Journal Logo


β-Adrenergic Stimulation Enhances Left Ventricular Diastolic Performance in Normal Subjects

Nixdorff, Uwe; de Mey, Christian*; Belz, Gustav G.*; Erbel, Raimund; Simon, Helge; Butzer, Raunhild*; Schroeter, Volkmar; Meyer, Jürgen

Author Information
Journal of Cardiovascular Pharmacology: April 1997 - Volume 29 - Issue 4 - p 476-484
  • Free


It is well accepted that transmitral inflow velocities determined by pulsed wave (PW) Doppler echocardiography provide a clinically useful parameter of left ventricular diastolic function. The accuracy has been validated with respect to more standard techniques (1). The method is non-invasive and easy to apply by all state-of-the-art ultrasound equipment. Impairment of left ventricular relaxation is known to reduce the rate of early diastolic filling (2) and could potentially reverse the early diastolic to atrial systolic filling ratio. In contrast, decreasing compliance or increasing chamber stiffness is associated with enhanced early filling and decreased filling during atrial systole (3). This gains in importance in the evaluation of a variety of cardiovascular diseases, such as coronary artery disease, hypertension, and left ventricular hypertrophy (3,4).

Although cardiovascular pharmacologic interventions such as adrenergic amines are said to alter hemodynamics by enhancing systolic performance, for instance in chronic heart failure, this hypothesis fails to explain many of the responses seen after the administration of these drugs. The observation that abnormalities of diastolic function are an earlier and more sensitive marker of heart failure than are abnormalities of systolic function suggests that inotropic and vasodilating drugs may act primarily by improving diastolic function (5). These considerations lead to the question of whether Doppler-derived diastolic flow velocity patterns may be a sensitive clinical and pharmocologic variable in quantifying changes of left ventricular performance. In our study, changes in cardiac function were provoked by intravenous infusion of dose-titrated isoprenaline, which stimulates cardiac β1- and β2-adrenoceptors directly (6) and is known to result in positive inotropy, chronotropy, and vasodilation (7).



The study population was composed of 10 healthy, nonsmoking male volunteers aged 23-31 years (mean age, 26.6 years) not taking any form of medication. All subjects were instructed extensively about the study and gave written informed consent. The study was approved by the local ethics committee. Results of PW-Doppler examinations were satisfactory in the subjects included in the study.

Study design

The volunteers were studied after an overnight fast and did not eat until completion of the day's protocol. All measurements were performed with the subjects lying in a relaxed, recumbent position. After we fixed the electrodes and introduced an indwelling venous catheter, the volunteers rested for 30 min, and two baseline recordings were taken 5 min apart. Thereafter in a randomized sequence, each volunteer received the first of two intravenous infusions, one containing active drug (isoprenaline) and one, placebo. The washout time between the two infusions was 60 min. PW-Doppler data were registered sequentially at each dose step, which was preceded by registrations of mechanical and electrical impedance cardiography, dual-beam Doppler echoaortography, and two-dimensional echocardiography. Methods and results of the latter four modalities were published elsewhere (7). The subjects, observers, and analysts were blinded to the treatment protocol.

Drug infusion

Isoprenaline (Isoprel; Wintrop, Geneva, Switzerland) was diluted with physiologic saline and 1 mg/ml vitamin C added as an antioxidant. Saline served as the placebo solution. The solutions were administered by using an automatic pump (Perfusor; Braun, Melsungen, Germany) at a speed of 10 ml/h minimum and up to a maximum of 300 ml/h. The respective isoprenaline doses were 0.1, 0.2, 0.4, 0.75, and 1.5 μg/min for a period of 15 min for each dose step. Cardiac function was measured during the last 10 min of each infusion step.

Heart rate and blood pressure

Heart rate (HR) was measured by lead II on the ECG recorded by the echo machine. Systolic and diastolic blood pressures (SBP and DPB) were measured by using an automatic oscillometric device (Dynamap; Critikon Inc., Tampa, FL, U.S.A.; DPB measured at Korotkoff-IV sounds).

