pulsatile flow and pressure in human systemic arteries

363

Pulsatile Flow and Pressure in HumanSystemic Arteries

Studies in Man and in a Multibranched Model of theHuman Systemic Arterial Tree

MICHAEL F. O'ROURKE AND ALBERT P. AVOLIO

SUMMARY This study seeks to explain mechanisms responsible for the contour of pressure and flowwaves and the pattern of vascular impedance in human systemic arteries. Pulsatile pressure and flowwere recorded from the ascending aorta of seven patients undergoing open heart surgery and from theascending aorta and other arteries of 17 patients at diagnostic catheterization. Ascending aorticpressure/flow relationships in the seven surgical patients were expressed as input impedance to thesystemic circulation. Pressure and flow wave contour and impedance results were interpreted with theaid of a multibranched model of the systemic arterial tree, whose parameters could be manipulated tosimulate different physiological and pathological conditions. Our data and data previously publishedon pressure and flow waves and their relationship in human subjects could be explained on the basisof two reflecting sites in the systemic circulation—one representing the resultant of all arterialterminations in the upper part of the body, and the other, some 1.5 times further away, the resultant ofall arterial terminations in the lower body. The concept of the arterial system as an asymmetric T tubewith two discrete ends has been advanced previously to explain the main features of pressure and flowwaves and their relationship in different experimental animals. This concept appears equally applicableto human subjects. Circ Res 46: 363-372, 1980

PULSE contour in any artery is determined by thetime course of ventricular ejection and by vascularproperties. Separation of cardiac from vascular ef-fects has been aided by determination of vascularimpedance (McDonald and Taylor, 1959; Mc-Donald, 1974), which permits characterization ofvascular properties downstream from the point ofmeasurement. Impedance determinations in var-ious arteries of different experimental animals haveshown great consistency and have been interpretedto show evidence of strong wave reflection in pe-ripheral vascular beds (McDonald, 1974; O'Rourkeand Taylor, 1966, 1967; Avolio et al., 1976). Suchstudies have suggested that the main features ofpressure and flow waves in major arteries are expli-cable on the basis of wave reflection between twofunctionally discrete reflecting sites in the upperand lower parts of the body (O'Rourke, 1967;McDonald, 1974; Avolio et al, 1976).

Studies of vascular impedance and of pulse wavecontour in man generally have been interpreted interms of peripheral wave reflection (Gabe et al,1964; Mills et al, 1970; Nichols et al, 1977; Pepine

From the School of Medicine, University of New South Wales, Sydney,Australia.

Michael F. O'Rourke, M.D., is an Associate Professor of Medicine atthe University of South Wales.

Address for reprints: Michael F. O'Rourke, M.D., CardiovascularDivision, Peter Bent Brigham Hospital, 721 Huntington Avenue, BostonMassachusetts 02115.

Original manuscript received February 27, 1976; accepted for publi-cation October 31, 1979.

et al, 1978), but there is no agreement as to theintensity of wave reflection or the site or sites ofreflection. Recent studies (Murgo et al, 1978a,1978b; Westerhof et al, 1979) have been interpretedin terms of one apparent reflecting site in the vicin-ity of the distal aorta, with variable reflection, rang-ing from strong to minimal, in different patientsand under different conditions.

We undertook this investigation to seek a com-prehensive explanation for the main features ofpressure and flow waves in major human systemicarteries. We wished to explain vascular impedanceand pulse contour in humans on the same basis aspreviously proposed for experimental animals.

In an attempt to overcome the difficulties ofobtaining data for humans under different condi-tions, we developed a digital computer model of thehuman systemic arterial tree; parameters of thismodel could be manipulated to simulate differentphysiological and pathological conditions. Thus, thepresent study comprises collection and analysis ofexperimental data for humans, interpretation ofdata for humans already published by others, andsimulation in the human arterial model.

Methods

Patients

Pulsatile arterial pressure and flow were mea-sured in the ascending aorta of seven patients at

by guest on April 8, 2018

http://circres.ahajournals.org/D

ownloaded from

http://circres.ahajournals.org/

364 CIRCULATION RESEARCH VOL. 46, No. 3, MARCH 1980

open heart surgery, immediately before startingcardiopulmonary bypass, and in systemic arteriesof 17 patients during diagnostic catheterization. Theexperimental procedures were considered and en-dorsed by the Cardiovascular Unit at St. Vincent'sHospital, and informed consent was obtained fromall participating patients. The subjects' ages were23-60 (average 47) years. None had clinical evi-dence of aortic or peripheral arterial disease. In thegroup of seven subjects in whom ascending aorticimpedance was determined, five had coronary ar-tery disease, one had mitral stenosis, and one anatrial septal defect; ages were 40-52 (average 47),and mean blood pressure was 62-118 (average 100)mm Hg.

For the seven subjects studied at surgery, flowwas measured with a cuff-type electromagnetictransducer applied to the exposed ascending aorta.Pressure was measured immediately distal to thetransducer through a 21-gauge needle inserted di-rectly through the aortic wall and with its bevelfacing the direction of flow, as in previous experi-ments on animals (O'Rourke and Taylor, 1967).The needle was connected to a Statham pressuretransducer by a short rigid catheter. The flow trans-ducer was connected to an E.M.I. (E.M.I. Co.)flowmeter instrument. Data were recorded on mag-netic tape.

