Kinetocardiogram, Phonocardiogram, and Arterial Pulse Waves During Acute Hemodynamic Changes By ROBERT C. DADDARIO, M.D., AND EDWARD D. FREIS, M.D. T HERE is currently a need for simple, atraumatic methods of assessing cardio- vascular status. The kinetocardiogram (KCG) and external recordings of arterial pulse waves represent possible approaches which have not yet been adequately explored. Both techniques have been applied in various cardiovascular disorders, but little information has been ob- tained on the effects of acute alterations in hemodynamics. Since it is possible to force the circulation in known directions by the ad- ministration of vasoactive drugs with specific hemodynamic actions, it seemed worthwhile to assess the effects of such acute changes on the KCG, carotid pulse contour, and the speed of transmission of the central arterial pulse wave. Such observations on the effects of known changes in hemodynamics should help clarify the interpretation of KCG and arterial pulse-wave tracings. Part I: The Kinetocardiogram Methods The subjects consisted of 28 males who were either normal vohmteers or patients without car- diovascular disease, selected from the wards of the Veterans Administration Hospital. No pa- tients were acutely ill and none were suffering from chronic debilitating diseases. The subjects ranged in age from 25 to 48 years with a mean age of 37.4 years. Each received one to three drugs dul-ing the recording session. When more than one drug was used in the same subject, the short acting agents, such as amyl nitrite and angiotensin II, were given first and then sufficient From the Veterans Administration Hospital and the Department of Xfedicine, Georgetown University School of Yledicine, \Vnshington, District of Colum- bia. This stlltly was ~IOIIC while Dr. Daddario was a trainee antler a grant from the National Heart In- stitute, U. S. l'ul)Iic IIealth Service. time was allowed for return of the KCG, pulse wave recordings, and blood pressure to control values before the next agent was administered. The kinetocardiograms were recorded accord- ing to the method of Eddleman' using twb pick- ups, one in the K, and the other in the K4 posi- tion. These are similar in location to the electrode placements V, and V, of the standard 12-lead electrocardiogram. Lead I of the electrocardio- gram was used as a time reference. Recordings wel-e made simultaneously through Sanborn am- plifiers and a direct writing multichannel oscillo- graph at a paper speed of 100 mm/set. Blood pressure was measured in the upper extremity by the auscultatory method. Following the control tracings, various vaso- active drugs were administered. Synthetic angio- tensin II or methoxamine were used to increase total peripheral vascular resistance. Amy1 nitrite or isoproterenol was administered to decrease total peripheral resistance and increase cardiac output and also myocardial contractility. Hex- amethonium was given to decrease myocardial contractility, cardiac output, and arterial pres- sure. All drugs except amyl nitrite (which was given by inhalation) were administered by slow intravenous infusion after dilution in 5% dex- trose solution. The infusion rates were regulated to obtain a significant change in arterial pres- sure as determined by frequent monitoring. Both the amplitude and the duration of the various components of the KCG were measured and expressed respectively as percentages of total c!.cle amplitude and cycle length. Changes were determined as differences in these percentages between the control and post-drug periods. Sta- tistical anal\-sis of the changes following ndmin- istration of the various drugs was carried out by the signed ranks method.2 Results Increased Left Ventriculur Pressure Load Angiotensin II and methoxamine increase total peripheral resistance without an increase in cardiac output. 3 They were used to impose an acute prcssnre load on the left ventricle. Dcspitc conritlcrahle cle\fations of both sys- tolic and diastolic blood pressures, avern$ing 424 DADDARIO, FREIS Chunees in becordial Movements Following Vasoactive Drugs Expresses I as Per Cent of Cycle Amplitude DlllgS MeaIl B.P. H.R. R change change Mr.n Nn No. ight ventricular movement* Table 1 Kl Systolic-retraction SVS. Dia. % _.__ -- _. _. _. _. (%) incr. deer. P Mran No. No. No. (96) incr. de&. unch. P _____ Angiotensin II 9 +47 +42 -18.9 +10.3 6 1 < 0.05 -11.5 3 6 0 ns hiethoxamine 9 +x $27 -26.1 - 8.8 3 6 ns + 3.0 6 3 0 ns .4myl nitrite 13t -27 -29 +44.5 + 1.0 4 6 11s + 4.0 6 3 1 ns Isoproterenol 14 +14 -35 $61.2 + 8.0 7 4 ns + 5.1 9 5 0 ns Hesamethonium 7 -17 -22 +17.9 - 1.8 4 3 ns +17.0 5 2 0 ns . *Not present in all subjects. atria1 waves, the average change being + 4.3%. The changes in the atria1 wave after either agent were not significant. The durations of CONTROL - ANGIOTENSIN tK, not obtained in three subjects after amyl nitrite. 