The Pulse Wave Velocity as an Early Indicator of
Atherosclerosis in Diabetic Subjects
By G. L.
WOOLAM,
AND
P. IL. SCH-NUR, B.S., C. VALLBONA, M.D.,
H. E. HOFF, M.D., PH.D.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
DESPITE exten-sive studies of the velocity
of the arterial pulse wave, relatively little practical use has been made of this expression of the functional status of the cardiovasecLlar system. It has lobug been recognized
that the velocity of a pulse wave through a
fluid-filled tube is, in part, a funetion of
the distensibility of the wall of the tube.
Investigations of the relationships between
pulse-wave velocity, pressure, tension, distensibility, and tube volume have been made
by Bramwell, Downing, and Hill' on segments of excised carotid artery and by
Hamilton, Remington, and Dow2 on mercuryfilled Gooch tubing and on the aorta of
cadavers. Factors that have been shown to
affect the velocity of the pulse wave through
the arterial tree are (a) degree of contraction of the smooth musele of the vessel,3 (b)
arterial wall elastieity,4' 5 (e) intra-arterial
diastolic pressure,4 (d) wall thickness,4-6 (e)
vessel tortuosity,6' 7 (f) diastolic diameter of
the vessel,5 (g) blood density,5 and (h) blood
flow velocity.5 In normotensive subjects, variations in the velocity of the propagation of
the pulse wave are due almost entirely to
elasticity alterations.
The theoretic implication that the transmission of the pulse wave is altered in diffuse atherosclerosis has been corroborated in
studies carried out in huiman subjects.8-10 In
a recent study, Simonson and Nakagawa"
observed that individuals with elinically evident coronary artery disease had higher
values of pulse wave velocity than healthy
subjects, and they also stressed the value of
measurements of the pulse wave -velocitv as
indicators of diffuse atherosclerosis.
The large incidence of premature vascular
deegeneration in diabetes has been well docunmented. Atheroselerosis has been reported as
the cause of death in more than 50 per cent
of the fatalities in diabetic subj ects.12
Furthermore, the reported incidenee of vascular complications in diabetes appears to
be progressively inereasing.'3 It would seem
pertinent, therefore, to study whether the
propagation of the pulse wave is altered in
inidividuals with diabetes mnellitus and, if so,
to deternmiine whether this alteration is due
to atherosclerotic changes in the structure
of the arterial wall.
Materials and Methods
The pulse wave velocity. was measured in 52
diabetic personis and the results were conmpared to
those obtained in a group of 87 healthv individuals. The miiethod of study was identical for
all the subjects.
The pulse wave veloeity was mneasured fromnl
simiiultaneous tracings of the arterial pulse detected in two points of the cardiovaseular systeml.
The transducer used to record the pressure pulse
w'as a piezoelectric crystal microphone of hihli
efficiency and high frequency response.'4 The
value of the piezoelectric erystal microphone for
recording the pressure pulse was recognized in
1941 by Miller and White,15 but they used a
plneumiatic coupling between the cup applied to
the pulsating, area and the microphone. This
coupling was the cause of a snmall timie lag between the occurrence of the pulse wave and its
reegistration. The crystal pick-up that we used
for this studv was applied directly to the skin over
the vessel, thereby eliminating any time lag and
the low efficieney of a pneumaatic coupling.
One pressure pulse pick-up was applied to the
wrist and held in position with an elastic bracelet. The second pick-up was held by the operator
over the carotid vessel near the site of the bifurcation. Thus, sinmultaneous tracings of the pressure
From the Departmeints of Physiology, Rehabilitation, and Pediatrics, Baylor University College of
Medicine, anid the Texas Institute for Rehabilitation
and Research, Houston, Texas.
Supported in part by grants HTS 5125 and OVR
318, U. S. Public Health Service.
Circulation, Volume XXV, March 1962
533
WOOLAM ET AL.
534
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
pulse at the right radial and at the right carotid
arteries could be obtained. All recordings were
nade with a Sanborn Twin-Viso direct-writing
instrumnent. The speed of the recording paper
was 10 cm. per second.