For the 17 subjects studied at diagnostic cathe-terization, data were obtained with an S.E. (S.E.Laboratories) catheter tip electromagnetic flowtransducer incorporating a lumen for measuringarterial pressure. The catheter was inserted in thebrachial artery and advanced, when possible, to theascending aorta. Flow was measured by an S.E.flowmeter. Pressure was measured with a Stathamstrain gauge manometer. Data were recorded onmagnetic tape.

We calibrated flow transducers in vitro by pass-ing physiological saline at known velocity over (forthe catheter) or through (for the cuff) the trans-ducer. Frequency response of the E.M.I, flowmetershowed amplitude flat to 30 Hz and a phase delayof 3.0°/Hz (Goodman, 1966). The S.E. instrumenthad a flat amplitude response to 30 Hz and a phasedelay of 2.0°/Hz (manufacturer's specifications).The manometer was calibrated against a column ofmercury and frequency response determined by thepressure-step or "pop" test; natural frequency ofthe system used in the operative studies was in therange of 50-70 Hz, and the damping coefficientaveraged 0.18. It was difficult to maintain an ade-quate frequency response for the manometer sys-tem through the S.E. flow catheter lumen; becauseof this and apparent motion artifact on ascendingaortic flow tracings, impedance was not determinedfrom records obtained at diagnostic catheterization.Sampling rate of waves recorded at surgery was100/sec.

The relationship between pulsatile pressure andflow in the ascending aorta was expressed as vas-

cular impedance by relating corresponding harmon-ics of simultaneously recorded pressure and flowwaves, as described elsewhere (O'Rourke and Tay-lor, 1966,1967). Vascular impedance is analogous toelectrical impedance and takes the form of a graphof modulus (amplitude of pressure -s- amplitude offlow) and phase (delay between pressure and flow)plotted against frequency. Allowance was made forfrequency characteristics of flowmeter andmanometer systems in calculation of vascularimpedance. Impedance was plotted only for the firsteight harmonics corresponding to a frequency rangeof 0-12 Hz. Impedance values were rejected if ob-tained from either pressure or flow harmonics whichwere considered to be within the noise level ofrecording instruments (<0.5 mm Hg for pressure,<1 cm/sec for flow).

The Arterial ModelThe multibranched model of the human arterial

system was developed on a digital computer (Avo-lio, 1976; Avolio and O'Rourke, 1976) (Fig. 1). Theconfigurations and dimensions of major arteries forthis model were obtained from anatomical texts.This model bore a close relationship to the analogcomputer model of Noordergraaf et al. (1963) butwas more realistic in its representation of the vas-cular bed, particularly that in the upper part of thebody. Values of wall thickness, diameter, and elas-ticity were identical to those used by Noordergraafet al. (1963) and by Westerhof et al. (1969). Prop-erties of the model with respect to pressure wavetransmission and pressure/flow relationship weredetermined through application of Womersley'stheory of fluid flow in elastic tubes (Womersley,1957) and uniform electrical transmission line the-ory. Computational techniques were virtually iden-tical to those used by Taylor (1966), the maindifference being that this model used anatomicallyrealistic values of arterial length, whereas Taylor'smodel generated arterial segments of randomlength within the computer.

When the general algorithm of this digital com-puter model was used with the same arterial dimen-sions and properties as used by Noordergraaf(1963), identical results were obtained as with Noor-dergraaf's analog technique. The general algorithmalso has been used with vessel dimensions obtainedfrom angiography to simulate the arterial system ofvarious experimental animals. There has been verygood agreement between these studies and the ex-perimentally recorded pressure and flow waves andderived impedance patterns (Avolio et al., 1976;Avolio 1976).

Results

Human SubjectsContour of pressure and flow waves recorded in

the ascending aorta, descending thoracic aorta, andbrachiocephalic artery were similar to those re-



ownloaded from


HUMAN PULSATILE PRESSURE AND FLOW/O'Rourke and Avolio 365

10cm

FIGURE 1 The multibranched model of the systemicarterial tree and (at right) its equivalent asymmetric Ttube model, both drawn to scale. The open circle in thebranched model represents the aortic valve.

mental animals also show the same distinctive dif-ference in flow patterns as in man between vesselssupplying the upper and lower parts of the body(O'Rourke, 1967; Avolio et al., 1976).

Input impedance to the systemic circulation wasdetermined from data obtained at cardiac surgery(Fig. 3). Modulus and phase of impedance in thesesubjects were similar to those found by others inman (Gabe et al., 1964; Patel et al., 1965; Mills etal, 1970; Nichols et al, 1977; Pepine et al, 1978;Murgo et al, 1978a; Westerhof et al, 1979). Char-acteristic impedance (the value of impedance mod-ulus about which fluctuations occur) was approxi-mately 600 dyne sec cm"3 or with aortic diameter

AA

100

BCA

ported by others for man. There was little differencein pressure pulse contour between the ascendingaorta and the proximal brachiocephalic artery andproximal descending aorta. In contrast, the brachi-ocephalic flow contour was quite different from thatrecorded in the ascending and descending aorta(Fig. 2). In the brachiocephalic artery, the peak offlow occurs early in systole, and this peak usually isfollowed by a plateau or notch on the systolicdownstroke of the wave some 50-150 msec beforeaortic valve closure. Similar patterns are seen inthe subclavian (Mills et al, 1970) and brachialarteries (Gault et al, 1966). Different flow patternsin arteries supplying the upper and lower parts ofthe human body have been noted previously andstressed by Mills et al. (1970).