47/42 mm Hg in nine subjects receiving angiotensin II, and 34/27 mm Hg in a similar number given methoxamine, there were no significant changes in the amplitudes of either the pre-ejection movement or the left ventricular thrust as recorded in the K, position (table 1). The duration of the thrust was also essentially unchanged when ex- pressed as a percentage of the cycle length. The systolic retraction which occurs during left ventricular ejection fell significantly after both drugs (table 1, figs. 1 and 2). The mean change was a reduction in systolic retraction of 13.9% (P < 0.05) of the total cycle ampli- tude following angiotensin II, and 15.3% (P = 0.01) after methoxamine. The right ventricular outward movement lvhich preceded \rentricular ejection as re- corded in the K, position increased slightly during angiotensin II infusion (mean + 10.3, P < 0.05) but showed no definite trend during mcthoxamine infusion. Systolic retraction de- creased in the K1 position after angiotensin II, but the change \vas not significant. It remained essentially unchanged after methoxamine. Atria1 waves were recorded in the K1 position in five subjects receiving angiotensin II. The atria1 wave increased in three subjects and decreased in two, the average change being + 8.3% of the total cycle amplitude. Atria1 waves \wre pwsent in three subjects \\-I~0 \vcre given mcthosamine. Two subjects had increased atria1 wa\.es and one had decreased EGG HEART SOUNDS KI K4 CAROTID FEMORAL ,4CUTE HEMODYNAMIC CHANGES 425 nt -- Prc-ejection Left ;ihernu:;icuicular Systolic movement' retraction Mean No. NO. NO. Mean No. NO. Mean NO. NO. NO. (%I incr. dew. unch. P ( % ) incr. dew. P (70) incr. decs. unch. P + 5.5 5 4 0 Il.5 + 2.0 5 4 - 2.3 4 2 0 xl.5 + 0.6 4 5 + 9.5 5 4 2 ns -29.3 1 12 +14.0 7 2 0 ns `-28.2 3 11 --14.8 0 3 1 ns -11.4 3 4 either the atria1 wave or the right ventricular movement were essentially unchanged. Increased Cardiac Output with Decreased Pressure Load Both amyl nitrite4 and isoproterenol" increase cardiac output and diminish total peripheral resistance. The increase in output is due primarily to increased heart rate rather than ECG HEART SOUNDS K4 CAROTID CONTROL MTHOXAMINE FEMORAL ns -13.9 1 7 1 < 0.05 ns -18.3 0 8 1 = 0.01 < 0.01 +25.2 11 2 0 < 0.01 = 0.01 +30.9 12 1 1 < 0.01 ns f29.2 6 1 0 < 0.05 _ to stroke volume. The ventricular stimulation produced by amyl nitrite is probably reflex in origin secondary to the diminished arterial pressure whereas isoproterenol has a direct inotropic effect on the heart. The mean de- crease in arterial pressure following amyl ni- trite inhalation was 27/29 mm Hg. After isoproterenol, systolic pressure increased by an average of 14 mm Hg while diastolic pressure decreased by 35 mm Hg. Eleven of 13 subjects receiving amyl nitrite and nine of 14 subjects who were given isoproteronol exhibited pre-ejection move- ments in the records taken in the K, position. In the amyl nitrite group there was no sig- nificant change in amplitude of this deflection, five increased, four decreased, and two exhibited no change (table I). In the nine subjects receiving isoprotcrenol, seven in- creased and tw`o decreased, the mean change being + 14.0% (P < 0.1) of the total cycle amplitude. The amplitude of the left ventricular thrust decreased significantly following both drugs. The mean reduction after amyl nitrite was 29.33 (P < O.Ol), and after isoproterenol it \vas 25.2% (P = 0.01). The duration of the thrrrst expressed as a percentage of the cycle length shortened slightly after both drugs, but these changes were not significant. Fol- lowing amyl nitrite, the mean decrease was 2.0% of the total cycle length with 11 of the 13 subjects showing this response. Jn the 14 palicnts rewiving isoprotcwnol, sewn sllo~wtl a decrease in duration, folrr an increase, and 426 DADDARIO, FREIS three remained unchanged, the mean change being - 1.2% of the total cycle amplitude. The degree of systolic retraction during the left ventricular ejection phase of the cardiac cycle, as recorded in the K, position, increased (table 1, frgs. 3 and 4). The increase in negative deflection averaged 25.2% (P < 0.01) after amyl nitrite inhalation, and 30.9% (P < 0.01) in the subjects receiving isopro- terenol. There were no significant changes in the recordings taken from the K, position following either amyl nitrite or isoproterenol. Decreased Left Ventricular Pressure Load and Ouiput Hexamethonium lowers arterial pressure pri- marily by reducing cardiac output, the total peripheral resistance remaining essentially un- ECG HEART SOUNDS KI K4 CAROTID CONTROL FEMORAL AMYL NITRITE Figure 3 Recordings before and offer rcdudion of blood pres- sure with umyl nitrite shotcirrg incrscwz in systolic re- trmtion and dccrcose in mugnitude of left vmtrictrlar thrust in the K, posifion recording of the RCG. Other tlotafions us in figure 1. CONTROL ECG HEART SOUNDS KI K4 CAROTID Figure 4 Records showing disappearance of P maximum and reduction in l/F ratio of the carotid pul~c wave fol- lowing infusion of isoproterenol. Other notations a.~ in figure 1. altered." The only significant change in the KCG following hexamethonium \vas an in- creased systolic retraction in the K, position (table 1, fig. 5). Th e mean value for the amplitude of the left ventricular thrust de- creased 11.4%, but the rqonse was too vari- able in the different subjects to be regarded as significant. Discussion The KCG abnormalities characteristically associated with chronic left ventricular over- load, as seen in patients nith aortic valvular disease or hypcrtcnsion, have been described by Da\+ and associates.' These chnnges con- sisted of an increase in the magnitude and duration of the left vcntricrrlar thrrrst recorded in the K, position. Davie arid associates were unnble to differentiate betw,ecn the possible effects of hypertrophy, dilatation, or increase in work of the left vcntr-icle as a cause of these abnormalities. In the prcscnt study acute left vcntj-icrrlar overloads were imposed by clIc\,ating total CircdsGon, Volume XXXIV, Stptembrr 1366 ACUTE HEMODYNAMIC CI1ANGES HEART SOUNDS K4 CAROTID FEMORAL Figure 5 Records showing reduction in P/F and I/F ratios of CONTROL HEXAMETHONIUM the carotid pulse uaoe during reduction of arterial pressure after infusion of hexametho,lium. Carotid- femoral transmission time