In all the subjects the measurements were made
between 2 and 5 p.m., at least 2 hours after miieals,
with the subject at a resting state (but not basal),
in the supine position, and at the resting expiratory level. Alterations in pulse wave velocity
due to extraneous factors'6-18 were muinimized by
observance of the above conditions. The arterial
blood pressure was obtained by the auscultatorv
method in the right arm with the subject in the
supine position before and after the recording
of the pressure pulses.
In order to calculate the velocity of the propagation of the pressure pulse wave, it was necessary to muake an accurate measurement of the
length of the segment of the arterial tree where
the pulse wave was recorded. This distance could
not be nmeasured directly, but it was computed
as the difference between two separate measurements. The first masurement was the distance
from the proximal tip of the right clavicle (the
usual point of bifurcation of the innominate artery) to the point where the transducer was applied over the right carotid artery. This distance
was measured along the right side of t.he neck
which was slightlv extended. The second measurement was that of the distance from the proximal
tip of the right clavicle to the place where the
transducer was applied at the stvloid process of
the radius. This distance was measured along
the course of the brachial and radial arteries with
the arm extended and the shoulder abducted 90
degrees. Bazett and Dreyer8 correlated measuremiients calculated in this fashion in the intact
cadaver with aetual measurements of the distance
betweeii two points of the arterial tree after
dissection. The error between the two mleasurements was of the order of 1 cm. or less.
The time of arrival of the pulse wave at the
point of recording was tak-en as the intercept of
the diastolic portion of the pressure pulse wave
and the ascending limub produced by the rapid
ejectioni of blood froom the ventr-icle. The intercept could always be clearly defined, and errors
introduced in attempting to determnine the transition of the presystolic ejection phase to the phase
of rapid ejection were thus avoided. Differences
in timiie of onset of the carotid and radial pulses
in the best five cyeles of each record were read
to the nearest 0.002 second and averaged.
The mnean velocity of the pulse wave in this
portion of the arterial tree was then calculated
by the formula, V-L/T (V=velocity in meters
per second; L=length of the vessels in mleters;
(and T=time of transmnission of the p-ulse wave
in seconds). The standard error of the measuremeat,19 calculated from two consecutive determinations in 10 healthy subjects was 0.06s meter per
second.
The pulse wave velocity was determined by this
method in 44 male and in ei,rbt femnale diabetic
subjects ranging in age from 5 to 75 years. The
adult patients were hospitalized at the Houston
Veterans Administration Hospital. The children
were staying in a summer camp of the Houston
Area Diabetes Association. All these patients
were taking either oral hypoglyeemic agents or
insulin, but thev did not receive other drugs known
to affect the cardiovascular system. None exhibited any evidence of arteriosclerosis or hypertension by clinical history and phvsical examination. The duration of diabetes in these subjects
as determined from the clinical history varied from
1 month to 42 years.
Eighty-seven healthy individuals, 73 males and
14 females, ranging in age from 4 to 56 years
were examined by the sa me technic as a control
group. The majority of these subjects were staff
physicians or emplovees of the Texas Institute for
Rehabilitation and Researeh and medical students
of Baylor University College of Medicine. There
was no evidence of hvpertension or diabetes in
these subjects on the basis of history. or recent
physical examinations.
Results
The statistical data pertaining to the pulse
wave velocity and the blood pressure of the
87 healthy subjects are presented in table 1.*
Table 2 indicates the data oni pulse wave
velocity anid blood pressure obtained in the
32 diabetic subjects. The duration of the
illness in each age group is also listed. In
each patient a calculation was made of the
ratio between the observed pulse wave velocitv anid the mnean value obtained in the
healthv subjects of the same age group
(Q =A1) "VT) The mean of the ratios between observed and predicted values in the
different age groups of the diabetic patients
are also presented in table 2. The probabilitv
*The individual values for eacbh healthy subject
and patient of this study have been deposited as
Document number 6913 wvith the ADI Auxiliarv
Publications Project, Photoduplication Service, Library of Congress, Washinigtoni 25, D.C. A copy
may be secured by citing tie Documiiient number ancd
by remitting 81.25 for plhotoprints, or $1.25 for 35
mm. microfilm. MIake checks payable to: Chief,
Photoduplication Service, Library of Congress.
