Br Heart J 1986;56:531-4

Mechanism of the dicrotic pulse

DAMON SMITH, ERNEST CRAIGE

From the Department of Medicine of the University of North Carolina, Chapel Hill, North Carolina, USA

SUMMARY The dicrotic pulse is an abnormal carotid pulse found in conjunction with certainconditions characterised by low cardiac output. It is distinguished by two palpable pulsations, thesecond of which is diastolic and immediately follows the second heart sound. In the course ofopen chest canine studies of the second heart sound, micromanometers and an electromagneticflow meter were used to study proximal aortic haemodynamic function in both strong and weakbeats. It was found that the incisural notch of the aortic pressure signal is not strongly dependenton the extent of left ventricular ejection, and is of essentially normal amplitude even in beatshaving greatly reduced aortic flow. In contrast, the magnitude of the systolic upstroke of theaortic pressure pulse is strongly determined by the magnitude of left ventricular ejection and isconsiderably reduced in weak beats. With low cardiac output the relative size of the incisuralnotch becomes exaggerated in comparison with the overall pulsation, thus creating the character-istic M shaped waveform of the dicrotic pulse.

The dicrotic pulse has long been described as aphysical sign associated with low cardiac output.' Asearly as 1863, with the development of Marey'ssphygmograph, the abnormality was documentedgraphically. This abnormal pulsation is character-ised by the presence of two distinct upstrokes, thesecond being diastolic and immediately followingthe second heart sound. This configuration is shownin fig 1. Flemming suggested in 1881 that the neces-sary conditions for the abnormality included a fastpulse, elastic vessels, feeble tension in the vessels,and a small amount of blood injected with each ven-tricular systole.2 In 1902, MacKenzie pointed outthat dicrotism is usually present in conditions ofweak ventricular contraction and relaxed vessels.3Dicrotism has also been found in association withcardiomyopathy.4 5 In a more recent paper from thislaboratory, Orchard and Craige reported that persis-tence of the dicrotic pulse is an important prognosticsign after open heart surgery.6The mechanism of the dicrotic pulse has been the

subject of debate. The importance of a competentaortic valve was suggested by the study of Orchardand Craige in which two of their patients lost ini-

Requests for reprints to Damon Smith, BME, School of Medicine,University of North Carolina, 338 Bumett Womack Building,229-H, Chapel Hill, NC 27514, USA.

Accepted for publication 21 August 1986

tially large dicrotic waves when a paravalvar aorticleak developed.6 The contractile state of the myo-cardium is also of fundamental importance. Orchardand Craige reported that in pulsus altemans it is theweaker beats that display the dicrotic pulse moreprominently. Some have suggested that dicrotism isa distortion of the waveform resulting from abnor-mal arterial resistance such that the dicrotic wavebecomes exaggerated.4As a part of open chest canine studies designed to

study the second heart sound, we have made obser-vations which we feel explain the genesis of this phe-nomenon.

Methods

Six mongrel dogs weighing from 20 to 35 kg wereanaesthetised with intravenous pentobarbitonesodium and ventilated by means of a mechanicalrespirator. The dogs were placed in the supine posi-tion and a median sternotomy was performed. Thepericardium was opened and sewn so as to create apericardial cradle for the heart. The proximal aorticroot was exposed by blunt dissection. An electro-magnetic flow probe of the cuff type (Carolina Med-ical Electronics) was placed around the proximalaortic root, just above the coronary sinus. The flowsignal was recorded on a direct current input chan-nel of a multichannel physiological recorder (Elec-tronics for Medicine VR12) with a bandwidth of

531

on 28 May 2018 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.56.6.531 on 1 Decem

ber 1986. Dow

nloaded from

532

120 Hz. The aortic and ventricular pressure signalswere detected through high fidelity micro-manometer tipped catheters (Millar Instruments).The aortic catheter was inserted through the rightexternal carotid artery, with its micromanometer tippositioned within 2 cm of the aortic valve. The ven-tricular catheter was inserted directly through theleft ventricular wall at the apex and positioned in theapical region of the ventricular chamber. The aorticand left ventricular pressure signals were recordedwith a frequency bandwidth of 250 Hz. In somecases the aortic pressure signal was also recordedwith a bandwidth of 25 Hz, and this low fidelity sig-nal is shown simultaneously with the high fidelityaortic pressure signal in figs 2, 3, and 4. The highand low fidelity aortic pressure tracings wereobtained from the same micromanometer tipped

lI lll l l il1 1 1 1llllll

PCG2nd LICS

PCG-Apex

Smith, Craigecatheter. In the recordings made during studies ofhaemodynamic failure, the high fidelity aortic andventricular pressure signals are calibrated, as indi-cated in figs 3 and 4. In contrast, the low fidelityaortic pressure signal is not calibrated and isadjusted between recordings so as to show a pul-sation of approximately constant amplitude. Therecordings were made during control conditions andduring heart failure subsequent to global hypox-aemia caused by interruption of the mechanicalrespirator.