increased from 68 to 76 msec. Other notations as in figure 1. peripheral resistance with angiotensin II or methoxamine. No significant cbnnges were noted in tither the magnitude or duration of the left ventricular thrust. The pre-ejection movement was similarly unaffected. These results are in contrast to the changes foulld by Davie and associates7 in patients vrith chronic overloads. This suggests that in pa- tients with long-standing hypertension and aortic valvular disease the increase in both amplitude and duration of the thrust is related not so much to the magnitude of the load but more specifically to its chronicity. The principal difference in the adjustment of the ventricle to a chronically imposed load as contrasted to an acutely imposed load is myocardial h>pcrtrophy. Davie and associates' did not find a significant car-relation bctwcen the degree of thrust abnormality and cardio- mcgaly as determined by x-ray. Ilovvevcr, it is \veIl known that ventricular h~pcrtrophy 427 in the absence of `dilatation often cannot be detected in the x-ray. On the other hand, reduction in arterial pressure with amyl nitrite and isoproterenol resulted in a significant decrease in the mag- nitude of the left ventricular thrust and a slight but insignificant shortening of its thrust, each expressed as a percentage of the total cycle amplitude and duration, respectively. Hexamethonium resulted in a reduction in the magnitude of the ventricular thrust in some l:atients, but the responses were too v-ariable to be regarded as significant. These results indicate that the magnitude of the thrust is relatively independent of increases in left ventricular load although reduction in pressure load may reduce the percentage magnitude but not the percentage duration of the thrust. In contrast, the changes in systolic retraction as a percentage of total cycle amplitude were consistent and significant with all of the acutely imposed alterations in left ventricular load. Systolic retraction decreased after angio- tensin II and methoxamine and increased after amyl nitrite, isoproterenol, and hexa- methonium. Thus, the changes in systolic retraction were inversely related to the pres- sure loads imposed on the left ventricle. Since systolic retraction is an inward movement of the chest wall occurring during the period of ventricular ejection, it is proba- bly a reflection of the decrease in ventricular volmne occurring during the phase of rapid vcntricrrlar ejection. An acute increase in the pressure load with angiotensin II reduces left ventricular emptying thereby increasing the end-diastolic volume of the ventricle.8 When ventricular volume is expanded, less shorten- ing of vwrtricular diameter \\511 be required to expel a given strolie vohime than u-hen \w~tricular volume is reduced. If this inter- prctation is correct, the magnitnde of the dovvnstroke beginning at the peak of the thrust and ending at the completion of the systolic retraction \vave should provide an indication of the approximate percentage chalrge in left ventricular v~olume during the period of rapid left vcntricrrl;u ejection. This suggestion Ob~iousIy needs to be invcstigatcd Irnder a 428 DADDARIO, FREIS wide variety. of clinical and experimental ated on various indices provided by simul- conditions before it can be accepted. * taneous recordings of the carotid and femoral Summary A variety of vasoactive drugs were employed to produce acute hemodynamic alterations and * their effects were determined on the kineto- cardiogram. The most consistent alteration was a change in the systolic retraction wave recorded at the I& or apex region. Imposition of a left ventricular overload using angiotensin II or methoxamine decreased the amplitude of the wave expressed as a percentage of the total cycle amplitude. Reduction of left ventricular pressure load, by using amyl nitrite, isoproterenol, or hexamet~lonium in- creased the degree of systolic retraction. These alterations in the lnagnit~lde of the downstroke from the peak of left ventricular thrust to the end of the systoIic retraction wave may reflect percentage changes in left ventricular v&me during this period. In contrast to patients with chronic hyper- tension or aortic valvuIar disease, the impo- sition of an acute left ventricular overload in norma subjects produced no significant change in the magnitude or duration of the left ventricular thrust expressed as percentages of the total cycIe amplitude and duration, re- spectively. These results supply additiona evidence that the duration of the thrust is a useful index of left ventricular hypertrophy. Part II: Changes in Arterial Pulse Waves Aside from the recording of changes in contour in aortic valvular and peripheral vas- cular occlusive disease, little use has been made of externally recorded arterial pulses in clinical medicine. Nevertheless, character- istic alterations in the shape of the carotid pulse occur with aging and hypertension.s* lo These appear to be related to the loss of arterial dister~sibil~ty associated with these con- ditions." In addition, Katz and Feil12 and Weissler and associates'" have shown that the carotid pulse can be used as an indicator of alterations in left ventricular dynamics. In the present study, the effects of acutely indnccd changes in cardiac output or total peripheral vascular resistance \vcre cvalu- pulse waves, heart sounds, and eiectrocardio- gram (ECG). Such studies might prove useful in the interpretation of alterations in these functions observed in different age groups and in the presence of cardiovascular disorders, Methods The subjects were the