Age-related changes in carotid artery flow and pressure pulses: possible implications for cerebral microvascular disease.
ABSTRACT We sought to establish the relation between the pulsatile components of pressure and flow waveforms in the carotid artery and their change with age.
Distention (pressure) and axial flow velocity waveforms were recorded noninvasively and simultaneously from the common carotid artery of 56 healthy subjects aged 20 to 72 years.
There was a close relation between the time intervals of pressure and flow waves: from foot to first shoulder or peak, to second shoulder or peak, and to incisura (r=0.97, P<0.0001 for each), which approximated the line of identity. The peak and nadir of flow velocity decreased with age, but late systolic flow augmentation increased substantially (1.6 times in the older group); this can be attributed to earlier wave reflection from the lower body. Pressure augmentation index (PAI) and flow augmentation index (FAI) increased similarly with age (PAI (%) = 0.84 x age - 26.6; FAI (%) = 0.75 x age + 11.9; both P<0.0001).
Arterial stiffening with aging increases carotid flow augmentation and can explain the increasing flow fluctuations in cerebral blood vessels. Measurement of carotid FAI may provide a gauge for risk of cerebral microvascular damage, just as PAI provides a gauge for risk of left ventricular hypertrophy and failure.
- SourceAvailable from: Alberto Avolio[show abstract] [hide abstract]
ABSTRACT: Pulse wave velocity (PWV) was measured by means of transcutaneous Doppler techniques in the aorta, right arm, and right leg of 480 normal subjects of both sexes in urban Beijing, China (age range 3 to 89 years, mean age 41 +/- 20.8 SD); supine blood pressure was recorded in the brachial artery of each subject with standard sphygmomanometric procedures. Serum cholesterol was determined in a subgroup of 79 subjects (age 17 to 85 years, mean 47 +/- 26 SD). PWV (y in cm/sec) was found to vary with age (x, years) at each of the three locations according to the following regression equations: aorta, y = 9.2x + 615, r = .673 (p less than .001); right arm, y = 4.8x + 998, r = .453 (p less than .001); right leg, y = 5.6x + 791, r = .630 (p less than .001). Systolic, diastolic, mean, and pulse pressures were found to increase with age. PWV also increased with mean supine blood pressure but was not related to serum cholesterol (average 4.49 +/- 0.11 [SEM], mmol/l). Compared with that of Western populations, serum cholesterol tended to be lower at all age groups, systolic pressure higher at ages over 35 years, and PWV higher at all ages. Because change in PWV is directly related to change in arterial compliance, these results indicate that aging and not concomitant atherosclerosis (known to be rare in Asian populations) is the dominant factor associated with reduced arterial compliance and increased left ventricular load in these subjects.Circulation 08/1983; 68(1):50-8. · 15.20 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Our objective was to validate a carotid artery tonometry-derived augmentation index as a means to estimate augmentation index (AI) of ascending aortic pressure under various physiological conditions. A total of 66 patients (50 men, 16 women; mean age, 55 years; range, 21 to 78 years; 44 in Taiwan and 22 in the United States) undergoing diagnostic catheterization were studied. Arterial pressure contours were obtained simultaneously from the right common carotid artery by applanation tonometry with an external micromanometer-tipped probe and from the ascending aorta by a micromanometer-tipped catheter at baseline (n = 62), after handgrip (n = 36), or after sublingual nitroglycerin administration (n = 17). The AI (expressed as percentage values) was calculated as the ratio of amplitude of the pressure wave above its systolic shoulder to the total pulse pressure. The carotid AI was consistently lower than the aortic AI, but the two were highly correlated at baseline and after both handgrip and nitroglycerin. Mean +/- SD and correlation coefficients were baseline (14 +/- 16, 28(+) +/- 17, .77), handgrip (18 +/- 19, 32(+) +/- 15, .86), and nitroglycerin (7 +/- 12, 18(+) +/- 13, .52). In addition, after adjusting for age, sex, height, blood pressure, heart rate, and study site, the changes of both AIs from baseline values with handgrip or nitroglycerin were highly associated such that the aortic AI could be approximated from the carotid