Patterns of pressure and flow waves in the prox-imal aorta and brachiocephalic artery of humansubjects were similar to those seen in experimentalanimals, although in general diastolic pressure andflow fluctuations were less marked in man. Experi-

100

DTA

0.5 sec

FIGURE 2 Typical flow waves recorded in the humanascending aorta (top) with a cuff-type electromagneticflowmeter and in the brachiocephalic artery (BCA) anddescending thoracic aorta (DTA), recorded with a cath-eter-tip electromagnetic flow transducer. Brachioce-phalic and descending thoracic aortic flow recorded insame subject under stable conditions within a 10-minuteperiod.



ownloaded from



4000

$ 3200c>»•o

2400OO

y 1600

0.

800

1189311843104009 0 0 37 3 0 36 4 7 259647 8 6 4 ( M o d e l )

4 6 8

FREQUENCY (Hz)10

<

60°

0°

FIGURE 3 Input impedance of the systemic circulationin seven subjects and of the multibranched model. Val-ues of peripheral resistance are given at upper left;values of mean arterial pressure (from above down) were(M)99, (V)117, (0)114, (9)62, (x)88, (T)118, (O)103mm Hg. Subjects, (V), ( • ) , (x), (•), (O), had coronaryartery disease and underwent coronary artery bypass.

28 mm, approximately 97 dyne sec cm"'. This is inthe range of 53-202 dyne sec cm"5 found by otherworkers. Humans show the same general patternsof ascending aortic impedance as experimental an-imals. However, characteristic impedance deter-mined from linear velocity and expressed as dynesec cm"3 appears to be greater in man and to showgreater variability from subject to subject; in addi-tion, fluctuations of impedance modulus appear tobe greater than in dogs and other animals.

The Arterial ModelThere was good agreement between ascending

aortic impedance in the model and results obtainedfrom subjects (Fig. 3). There was a minimal valueof modulus at 3.0 Hz and a maximal value at 7.8Hz. Impedance at the origin of the descending aortawas similar to that in the ascending aorta; at theorigin of the brachiocephalic artery, the minimumand maximum of the impedance modulus occurredat 4.8 and 9.0 Hz, respectively (Fig. 4A).

The distance to peripheral reflecting sites in theupper and lower parts of the body was calculatedfrom brachiocephalic and descending thoracicimpedance, respectively, assuming the minimalvalue of modulus to be reached at a frequencycorresponding to one-quarter wave length from afunctionally discrete reflecting site (McDonald andTaylor, 1959; McDonald, 1974). Wave velocity [cal-culated from the Moens Korteweg formula (Mc-Donald, 1974)] was 514 and 478 cm/sec, respec-

1500

1000

d.s.cm"3

500

I

tAADTABCA

yVCK

0

rads

-1.0

4 8

V

H i 12

1500

1000

d.s.cm

500

Iii\

Multibranched

Asymmetric T

Model

Model

FIGURE 4 A: Vascular impedance in the multi-branched model for ascending aorta, (AA), DTA, andBCA. B: Ascending aortic impedance in the multi-branched model and at the input of the asymmetricuniform T tube with the following characteristics—shortlimb, 26 cm; long limb, 40 cm; wave velocity in shortlimb, 514 cm/sec, and long limb, 478 cm/sec; attenua-tion, 50% per wavelength; reflection coefficient, 0.7 ateach end.



ownloaded from


HUMAN PULSATILE PRESSURE AND FLOW!/O'Rourke and Avolio 367

tively, in the brachiocephalic artery and descendingaorta. These values were used to calculate the dis-tance to reflecting sites as 26 and 40 cm, respectively[514 -H (4 X 4.8) and 478 -s- (4 X 3.0)], from theorigin of the brachiocephalic artery and of the de-scending thoracic aorta.

Thus, these impedance studies suggested thatthe complex arterial network of the model be-haves—to a first approximation—like an asymmet-rical T tube with a short limb 26 cm long, whichrepresents all arteries supplying the upper part ofthe body, and a long limb 40 cm long, which repre-sents the aorta and the other arteries supplying thelower part of the body (Fig. 1). The input impedanceto this simple asymmetric T model was similar tothat of the multibranched anatomical model (Fig.4B). Data from Mills et al. (1970) support thisasymmetric T model with short-to-long limb ratioof approximately 1:1.5. They calculated the dis-tance-to-upper body reflecting site to be 29 cm fromthe brachiocephalic root and the distance to lowerbody reflecting site to be 31 cm from the middes-cending thoracic aorta at T7 level. Allowing 10 cmaround the aorta arch and down from T4 to T7level, the lower body reflecting site would have been41 cm from the origin of the descending aorta. (Millset al., 1970 had calculated the distance to the pe-ripheral reflecting site as 44 cm from the ascendingaorta.) The asymmetric T tube suggested by dataof Mills et al., thus, had a short limb of 29 cm anda long limb of about 41 cm.