same as those de- scribed in Part I, the KCG and pulse waves be- ing recorded simultaneously, The arterial pulse- wave transducer has been described previously.iq It consists essentially of a water-filled chamber sealed at one end with a compliant plastic mem- brane, and at the other by a metal diaphragm, on which two strain gauges are mounted. One of these transducers was used to record the carotid pulse. It was attached to the subject's neck with an adjustable clamp previously de- scribed.15 Another similar transducer used to re- cord the femoral pulse was attached to a rigid support overlying the femoral triangle. The trans- ducer then was lowered over the artery and clamped in place at the overhead support. The measurements taken on the carotid pulse were based on the relative heights of three inflec- tions, two positive inflections or maxima occurring during systole followed by a negative inflection, the in&ma. A Iine was drawn connecting the foot points at the beginning and end of a pulse cycle and perpendiculars were dropped from the three inflections. The ratio of the height of the second to that of the first maxi- mum (P/F ratio) and of the height of the incis- ma to the first maximum (I/F ratio) were caf- cuiated. The difference in pulse-wave transmission time between the carotid and femoral arteries was determined by using a magnifying lens as fol- lows: The steep ascending upslope of the wave was extrapolated downward until it intersected a her-izontal line drawn between the foot points at the beginning and end of the cycle. The point of intersection was taken as the onset of the wave. The time difference between the onset of the carotid and femoral pulses during the same cardiac cycle was taken as tire carotid-femoral transmission time difference. The ejection time was measured from the time of onset of the carotid wave, as defined above, to the minimum point of the incisur-a. The ejection time index (ETI) was calculated by the method of tireissler and asso&ntes'G which nor- malizes the ejection time with respect to heart rate. The isovolumie contraction time ( ICT) was determined by measuring the time between the beginning of the first and second hrart , ACUTE HEMODYNAMIC CHANGES 429 sounds and subtracting from this the ejection time. Low-frequency and low-amplitude vibra- tions preceding the first and second heart sounds were disregarded," the onset of each sound be- ing taken as the first high frequency vibration of amplitude at le.ast twice that of the background noise level. With respect to the first heart sound this point in time coincides with the onset of the steep rise in left ventricular pressure and does not take into account the slow initial phase of left ventricular contraction which lasts 10 to 20 msec. Thus ICT may be underestimated by this method but is probably satisfactory for compara- tive purposes. In some cases, following isoprotere- nol, the second heart sound was not clearly de- lineated. In the cases where it was recognizable after isoproterenol, the time interval between the beginning of the second heart sound and the carotid incisura remained unchanged from the control. Therefore, in cases of doubt, the onset of the second heart sound following isoproterenol was determined by subtracting the contr-ol time interval from the carotid incisura. The heart sounds were recorded with a San- born dynamic microphone and amplifier using a high-pass filter of 12 db per octave, and a nom- inal frequency cut-off of 100 cycles per second. The amplitude of the first heart sound was measured from peak to peak in millimeters, and the postdrug results were expressed as a per- centage of the controlZrs Tension periodlg was measured from the time of onset of the QRS complex to the onset of the second heart sound, and from this interval the ejection time was sub- tracted. The Q-S, interval was taken as the time between the beginning of QRS and the onset of the first heart sound. The average of three con- secutive pulse cycles was used in all of the above measurements. R.esults The externally recorded carotid pulse char- acteristically displayed two positive inflections or maxima during systole. The first (F maxi- mum) was temporally related to the anacrotic bend of the aortic pressrue pulse and to peak blood velocity, while the second (P maxi- mum) \\-as related to peak aortic pressure.20 The ratio of the heights of P to F was previ- ously found to increase with age and hyper- tension97 lo* 2o The ratio of I, the height of the incisura, to F, also increased. Angiotensin II and mcthoxnmine elevate peripheral vascular resistance thereby raising both systolic and diastolic pressure." The ratios of P/F and I/F \vcre significantly increased 430 by both agents (table 2, figs. 1 and 2). By contrast, amyl nitrite and isoproterenol de- crease total peripheral resistance,4v 6 and these agents produced a significant decline in P/F and I/F ratios (table 2, figs. 3 and 4). Hex- amethonium lowers blood pressure primarily by reducing cardiac output rather than by lowering peripheral resistance. This drug also produced a decrease in P/F and I/F ratios (f45 5). The interval between the foot points of the carotid and the femoral pulses was designated as the carotid-to-femoral transmission time difference and was used to serve as an index of pulse-wave velocity changes occurring in the aorta in response to the various vasoactive agents. The transmission time difference in- creased significantly following amyl nitrite and hexamethonium but not after isoproterenol (table 2). Both angiotensin II and methoxa- mine decreased the transmission time