Doppler echocardiography

A phased-array sector scanner (Toshiba SSH 140; Toshiba Medical Co., Tokyo, Japan) with a 3.5-MHz transducer was used for PW-Doppler ultrasound examination. Each subject was examined in supine position (lateral recumbent position not possible because of further instrumentation) (7) during unforced end expiration to provide a standard, including a stable signal unaffected by respiration. Mitral inflow velocity was recorded from the apical window [right anterior oblique (RAO) projection], where the cursor was positioned at the leaflet tips of the opened mitral valve. The greatest velocity of diastolic flow was obtained by visual and auditory guidance. Care was taken to obtain the smallest possible angle between the assumed blood-flow direction and the cursor at the sampling space. This angle was estimated to be zero or <20° in each subject. All signals were recorded on videotape.

Quantitative evaluation

To minimize intraobserver variability in the Doppler measurements, tracings of at least three cardiac cycles with the highest velocity of early diastolic filling were analyzed and averaged. Peak transmitral flow velocities were measured in meters/second at the darkest point of the spectral wave form (peak modal velocity). Both the peak flow velocity and flow velocity-time integral of early left ventricular filling (VmaxE and VTI E) and late (atrial) ventricular filling (VmaxA and VTI A), respectively, were measured. The ratios of the peak velocities (E/A Vmax) and flow velocity-time integrals (E/A VTI) were derived from these measurements. Further, the duration from onset to peak early diastolic velocity (acceleration time, Tacc = DE; Fig. 1), the duration from peak early diastolic velocity to the time when flow velocity returned to baseline (deceleration time, Tdec = DF; Fig. 1), and the diastolic filling time (FillT = DG, Fig. 1) were determined. In subjects in whom diastasis was not well defined between the early and late diastolic filling peaks, the onset of flow was defined as the first increase in flow velocity after the plateau or downslope of the E peak. Although flow velocities could be identified and quantified at all time points and in all subjects, measurements of velocity-time integrals and diastolic filling times could not always be judged to be reliable, especially at higher doses. In those instances, the missing values (up to three of 10) were replaced by model adjusted estimates derived from the analysis of variance (ANOVA; see the following). Quantitative measurement and analysis was performed offline by a computer workstation (EchoCom; Individual Software, Fulda, Germany).

FIG. 1
FIG. 1:
Transmitral Doppler velocity profile with its temporal relation to the ECG. E (VmaxE), early diastolic filling velocity; A (VmaxA), late (atrial) diastolic filling velocity; VTI E/VTI A, velocity time integral of early/late (atrial) filling; DE (Tacc), acceleration time; EF (Tdec), deceleration time; DG (FillT), diastolic filling time.

Statistical methods

The arithmetic changes of the variables from the preinfusion baseline values were fitted to a general linear model with effects for subject and treatment (ANOVA). This simplified model was choosen after exclusion of eventual period effects. The treatments were contrasted at each time point by the difference of the model adjusted means for isoprenaline minus those for placebo and the corresponding smallest 95% confidence intervals of the treatment differences. The contrasts are presented graphically by the time course of the model adjusted means for placebo and the point estimates of the mean treatment differences. The latter are flagged by the corresponding 95% confidence interval. The treatment differences reach statistical significance when the confidence interval flags do not intersect the zero line. The contrasts are presented mainly for descriptive purposes. No adjustment was made for multiplicity. The treatment effects at the highest dose step were furthermore analyzed by an analysis of covariance (ANCOVA), which included the relative changes of HR, SBP, and DBP as covariates (8).


The values obtained at baseline recordings before administration of isoprenaline and placebo are shown in Table 1. There were no significant differences between these values. A typical example of this series in one subject is illustrated in Fig. 2. In Figs. 3-6, differences with respect to placebo are depicted for the various titration doses of isoprenaline infusion. HR, SBP, and DBP were registered twice at each isoprenaline dose step, one at 5 min and one at 15 min, demonstrating the stability of the effects of isoprenaline at each dose level for the entire period.