Figure 5 shows waves synthesized at the origin ofthe brachiocephalic artery and descending thoracicaorta when the model was activated by an ascend-ing aortic flow wave recorded in one subject. Thesebrachiocephalic and descending aortic waves werealmost identical to those recorded by us and others(Gabe et al., 1969; Mills et al, 1970) in man. Thebrachiocephalic flow wave showed the sameshortening of systolic flow as is noted in man. Themost likely explanation for the difference betweenbrachiocephalic and descending aortic flow patternsis the different distance to arterial terminations—and so to functional reflecting sites—in the upperand lower parts of the body. To test this hypothesis,all arterial segments in the upper part of the bodywere doubled in length and, in this model, activatedby the same ascending aortic flow wave. This re-sulted in a loss of the characteristic brachiocephalicflow contour so that it resembled that in the prox-imal descending aorta (Fig. 5).

The effect of change in pulse wave velocity wassimulated in the model by altering elastic constantsof the arterial segments while maintaining all otherparameters constant. Such changes had little effecton the descending thoracic aortic flow patterns butled to considerable alteration in the brachiocephalicflow wave (Fig. 6). An increase in pulse wave veloc-ity led to further shortening of the initial brachioce-phalic flow peak, whereas a decrease in pulse wavevelocity resulted in lengthening of the period of

100

o

100i

AA

cm/sec

VBCA

100-1

cm/secDTA

0.5 sec

FIGURE 5 Flow waves generated in the brachioce-phalic artery and upper descending thoracic aorta ofthe model with ascending aortic flow wave (top) as input.Dotted lines indicate altered flow patterns when lengthsof all arterial segments in the upper part of the bodywere doubled. For explanation see text.

forward systolic flow and appearance of a promi-nent secondary flow wave in diastole. Changes sim-ilar to these were brought about by increasing anddecreasing the duration of ventricular ejection withpulse wave velocity held constant (Fig. 7). Thesealterations in brachiocephalic flow contour withreduction in pulse wave velocity or ejection periodare virtually identical to those noted in man whenwave velocity was reduced and ejection durationshortened during the Valsalva maneuver. (Mills etal., 1970).

Pressure waves synthesized in the model for thesame ascending aortic flow wave were very similarto those recorded in corresponding arteries in man.The changes brought about by altered pulse wavevelocity and ejection period also were similar tothose reported in man. Under normal circum-stances, a diastolic wave was not apparent in theproximal aortic or brachiocephalic pressure pulse.However, when pulse wave velocity was reduced, adiastolic wave became apparent in these vessels. Aswave velocity then was increased progressively, thediastolic wave became less obvious, whereas ampli-tude of the pressure wave increased in late systole.These changes in wave contour were attributableto early return of reflected waves before aortic valveclosure under normal circumstances and late returnafter aortic valve closure when wave velocity was



ownloaded from


368 CIRCULATION RESEARCH

BRACHIOCEPHALIC ARTERY

VOL. 46, No. 3, MARCH 1980

50

100DESCENDING THORACIC AORTA

FIGURE 6 Changes in brachiocephalic artery (above) and descending thoracic aortic (below) flow waves with increasein pulse wave velocity by -J2 (left) and decrease in wave velocity by V2 (right). Control waves in center.

x v z

100-

oJ

140

mm Hg

90

X Y Z

110

r\

X Y Z

/

1\\\

f

II

AA

BCAP

BCAF

Y Z

0.5 sec

FIGURE 7 Effects of change in ejection period on pres-sure and flow waves in the brachiocephalic artery of themultibranched model with systolic ejection 355 msec(left) and 222 msec (right). Ascending aortic flow (top)was input and brachiocephalic pressure (center) andflow (bottom) were generated in the model. The line XXcorresponds to the peak of ventricular ejection; the lineYY corresponds to the calculated return of peak reflec-tion from the upper part of the body (26 X 2/514 = 0.10sec after peak flow); and the line ZZ to the calculatedreturn of peak reflection from the lower body (40 X 2/478= 0.17 sec after peak flow).

reduced. This explanation also holds for the differ-ence in pressure wave contour with different periodsof ejection. The ejection period was altered in themodel (Fig. 7), with all arterial properties (andtherefore vascular impedance and wave velocity)held constant. Under control conditions, the broad-ened late systolic peak is attributable to wave re-flection from upper and lower body sites precedingaortic valve closure, whereas when ejection periodis shortened, the diastolic wave becomes apparentand is attributable to the lower body reflected wavereturning after aortic valve closure. The calculatedtime of reappearance of reflected waves from upperand lower body sites is indicated in Figure 7; thesecorrespond well with the fluctuations in brachioce-phalic flow and appear to account for changes inthe pressure wave.