AI with appropriate regression equations. The high correlations and predictable changes after interventions between the central AI and those estimated from noninvasive carotid tonometry suggest that this technique may have wide applicability for many cardiovascular studies.Hypertension 03/1996; 27(2):168-75. · 6.87 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The study was designed to test whether aortic input impedance and left ventricular power output can be accurately assessed noninvasively. Aortic input impedance describes both the pulsatile and nonpulsatile artery load encountered by the left ventricle. Until now, this measure of afterload has only been determined by invasive techniques. The aortic pressure wave was estimated by recording the calibrated carotid artery pressure wave noninvasively with use of a micromanometer-tipped probe by the technique of applanation tonometry. Flow was determined with pulsed wave Doppler measurement of ascending aortic velocity profile and aortic diameter. In 18 subjects undergoing cardiac catheterization, invasive measurements were taken to assess the accuracy of noninvasive data. In 17 other subjects noninvasive measurements were taken on different days to assess the reproducibility of results. Noninvasive pressure measurements correlated well with invasive data: systolic pressure (mm Hg), noninvasive 126 +/- 28 versus invasive 127 +/- 28, r = 0.96, p less than 0.001; diastolic pressure (mm Hg), noninvasive 71 +/- 10 versus invasive 66 +/- 7, r = 0.60, p less than 0.02; augmentation index (%), noninvasive 23.9 +/- 9.3 versus invasive 30.7 +/- 11.9, r = 0.87, p less than 0.001. Doppler-measured cardiac output was closely correlated with invasively measured flow (liters/min): Doppler, 5.3 +/- 1.2 versus invasive, 5.5 +/- 1.3, r = 0.98, p less than 0.001. Impedance and left ventricular power variables calculated from noninvasive and invasive techniques were also closely related: systemic vascular resistance (dynes.s.cm-5), noninvasive 1,479 +/- 488 versus invasive 1,502 +/- 498, r = 0.91, p less than 0.001; characteristic impedance (dynes.s.cm-5), noninvasive 137 +/- 52 versus invasive 136 +/- 79, r = 0.92, p less than 0.001; pulsatile power (mW), noninvasive 249 +/- 94 versus invasive 291 +/- 103, r = 0.91, p less than 0.001; mean power (mW), noninvasive 1,107 +/- 319 versus invasive 1,144 +/- 266, r = 0.93, p less than 0.001. Repeated measures of impedance variables and power output showed coefficients of variation of less than 9%. Measurement of noninvasive impedance by this technique provides an accurate and repeatable assessment of mean and pulsatile cardiac load.Journal of the American College of Cardiology 11/1992; 20(4):952-63. · 14.09 Impact Factor
Kozo Hirata, Toshio Yaginuma, Michael F. O'Rourke and Masanobu Kawakami
for Cerebral Microvascular Disease
Age-Related Changes in Carotid Artery Flow and Pressure Pulses : Possible Implications
Print ISSN: 0039-2499. Online ISSN: 1524-4628
Copyright © 2006 American Heart Association, Inc. All rights reserved.
is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
2006;37:2552-2556; originally published online August 31, 2006;
World Wide Web at:
The online version of this article, along with updated information and services, is located on the
is online at:
Information about subscribing to
Information about reprints can be found online at:
document. Permissions and Rights Question and Answer process is available in the
Request Permissions in the middle column of the Web page under Services. Further information about this
Once the online version of the published article for which permission is being requested is located, click
can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Requests for permissions to reproduce figures, tables, or portions of articles originally published
by guest on June 12, 2013 http://stroke.ahajournals.org/Downloaded from
Age-Related Changes in Carotid Artery Flow and
Possible Implications for Cerebral Microvascular Disease
Kozo Hirata, MD; Toshio Yaginuma, MD; Michael F. O’Rourke, MD, DSc; Masanobu Kawakami, MD
Background and Purpose—We sought to establish the relation between the pulsatile components of pressure and flow
waveforms in the carotid artery and their change with age.
Methods—Distention (pressure) and axial flow velocity waveforms were recorded noninvasively and simultaneously from
the common carotid artery of 56 healthy subjects aged 20 to 72 years.