Ao25Hz

Ao

LV

ECG

Fig 2 Recording obtainedfrom open chest canine studyduring pulsus alternans. The weak beat shows no diminutionof its incisural downstroke (arrow) and rebound, although itssystolic upstroke is markedly diminished. Ao 25 Hz, lowfidelity aortic pressure signal; Ao, high fidelity aorticpressure signal; LV, left ventricular pressure; ECG,electrocardiogram.

Ao25Hz r

JlA ECG

lI I I I I I I I I I I

Fig 1 Recording of the carotid pulse and simultaneousphonocardiograms obtainedfrom a patient with dilatedcardiomyopathy in the non-invasive cardiac graphicslaboratory. The carotid tracing shows a dicrotic pulse with itsdistinguishing M shaped configuration ofwhich the secondcomponent is diastolic. PCG 2nd LICS, phonocardiogramfrom second left intercostal space; PCG-Apex,phonocardiogram from apex; CAR, carotid tracing; Dic,the dicrotic wave; ECG, the electrocardiogram.

Fig 3 Recording obtainedfrom open chest canine studyduring control condition. The aortic pressure waveformmorphology is normal. Ao 25 Hz, low fidelity uncalibratedaortic pressure signal; Ao, high fidelity calibrated aorticpressure signal; LV, left ventricular pressure.

on 28 May 2018 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.56.6.531 on 1 Decem

ber 1986. Dow

nloaded from

Mechanism of the dicrotic pulseResults

Three of the dogs showed pulsus alternans, withalternating strong and weak beats. In addition, theweak beats showed slight alteration of strength, withsome beats exceeding aortic pressure and others notexceeding aortic pressure. This fortuitous variationafforded an "experiment of nature" in which theincisural notch could be studied. An example isshown in fig 5. This figure shows a sequence of threeweak beats numbered 1, 2, and 3. These numberedbeats are shown to increase in strength, with beatnumber 2 being larger than number 1, and number 3

Ao 25Hz

75

50 Ao

Fig4 Recording obtainedfrom same dog as infig3, nowwith heartfailure. The low fidelity uncalibrated aorticpressure signal (Ao 25 Hz) now shows dicrotism. Ao, highfidelity calibrated aortic pressure signal; LV, left ventricularpressure.

II II -1-11....1....1....1....1 LA....1I...1....1....1....1....1..1 1 1....I....I...

Fig 5 Recording obtainedfrom open chest canine studyduring pulsus alternans. Wedk beats numbered 1, 2, and 3grow progressively in peak systolic pressure. Weak beats 2and 3 result in incisural notches of the aortic pressure signal(arrows). Ao, high fidelity aortic pressure signal; Ao dP/dt,time derivative of aortic pressure; LV, left ventricularpressure; AoF, proximal aortic bloodflow; ECG,electrocardiogram.

533larger than number 2. In weak beat number 1 noaortic pressure pulse is produced. Therefore, theaortic pressure of beat number 1 shows no incisuraldownslope or rebound. During this beat there is noforward or retrograde proximal aortic flow seen inthe flow signal. In weak beat number 2 an incisuraldownslope (arrow) and rebound is clearly demon-strated, even though essentially no forward aorticblood flow was detected in the flow signal. In weakbeat number 3 the size of the incisural notch (arrow)is almost as large as that of the strong beats of thefigure, even though the systolic upstroke of the pul-sation is insignificant. Note also that the extent ofretrograde aortic blood flow is as large in weak beatnumber 3 as it is in the strong beats, even though theextent of forward systolic blood flow is a very smallpercentage of that obtained in the strong beats.

Figure 2 shows simultaneously the high and lowfidelity aortic pressure signals in a sequence thatincludes a weak beat due to pulsus alternans. Theweak beat is characterised by a considerable reduc-tion in the size of the overall aortic pressure pulse.The size of the incisural downslope (arrow) andrebound is, however, of the same size as that of thestrong beats, again indicating the insensitivity of themagnitude of the incisural notch to the strength ofsystole.

