP (R⫽0.21, P⫽0.007)
and weight (R⫽0.26, P⫽0.001) and of lnEh with MAP
(R⫽0.19, P⫽0.019). In male subjects, a general linear model
that included hypertension, MAP, a MAP-hypertension interaction term, and weight demonstrated reduced aortic diameter
(P⫽0.048) and reduced the slope of the relationship between
MAP and diameter (P⫽0.037) in hypertensives. At the overall
average MAP (112 mm Hg) and weight (88.8 kg), estimated
diameter was significantly lower in hypertensive versus normotensive men (mean⫾SEM, 2.87⫾0.06 versus 3.78⫾0.38 cm,
respectively; P⫽0.020). A similar model did not reveal a
significant difference in diameter between hypertensive and
normotensive women (mean⫾SEM, 2.48⫾0.08 versus
2.72⫾0.48, respectively; P⫽0.621), although effective diameter
adjusted for weight and MAP remained significantly lower in
women compared with men (mean⫾SEM, 2.70⫾0.07 versus
2.94⫾0.06, respectively; P⬍0.001). In models for lnEh that
included MAP, hypertension, and an interaction term, neither
hypertension nor the interaction term was significant in men or
women (all P⬎0.500), indicating that at comparable pressure,
lnEh did not differ in normotensive and hypertensive subjects of
either gender.
Hemodynamic correlates of PPP and CPP were assessed
separately for men and women by stepwise forward linear
regression analysis. Candidate predictor variables included mea-
1596
Circulation
September 30, 2003
CPP adjusted for MAP and weight and
Zc adjusted for weight in normotensive
and hypertensive women and men. Values shown correspond to
MAP⫽112 mm Hg and weight⫽84 kg.
Values are mean⫾SEM.
Downloaded from http://ahajournals.org by on June 14, 2020
sures of central pressure-flow (Zc, peak aortic flow, time to peak
flow) and pressure-volume (TAC, stroke volume) interactions,
wave propagation (carotid-brachial, carotid-radial, carotidfemoral PWVs), wave reflection (augmentation index, wave
amplification, global reflection coefficient, systolic ejection
period), steady-flow load (peripheral resistance), and heart rate.
Augmentation index was used in the model for CPP, whereas
true waveform amplification was used in the model for PPP. The
final models and the contribution of each predictor variable to
the observed difference in PP between control and hypertensive
subjects are presented in Table 5. These analyses demonstrated
that Zc accounted for a substantial component of the difference
in PP between normal and hypertensive subjects, especially in
women. Indices of wave amplification and reflection (women)
and TAC (men) accounted for much of the remainder of the PP
difference (Table 5).
Discussion
This study evaluated abnormalities in CPP and PPP and
arterial stiffness in subjects with uncomplicated systolic
hypertension compared with normotensive subjects of comparable age and gender. We demonstrated that CPP and PPP
were elevated out of proportion to the increase in MAP in
subjects with hypertension. Increased PP was associated with
a marked increase in Zc, which was independent of MAP,
along with increases in central (carotid-femoral) and peripheral (carotid-brachial, carotid-radial) PWV, which were not
independent of MAP. This pattern of elevation of Zc out of
TABLE 5. Correlates of Increased PP and Their Contribution to Differences in PP
Between Hypertensive and Control Subjects
Women
Men
B
P
⌬PP
0.27
⬍0.001
5.2
B
P
⌬PP
PPP
True amplification, %
Total arterial compliance, mL/mm Hg
Timing of peak Q, ms
䡠䡠䡠
䡠䡠䡠
0.15
⬍0.001
0.9
䡠䡠䡠
⫺13.05
⬍0.001
6.7
⬍0.001
1.2
0.57
0.027
1.6
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
0.70
䡠䡠䡠
⬍0.001
䡠䡠䡠
⫺0.6
0.16
⬍0.001
Carotid-femoral PWV, m/s
䡠䡠䡠
0.52
䡠䡠䡠
0.004
䡠䡠䡠
2.3
⫺0.27
Pb/Pf amplitude ratio, %
0.55
⬍0.001
2.7
䡠䡠䡠
Peak flow, mL/s
0.24
⬍0.001
⫺0.7
Stroke volume, mL
䡠䡠䡠
0.27
䡠䡠䡠
⬍0.001
䡠䡠䡠
22.3
Characteristic impedance, DSC
Total ⌬PP, mm Hg
31.9
7.9
17.8
CPP
Augmentation index, %
0.55
⬍0.001
8.0
0.40
⬍0.001
3.3
Total arterial compliance, mL/mm Hg
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
⫺10.96
⬍0.001
5.6
Timing of peak Q, ms
䡠䡠䡠
⬍0.001
䡠䡠䡠
3.1
⫺0.30
⬍0.001
1.4
Pb/Pf amplitude ratio, %
䡠䡠䡠
0.62
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
Peak flow, mL/s
0.25
⬍0.001
⫺0.8
Stroke volume, mL
䡠䡠䡠
0.30
䡠䡠䡠
⬍0.001
䡠䡠䡠
25.1
䡠䡠䡠
0.71
䡠䡠䡠
⬍0.001
䡠䡠䡠
⫺0.6
0.20
⬍0.001
Characteristic impedance, DSC
Total ⌬PP, mm Hg
35.4
9.8
19.5
PPP model R2⫽0.95 in women and 0.82 in men. CPP model R2⫽0.93 in women and 0.80 in men.