Results—There was a close relation between the time intervals of pressure and flow waves: from foot to first shoulder or
peak, to second shoulder or peak, and to incisura (r?0.97, P?0.0001 for each), which approximated the line of identity.
The peak and nadir of flow velocity decreased with age, but late systolic flow augmentation increased substantially (1.6
times in the older group); this can be attributed to earlier wave reflection from the lower body. Pressure augmentation
index (PAI) and flow augmentation index (FAI) increased similarly with age (PAI (%)?0.84?age?26.6; FAI
(%)?0.75?age?11.9; both P?0.0001).
Conclusions—Arterial stiffening with aging increases carotid flow augmentation and can explain the increasing flow
fluctuations in cerebral blood vessels. Measurement of carotid FAI may provide a gauge for risk of cerebral
microvascular damage, just as PAI provides a gauge for risk of left ventricular hypertrophy and failure. (Stroke. 2006;
Key Words: carotid flow waveforms ? flow augmentation ? pulse wave encephalopathy ? wave reflection
pulsations. “Aortic” pulse wave velocity is determined from
the delay of the wave foot between carotid and femoral
sites.1,2The carotid pressure wave has been used as a
surrogate of the ascending aortic pressure wave in studies of
ascending aortic impedance, with change in age, in normal
populations, and in the presence of disease.3–6The carotid
pressure wave is also interpreted independently under differ-
ent pathological conditions from change in its late systolic
In contrast to the carotid pressure waveform, the clinical
significance of the pulsatile component of the carotid flow
wave is not fully understood, even though it is frequently
recorded as a part of population studies or for clinical
investigation of possible carotid disease. Although it is
accepted that upper-body flow waveforms are completely
different from wave patterns in the ascending aorta and in
lower-body arteries, such waves are known to show consid-
erable variability in normal human subjects2,9; their relation
with the carotid pressure wave and with subsequent disease
has not been considered.
ore often than not, noninvasive measures of arterial
stiffness in humans entail analysis of carotid artery
The present study was undertaken to establish the relation
between pulsatile components of the pressure/diameter wave
in the carotid artery and the accompanying flow wave in
normal subjects at different ages.
Seventy healthy, unmedicated volunteers were recruited during a
3-month period at a health awareness center at Omiya Medical
Center, Saitama, Japan. Fourteen subjects were excluded after
interview and examination on the basis of a history of hyperten-
sion, diabetes mellitus, hypercholesterolemia, or any cardiovas-
cular disease or from finding of a seated blood pressure after rest
of ?140/90 mm Hg, a fasting glucose value ?110 mg/dL, a
cholesterol value ?240 mg/dL, or some combination thereof. Fifty-
six subjects (mean?SD age, 48.2?14.3 years) were entered into the
study (Table 1). Informed consent was obtained from all subjects
according to institutional guidelines.
Blood Pressure Measurement
Brachial blood pressure was measured on the left arm in the supine
position, before carotid distention (pressure) and flow velocity
waveforms were recorded. Brachial mean blood pressure was calcu-
lated as diastolic blood pressure (DBP)?(0.33?pulse pressure).
Received May 22, 2006; accepted June 13, 2006.
From the Jichi Medical University Omiya Medical Center (K.H., M.K.), Saitama, Japan; Shizuoka City Hospital (T.Y.), Shizuoka, Japan; and
UNSW/St. Vincent’s Clinic (M.F.O’R.), Sydney, Australia.