DiscussionThe heart occupies an eccentric position in the

mammalian body and so is closer to arterial termi-nations in the upper part of the body than to arterialterminations in the lower body. This arrangementin dogs, sheep, and rabbits has been shown to resultin two functionally discrete reflecting sites beingpresented to the heart—one relatively close andrepresenting the resultant of all arterial termina-tions in the upper part of the body, and the otherfurther away and representing the resultant of allarterial terminations in the lower part (O'Rourke,1965; O'Rourke and Taylor, 1967; O'Rourke, 1967;McDonald, 1974, Avolio et al., 1976). The mam-malian arterial system has been likened to an asym-metric T whose shorter limb represents the upperbody vasculature and the longer limb the aorta andlower body vasculature (McDonald, 1974). In dogs,the ratio of short-to-long limbs of the T is approxi-mately 1:2 (O'Rourke and Taylor, 1967; O'Rourke,1967), whereas in rabbits and sheep, it is approxi-mately 1:3 (O'Rourke, 1965; Avolio, 1976; Avolio et



ownloaded from



al., 1976). The presence of two functionally discretereflecting sites in the canine systemic circulationwas originally suggested by McDonald (1974) as apossible explanation of pressure wave transmissionproperties. This concept was further developed byTaylor and O'Rourke, whose studies of vascularimpedance showed evidence of a single functionallydiscrete reflecting site in most systemic arteries butnot in the ascending aorta (Taylor, 1966; O'Rourkeand Taylor, 1966, 1967; O'Rourke, 1965, 1967). As-cending aortic impedance patterns were explainedon the basis of two reflecting sites in the upper andlower parts of the body, presented by brachioce-phalic and subclavian arterial beds on the one handand the descending aorta and lower body vascula-ture on the other. This explanation was supportedby determinations of vascular impedance in theseindividual arteries and in the ascending aorta whenthey were occluded.

The concept of two separate reflecting sites inthe systemic arterial tree explained not only as-cending aortic impedance patterns, but also thereciprocal diastolic fluctuation of flow in the bra-chiocephalic artery and descending aorta. It alsoexplained the reciprocal diastolic pressure fluctua-tion in the upper and lower parts of the body, abouta "node" in the proximal descending aorta. Thisconcept modified the thesis presented by Frank(1905) and by Hamilton and Dow (1939), who con-sidered that wave reflection occurred between asingle peripheral vascular reflecting site and theaortic valve; this, however, could not explain thereciprocal flow oscillations which were subse-quently demonstrated in upper and lower bodyvessels or the proximal location of the thoracicpressure "node."

Since the human arterial system has the samegeneral anatomical arrangement as other mammals,one would expect that the same general principlesfor wave reflection as found in animals also wouldapply in man. There is, however, no general agree-ment that this is the case. Pressure and flow dataand impedance determinations in man have beeninterpreted to show definite wave reflection (Millset al., 1970; Nichols et al., 1977; Pepine et al., 1978),minimal or no wave reflection (Murgo et al., 1978a),or that wave reflection disappears under some con-ditions including the Valsalva maneuver (Murgo etal., 1978b).

We believe that the general principles do applyand indeed provide the only comprehensive expla-nation for our own and other published data.

One of the sources of confusion in man has beenthe variability of published data—not only betweendifferent series but also in the same report. Some ofthe variability is attributable to technical problems,particularly difficulties in measuring flow in theproximal aorta with the earlier electromagneticcatheters, which were prone to whip and artifact.Artifact appeared to be present in all our records

taken from the ascending aorta with a catheter-tipelectromagnetic flowmeter and was the reason forthese data not being used for impedance determi-nation. However, most of the variability in pub-lished results appears to be real and is attributableto differences in aortic compliance between differ-ent subjects. Variability of aortic compliance inman is well known and indeed has thwarted at-tempts to measure stroke volume from the pressurepulse (Remington et al., 1948).

Vascular Impedance

In the human ascending aorta, there is usuallymore fluctuation of impedance modulus with fre-quency than in experimental animals. Most re-ported studies have shown a maximal value ofimpedance modulus at 6-10 Hz following an initialminimum at 3-5 Hz. This finding has been inter-preted to indicate wave reflection from a singlefunctionally discrete peripheral reflecting site. Thefluctuations of impedance modulus with frequencyhave been most marked in patients with hyperten-sion and known arterial disease (Nichols et al.,1977) and in patients with aortic pressure wavessuggestive of arterial degeneration (Murgo et al.,1978a; O'Rourke et al, 1968; Puls and Heizer, 1967;Freis et al., 1966). These fluctuations have beenleast marked in normal subjects (Nichols et al.,1977), in patients with a normal aortic diastolicpressure wave (Murgo et al., 1978), and during theValsalva maneuver (Mills et al, 1970; Murgo et al,1978b; Westerhof et al, 1979).

The model suggests an explanation for thesefindings. As a consequence of relatively lonf armsand a relatively short trunk, there is in man arelatively greater distance to reflecting sites in theupper part of the body than in experimental ani-mals. The ratio of short-to-long limbs in the asym-metrical T model proposed for man is approxi-mately 1:1.5 as against 1:2 for dogs and approxi-mately 1:3 for sheep and rabbits. Thus, in man,there is less interaction of reflected waves fromboth sites, greater fluctuations of impedance mod-ulus in the ascending aorta, and a maximum ofmodulus closer to twice the frequency of the firstminimum than in experimental animals.

The greater fluctuations in impedance modulusin patients with known or apparent arterial disease(Nichols et al, 1977; Murgo et al, 1978a) may bedue to relatively greater wave velocity in the aortaand lower limb arteries than in the upper limbvessels. Changes in wave velocity with age appearto be more marked in the lower body arteries(O'Rourke et al, 1968; O'Rourke, 1970a). Such dif-ferential changes in wave velocity would result inreflected waves returning almost simultaneouslyfrom the upper and lower parts of the body andmimicking a single peripheral reflecting site. Dis-appearance of impedance modulus fluctuations dur-ing the Valsalva maneuver (Mills et al, 1970; Murgo



ownloaded from



et al., 1978b) is attributable to lower wave velocityin the thoracic and abdominal aorta with laterreturn of reflected waves from the lower body(Kroeker and Wood, 1956).