Mean baseline values obtained before placebo and before isoprenaline administration
FIG. 2
FIG. 2:
Representative example of one healthy subject. Doppler profiles before infusion (pre drug 2) and at the last titration step (highest dose of isoprenaline: 1.5 μg/min) presented for isoprenaline infusion at the top and for placebo at the bottom. VmaxE, early diastolic filling velocity; VmaxA, late (atrial) diastolic filling velocity; HR, heart rate.
FIG. 3
FIG. 3:
Time course of the model adjusted mean (n = 10) changes from preinfusion baseline for heart rate (HR), systolic (SBP), and diastolic blood pressure (DBP) during the control infusion (broken line) and time course (relative to dose, at bottom) of the mean difference (+95% confidence interval) for isoprenaline (continuous line) relative to placebo. The isoprenaline effects reached statistical significance at alpha = 0.05 when the confidence interval did not intersect with the 0-line.
FIG. 4
FIG. 4:
Diastolic filling velocities: Time course of the model adjusted mean (n = 10) changes from preinfusion baseline for early diastolic filling velocity (VmaxE), late (atrial) diastolic filling velocity (VmaxA), and the ratio of VmaxE and VmaxA (E/A Vmax) during the control infusion (broken line) and time course (relative to dose, at bottom) of the mean difference (+95% confidence interval) for isoprenaline (continuous line) relative to placebo. The isoprenaline effects reached statistical significance at alpha = 0.05 when the confidence interval did not intersect with the 0-line.
FIG. 5
FIG. 5:
Diastolic velocity time integrals: Time course of the model adjusted mean (n = 10) changes from preinfusion baseline for early diastolic velocity time integral (VTI E), late (atrial) diastolic velocity time integral (VTI A), and the ratio of VTI E and VTI A (E/A VTI) during the control infusion (broken line) and time course (relative to dose, at bottom) of the mean difference (+95% confidence interval) for isoprenaline (continuous line) relative to placebo. The isoprenaline effects reached statistical significance at alpha = 0.05 when the confidence interval did not intersect with the 0-line.
FIG. 6
FIG. 6:
Diastolic time intervals: Time course of the model adjusted mean (n = 10) changes from preinfusion baseline for acceleration time (Tacc), deceleration time (Tdec), and diastolic filling time (FillT) during the control infusion (broken line) and time course (relative to dose, at bottom) of the mean difference (+95% confidence interval) for isoprenaline (continuous line) relative to placebo. The isoprenaline effects reached statistical significance at alpha = 0.05 when the confidence interval did not intersect with the 0-line.

Heart rate and blood pressure

According to the well-known effects of isoprenaline on the cardiovascular system, HR and SBP increased (Fig. 3) but reached statistical significance only at the relatively high dose of 0.75 μg/min. DBP decreased nonsignificantly (Fig. 3).

Diastolic filling velocities and velocity-time integrals

Isoprenaline induced a dose-dependent increase in VmaxE and VmaxA(Fig. 4). The changes were larger for VmaxE. These effects were related to infusion time in a linear fashion; their relation to dose thus was log-linear. The regression was dose independent for VmaxA, but its steepness increased at higher doses for VmaxE. Both indices proved highly sensitive to isoprenaline, as statistically significant changes were detectable at the lowest dose (0.1 μg/min). Similar results were observed with the VTI indices, but VTI A tended to be more variable and less reliable (Fig. 5). The ratios of the early and late filling parameters were not consistently altered.

Diastolic filling time intervals

Tacc increased beyond the second dose step (≥0.2 μg/min), but the changes rapidly reached a ceiling and permitted little distinction of the various dosage steps. The variability of these changes was relatively large, and the estimates thus lacked precision. Tdec was shortened significantly at the lowest dose (0.1 μg/min), and this effect increased in log-linear fashion on further increase of dose. All doses but the second (0.2 μg/min) shortened FillT (Fig. 6).


Our study investigated the influence of β-adrenergic stimulation induced by isoprenaline on Doppler measurements of left ventricular filling variables in young, normal men. The results indicate an adrenergically mediated enhancement of diastolic performance as reflected by an increase of VmaxE, VmaxA, VTI E, VTI A, and a shortening of Tdec and FillT. Although the sample was small, most of these changes reached statistical significance at the lowest dose of isoprenaline. These high sensitivities are noteworthy because so far the only fully established adrenergic cardiovascular effects have been positive inotropy or enhancement of systolic performance (6,7,9). Evidence of our data is further strengthened by comparing noninvasive estimates of systolic function evaluated at the same time in the same population published previously (7). End-systolic volume and ejection fraction derived by two-dimensional echocardiography were altered in inotropic stimulation only at a dose of ≥0.4 μg/min, which was 4 times the dose at which diastolic filling increased. Although dual-beam Doppler echocardiography and impedance cardiography do not reach the sensitivity of diastolic filling, systolic time intervals (preejection period and electromechanical systole) were comparable. However, confidence intervals were higher at every titration step, and the time curves of sequential measurements were not so steep (7). Thus stimulation of β-adrenergic receptors with low-dose isoprenaline seems to exert greater effects on myocardial relaxation than on contraction in healthy subjects. Recently similar findings were reported by other investigators (9); they are also in keeping with previous experimental studies of strips of guinea pig atria (10) and isolated cardiac muscle of the dog (11) in which a lusitropic effect occurred at concentrations that did not stimulate any inotropic effect. In intact hearts, β-adrenergic stimulation with isoprenaline also improved parameters of cardiac muscle relaxation (12).