DSC indicates (dyne䡠s)/cm5. Variables with ellipses (䡠 䡠 䡠) did not enter the stepwise model.
Mitchell et al
Downloaded from http://ahajournals.org by on June 14, 2020
proportion to elevation of PWV at comparable MAP was
suggestive of reduced aortic diameter in hypertensives, which
was confirmed by analysis of effective diameter calculated
from the water hammer equation. In contrast, aortic wall
stiffness (Eh) did not differ from control when evaluated at
comparable MAP. Abnormal Zc accounted for much of the
difference in PP between hypertensive and normotensive
subjects, especially in women, with TAC in men and measures of wave reflection in women emerging as the next
largest contributor. These data suggest that an abnormality in
proximal aortic diameter, rather than a difference in the
material properties of the aortic wall, may be responsible for
the abnormal aortic pressure-flow relationship and increased
CPP and PPP in subjects with systolic hypertension.
A number of previous studies have used invasive methods to
evaluate aortic input impedance in hypertensive patients undergoing cardiac catheterization and have presented conflicting
results,14,29 –33 possibly because subjects differed considerably
among studies. Subjects in the studies by Ting et al32,33 were
ethnic Chinese and were relatively young (mean age, 32 years).
Like many previous studies, elevated DBP was required, thus
systematically eliminating subjects with widest PP and presumably stiffest aortas. This may explain the low value for Zc that
was reported for hypertensives. Merillon et al29 studied a
younger group of hypertensives (group mean ages, 38 and 32
years for normotensive versus hypertensive) and found no
difference in Zc; however, a later study in slightly older subjects
(mean ages, 42 and 46 years, respectively) revealed increased Zc
in hypertensives.30 The study by Nichols et al,14 which more
closely parallels ours in terms of patient characteristics (although
their subjects were still a decade younger), found increased Zc in
the hypertensive subjects. Using a noninvasive approach, we
have extended the foregoing observations to a cohort of relatively healthy normotensive and hypertensive subjects with no
evidence of atherosclerosis and no indication for cardiac catheterization. Consistent with the current emphasis on use of SBP to
diagnose and stage hypertension,34 we did not specify a minimum entry criterion for DBP in the hypertensive group. Using
this approach, we found a marked elevation of Zc in hypertensives. Furthermore, by analyzing differential changes in PWV
and Zc, we showed that aortic geometry rather than wall
properties may be involved in the abnormal pulsatile hemodynamics of hypertension.
Consistent with previous observations,15–17 after adjusting for
differences in mean arterial pressure, we found no differences in
the stiffness of peripheral (brachial or radial) arteries in hypertensives and control subjects. There was a disproportionate
contribution of wave reflection to CPP and PPP in women,
which may be a manifestation of shorter stature.35 Early systolic
loading caused by increased Zc may explain reduced peak flow
and stroke volume and increased heart rate in women, which
may in turn account for the reduced dependence of PP on TAC.