Figures 3 and 4 were recorded from the same dogand indicate the characteristic alteration in the aorticpressure signal from the proximal aorta which weobserved when the dogs experienced haemodynamicfailure. Figure 3 shows the control condition, inwhich the aortic pressure signals display normalconfiguration. The incisural downslope andrebound are a small percentage of the overall pres-sure pulsation. Figure 4 shows the results obtainedduring haemodynamic failure. The overall aorticpressure pulse indicated in the calibrated highfidelity aortic pressure signal has reduced to approx-imately 60% of its size in fig 3. The size of the inci-sural notch, however, is essentially unchanged com-pared with that of fig 3. The amplifier gain of the lowfidelity aortic pressure signal in fig 4 has beenincreased between recordings so that the size of theoverall pulsation appears as large as in fig 3. Whenthis is done, the low fidelity aortic pressure signaldisplays the characteristic dicrotic pulse waveform.

Discussion

In recording the carotid or other peripheral arterialpulsations in the non-invasive cardiac graphicslaboratory it is widespread practice to adjust theamplifier gain of the recording equipment so as to

obtain a tracing of convenient amplitude. Thismeans that the smaller pulse associated with hae-

...I....I....I....I....I....I....I....I... 11 ....I....I....I....I....L.I.1 ...

on 28 May 2018 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.56.6.531 on 1 Decem

ber 1986. Dow

nloaded from

534 Smith, Craigemodynamic failure is routinely recorded with largergain than is the normal pulse. This practice is appro-priate in that no attempt is usually made to quantifythe pulse, and it is the waveform configuration andtiming that are of principal concern. We thereforeinclude in our tracings a low fidelity uncalibratedaortic pressure signal for which we adjusted theamplifier gain between recordings as the dog devel-oped haemodynamic failure. There are obviousdifferences between this low fidelity filtered proxi-mal aortic pressure signal and the non-invasivelyobtained carotid pulse. One important difference isthat this signal obtained from the proximal aorta isnot subject to the distortion that occurs as the pulsetravels down the arterial tree. Our purpose forincluding the low fidelity pressure signal is todemonstrate that the characteristic dicrotic waveconfiguration can be detected in the proximal aortaand is not dependent on modification of the pulse asit travels to the peripheral circulation.

It should be noted that the configuration of thearterial pulse is not dependent on the actual mag-nitudes of its various components but instead isdetermined by the relative contributions of these.We hypothesise that the development of dicrotism,which appears to involve an increase in amplitude ofthe dicrotic wave component, is not actually theresult of an exaggeration of the dicrotic wave butrather of a reduction of the systolic upstroke of thepulsation in conjunction with an essentially normalamplitude of the incisural downslope and rebound.This results in the characteristic M shapedconfiguration.The ability to palpate the secondary component of

the dicrotic pulse appears to be a consequence of thecharacteristics of the tactile sense. If two pulsationsoccurring closely in time are applied to the fingers,the ability to detect the second pulsation depends toa large extent on the relative amplitudes of the twopulsations. If the first pulsation is much more prom-inent than the second, the secondary pulsation willnot be perceived. If the amplitude of the first pul-

sation is then reduced so as to more nearly equal theamplitude of the second pulsation, the second pul-sation can be palpated even though it has notchanged in amplitude. The auscultatory analogue ofthis phenomenon is well known, in that a loud soundor murmur may interfere with the appreciation of animmediately following softer event.

Conclusion

Whereas in the past cardiologists have interpretedthe dicrotic pulse as the result of an exaggeration ofthe dicrotic wave, the actual basis for the unusualwaveform appears to be a reduction of the overallpressure pulse in conjunction with an essentiallynormal magnitude of the incisural downslope andrebound. This causes the pulse to show a consid-erable increase in the relative contribution from theincisura, resulting in the characteristic M shapedwaveform. Although we cannot exclude an addi-tional involvement of transmission distortion as thepulse travels peripherally, we have demonstratedthat the characteristic waveform of dicrotism can bedetected in the proximal aorta immediately abovethe aortic valve.

References

1 O'Rourke RA. Physical examination of the arteries andveins (including blood pressure determination). In:Hurst JW, ed. The heart. New York: McGraw Hill,1986:145.

2 Flemming WJ. Pulse dicrotism. Journal of Anatomyand Physiology 1881;15:278-91.

3 MacKenzie J. The study of the pulse, arterial, venous andhepatic and of the movement of the heart. London:Young J Pentland, 1902:25.

4 Ewy GA, Rios JA, Marcus FI. The dicrotic arterialpulse. Circulation 1969;39:655-61.

5 Meadows WR, Draur RA, Osadjan CE. Dicrotism inheart disease. Am Heart J 1971;82:596-608.

6 Orchard RC, Craige E. Dicrotic pulse after open heartsurgery. Circulation 1980;62:1107-14.

on 28 May 2018 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.56.6.531 on 1 Decem

ber 1986. Dow

nloaded from