The concept of the systemic arterial system asan asymmetric T thus appears to provide an expla-nation for our own and all published results ofascending aortic impedance in man under all con-ditions. The explanation of different patterns onthe basis of difference in interaction of reflectedwaves is the same as previously offered for differentexperimental animals and seems more logical thanthe alternative explanation of difference in intensityof peripheral reflection.

Flow PatternsThe most convincing evidence of differential re-

flection from upper and lower body sites comesfrom consideration of the flow patterns in majorarteries supplying the upper and lower parts of thebody. It is acknowledged that these differ markedlyin man (Mills et al., 1970) as well as in experimentalanimals. Flow patterns at a vascular junction aremore affected by wave reflection than pressure pat-terns, since reflected waves subtract from flow inone direction while adding to flow in the other,whereas reflection adds equally to pressure in allbranches of the junction (Mills et al., 1970). Whenthe junction is between two vessels of unequaldiameter (as that of the brachiocephalic artery andaortic arch), reflected waves are more obvious inthe narrow vessel (Mills et al., 1970).

Mills et al. described three types of flow patternsin the brachiocephalic and subclavian arteries andreferred to these as type I, type II, and type IIIpatterns. These three waves correspond to the three

brachiocephalic flow waves generated in the modelfor the same ascending aortic wave when pulsewave velocity or ejection duration was altered (Figs.6 and 7). The three wave patterns from Mills' paperare redrawn in Figure 8. The timing of reflectedwaves from upper and lower body is indicated bytriple arrows and the incisura caused by aortic valveclosure by a single arrow. The reflected wave fromthe upper part of the body decelerates flow in thebrachiocephalic artery and tends to shorten theperiod of forward flow; the reflected wave from thelower part of the body accelerates brachiocephalicflow in late systole or early diastole. The type Iwave is explained to result from early return ofreflected waves from upper and lower body whichprecedes aortic valve closure, type II to result fromthe lower body reflected wave occurring during theincisura, and type III to result from this wavefollowing the incisura. Mills et al. offered this expla-nation for the type III but did not go further toexplain type I and type II which were the mostcommonly observed patterns. This appears strange,since they had shown that brachiocephalic imped-ance was similar irrespective of brachiocephalicflow contour. Mills and colleagues showed the tran-sition of type I through type II to type III brachi-ocephalic flow patterns during the Valsalva maneu-ver, with return to type I afterward. Gault et al.(1966) had shown similar changes in brachial arteryflow patterns when ejection duration was shortenedafter a premature ventricular extrasystole. Allchanges are easily explained on the basis of de-creased wave velocity in the aorta and decreasedejection period during the Valsalva maneuver, withdifferent timing of upper and lower body reflectedwaves in relation to aortic valve closure.

m

m m

FIGURE 8 Types I, 11, and HI brachiocephalic flow waves redrawn from the paper by Mills et al. with suggestedmechanism outlined below. The single arrow represents aortic valve closure and the triple arrow represents the timingof wave reflection from the upper part of the body and later wave reflection from the lower part of the body. The dottedline indicates the expected wave pattern in the absence of an incisura caused by backflow through the aortic valve.



ownloaded from



Pressure PatternsIn experimental animals, pressure wave contour

in different arteries is readily explained on the basisof wave reflection between two functionally discretereflecting sites in the upper and lower parts of thebody. This explanation has not been as obvious inman, because man often does not show the sameevidence of wave reflection—a prominent diastolicwave (Porje, 1946; Salans et al., 1951; Kroeker andWood, 1956; O'Rourke et al., 1968) as is regularlyseen in dogs, rabbits, sheep, and other experimentalanimals (Hamilton and Dow, 1939; O'Rourke, 1967;McDonald, 1974), except the guinea pig (Avolio etal., 1976). Children regularly show the same pat-terns as these animals, but as age advances, thediastolic wave tends to disappear (O'Rourke et al.,1968). In the elderly, in patients with hypertensionand arterioscelerosis, the pressure wave is similarin all arteries with a late systolic peak, and withpressure falling exponentially from this peak, inter-rupted only by the incisura which represents aorticvalve closure (Salans et al., 1951; Freis et al., 1966;O'Rourke et al., 1968; O'Rourke, 1970b).

These changes with age and disease are attrib-utable to high arterial wave velocity with reflectedwaves from upper and lower body fusing with eachother and with the initial wave generated by ven-tricular ejection (O'Rourke, 1970b). Although re-flected waves are not apparent as such in pressurewaves recorded in human arteries under most con-ditions, they do become prominent when wave ve-locity falls or duration of ejection shortens (Kroekerand Wood, 1956; Greenfield et al., 1967; Mills et al.,1970). In other conditions in man, such as shockwhen hypotension and a short ejection period arecombined, the diastolic waves become extremelyprominent (O'Rourke, 1970a). Indeed, such exag-gerated diastolic waves in patients with cardiogenicshock can cause difficulties in timing of arterialcounterpulsation (personal observations).