Several hypotheses have been proposed to explain increased
central aortic stiffness in patients with hypertension. In animal
models of hypertension, increased deposition of collagen and
elastin has been observed in the aortic wall,36 suggesting that
changes in wall stiffness may provide the basis for abnormal
vessel properties. However, our analysis of the pattern of change
of Zc and PWV (ie, MAP-adjusted abnormality in Zc exceeds
Hypertension and Aortic Properties
1597
abnormality in PWV) suggests that reduced vessel diameter may
play an important role in the genesis of increased functional
stiffness of the aorta. Although our estimate of diameter is based
on a functional rather than an image-based assessment, our mean
values, if converted to areas, are comparable to those previously
reported by several groups,29,31,33,37,38 most of whom also found
no difference in proximal aortic dimension at ambient pressure
despite much higher MAP in the hypertensive group. An
evaluation of determinants of aortic root diameter in participants
in the Framingham Heart Study provides additional strong
evidence that aortic diameter may be abnormally small in
patients with systolic hypertension. These investigators demonstrated that aortic root diameter, when evaluated as a continuous
variable, was inversely related to SBP and brachial PP.39
Furthermore, aortic root dilatation, when assessed as a categorical variable (⬎95th percentile), was also inversely related to
SBP and PP. Women had lower aortic root diameter than men
after body size had been accounted for in that study, which is
consistent with our finding of reduced effective diameter, higher
Zc, and higher PP despite lower peak flow in normotensive and
hypertensive women. Thus, image-based methods used in previous studies and physiological measures used in our study both
indicate that aortic diameter is reduced rather than increased in
patients with hypertension.
The basis for abnormal pressure-flow and pressure-diameter
relationships in our hypertensive subjects remains speculative.
Arterial diameter is influenced by the interaction of local and
systemic forces. Increased aortic tone or myocyte hypertrophy in
the setting of hyperactivity of the sympathetic nervous system or
in response to elevated MAP (myogenic tone) may contribute.
Alternatively, abnormal endothelial function may be involved in
the functional imbalance between aortic flow and diameter that
we have described. It is well accepted that the endothelium is
capable of modulating smooth muscle mass and tone and
therefore diameter in large and small blood vessels; this modulation may be impaired in the aorta of hypertensives. Finally, a
primary abnormality in aortic diameter or resting smooth muscle
tone may be involved.
Zc has a strong dependence on vessel diameter (to a power of
2.5), whereas PWV has a square root dependence.20 Changes in
vessel tone therefore have an amplified effect on Zc and PP and
a moderated effect on PWV, resulting in an expected 5-fold
greater change in Zc than PWV with modest changes in vessel
diameter, which is comparable to what has been observed
experimentally.40 Even minor alterations in the regulation of
vascular tone can thus have a major effect on Zc41,42 and PP.
These observations, coupled with findings from our recent study,
which demonstrated that aortic properties can be favorably
modified after a relatively short (12-week) intervention,23 are
consistent with the hypothesis that at least a component of
abnormalities in large artery function is dynamic and reversible
rather than structural and irreversible. These findings suggest
that an earlier hypothesis that aortic stiffness in hypertension is
attributable to accelerated fragmentation of elastin, leading to a
dilated aorta with a stiffer wall, should be critically reevaluated.9
Instead, our study indicates that reduced aortic diameter with
unchanged wall properties contributes to elevated PP and may
represent a mechanism rather than a result of systolic hypertension.
1598
Circulation
September 30, 2003
The limitations of our study should be underscored. We
excluded patients with clinical evidence of atherosclerosis,
although we did not perform a cardiac catheterization on each
patient. Therefore, we cannot exclude the possibility of subclinical atherosclerotic disease. Furthermore, although we had welldefined criteria for inclusion of normal control subjects with
respect to the presence of classic risk factors (glucose, lipids), we
did not measure these values in the context of the present study.
Therefore, we cannot exclude the possibility that intranormal
variation in glucose or lipid levels contributed to the observed
differences in vessel stiffness.
In summary, we have shown that PP is elevated out of
proportion to the increase in mean arterial pressure in patients
with systolic hypertension because of increased Zc and reduced
effective aortic diameter. Elevated Zc accounted for nearly half
of the excess PP in men and more than two thirds of the excess
PP in women with hypertension. Because elevated PP is an
important risk factor for cardiovascular end points, central aortic
stiffness represents an attractive new target in the treatment of
systolic hypertension.