Circulation, Volume XXV, March 1962
PULSE WAVE VELOCITY
535
Table 1
Healthy Subjects
1-10
8
11-20
11-20
13
2 1-30
32
Range
VH
a-
5.51-8.46
6.75
0.97
6.40-8.62
7.51
0.92
6.00-9.00
7.78
0.89
6.22-10.19
7.86
1.00
7.38-11.00
8.80
1.00
7.25-10.91
9.04
1.17
Systolic
blood presure
(mm. Hg)
Range
90-100
98
8.14
106-137
121
8.66
110-140
123
7.06
106-140
126
105-140
128
9.99
116-142
Diastolic
blood pressure
(mm. Hg)
Rainge
58-72
64
6.31
70-82
76
3.67
72-88
77
64-95
77
8. 28
48-90
79
80-92
Age (years)
Number of subjects
Pulse wave
velocity (VH)
(m/see.)
x
Ra
x
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
values for the differences between the mean
pulse wave velocities of the diabetic patients
and the healthy subjects are indicated in
the table. There was no significant correlation (r = 0.07) between the duration of
diabetes and the degree of increase in pulse
wave velocity (as determined by Q). When
the individual values of pulse wave velocity
were correlated with the duration of illness
a higher, but not significant negative correlation was obtained (r =-0.41).
The coefficient of correlation between pulse
wave velocity and the systolic blood pressure
was 0.45 for the healthy subjects and 0.28
for the diabetic persons. The correlation coefficients for diastolic blood pressure and
pulse wave velocity in the healthy and diabetic groups were 0.46 and 0.40, respectively.
Discussion
The Pulse Wave Velocity in Healthy Subjects
The values of pulse wave velocity in
healthy subjects have been established by
several workers3' 8, 20 who utilized technics
similar to the one we used in this study.
Hallock in 1934 studied the pulse wave velocity in the radial artery and in the aorta
in 591 subjects.20 The values that he reported for the radial pulse wave velocity are
similar to those of this study, but the mean
values of all our groups are consistently
higher than those of Hallock's study. This
is most manifest in the younger age groups.
A test of significance of this discrepancy is
precluded because of the unavailability of
Circulation, Volume XXV, March 1962
5.50
31-40
16
9.86
41-50
12
11.10
51-60
6
132
9.52
83
4.50
the individual values for each one of the
subjects in Hallock's study.
A significant increase of the pulse wave
velocity with age was found in the studies of
Hallock20 and Bramwell.3 Bramwell reported, however, that the rate of increase in
pulse wave velocity was smaller in the higher
age groups. This is in contradiction with the
values reported by Hallock and it is probably
an invalid conclusion, since the size of the
sample of Bramwell's study in the older individuals is smaller than that of Hallock. Our
sample beyond the age of 50 is also very small
and for this reason we could not obtain statistical values beyond this age. However,
from the linear regression equation obtained
with the values on healthy subjects below 50
we were able to predict the values for subjects beyond 50 years of age (fig. 1). Our
predicted values are in complete agreement
with those observed by Hallock in subjects
of comparable ages. The slope of the regression equation reflects the increase in pulse
wave velocity that is expected when the
arteries become less distensible due to structural changes that take place with age.
The Pulse Wave Velocity in Diabetic Subjects
The pulse wave velocity of the diabetic
subjects was higher than that of the healthy
subjects in all but one age group. This is
reflected in the value of the ratio between
the observed pulse wave velocity in diabetic
subjects and that predicted for the corresponding decade of age (Q). The mean value
536
~ .~ I
c
so
*
o
',1
C5;
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
st:2 X
.AC
EHi
~
~~~
ce~
~
~ ~
i
C
i'
cCc!