The arterial pressure wave in man and thechanges seen with ageing, with disease, and withalteration of blood pressure and ejection periodthus are readily explained on the basis of wavereflection between two functionally discrete sites inupper and lower parts of the body.

We present here an explanation of arterial pulsecontour in man on the basis of two functionallydiscrete reflecting sites in the upper and lower partsof the body. This elaborates the original concept ofwave reflection propounded by Otto Frank over 70years ago and subsequently refined by Hamilton,Dow, Remington, Wood, McDonald, and others.We do not claim that such a concept explains allfeatures of the pulse—effects of subsidary reflectionsites, including the aortic valve, obviously play somerole as well. Further, other factors besides reflectionalso modify the arterial pulse, including nonuniformelasticity and attenuation and dispersion of thewave in travel. We simply draw attention to amechanism which appears to be responsible for the

major features of pulse wave contour hoping that,if this can be confirmed, further attention can begiven to the other factors which are responsible foralterations of arterial pulse contour in cardiac andvascular disease.

ReferencesAvolio A (1976) Hemodynamic studies and modelling of the

mammalian arterial system. Ph.D. thesis, University of NewSouth Wales

Avolio A, O'Rourke MF (1976) Application of a multibrancheddigital computer model to hemodynamic studies of the mam-malian arterial system (abstr). Excerpta Med Int Congr Ser405: 2

Avolio A, O'Rourke MF, Mang K, Bason PT, Gow BS (1976) Acomparative study of pulsatile arterial hemodynamics in rab-bits and guinea pigs. Am J Physiol 230: 868-875

Frank O (1905) Der puls in den arterien. Z Biol 46: 441-553Freis ED, Heath WC, Luchsinger PC, Snell RE (1966) Changes

in the carotid pulse which occur with age and in hypertension.Am Heart J 71: 757-765

Gabe IT, Karnell J, Porje IG, Rudewald B (1964) The measure-ment of input impedance and apparent phase velocity in thehuman aorta. Acta Physiol Scand 61: 73-84

Gabe IT, Gault JH, Ross J Jr, Mason DT, Mills CJ, ShillingfordJP, Braunwald E (1969) Measurement of instantaneous bloodflow velocity and pressure in conscious man with a cathetertip velocity probe. Circulation 40: 603-614

Gault JH, Ross J Jr, Mason DT (1966) Patterns of brachialarterial flow in conscious human subjects with and withoutcardiac dysfunction. Circulation 34: 833-848

Goodman AH (1966) A transistorised square-wave electromag-netic flowmeter. J Med Biol Eng 7: 115-132

Greenfield JC, Cox RL, Hernandez RR, Thomas C, Schoon-maker FW (1967) Pressure-flow studies in man during theValsalva maneuver with observations on the mechanical prop-erties of the ascending aorta. Circulation 35: 653-661

Hamilton WF, Dow P (1939) An experimental study of thestanding waves in the pulse propagated through aorta. Am JPhysiol 125: 48-59

Kroeker EJ, Wood EH (1956) Beat-to-beat alterations in rela-tionship of simultaneously recorded central and peripheralarterial pressure pulses during Valsalva maneuver and pro-longed expiration in man. J Appl Physiol 8: 483-494

McDonald DA (1974) Blood Flow in Arteries, ed 2. London,Arnold

McDonald DA, Taylor MG (1959) The hydrodynamics of thearterial circulation. Prog Biophys Biophys Chem 9: 107-173

Mills CJ, Gabe IT, Gault JH, Mason DT, Ross J Jr, BraunwaldE, Shillingford JP (1970) Pressure-flow relationships and vas-cular impedance in man. Cardiovasc Res 4: 405-417

Murgo JP, Giolma JP, Alter BR, Westerhof N (1978a) Ascendingaorta pressure waveshape—relationship to aortic input imped-ance (abstr). Circulation 57: 159

Murgo JP, Giolma JP, Westerhof N (1978b) Manipulation of theaortic pressure and flow wave reflections with the Valsalvamaneuver (abstr). Circulation 57: 76

Nichols WW, Conti CR, Walker WE, Milnor WR (1977) Inputimpedance of the systemic circulation in man. Circ Res 40:451-458

Noordergraaf A, Verdouw PD, Boom HBK (1963) Analog com-puter in a circulation model. Prog Cardiovasc Dis 5: 419-439

O'Rourke MF (1965) Pressure and flow in arteries. M.D. thesis.University of Sydney

O'Rourke MF (1967) Pressure and flow waves in systemic arter-ies and the anatomical design of the arterial system. J ApplPhysiol 23: 139-149

O'Rourke MF (1970a) Influence of ventricular ejection on therelationship between central aortic and brachial pressure pulsein man. Cardiovasc Res 4: 291-300

O'Rourke MF (1970b) Arterial hemodynamics in hypertension.Circ Res 26 (suppl 2): 123-133



ownloaded from



O'Rourke MF, Taylor MG (1966) Vascular impedance of thefemoral bed. Circ Res 18: 126-139

O'Rourke MF, Tavlor MG (1967) Input impedance of systemiccirculation. CircRes 20: 365-380

O'Rourke MF, Blazek JV, Morrells CL Jr, Krovetz LJ (1968)Pressure wave transmission along the human aorta. Changeswith age and in arterial degenerative disease. Circ Res 23:567-579