The changes described here reflect an isoprenaline-induced enhancement and facilitation of diastolic filling occurring with an increase of HR and SBP and decrease of DBP plus overall inodilatory changes of systolic performance (7). Inclusion of the relatively large changes of HR and SBP as covariates in the ANOVA blunts the treatment effects on the majority of diastolic indices. At the highest dose of isoprenaline, statistical significance is lost for all variables except VmaxE (p value for treatment effect based on ANOVA: 0.0001 vs. 0.0087 based on the ANCOVA). In contrast, the impact of the change of DBP was relatively small, as Korotkoff-IV rather than Korotkoff-V auscultatory sounds were used to detect the diastolic pressure level (13); its effect in the model in consequence was negligible. It thus cannot be excluded that the observed changes are mechanistically interrelated. Although the impact of the covariates does not prove their biologic role or contribution, it possibly merely reflects their collaterality (i.e., an apparent association with or without mechanistic relevance). A study by Harrison et al. (14). demonstrated unchanged VmaxE values during incremental right atrial pacing in normal subjects. Udelson et al. (15) compared isoprenaline infusion with atrial pacing in a population of patients with known diastolic dysfunction caused by hypertrophic cardiomyopathy. At similar HR, only β-adrenergic stimulation with isoprenaline favorably modified diastolic performance associated with improved diastolic distensibility (i.e., an increase in peak negative dP/dt (first derivative of change in LV systolic pressure over time) as well as a decrease in the time constant T1/2 and time to peak negative dP/dt). The active process of relaxation might be the result of a β1-mediated increase in intracellular levels of cyclic adenosine 3′,5′-monophosphate, which accelerates the rate of calcium reuptake by the sarcoplasmic reticulum and thus stimulates the myocardial deactivation process (16), thereby speeding the detachment of crossbridges (17). There may be a decrease of the sensitivity of troponin to calcium, which may to some extent explain the greater effect on relaxation than on inotropy. These mechanisms have to be an intrinsic effect of adrenergic stimulation, because only isoprenaline and not CaCl2 alters relaxation (18). Another mechanism at higher doses of isoprenaline working in our setting might be attributed to experimental findings described by Housmans et al. (19). Ventricles fill themselves by sucking blood from the arterial reservoir, an action caused by elastic recoil, which increases once end-systolic volumes start to decrease.

Clinical pharmacologic investigations require reliable noninvasive methods to describe changes in systolic and diastolic cardiovascular performance induced by drugs. For systolic function, the systolic time intervals have been shown to be sensitive and valid (20). For diastolic function, the simple noninvasive technique of transmittal PW-Doppler echocardiography, with its high sensitivity, is a very suitable method of doing this. As detectable alterations of diastolic performance temporarily precede those of systolic function, Doppler variables might be superior to those investigational methods that measure only systolic performance. Although reports of other adrenergic drugs, such as dobutamine (21), epinephrine (22), and xamoterol (23) exist, caution is required in the assessment of the results with regard to the sensitivity of the method to exclude adrenergic effects or detect a blunting rather than an enhancement of cardiac performance. Both the positive inotropic effects of isoprenaline and the negative inotropic effects of antiarrhythmic substances such as chinidine could be measured reliably by Doppler echocardiography of aortic flow pattern (24). The opposite effects on diastolic velocity pattern were reported in the comparison of methoxamine and isoprenaline infusions (25). Further prospective studies would be required to prove that this is a sensitive method of detecting changes in both directions.