Correspondence to M.F. O’Rourke, Suite 810, St. Vincent’s Clinic, 438 Victoria St, Darlinghurst, Sydney, NSW 2010, Australia. E-mail M.ORourke@
© 2006 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org DOI: 10.1161/01.STR.0000242289.20381.f4
by guest on June 12, 2013 http://stroke.ahajournals.org/Downloaded from
Measurements of Carotid Pressure and Flow
The distention and flow velocity waves of the right common carotid
artery were measured simultaneously with an ultrasound QFM1100
system (Hayashi Electric Co, Kawasaki, Japan), after the subjects
had rested in a supine position for 10 minutes. As previously
reported, this device determines the flow velocity waveform by the
continuous-wave Doppler method (frequency of probe, 5 MHz) and
the arterial internal diameter by the phase-locked echo-tracking
method (frequency of probe, 7.5 MHz), with good accuracy and
reproducibility.10,11Carotid distention waveforms were calibrated by
brachial mean blood pressure and DBP to obtain carotid pressure
values and were regarded as the carotid pressure waveform.5,12,13
Measured invasively, carotid distension and pressure waveforms are
virtually identical.14Tonometric measures of carotid pressure can be
inaccurate,15so that the distension waveform is preferred as a
noninvasive surrogate of carotid pressure.12,13
At least 5 carotid distention (pressure) and flow velocity wave-
forms obtained from a stable area of the record were analyzed to
determine the following parameters (Figure 1): peak systolic flow
velocity (Vs); end-diastolic flow velocity (Ved); peak flow velocity
of the secondary rise in the common carotid flow velocity waveform
(Vsr); and mean flow velocity (Vm). Pressure augmentation index
(PAI) of the distention (pressure) waveform was defined by the
following formula as reported by Laurent et al13: (P2?P1)/(P2?P0),
in which P2 is peak lumen diameter, P0 is minimal (end diastolic)
lumen diameter, and P1 is the lumen diameter at the inflection point
of the carotid distention (pressure) waveform. Flow augmentation
index (FAI) was defined similarly as (Vsr?Ved)/(Vs?Ved). Like
PAI, it is an index related to the amplitude and timing of wave
reflection.16?tp and ?tf were defined respectively as the time
between the wave foot and the peak of the initial rise in the carotid
distention (pressure) waveform (P1) and the flow velocity waveform
(Vs). ?TP and ?TF were defined as the time between the wave foot
and the peak of the secondary rise in the common carotid distention
(pressure) waveform (P2) and in the common carotid flow velocity
waveform (Vsr), respectively. EDpand EDfwere defined as the time
between the wave foot and the incisura of the carotid distention
(pressure) and flow waveforms, respectively. Flow volume was
calculated from the flow velocity and the corresponding internal
diameter of the vessel wall.
Statistical analysis was performed with StatView (version 5.0,
Abacus Concepts, Inc, Berkeley, Calif). The relation between 2
variables was determined by Pearson’s correlation coefficient. A
difference was considered significant at P?0.05. All data are
expressed as mean?SD.
Apart from age, which ranged from 20 to 72 years, and sex,
the population was homogeneous (Table 1). There was no
Figure 1. Definitions of measurements obtained from pressure
(top) and flow (bottom) waveforms in the common carotid artery.
Figure 2. Representative pressure (top) and flow velocity (bottom) in a healthy 23-year-old subject (A) and in a healthy 65-year-old
for Basic Characteristics vs Age
Basic Characteristics and Linear Regression Data
Body surface area, m2
Brachial systolic BP, mm Hg
Brachial DBP, mm Hg
pressure, mm Hg
Mean blood pressure, mm Hg
Heart rate, beats/min
NS indicates not significant.
Hirata et al Aging Changes in Carotid Pressure and Flow
by guest on June 12, 2013 http://stroke.ahajournals.org/Downloaded from
change in brachial systolic blood pressure or DBP with age,
but there was a small change in brachial pulse pressure and
The carotid pressure waveform was completely different
from the carotid flow waveform, and there were substantial
changes in both with age (Figure 2). The pressure waveform
was similar to that previously described2–7,17,18for the carotid
artery and proximal aorta, whereas the carotid flow wave was
different from that seen in other arteries with respect to 3
features: a sharp early systolic peak, a second late systolic
peak, and maintained high flow velocity at the end of diastole
(Figures 1 and 2). Despite such differences between pressure
and flow waveforms, there were close associations between
components of the waves and consistent patterns of change in
both with aging.
Associations between features of the pressure and flow waves
were so close and consistent as to indicate a functional
relation (Figure 3A). The time from wave foot to peak flow
(?tf) was the same as the time from wave foot to first pressure
inflection or peak (?tp), and the time to second systolic flow
peak (?TF) was the same as that to late systolic pressure
inflection or peak (?TP). The time to end systole was
likewise the same in pressure and flow waves, as gauged from
the wave foot to the sharp dip in pressure or flow wave, which
marked aortic valve closure (EDpand EDf).