Patel DJ, Greenfield JC Jr, Austen WG, Morrow AG, Fry DL(1965) Pressure-flow relationships in the ascending aorta andfemoral artery of man. J Appl Physiol 20: 459-463

Pepine CJ, Nichols WW, Conti CR (1978) Aortic input imped-ance in heart failure. Circulation 58: 460-465

Porje IG (1946) Studies of the arterial pulse wave particularly inthe aorta. Acta Physiol Scand 13[suppl 42]: 1-68

Puls RJ, Heizer KW (1967) Pulse wave changes with ageing. JAm Geriatr Soc 15: 153-165

Remington JW, Noback CR, Hamilton WF, Gold JJ (1948)

Volume elasticity characteristics of the human aorta andprediction of the stroke volume from the pressure pulse. AmJ Physiol 153: 298-308

Salans AH, Katz LN, Graham GR, Gordon A, Elisberg El,Gerber A (1951) A study of the central and peripheral arterialpressure pulse in man. Circulation 4: 510-521

Taylor MG (1966) The input impedance of an assembly ofrandomly-branching elastic tubes. Biophys J 6: 29-51

Westerhof N, Bosmann F, De Vries CJ, Noordergraaf A (1969)Analog studies of the human systemic arterial tree. J Biomech2: 121-143

Westerhof N, Sipkema P, Elzinga G, Murgo JP, Giolma JP(1979) Arterial impedance. In Quantitative CardiovascularStudies, edited by HWC Hwang, DR Gross, DJ Patel. Balti-more, University Park Press pp 111-150

Womersley JR (1957) The mathematical analysis of the arterialcirculation in a state of oscillatory motion. Wright Air Devel-opment Center, Technical Report. WADC-TR, pp 56-614

Membrane Electrical Effects of Histamine onVascular Smooth Muscle of Canine Coronary

Artery

DAVID R. HARDER

SUMMARY This study was undertaken to determine some of the electrophysiological effects ofhistamine on the coronary arterial vascular smooth muscle of the dog. Transmembrane potentials wererecorded from small (<500 /im o.d.) isolated canine coronary arteries with glass microelectrodes filledwith 3 M KC1. Histamine (10~6 M) increased the resting membrane potential (Em) from -55 to —64 mVand reduced input resistance from 9.8 to 4.0 mO. These effects of histamine were abolished when Mn2+

(1 mM) was added to block Ca2+ influx. The amplitude, maximal rate of rise, and frequency of the Ca2+-dependent action potential induced by tetraethylammonium ion (TEA) increased in the presence ofhistamine in a dose-dependent manner (10~7 to 10~6 M). Also, the effect of histamine on the TEA-inducedaction potential was inhibited by the Hi antagonist pyrilamine maleate (10"' M). When tension wasrecorded from helically cut strips of coronary arteries, variable results were obtained upon addition ofhistamine; i.e., some preparations showed no change in tension and others, a small increase. Whenhistamine was added to this preparation in the presence of TEA, tension increased to 60% of themaximum contraction induced by K+. These findings suggest that histamine increases the Ca2+ inwardcurrent in coronary arterial smooth muscle. The hyperpolarization induced by histamine may be dueto an increased K+ conductance that is mediated by an increased Ca2+ influx, since inhibition of Ca2+

influx by Mn2+ prevented the hyperpolarization. Circ Res 46: 372-377, 1980

THE effects of histamine on vascular smooth mus-cle (VSM) are highly variable and range from ex-citation to inhibition to no effect (Broadley, 1975;Carrier, 1965; Reite, 1972; Altura and Altura, 1974;Hudgins and Weiss, 1968). When injected into thebeating dog heart, histamine causes a marked in-

From the East Tennessee State University, College of Medicine,Johnson City, Tennessee.

Supported by Veterans Administration Grant 1A (74) 111-430100, andby a grant from the Tennessee Affiliate of the American Heart Associa-tion.

Address for reprints: David R. Harder, Ph.D., Department of Physi-ology, College of Medicine, East Tennessee State University, JohnsonCity, Tennessee 37601.

Received June 25, 1979; accepted for publication November 26, 1979.

crease in coronary blood flow (Giles et al., 1977)suggesting a vasodilation. However, when added tothe bathing solution of isolated coronary arteries ofdog (Carrier, 1965), swine, and man (Smith et al.,1951), histamine causes a vasoconstriction. Thus,there is controversy as to the direct mechanism ofaction of histamine on coronary arterial smoothmuscle.

There are only a few reports on the effect ofhistamine on electrophysiological characteristics ofVSM. Histamine increases spike frequency in rabbitmesenteric vein (Cuthbert and Sutter, 1965; Somlyoand Somlyo, 1968) and depolarizes rabbit pulmo-nary artery (Somlyo and Somlyo, 1968). There are



ownloaded from


M F O'Rourke and A P Avoliomultibranched model of the human systemic arterial tree.

Pulsatile flow and pressure in human systemic arteries. Studies in man and in a

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1980 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.46.3.3631980;46:363-372Circ Res.

http://circres.ahajournals.org/content/46/3/363.citationWorld Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Research Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:



ownloaded from

http://circres.ahajournals.org/content/46/3/363.citation

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circres.ahajournals.org//subscriptions/


pulsatile flow and pressure in human systemic arteries

Documents