Methodologic considerations and study limitations

Diastolic filling is a complex phenomenon that is influenced by multiple factors. Our study was limited to measures of transmitral filling velocities, velocity-time integrals, and filling-time intervals. These noninvasive variables are "pattern recognition" indices that cannot be precisely related to invasively derived measurements (26). No intracardiac pressures, including left atrial pressure by transseptal puncture, were measured in our study because we found it unacceptable to conduct such invasive studies in a healthy study group. However, our data show adrenergically mediated, dose-dependent differences with respect to baseline values and to double-blinded placebo values and independence of Vmax-E from HR and BP (ANCOVA) in the short term, which makes them relatively independent of hemodynamically derived indices. In terms of measurement reproducibility, Doppler indices of diastolic function have been shown to be excellent, especially the peak and time-velocity integrals (27). The higher variability of acceleration and deceleration parameters reported by others (27) could be confirmed by lower significances and higher variations in our data. The uncertainty in selecting the exact onset and termination of diastolic flow because of the width of the Doppler spectral tracing can introduce some error. Only volunteers of a homogeneous age group were investigated; however, an age-associated decline in HR, SBP, and DBP, ejection fraction, and cardiac output is known to be a response to β-adrenergic stimulation with isoprenaline in healthy men (28). Postural changes were avoided 30 min before and during the whole investigation, as Suzuki et al. (29) demonstrated significant decreases in peak filling rate when normal subjects simply moved from the supine to the upright position. The location of the sampling site may influence the absolute measures (30); therefore we used a constant location at the leaflet tips without any major angle to blood-flow direction.

Acknowledgment: The study was supported by a grant from the Zentrum für Kardiovaskuläre Pharmakologie, ZeKaPha GmbH, Mainz, Germany, and by Toshiba Medical Systems GmbH, Neuss, Germany.