There was also a close association (Figure 3B) between
carotid PAI as measured conventionally2,7,17and FAI, ex-
pressed as the amplitude of the difference between Vsr and
Ved, divided by the amplitude of the flow velocity pulse
(FAI; see Figure 1). PAI is accepted as a measure of wave
reflection from peripheral sites.2–8,17Similarities of FAI to
PAI indicate that the fluctuations in flow are likewise a
consequence of wave reflection from peripheral sites.
In this healthy cohort (Table 2), carotid systolic and pulse
pressure increased with age. The ?tp, ?tf, ?TP, and ?TF were
decreased, but EDpand EDfdid not change with age. There was
a progressive decrease in peak and end-diastolic flow velocity
expressed as flow volume (Table 2). In contrast to the falls in
peak and diastolic velocity with age, there was no change in Vsr
but an increase in Vsr?Ved. This was greater still when
flow waveforms with age were accompanied by increase in both
PAI and FAI (Table 2 and Figure 4).
Data presented here provide a firm basis for explaining the
shape of both flow and pressure waveforms in the carotid
artery of normal human subjects. Explanations support those
already provided for the innominate and brachial arteries.2,9,19
Figure 3. A, Relation between time-based indices derived from flow velocity (ordinate) and from pressure waveforms (abscissa).
Bottom left (closed circles); ?tf and ?tp, center (open rectangles); ?TF and ?TP, upper right (closed triangles); EDfand EDp.
Regression line for ?tf and ?tp: y?0.99x?0.74, P?0.0001. Regression line for ?TF and ?TP: y?0.99x?4.52, P?0.0001. Regres-
sion line for EDfand EDp: y?0.95x?17.10, P?0.0001. All values in ms. B, Relation between FAI (ordinate) and PAI (abscissa).
Regression line: y?0.97x?0.35.
TABLE 2. Flow and Pressure Waveform Indices of the Common Carotid Artery of the 56 Healthy Subjects
?40 14 101.8?9.829.3?6.45.87?0.57 68.6?7.7 1194.8?230.418.5?2.4 305.8?83.1 26.0?2.9440.6?95.633.3?5.1
40–5930 110.5?12.3 35.9?8.0 6.64?0.77 50.7?7.2 1043.2?262.615.4?6.4 313.1?60.224.2?4.2 493.9?91.032.2?4.9
?6012 109.3?14.241.1?5.7 6.71?0.7647.2?8.91010.8?257.413.5?2.7289.8?84.423.7?4.4504.7?151.333.1?6.5
All 56108.1?12.535.4?8.26.46?0.79 53.7?11.7 1074.2?259.315.7?3.1306.3?71.1 24.4?3.9 482.9?108.3 32.7?5.3
Abbreviations are defined in the text and expressed as mean?SD. n, No. of subjects investigated.
Symbols express significance of correlation between age and each parameter: *P?0.05; †P?0.01; ‡P?0.0001.
by guest on June 12, 2013http://stroke.ahajournals.org/Downloaded from
The complex patterns of flow in upper-body arteries belie the
simple impedance patterns that result from a comparison of
frequency components of pressure and flow waves (supple-
mental Figure I, available online at http://stroke.ahajournals.
org), which imply the influence of a functionally discrete
reflecting site in the vascular bed beyond.2The unusual
patterns of flow in upper-body arteries can be explained on
the basis of an early return of wave reflection from local
upper-body sites and their interaction with later wave reflec-
tion coming in the opposite direction from sites in the lower
part of the body2,9,18(supplemental Figure II, available online
at http://stroke.ahajournals.org). The effects of aging are
readily explained on the basis of an earlier return of wave
reflection from the lower body, as a consequence of aortic
degeneration and increased aortic pulse wave velocity. The
only difference in carotid compared with innominate and
brachial flow2,9,19(Figures 1 and 2) is the continued high
velocity in end diastole, which is attributable to the vasodi-
lated brain being the major organ supplied by this artery.