signifi- cantly. The ejection time measured as the interval from the foot of the pulse wave to the incisura increased after angiotensin II and methoxa- mine and decreased following amyl nitrite, hexamethonium, and isoproterenol (table 3). These changes in ejection time were approxi- mately inversely related to the alterations in heart rate produced by these agents. The ejection-time index which normalizes the ejection time with respect to heart ratelO remained essentially unchanged following methoxamine, angiotensin II, and isoprotenerol (table 3). It increased after amyl nitrite (P < 0.01) and decreased slightly following hexamethonium (P < 0.10). The iso\~olnmic contraction time (ICT), measured by subtracting the ejection time from the interval bet\\-een the t\vo heart sounds increased after angiotensin II and decreased after amyl nitrite and isoproterenol (table 3). There M'as no significant change after methox- amine or hexamethonium. The quotient of diastolic pressure divided by ICT (DP/ICT) provides an index of the mean rate of the left ventricular pressure rise during the isovolumic contraction period." As calculated by this method, the rate of left \:entricular pressure DADDARIO, FREIS ACUTE HE14ODYNAMIC CIIANGES rise for the 28 subjects during the control pe- riod averaged 2,043 k 1,280 mm Hglsec. This index increased 19.5% (P < 0.10) after isoproterenol and decreased 20.3% (P < 0.05) in the subjects given hexamethonium. There were no significant changes in this index fol- lowing amyl nitrite, angiotensin II, or methox- amine. The measurements of diastolic pres- sure by the auscultatory method during the peak action of isoproterenol probably were falsely low since the hemodynamic changes associated with this agent led to a persistence of the Korotkoff sounds. If the diastolic pres- sure had been recorded directly, the ratio of the latter to ICT probably \\Tould have been greater. Tension period is similar to ICT except that it also includes the period beginning at the time of onset of the QRS complex of the electrocardiogram. Tension period increased slightly after angiotensin II and methoxamine and decreased considerably after both amyl nitrite and isoproterenol (table 3). Tension period did not change significantly following hexamethonium. The Q-S, interval, which is the period be- tween the beginning of QRS and the first heart sound, remained mlchanged after angio- tensin II and methoxamine. It decreased after both amyl nitrite and isoprotorenol, the change being more marked with the latter drug (table 3, fig. 6). Following hexamethonium, Q-S, increased. The amplitude of the first heart sound in- creased after amyl nitrite and isoproterenol with the greater change (mean +284%) oc- curring after isoprotcrenol (table 3 and fig. 6). The amplitude decreased after hexametho- nium and \vas insignificantly changed follow- ing angiotensin II and mcthoxamine. Discussion Aortic Wall Stiffness The changes in relative amplitudes of the two systolic maxima observed in the carotid pulse produced by vasoactive drugs, have been described in a previous rcport.`0 The height of the first positive inflection, or F max- imum, is related to the acceleration of the blood in the central arterial system during Cirrulorion. Volame XXXIV, Sep:ember 1966 431 early ventricular ejection, The height of the second or P maximum is related to the input impedance of the arterial system.20 The pres- ent study confirms and extends the former ob- servations. Elevation of input impedance by angiotensin II and methoxamine increased the second maximum probably by restricting sys- tolic runoff with resulting distention of the large arteries. 22 The incisura, which was in- scribed soon after the second maximum, rose in association with the latter. These changes resulted in elevation of P/F and I/F ratios. Reduction of input impedance and wall ten- sion with the vasodilator drugs used in this study had an opposite effect on the P/F and I/F ratios. The velocity of the arterial pulse wave has long been used as an indicator of the changes in arterial elasticity occurring with age and cardiovascular disease.23* 24 Past attempts to x - .-ii.-- I x -a-- II Figure 6 Changes in the Q-S, interval and in the amplitude of the first heart mund follottii~g curious drug-indqced hernodynamic alterations. 432 DADDARIO, FREIS characterize the elasticity of the aorta in pre- cise quantitative terms from pulse-wave ve- locity data have not been accepted because of many indeterminable factors such as varia- tions in wall structure in different portions of the aorta, difficulties in accurately measuring the vessel length between the pickups, visco- elastic properties of the arterial walls, and other variables. Nevertheless, the available evidence suggests that pulse-wave velocity may serve as an approximate index of arterial elasticity of sufficient accuracy to be useful clinically.23~ 24 In the present studies the carotid-femoral transmission time difference was used to deter- mine whether the speed of propagation of the pulse wave through the aorta would change in the expected direction with alterations in aortic wall stiffness. Both angiotensin II and methoxamine increase aortic wall tension as indicated by the rise in systolic and diastolic pressures. As would be expected under such circumstances, the transmission time difference was consistently shortened. On the other hand, lowering of both systolic and diastolic pres- sures with amyl nitrite or hexamethonium de- creased aortic wall tension and lengthened carotid-femoral transmission time, again dem- onstrating that changes in the stiffness