1
L
t
,
i
VC~~
i
i
C-
~
~
C
CA)
00r,i
~
~
~
C)
~~~~~~~~CC
p
C)
~ ~
I
C
.°c
c
-
O
CCX
ii
4-1,
5C
Q)~~~~~~~C
IKi
:i -i ^--.3
~~~
Ic I,>
Z-
~
C),
-
o
of this ratio in all but onie of the age groupps
greater thani onie. The elevationi was
statistically significant at the level indicated
in each of the groups. It is possible that
the arteries of the diabetic subjects were
niore tortuous tlhani those of the healthy subje(cts, in which case the inc;reased values of
pulse wvave velocity- in this stuidy would take
on oreater significance, since tortuiositv slows
the actual transimiission rate( as well as the
caleulated rate because of underestimnation of
the vessel length. Since the diabetic sul)jects of this studyd had nio clinical evidlencc
of atheroselerosis, we may conclude that their
high values of pulse wave velocity were indicative of ani inicipienit process of diffuse atheroselerosis anid that the measur-ement of thcp)ulse wave velocity broughit this in evidence
earlier than the classical siglns an(d svmiiptonis. Difficulties in establishing definiite
proof of this hypothesis are obviouis. Serial
neasurements of the pulse wave velocity in
diabetic subjects are needed to establisl
whether there is a high correhlition between
ani inereased pulse wave velocitv and an
earlv developmnent of atheroselerosis. This
outline of investigation ha.s been advocated
by Lax, Feinberg, and Cohen2' anid it is inI
agreement ith Rushmer 's sggestions for
was
s~~Iz
H
;
~
~
00
cCI
WOOLAM ET AL.
z
investigating
cardiovascular
.fui,nctioni
in
health and in disease.22
Lax Feiinberg, and Collen21, 3 have showi
alterations in the conltour of the pressure
p)ulse Awave in individuals with clinical evilellee of atherosclerosis.
They observedl
simnilar changes in older persons, in hypertensive patienits, and in hiealtlhv suLbjects after
the admlinistrationi of niorepiniephrine. These
alterations in the contour of the pressure pulse
wave usually inelude diminution or disappearance of the dicrotic wave. They suggested that this change wAas due to rigiditv of
the vascular wall in the ease of aged indiv.iduals and atherosclerotic subjects and to an
inerease(d vascular tone in the hypertensive
patients alnd after administration of niorepineplhrine. In a subsequent publication, Lax
and Feinberg24 deimonstrated the saime chaniges
in a high percentage of younig diabetic subCirculation,
Volume
XXV, March, 1962
537
PULSE WAVE VELOCITY
Pulse
Wave
Velocity
(m/se.)
20
0
is
_-
16
0
14
0
o
12 _-
0
0
.0
0
0
0
lo0
X °o
Me0
*%
_-~~~~~
8
Voue Fo HeySb
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
61
ects (With al
2
0A l
Individual Values For Diabetics
°
4
---
Regression Line, Healthy Subjects of This Study
Regression Line, Healthy Subjects of Hallocks Study
I
I
I
I
I
I
E
10
20
30
40
50
60
70
80
AGE (Years)
Figure 1
Graphic presentation of the pulse icvae relocity ini diabetic subjects. The solid lilue
expresses the regression equation of the values for the healthy subjects of this study
The dotted line represents the regression equation of the values obtained in healthy
subjects in Hallock's study.20
jects who had no clinical evidence of atherosclerosis and in "normal" subjects with no
evidence of diabetes or atherosclerosis, but
with family history of diabetes. Although we
observed disappearance of the dicrotic wave
in 12 of our diabetic subjects, it was not
found as consistelntly as in the above-mentioned studies. In general, however, the amplitude of the dicrotic wave was smaller in
the diabetic subject (fig. 2).
Influence of the Duration of the Disease
The results of this study failed to indicate
positive correlation between the duration of the disease and the ilerease of the
pulse wave velocity. Indeed, we have observed high pulse wave velocities in very
young diabetic persons. It should be recognized, however, that the duration of the illness, as it was expressed in this study, may
any
Circulation, Volume XXV, March 1962
not be a precise measuremelnt of the time at
which the altered nmetabolic forces began to
exert their influence. Furthermore, information was not available to us as to the
degree of control that was obtained in these
subjects under the variety of therapeutic
regimens to which they had been subjected.