Acknowledgment
This study was funded by a grant from Bristol-Myers Squibb
Pharmaceutical Research Institute.
References
Downloaded from http://ahajournals.org by on June 14, 2020
1. Madhavan S, Ooi WL, Cohen H, et al. Relation of pulse pressure and
blood pressure reduction to the incidence of myocardial infarction.
Hypertension. 1994;23:395– 401.
2. Fang J, Madhavan S, Cohen H, et al. Measures of blood pressure and
myocardial infarction in treated hypertensive patients. J Hypertens. 1995;13:
413–419.
3. Domanski MJ, Davis BR, Pfeffer MA, et al. Isolated systolic hypertension:
prognostic information provided by pulse pressure. Hypertension. 1999;34:
375–380.
4. Domanski MJ, Norman J, Wolz M, et al. Cardiovascular risk assessment
using pulse pressure in the first National Health and Nutrition Examination
Survey (NHANES I). Hypertension. 2001;38:793–797.
5. Saba PS, Roman MJ, Pini R, et al. Relation of arterial pressure waveform to
left ventricular and carotid anatomy in normotensive subjects. J Am Coll
Cardiol. 1993;22:1873–1880.
6. Roman MJ, Ganau A, Saba PS, et al. Impact of arterial stiffening on left
ventricular structure. Hypertension. 2000;36:489 – 494.
7. Gimbrone MA, Topper JN, Nagel T, et al. Endothelial dysfunction, hemodynamic forces, and atherogenesis. Ann N Y Acad Sci. 2000;902:230 –240.
8. Lyon RT, Runyon-Hass A, Davis HR, et al. Protection from atherosclerotic
lesion formation by reduction of artery wall motion. J Vasc Surg. 1987;5:
59 –67.
9. O’Rourke M. Mechanical principles in arterial disease. Hypertension. 1995;
26:2–9.
10. Laurent S, Boutouyrie P, Asmar R, et al. Aortic stiffness is an independent
predictor of all-cause and cardiovascular mortality in hypertensive patients.
Hypertension. 2001;37:1236 –1241.
11. Meaume S, Benetos A, Henry OF, et al. Aortic pulse wave velocity predicts
cardiovascular mortality in subjects ⬎70 years of age. Arterioscler Thromb
Vasc Biol. 2001;21:2046 –2050.
12. Mitchell GF, Pfeffer MA, Finn PV, et al. Equipotent antihypertensive agents
variously affect pulsatile hemodynamics and regression of cardiac hypertrophy in spontaneously hypertensive rats. Circulation. 1996;94:2923–2929.
13. Avolio AP, Deng FQ, Li WQ, et al. Effects of aging on arterial distensibility
in populations with high and low prevalence of hypertension: comparison
between urban and rural communities in China. Circulation. 1985;71:
202–210.
14. Nichols WW, O’Rourke MF, Avolio AP, et al. Ventricular/vascular interaction in patients with mild systemic hypertension and normal peripheral
resistance. Circulation. 1986;74:455– 462.
15. Boutouyrie P, Laurent S, Benetos A, et al. Opposing effects of ageing on
distal and proximal large arteries in hypertensives. J Hypertens Suppl. 1992;
10:S87–S91.
16. Hayoz D, Rutschmann B, Perret F, et al. Conduit artery compliance and
distensibility are not necessarily reduced in hypertension. Hypertension.
1992;20:1–6.
17. Laurent S, Hayoz D, Trazzi S, et al. Isobaric compliance of the radial artery
is increased in patients with essential hypertension. J Hypertens. 1993;11:
89 –98.
18. Learoyd BM, Taylor MG. Alterations with age in the viscoelastic properties
of human arterial walls. Circ Res. 1966;18:278 –292.
19. van der Heijden-Spek JJ, Staessen JA, Fagard RH, et al. Effect of age on
brachial artery wall properties differs from the aorta and is gender dependent:
a population study. Hypertension. 2000;35:637– 642.
20. Mitchell GF, Tardif JC, Arnold JMO, et al. Pulsatile hemodynamics in
congestive heart failure. Hypertension. 2001;38:1433–1439.