Similar high-velocity flow at end diastole is seen in the renal
artery and in a peripheral artery during local vasodilation of
its vascular bed.2,20
Carotid tonometry is widely used to assess the effects of
wave reflection through measurement of PAI.2–4,6–8,12,21The
major problem in this measurement is detection of an inflec-
tion point on the wave that corresponds to the peak of carotid
flow.2–8,12,21–23Use of the method reported here simplifies
and increases the accuracy of measuring pressure augmenta-
tion, because the peak of the flow wave is unmistakable,
whereas the shoulder of the pressure wave is often indistinct,
especially as measured by carotid tonometry.23
The effects of carotid (and aortic) PAI with age are well
known and include increased left ventricular load, left ven-
tricular hypertrophy, left ventricular failure, and increased
severity and extent of coronary atherosclerosis.2,24Data
presented here show that this increased pressure augmenta-
tion with age is accompanied by increased late systolic flow
augmentation in the carotid artery as well (Figure 4 and Table
2). This is not immediately apparent on inspection of the
carotid flow velocity waveforms (Figure 2), nor through
consideration of difference between peak and nadir of the
flow velocity waveforms (Vs?Ved significantly decreased
with aging (P?0.0001); data not shown). Carotid dilation
with normal aging has already been reported, and the mag-
nitude was similar with our study result.25This and heart rate
increases are attributable to a velocity decrease in this study.
In the presence of microvascular rarefaction with age,
velocity fluctuations would be expected to increase in the
microvasculature, even if volume carotid flow pulsations
were unchanged. This has been confirmed.26,27However,
there is another factor as well. The early systolic flow peak,
a high-frequency component of the wave, attenuates in
arteries leading down to the cerebral microvasculature.2,28
In fact, fluctuations of flow in cerebral veins and capillar-
ies are of low frequency and, though delayed, appear to
correspond to late systolic flow augmentation in the
carotid or cerebral arteries rather than to an early systolic
flow peak with respect to timing and amplitude (ie, to
Vsr?Ved, not to Vs?Ved).29,30An increased amplitude of
such cerebral venous flow pulsations has been noted in older
persons and in those with vascular dementia27,29,30and has
been related to “pulse wave encephalopathy.”26,29,31
It has been shown that cerebral microvascular disease is
increased in the presence of aortic stiffening,32and it has been
hypothesized that such microvascular disease is caused by
increased pulsatile shear forces in these vessels from in-
creased pulsatile flow velocity.20Such a view is based on the
similarity of cerebral vascular lesions in humans33and exper-
imental animals34with aging and hypertension and with
lesions in human lungs when pulsatile blood flow is increased
over many years as a consequence of congenital heart disease
with left-to-right shunt.35It is also based on the calculated
yield stress of endothelial cells, which approaches values for
disruption in the presence of high pulsatile flow,2,36together
with an increase in permeability in the endothelial cells
Figure 4. Relation between FAI (solid circles, top) and carotid
PAI (open diamonds, bottom) with age. Regression line for FAI:
y?0.75x?11.9. Regression line for PAI: y?0.84x?26.6.
TABLE 2. Continued
639.0?178.114.7?5.0 333.3?121.7133.2?21.3132.0?18.7 256.1?36.6 260.6?39.2322.3?26.6321.4?23.9
751.5?136.617.0?3.5 438.5?97.5 103.2?22.9 103.8?22.0223.2?30.2 223.1?29.1308.7?27.8310.7?29.2 18.5?9.351.6?13.4
813.6?270.719.6?4.6 523.8?193.997.5?11.7 97.3?12.0215.8?37.5215.9?35.8316.0?28.0316.0?25.325.5?8.0 59.1?11.2
736.7?189.317.0?4.4 430.5?142.7 109.5?22.6109.4?21.7229.8?36.3231.0?37.1313.6?27.6314.5?27.013.8?15.5 47.9?16.6
Hirata et alAging Changes in Carotid Pressure and Flow
by guest on June 12, 2013http://stroke.ahajournals.org/ Downloaded from