Students of diabetes have lolng debated the
exact role that the control of carbohydrate
netabolism in diabetes may have on the
eventual development of degenerative vaseular complications.25 This remains an unsettled
question because of the problems involved in
the evaluation of the intervening factors. A
clarification of this point awaits the collection and analysis of data of longitudinal
studies carried out in a sufficient number of
diabetic patients. If such longitudinal studies
are indeed warranted, it is the coliclusion of
WOOLAM ET AL.
538
References
C.7
-v-1-1I----L.~~~~~W
A) HEALTHY
PERSN
B) DIABETIC PATIENT
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Figure 2
]
is
i,
IF
o/ a healthy person (A)
radialZ J)ulse tr'aing
presutRCted wi ith tha:t of1' a dlbelica patient (B)
order to illustrate the significant decrease in
tOll f tle littutieu ware (d.) obaseured in
pJ)lit
the dib etic patient.
this .stiulv fat T)eriodic etaluations of the
pulse W .\ eveloeity sh.oull lielp the ('liliciall
to etcet earl>v atherosclerotic chlallnges. Corre
lation of tIle infornaitioll tillis gathered witi
reinets; )of other physiological anld bijo
'lIl(51X1
lllllical xvairiables may )erinlit a better tinleal(lilg of thie pallogellesis of cardiovascular
ts1i
colip)licatiOJ15 oifdacts
Summary
Tple arterial pullse
wave velocity N-as iieasured ill 87 healthy individuals and in 52
diabetic patielits who did niot have clinieal
evidene,e of athierosel erosis or hypertension.
A significant inerease in the piulse wave veloeilv wsas fornud in~ tile diabetic subjects. This
ilcleoase was not correlated with tlhe duilration
of their illness as juidged from elinical hist-orv. It is verv likely tlhat tle higoh values of
ave velocity were indicative of an in,
pulse w:tN
c;l)ienlt process of diffuse atherosclerosis an(l
thiaft tlhe measurement of the puLlse vave yelocity brougliht this into evidence earlier than
the classie signs a.nd syimp)toims. Serial ineasuremnents of the pulse wave velocity in diahietie subjects may permiit earlv detection of
atherosclerosis in these patients.
1. B\wELE, J. CA., DEOLWNING, A. C., AND HILL,
A. V.: The ef`fect of blood pressure on the
(InimIlia :artery. hleart 10:
exteinsibility of tile
289, 1.923.
2. H-ITrTO., W. P. rEMINGTON, J. N. A'
ANGO1)0
P.: The dcetenniaiation of the propagation.
Aelocity of the arterial pulse wave. Am. J.
P1l0siol. 144: 521, 1945.
B.l\MAELL, J.. C.: Trlie I Lmleiall lectiires o01 tile
ar.ter.ial pulse ill lIealtll am) (disea-As (1, 1r, 1 1 1).
L.aiieet 2: 239. .30o, .366, 1937.
4. SrltiEEL, J. M.: ltelpretatiOn of arterial elastlcitv fioti neiasuremlients of piulse naT¢2ve
xelo(it!i. Ami1. llerlIt .1. 14: 452, 1937.
aj. 1axR.\P, J. C., OGIuEN"J, E., XN_LD McPrEnRS0N,
C.: Iiiile{li(:itC variationlshi pulse wave
C.
lin iIl short iiniiforiii
veloeity eatived L) sad
J)arts of the lteri.l tr.ee. A-la. J. Physiol.
197: 432, 1959.
6., SAINDs, ..: Studies in pulse wave velocity. III.
Pulse waxe xelocity in. pathological co-nditionis.
Ami. J. Phlysiol. 71: 519, 1925.
7 TuR.-P, R. Hf., XN'l) 1{E£OR.RMANN, (G.: Pulse wave
velocity un.der varying conlitions in normal
and abnormal human cardiovascular system.
J. Clin. Tnlvest. 4: 430, 1927.
8. BAZETT,TT. C., AN-D DRli-YE,R NT. B.: Measurements
of pullse na:lve veleocity. Amn. J. Physiol. 63:
94, 1922.
9. HAxNES, F. W., Emms, L. B., AND WEISS, S.:
Puilse wave velocit) anld arterial elastieit.y
iniarter.ial h>pertension, arterioselerosis, and
related eoilditions. Aml. Hearit J. 11: 3185,
1936.