21. Mitchell GF, Pfeffer MA, Westerhof N, et al. Measurement of aortic input
impedance in the rat. Am J Physiol. 1994;267:H1907–H1915.
22. Kelly R, Fitchett D. Noninvasive determination of aortic input impedance and
external left ventricular power output: a validation and repeatability study of
a new technique. J Am Coll Cardiol. 1992;20:952–963.
23. Mitchell GF, Izzo JL, Lacourcière Y, et al. Omapatrilat reduces pulse
pressure and proximal aortic stiffness in patients with systolic hypertension:
results of the Conduit Hemodynamics of Omapatrilat International Research
Study. Circulation. 2002;105:2955–2961.
24. Murgo JP, Westerhof N, Giolma JP, et al. Aortic input impedance in normal
man: relationship to pressure wave forms. Circulation. 1980;62:105–116.
25. Safar ME, Blacher J, Pannier B, et al. Central pulse pressure and mortality in
end-stage renal disease. Hypertension. 2002;39:735–738.
26. Dujardin JP, Stone DN. Characteristic impedance of the proximal aorta
determined in the time and frequency domain: a comparison. Med Biol Eng
Comput. 1981;19:565–568.
27. Lucas CL, Wilcox BR, Ha B, et al. Comparison of time domain algorithms
for estimating aortic characteristic impedance in humans. IEEE Trans
Biomed Eng. 1988;35:62– 68.
28. Westerhof N, Sipkema P, Van Den Bos GC, et al. Forward and backward
waves in the arterial system. Cardiovasc Res. 1972;6:648 – 656.
29. Merillon JP, Motte G, Fruchaud J, et al. Evaluation of the elasticity and
characteristic impedance of the ascending aorta in man. Cardiovasc Res.
1978;12:401– 406.
30. Merillon JP, Masquet C, Dahan M, et al. Changes in left ventricular performance during chronic pressure or volume overload: importance of physical
properties of the arterial system. J Cardiovasc Pharmacol. 1985;7:S36–S40.
31. Merillon JP, Motte G, Masquet C, et al. Relationship between physical
properties of the arterial system and left ventricular performance in the course
of aging and arterial hypertension. Eur Heart J. 1982;3(suppl A):95–102.
32. Ting CT, Brin KP, Lin SJ, et al. Arterial hemodynamics in human hypertension. J Clin Invest. 1986;78:1462–1471.
33. Ting CT, Chen JW, Chang MS, et al. Arterial hemodynamics in human
hypertension: effects of the calcium channel antagonist nifedipine. Hypertension. 1995;25:1326 –1332.
34. Izzo JL Jr, Levy D, Black HR. Importance of systolic blood pressure in older
Americans. Hypertension. 2000;35:1021–1024.
35. Smulyan H, Asmar RG, Rudnicki A, et al. Comparative effects of aging in
men and women on the properties of the arterial tree. J Am Coll Cardiol.
2001;37:1374 –1380.
36. Wolinsky H. Long-term effects of hypertension on the rat aortic wall and
their relation to concurrent aging changes: morphological and chemical
studies. Circ Res. 1972;30:301–309.
37. Isnard RN, Pannier BM, Laurent S, et al. Pulsatile diameter and elastic
modulus of the aortic arch in essential hypertension: a noninvasive study.
J Am Coll Cardiol. 1989;13:399 – 405.
38. Bouthier JD, De Luca N, Safar ME, et al. Cardiac hypertrophy and arterial
distensibility in essential hypertension. Am Heart J. 1985;109:1345–1352.
39. Vasan RS, Larson MG, Levy D. Determinants of echocardiographic aortic
root size: the Framingham Heart Study. Circulation. 1995;91:734 –740.
40. Cox RH. Pressure dependence of the mechanical properties of arteries in
vivo. Am J Physiol. 1975;229:1371–1375.
41. Dujardin JP, Stone DN, Paul LT, et al. Response of systemic arterial input
impedance to volume expansion and hemorrhage. Am J Physiol. 1980;238:
H902–H908.
42. Stone DN, Dujardin JP. Changes in smooth muscle tone influence characteristic impedance of the aorta. Am J Physiol. 1984;246:H1–H7.