1. Rokey R, Kuo LC, Zoglubi WA, Limacher MC, Quinones MA. Determination of parameters of left ventricular diastolic filling with pulsed Doppler echocardiography: comparison with cineangiography. Circulation 1985;71:543-50.
2. Fioretti P, Brower RW, Muster GT, Serruys PW. Interaction of left ventricular relaxation and filling during early diastole in human subjects. Am J Cardiol 1980;46:197-203.
3. Stoddard MF, Pearson AC, Kern MJ, Ratcliff J, Mrosek DG, Labovitz AJ. Left ventricular diastolic function: comparison of pulsed Doppler echocardiographic and hemodynamic indexes in subjects with and without coronary artery disease. J Am Coll Cardiol 1989;13:327-36.
4. Gardin JM, Drager JI, Rohon MK. Doppler evaluation of left ventricular filling in mild and severe hypertension. J Am Coll Cardiol 1986;7:185-92.
5. Packer M. Diastolic function as a target of therapeutic interventions in chronic heart failure. Eur Heart J 1990;11(suppl C):35-40.
6. Brodde DE. β1 and β2 Adrenoceptors in the human heart: properties, function, and alterations in chronic heart failure. Pharmacol Rev 1991;43:203-42.
7. de Mey C, Belz GG, Nixdorf U, et al. Relative sensitvity of four noninvasive methods in assessing systolic cardiovascular effects of isoproterenol in healthy volunteers. Clin Pharmacol Ther 1992;52:609-19.
8. Evans SJ, Mills P, Davison J. The end of the p value? Br Heart J 1988;60:177-80.
9. Vlarkson PBM, Wheeldon NM, Macleod C, MacLennan M, Mac-Donald T. Systolic and diastolic effects of beta-adrenergic stimulation in normal humans. Am J Cardiol 1995;75:206-9.
10. Kenakin TP, Ambrose JR, Irving PK. The relative efficiency of β-adrenoceptor coupling to myocardial inotropy and diastolic relaxation: organ-selective treatment for diastolic dysfunction. J Pharmacol Exp Ther 1991;257:1189-97.
11. Morad M, Rolett EL. Relaxing effects of catecholamines on mammalian hearts. J Physiol 1972;224:537-58.
12. Karliner JS, LeWinter M, Makler F, Engler R, O'Rourke RA. Pharmacologic and hemodynamic influence on the rate of isovolumic left ventricular relaxation in the normal conscious dog. J Clin Invest 1977;60:511-21.
13. De Mey C. Method specificity of the auscultatory estimates of the inodilatory reduction of diastolic blood pressure based on Korotkoff-IV and -V criteria. Br J Clin Pharmacol 1995;39:485-90.
14. Harrison MR, Clifton D, Penell AT, DeMaria AN, Cater A. Effect of heart rate on left ventricular diastolic transmitral flow velocity pattern assessed by Doppler echocardiography in normal subjects. Am J Cardiol 1991;67:622-7.
15. Udelson JE, Cannon RO III, Bacharach SL, Rumble TF, Bonow RD. β-Adrenergic stimulation with isoproterenol enhances left ventricular diastolic performance in hypertrophic cardiomyopathy despite potentiation of myocardial ischemia: comparison to rapid atrial pacing. Circulation 1989;79:371-82.
16. Colan SD, Borow KM, Neumann A. Effects of loading conditions and contractile state (methoxamine and dobutamine) on left ventricular early diastolic function in normal subjects. Am J Cardiol 1985;55:790-6.
17. Katz AM, Bailin G, Kirchberger MA, Tada M. Regulation of myocardial cell function by agents that increase cyclic AMP production in the heart. In: Fishman AP, ed. Heart failure. Washington: Hemisphere Publishing, 1978:11-28.
18. Leone BJ, Lehor J-J, Drrington KL, Foex P. Left ventricular regional function and relaxation: effects of inotropic stimulation. J Cardiothor Vasc Anesth 1992;6:578-85.
19. Housmans PR, Goethals MA, Paulus WJ, Brutsaert DL. Comments on pressure-diameter relations during early diastole in dogs: incomparability with the concept of passive left ventricular filling and negative diastolic pressure in the intact canine right ventricle: evidence of diastolic suction. Circ Res 1982;50:443-4.
20. Beiz GG. Systolic time intervals: a method to assess cardiovascular drug effects in humans. Eur J Clin Invest 1995;46(suppl I):35-41.
21. El-Said E-SM, Roelandt JRTC, Fioretti PM, et al. Abnormal left ventricular early diastolic filling during dobutamine stress Doppler echocardiography is a sensitive indicator of significant coronary artery disease. J Am Coll Cardiol 1994;24:1618-24.
22. Johannessen K-A, Cerqueira M, Veith RC, Stratton JR. Influence of sympathetic stimulation and parasympathetic withdrawal on Doppler echocardiographic left ventricular diastolic filling velocities in young normal subjects. Am J Cardiol 1991;67:520-6.
23. Pouleur H, Rousseau MF, Hasset C, Marlow HF, Charlier AA. Left ventricular sensitivity to β-adrenoceptor-stimulating drugs in patients with ischemic heart disease and varying degrees of ventricular dysfunction. Circ Res 1987;61(suppl I):1-91-5.
24. Menzel T, Mohr-Kahaly S, Epperlein S, et al. Determination of systolic left ventricular time intervals using the Doppler-technique: echocardiographic analysis of aortic flow pattern. Drug Res 1996;46:228-34.
25. Meyer TE, Casadei B, Pipilis A, et al. The interaction between early and atrial filling of the left ventricle: effects of heart rate, atrioventricular delay, isoproterenol and methoxamine. Am J Noninvas Cardiol 1993;7:195-203.
26. Plotnick GD, Vogel RA. Noninvasive evaluation of diastolic function: need for hemodynamically and clinically relevant variables. J Am Coll Cardiol 1989;13:1015-6.
27. Galderisi M, Benjamin EJ, Evans JC, et al. Intra- and interobserver reproducibility of Doppler-assessed indexes of left ventricular diastolic function in a population-based study (the Framingham Heart Study). Am J Cardiol 1992;70:1341-6.
28. Stratton JR, Cerqueira MD, Schwartz RS, et al. Differences in cardiovascular responses to isoproterenol in relation to age and exercise training in healthy men. Circulation 1992,86:504-12.
29. Suzuki T, Sato K, Aoki K. Influence of postural change on transmittal flow velocity profile assessed by pulsed Doppler echocar-diography in normal individuals and in patients with myocardial infarction. Am Heart J 1990;120:110-5.
30. Dittrich HC, Blauchard DG, Wheeler KA, McCann HA, Donaghey LB. Influence of Doppler sample volume location on the assessment of changes in mitral inflow velocity profiles. J Am Soc Echocardiogr 1990;3:12-20.

β-Adrenergic stimulation; Isoprenaline; Normal subjects; Doppler echocardiography; Left ventricular diastolic performance

© Lippincott-Raven Publishers