of the aortic wall were reflected in the speed of pro- pagation of the pulse wave. The inconsistent changes which occurred in the pulse-wave transmission time during iso- proterenol may reflect opI3osing influences on the pulse-wave velocity. This drug de- creased diastolic pressure which would reduce the speed of transmission. However, aortic blood velocity rises considerably after iso- proterenol wrhich, as pointed out by Bramwell and IIi1k2" will cause an equal increase in the velocity of the pulse wave. The present study employed the method of external pulse-wave recording using drugs as hemodynamic forcing functions. One purpose was to determine whether certain character- istics of these waves could be used as indices of the extent of changes in central arterial wall structure which are known to occur with advancing age.", " The results indicate that both contour changes in the carotid pulse wave and transmission time through the aorta can be used as indices of alterations' in cen- tral arterial wall stiffness. Furthermore, both indices have been shown to change character- istically with advancing age although the cor- relation shows considerable spread.`O* 23, 24 The use of several criteria of central arterial wall stiffness recorded simultaneously, such as carotid pulse contour and carotid-femoral transmission time differences, should increase the reliability of the method. Cardiac Function Another purpose of this study was to deter- mine whether externally applied transducers can -be used as a measure of cardiac perfor- mance. Weissler and associates13 found a re- duction in left ventricular ejection time index (ETI) in normal subjects given digitalis which they attributed to the inotropic effects of the drug. In the present study, ET1 re- mained unchanged except for a slight increase following amyl nitrite, and decrease after hexamethonium. These small variations may be due to alterations in stroke volume follow- ing use of these agents. Stroke volume falls after hexamethoniumG and may rise after amyl nitrite, although the increase in cardiac output produced by amyl nitrite and isoprotercnol is due primarily to an elevation in heart rate rather than in stroke volume.4p 5 The ejection- time index did not seem to provide an ac- curate reflection of ventricular contractility in the prcscnt study since the index was insig- nificantly altered after isoproterenol, a drug which considerably augments ventricular pow- er. Following digitalis, the contraction is not only more powerful but also is more sus- tained whereas after administration of isopro- tcrenol contractile force is increased but the contraction period is shortened. Thus, ET1 by itself does not appear to express changes in myocardial contractility under all circum- stances. The isovolumic contraction time (ICT) re- fers to the interval bctlveen the closure of the mitral valve and the opening of the aortic valve. Frank and Kinlaw," using the same in- direct method employed in the present study, C:rruldion, Volume .XXXlV. Sep'omber I.966 ACUTE HEMODYNAMIC CHANGES 433 found an average ICT of 49 msec in normal subjects with a standard deviation of con- secutive cycle variation of 4.0 msec. In the present series ICT averaged 38.3 msec. The difference may be due at least in part to the somewhat lower diastolic pressure of 72.1 mm Hg in the present series as contrasted to an average value of SO.7 mm Hg in Frank and Kinlaw's subjects.`7 Katz and Feil,r2 who also employed the same method, concluded that ICT was an index of the velocity of ventricu- lar contraction. Reeves and associates,2s using the left ventricular and aortic pressure pulses, also related ICT to myocardial contractility. If ventricular power remains unchanged, ICT should rise or fall with corresponding changes in aortic diastolic pressure since ejec- tion begins at the moment that left ventricular pressure exceeds aortic pressure. ICT in- creased with angiotensin II and methoxamine and decreased following amyl nitrite and isoproterenol. ICT did not shorten significant- ly after hexamethonium despite a reduction in diastolic pressure suggesting that ventric- ular power had decreased under the influence of ganglionic blockade. An attempt was made to find a regression equation that would nor- malize ICT with respect to diastolic pressure because of the apparent relationship between diastolic pressure and ICT. Holvever, the correlation coefficient between diastolic pres- sure and ICT in the 26 subjects during the control period was only 0.47. This degree of scatter was considered too great to make such normalization useful. The ratio of diastolic pressure to ICT (DP/ICT) should provide an index of the mean rate of rise of left ventricular pressure during the initial phase of ventricular contrac- tion. Both Reeves and associatesz8 and Rush- merzu have indicated that the maximum rate of pressure rise in the Iyft ventricle (maxi- mum dp/dt) is closely correlated \vith other indicators of myocartlial contractility. Landry and Goodyer" have shown recently in dogs that DP/ICT correlated closely with maxi- mrun dp/dt measrucd directly in the left vcn- tr-icle. Their avrragc control \,ahre \vas npllrox- imately 2;390 mm Hg/scc by both methods. They also observed changes in the rate of rise of ventricular pressure after beta-adrcnergic stimulation, and use of a ganglion blocking drug and methoxamine that were similar to those observed in the present studies in man. Gleason and BraunwaId30 measured the max- imum rate of left ventricular pressure rise directly in man. These investigators also found that it increased significantly