10. RxWSON, F. L.: The 5ise of the clectrokymograph
for determination of pulse wave velocity. J.
Lab. & Clin. M-ed. 37: 974, 1951.
11. SirONsoN-, E., AND NAKAGAWA, K.: Effect of
age on pulse wave velocity and aortic ejection
time in healthxy men and in men with coronary
heart disease. Circulation 22: 1216, 1960.
1 2. DITNCXN, G. C.: Diabetes Miellitus-Principles
and(1 Treatmjienit. Philadelpa, W. B. Saunders
& Co., 1951.
13. KRnAMERn D. W., AND PERILSTEIN, P. K.: Periphcral vascular complications in diabetes mellitus.
Dia.betes 7: 384, 193g8.
14. GEDDEs, L. A., ANlD hIo, H. E.: Graplhie recordiing of the pressuri.e pulse nave. J. Appl.
Pliysiol. 15: 959, 1960.
15. MIrLER, A., AND WHITE, P. D.: Crystal microphone for plilse waxve recording. Am. Heart J.
21: 504, 1941.
16. HICKZSoN, S., AND MCSWINEY, B. A.: Effect of
respiration oin pulse wave velocity. J. Pllysiol.
Soc. Proc. V, 1924.
1 7. BIvEERHOIL, 0.: Studles
oii
the x-elo(ity of
Circulation, Volume XXV, March 1962
539
PULSE WAVE VELOCITY
18.
19.
20.
21.
transmission of the pulse wave in normal
individuals. Acta med. scandinav. 67: 203, 1927.
KLIP, W.: Difficulties in the measurements of
pulse wave velocity. Anm. Heart. J. 56: 806,
1958.
SNEDECOR, G.: Statistical Methods. Ed. 4. Ames,
Iowa. The Iowa State College Press. 1955,
p. 75.
HALLOCK, P.: Arterial elasticity in man in
relation to age as evaluated by the pulse wave
velocity method. Areh. Tnt. Med. 54: 770, 1934.
LAX, H., FEINBERG, A. W., AND COHEN, L. B. M.:
Studies of the arterial pulse wave. I. The
normal pulse wave and its modification in the
22.
23.
24.
23.
presencee of human arter ioselerosis. J. Chron.
Dis. 3: 618, 1956.
RUSHMER, R. F., AND SMITH, 0. A., JR.: Cardiac
control. Physiol. Rev. 39: 41, 1959.
FEINBERG, A. W., ANTD LAX, H.: Studies of the
arterial pulse wave. Circulation 18: 1125, 1958.
LAX, H., AND FEINBERG, A. W.: Abnormalities of
the arterial pulse wave in young diabetic
subjects. Circulation 20: 1106, 1959.
DAYSOG, A., JR., DOBSON, H. L., AND BRENNAN,
J. C.: Reaial glonmerular and vascular lesionis
ill prediabetes and ill diabetes miiellitus: A
study basedl on renal biopsies. Anan. Tlat. Med.
54: 672, 1961.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
.cV.
I have three personal ideals. One, to do the day's work well and not to bother about
tomorrow . . . The second ideal has been to act the Golden Rule, as far as in nme lay,
toward my professional brethren and toward the patients comlllitted to my care. And
the third has been to cultivate such a nmeasure of equanimity as would enable me to
bear success with humility, the affection of my friends without pride, and to be ready
when the day of sorrow and grief came to meet it with the courage befitting a man.SIR WILLIAM OSLER. Aphorisms From His Bedside Teachings and Writings. Edited by
William Bennett Bean, M.D. New York, Henry Schuman, Inc., 1950, p. 84.
Circulation, Volume XXV, March
1962
The Pulse Wave Velocity as an Early Indicator of Atherosclerosis in Diabetic
Subjects
G. L. WOOLAM, P. L. SCHNUR, C. VALLBONA and H. E. HOFF
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Circulation. 1962;25:533-539
doi: 10.1161/01.CIR.25.3.533
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1962 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/content/25/3/533
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles
originally published in Circulation can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column of
the Web page under Services. Further information about this process is available in the Permissions
and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/