following iso- proterenol but remained rmchanged after mcthoxamine. In the control state, they found that maximum left ventricular dp/dt varied between 541 and .1,696 mm Hg/sec. This range is considerably lower than is indicated by the mean value of 2,043 -C 1,280 mm HG/sec found in the present study, obtained by the indirect method. Their patients had valvular lesions or septal defects and, there- fore, probably did not have entirely normal ventricular dynamics. Another factor produc- ing the higher values in the present series is that the end-diastolic pressure in the left ven- tricle was assumed to be zero. The onset of the first heart sound also is sometimes indistinct due to low frequency vibrations which precede the closure of the mitral valve. This difficulty can be avoided by measuring total systole including electrical systole from the onset of the QRS complex to the beginning of the second heart sound. The difference between this interval and the ejec- tion period as determined from the carotid pulse has been called the tension period.`" The latter changed in the same direction as ICT in all instances except that the magnitude of the change \yas not as great. Thus, although easier to dctcrmine, tension period is a Icss sensitive index of changes in the pre-ejection phase of sy-stole than is ICT. In the a?,sence of mitral stenosis, the Q-S, in- terval may provide a uscflrl index of coutrxctil- ity. It shortened considerably after administra- tion of isoproterenol, moderately following amyl nitrite, incrcnsed \tTith hexamethonium. a11d rcrrrnined unchalrgcd following augioten- sin II and methoxamine. The latter t\vo agents are believed to have little or no inotropic caffects 011 the heart. Isoprotcr(no1 and ilnyl 434 DADDARIO, FREIS nitrite stimulate adrenergic activity, the for- mer directly, and the latter through reflex action, while hexamethonium blocks sympa- thetic nerve transmission. Sakamoto and associates,18 who used the directly measured maximum dp/dt as an index of left ventricular contractility in closed chest dogs, found that percentage changes in the peak-to-peak amplitude of the first sound var- ied directly with changes in ventricular con- tractility induced by a variety of procedures and pharmacological agents. The results of the present study in man are consistent with their conclusion and suggest that changes in ven- tricular contractility can be estimated by this simple method. Unfortunately, due to chest Configurations, degrees of adiposity and other factors, variations occur in the transmission of the heart sounds among different subjects making comparisons between subjects invalid by this method. However, in studies of drug effects or other acute procedures on ventricu- lar contractility in the same individual, the amplitude changes in the first heart sound may pro\Gde an attractively simple method for assessing this important cardiodynamic variable. Summary The effects of drugs that produce known hemodynamic alterations were assessed on ex- ternally recorded pulse waves, heart sounds, and the relationship of these to each other and to the electrocardiogram. The shape of the carotid pulse and the caro- tid-femoral pulse-wave transmission time dif- ference showed changes that could be related to alterations in central arterial distensibility. The ratio of the second to the first positive in- flection, and of the incisura to the first positive inflection of the carotid pulse wave, increased following use of drugs \vhich raised mean ar- terial pressure and decreased after those \qThich lo\\.ercd blood pressure. The carotid-femoral transmission time difference decreased with vasopressor agents, increased \r-ith amyl nitrite and hexamethonium, and showed no consis- tent change after isoproterenol. Left \.entricuIar ejection time, measured from the carotid pulse, changed in relation to heart rate, increasing with cardiac slowing and decreasing when the rate accelerated. The ejection time index remained unchanged ex- cept for a slight increase following amyl ni- trite and a decrease after hcxamethonium. Alterations in isovolumic contraction time approximately paralleled changes in diastolic pressure. The ratio of diastolic pressure to isovolumic contraction time provides an in- cl~x of the rate of rise of left ventricular pres- sure. The quotient of diastolic pressure divid- ed by isovolumic contraction time increased with adrenergic stimulation (isoproterenol) aud decreased after ganglion blockade. It was not significantly changed by angiotensin II, methoxamine, or amyl nitrite. The interval between the onset of QRS and the beginning of the first heart sound (Q-S,), shortened considerably after isoproterenol and moderately following amyl nitrite. It increased after hexamethonium and remained essentially unchanged following angiotensin II or methox- amine. These results suggest that the Q-S, interval may reflect ventricular contractility in the absence of mitral valvular disease. The amplitude of the first heart sound ap- peared to be a sensitive indicator of ventricu- lar contractility increasing with isoprotcrcnol and amyl nitrite, decreasing \l.ith hexametho- nium and remaining unchanged following use of angiotensin II or methoxamine. Conclusions: Parts I and II The present studies indicate that external transducers may provide important informa- tion on structural and functional alterations in the cardiovascular system. The following in- dices appear to merit further study: 1. Percentage change in left ventricular vol- lime during the interval from the onset to the peak rate of left ventricular ejection. This is estimated in the K, position of the kinctocar- diogram from the magnitude of the down- stroke beginning at the peak of the left \Tentricular thrust to the end of the systolic retraction \vave expressed as a percentage of the total cycle amplitude. 2. Left ventricular hypertrophy from the magnitlltle and duration of the left vc'ntriclllar thrust. CirruLfion, Volume XXXIV. Stpicmber 1966 ACUTE HEMODYNAMIC CHANGES 435 3. Left ventricular contractility either from the Q-S, interval or the estimated mean rate of rise of left ventricular pressure as derived from the ratio of diastolic pressure to the isovolumic contraction time (DP/ICT). Changes in contractility in the same individual due to drugs or other factors may be estimated simply from percentage changes in the ampli- tude of the first heart sound. 4. Central arterial distensibility from the contour of the carotid pulse as well as the carotid-femoral transmission time difference. Acknowledgment The authors wish to thank Mr. Joseph C. Strong for his valuable technical assistance. References 1. EDDLEMAN, E. E., JR.: The kinetocardiogram- ultra low frequency precordial movements. In Cardiology, snppl. 1, edited by A. A. Luisada. New York, Blakiston-McGraw, 1963, p. 3. 2. MAINLASD, D.: Elementary hledical Statistics. Philadelphia, W. B. Saunders Co., 1963, p. 273. 3. i\v~*no, D. M.: Phxmacologic approach to the treatment of shock. Ann Intern Med 62: 1050, 1965. 4. PEHLOFF, J. K., COLVIN, J., AND DE LEON, A. C.: Systemic hemodynnmic effects of amyl nitrite in normal man. Amer J3eart J GF: 460, 1963. 5. DOLXE, H. T., J,OHD, J. D., AND SANDLER, H.: Cardio\ nscular effects of isoproterenol in nor- mal subjects and subjects with congestive heart failure. Amer Heart J GO: 94, 1960. 6. FREIS, E. D., ET AL.: Hemodynnmic effects of h).potensive drugs in man: III. Heuamethoni- um. J Clin In\.est 32: 1285, 1953. 7. D.~vIE, J. C., LANGLEY, J. O., DODSON, W. H., .~ND hDLEXlAN, E. E.: Clinical and kineto- c;u-dingraphic studies of paradoxical precor- dial motion. Amer Heart J 63: 775, 1962. 8. I+xsrEIN, J. W., AXD \~'mnr~r~c, M. G.: Com- parative effecti of norepillephrine and nngio- tensin on left \,cntricular performance in in- tat: dogs. Circulation 28: 714, 1963. 9. FJUART, J.: La morphologic du sphygmogramme cnrotidien dans l'nrterisoclerose: Etude prelimi- naire. Acta Cardiol (Brux) 15: 557, 1960. 10. Dosiras, A. S., TAYI.OR, H. L., AND KEYS, A.: Cnroticl pressure plcthysmograms: Effects of age, diastolic blood pressure, relative body Lveight and physical activity. Arch Krcislnuf- fol-sch 3F: 49, 1961. 11. JJALLo~K, P., AND BENSON, I. C.: Studies on Ciriularion, Volrrms XXXIV, September 1966 the elastic properties of isolated human aorta. J Clin Invest 16: 595, 1937. 12. KATZ, L. N., AND FEIL, H. S.: Dynamics of auricular fibrillation: Ventricular systole: Arch Intern Med 32: 672, 1923. 13. \VEISSLER, A. M., GAINYL, W. G., GRODE, H. E., COHEN, S., AND SCIIOESFELD, C. D.: Effect of digitalis on ventricular ejection in normal hu- man subjects. Circulation 29: 721, 1964. 14. DAVIS, M., GILMORE, B., AND FREJS, E. D.: Im- proved transducer for external recording of arterial pulse waves. IRE Trnns Biomed Electronics 10: 173, 1963. 15. GILXIORE, B. L., AND FREIS, E. D.: Effect of amyl nitrite on the height of the pulse wave incisura in atherosclerotic patients. Angiology 15: 219, 1964. 16. M'EJSSLER, A. M., HARRIS, L. C., AND WHITE, G. D.: Left ventricular ejection time index in man. J Appl Physiol 18: 919, 1963. 17. COUKJHAN, T., ~IESSER, A. L., RAPPAPORT, E. E., AND SPR.AGUE, H. B.: Initial \&rations of the first heart sound. Circulation 3: 730, 1941. 18. SAKAA~OTO, T., KuSKAWA, B., hfAcCAh=N, D. M., AND LUISADA, A. A.: Hemodynamic deter- minants of the amplitude of the first heart sound. Circulation Research 16: 45, 1965. 19. BLU~WEWGER, K., .~XD ~IEINERS, S.: Studies of cardiac dynamics. In Cardiology, vol. 2, part 4, edited by A. A. Luisada. New York, Blakis- ton-McGraw, 1959, p. 372. 20. FREEIS, E. D., HEATH, W. C., J,UCHSINGER, P. C., AND SNELL, R. E.: Changes in the carotid pulse with age and hypertension. Amer Heart J 71: 757, 1966. 21. LANIIRY, A. B., JR., AND GOOIIYER, A. V. N.: Rate of rise of left ventricular pressure: In- direct measurement and physiological signifi- cance. ,4mer J Cardiol 15: G60, 1965. 22. PETFRSON, L. H.: Dynamics of pulsatile Idood flow. Circulation Research 2: 127, 1954. 23. J~ALLOCK, P.: Arterial elasticity in man in rela- tion to age as evaluated by the pulse wave velocity method. Arch Intern Med 54: 770, 1934. 24. JI~YZES, F. W., ELLIS, L. B., ASD WEISS, S.: Pulse \vave velocity and artcl-ial elasticity in aI-tczrinl h>-pertcnsion, arteriosclel-osis and re- lated conditions. i\mer Heart J 11: 385, 1936. 25. BRAMU.ELL, J, C., AXD HILL, A. V.: \'clocity of the pulse wave in man. Proc Roy Sot (Biol) 93B: 298, 1922. 26. ~IOHET, P. R.: Modifications de l'elnsticiti? ar- tel-ielle avec l'age. Bib1 Cardiol 15: 40, 1964. 27. liR.sTi~, hl. N., AWD KISLAW, W. B.: Indirect nx~a\urenxfnt of isovohunetric contraction time and tension period in normal sul)jcc,ts., Amer J Cnrdiol 10: 800, 1962. 436 DADDARIO, FREIS 28. REEVES, T. J., HEWER, L. L., JOSES, W. B., 29. RUSIIMER, R. F.: Initial ventricular impulse, po- COGHLAN, C., PRIETO, G., AND CARROLL, J.: tential key to cardiac evaluation. Circulation Hemodynamic determinants of the rate of 29: 268, 1964. change in pressure in the left ventricle during 30. GLEASON, W. L., AND RRAUNWALD, E.: Studies isometric contraction. Amer Heart J 60: 745, on the first derivative of the ventricular pres- 1960. sure in man. J Clin Invest 41: 80, 1962.