| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Division of Cardiology, Johns Hopkins University School of Medicine; and 2Laboratory of Cardiovascular Science, Gerontology Research Center, National Institute on Aging, Baltimore, Maryland 21224
Submitted 28 December 2002 ; accepted in final form 23 September 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
left ventricular deformation; torsion; untwist; noninvasive imaging; left atrial pressure
It has been presumed that these age-associated changes in filling dynamics result primarily from a slowing of the rate of LV pressure decline, manifested in the intact heart by a prolongation of the time constant of isovolumic pressure fall (
) (42). Prolonging would lengthen IVRT and reduce early diastolic suction (41). Reduced left atrial (LA) emptying in early ventricular diastole would result in greater atrial stretch in later diastole, advancing the atrium along its Starling curve, and causing the augmented A wave. Indeed, the filling pattern in the elderly is generally referred to as that of "slowed relaxation" (42).
However, a recent catheterization study (44) in humans failed to detect an age-related decline in the rate of LV pressure fall, suggesting that other factors may contribute to the altered filling pattern seen with aging. Downes et al. (6) implicated increased LV stiffness as the major factor. However, whereas an isolated increase in LV stiffness would reduce the E wave (41), this is generally associated with a rise in LA pressure, which has the opposite effect on E wave velocity (20, 38). Moreover, increasing LV stiffness should result in a more rapid rise in ventricular pressure as filling occurs, more quickly diminishing the atrioventricular gradient and thereby shortening (29), rather than lengthening, DT.
Computer model (32, 39, 41), animal (19), and clinical (27) studies have all shown that E is determined by the atrioventricular (AV) pressure gradient rather than simply the rate of LV pressure decline and that the LA pressure is an important component of that gradient. Because the mitral valve opens during the slow phase of LV pressure decline, a small change in LA pressure could also produce a significant prolongation of IVRT. Therefore, we used noninvasive methods to study the relative influence of the rate of LV pressure decline and LA pressure on early diastolic filling in normal subjects encompassing a broad age range.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Volunteers for this study represented a cross-sectional sample from the Baltimore Longitudinal Study of Aging. All were healthy, ambulatory, and community dwelling and were screened for coronary disease by the absence of significant resting Q waves and ST depression on ECG, either at rest or during maximal treadmill exercise. Subjects were excluded if seated blood pressure (BP) exceeded 160/90 mmHg. There were 63 women and 59 men aged 54 ± 20 yr, ranging from 21 to 92 yr. The study was approved by the Johns Hopkins Bayview Medical Center Institutional Review Board.
Assessment of LV Filling
All 122 subjects underwent Doppler echocardiography to measure E (in cm/s), velocity time integral (VTI; in cm), E wave DT (in ms), A wave velocity (in cm/s), and IVRT (in ms).
Assessment of LV Pressure Decline
Each subject also underwent cardiac MRI and/or color M-mode echocardiography. Pressure decline was quantified in each subject using at least one of the following two load-insensitive techniques.
Recoil rate by tagged MRI (n = 73 subjects). LV torsion, the counterclockwise twist of the LV apex with respect to the base, develops during ejection but recoils largely during isovolumic relaxation (4, 47) before mitral valve opening. Recoil rate during isovolumic relaxation reflects the rate of pressure decay and thus correlates with peak negative dP/dt (47) and with (5) without an influence of LA pressure (5). Torsion and recoil can be measured using tagged MRI, a method for noninvasive tracking of motion of specific myocardial points (2, 48).
Propagation rate of the wave of LV filling by color M-mode echocardiography (n = 63 subjects). This wave velocity reflects intraventricular base-to-apex pressure gradients at the time of mitral valve opening and has been shown to be a preload-insensitive measure of LV pressure decline (12). Although the mitral annular velocity was measured, this was not used as an index of pressure decline in this study, because it is preload dependent when relaxation is normal (8, 21).
Assessment of LA Pressure
In subjects undergoing color M-mode echocardiography, LA pressure was indexed using four validated methods. First, the ratio of E wave height to the propagation rate of the wave of LV filling (Vp) was calculated. This parameter correlates directly with LA pressure (11, 33), even in normal subjects (9). Second, pulmonary vein flow was measured, and the ratio of systolic to total flow velocity was calculated. This parameter correlates closely and inversely with LA pressure (1, 15, 36). Third, the DT of pulmonary vein diastolic flow was measured. This variable correlates inversely with LA pressure, even more closely than does the wedge pressure (23). Finally, we used tissue Doppler echocardiography to determine the myocardial E wave (Em). E/Em correlates directly with LA pressure in patients with abnormal ventricles but inversely in subjects with normal ventricular function and relaxation (8, 21). Therefore, a lower LA pressure would be characterized by reduced E/Vp and increased pulmonary vein systolic fraction, pulmonary vein diastolic DT, and E/Em.
MRI Methods
Myocardial tagging was used in these experiments to measure torsion and its rate of recoil (48). Tags are noninvasive markers that are imprinted on the myocardium by selective radiofrequency saturation of planes perpendicular to the imaging planes; they change the magnetization of the protons in the tagged plane compared with the neighboring nontagged regions, resulting in a difference in signal intensity. When placed at end diastole and then imaged throughout the cardiac cycle, tags reveal the deformation and displacement of the myocardium on which they were placed. Tags may be positioned in a radial pattern (48), which is ideal for the measurement of torsion, or in a grid pattern (2), which is commonly used for calculation of a full strain field.
Tagged MRI was performed on a Siemens 1.5-T Vision System or Resonex 0.38-T Rx 4000. Briefly, the image acquisition included a cine gradient echo sequence with radial tissue tagging, with a temporal resolution of 20 ms. Experiments performed on the Siemens Imager used breathold-segmented k-space sequences and view sharing. After serial scout images, three short-axis images, positioned at the base, mid-LV, and apex, each with eight radial tags, were acquired. Tagging and imaging were triggered by the R wave of the ECG: tags were placed at end diastole, and a series of images was acquired thereafter. Image processing and data analysis were performed using a special software package (Cardiology Image Processing System, Johns Hopkins University). The tag-endocardium intersection points on the basal and apical short-axis images were identified manually and digitized (8 points/slice). Angular rotation of the base and apex was calculated using the average rotation of the eight individual endocardial tag intersections. Torsion at each time in the cardiac cycle was calculated as the difference between the clockwise rotation at the base and the counterclockwise rotation at the apex, yielding a measure of the total base-to-apex deformation. Recoil, the directional reversal of systolic torsion during diastole, was expressed as a percentage of maximum systolic torsion: Recoil = 100 x Tort/Tormax, where Tort is torsion at time t and Tormax is the maximum systolic torsion. The recoil rate was defined as the slope of the linear regression of recoil versus time during the first 100 ms after peak torsion (i.e., using data from 6 images). Calculated in this way, the recoil rate is normalized for peak torsion and is thus expressed as "percent recoil" per millisecond.
Echocardiography
Subjects were placed in the left lateral decubitus position, and standard echocardiographic views were recorded using an Agilent Technologies Sonos 5500 or a Hewlett-Packard Sonos 2500 system and measured using the Sonos echo machines and an off-line Microsonics Image-Vue analysis system. Transmitral Doppler flows were recorded at the leaflet tips. Tracings for IVRT were recorded using continuous wave Doppler, and IVRT was measured as the interval between the aortic valve closure click and the onset of transmitral flow. The duration of systole was measured between the Q wave onset and aortic valve closure. Color M-mode imaging was derived from the four-chamber view, and Vp was defined as the slope of the first aliasing flow, measured between the mitral valve and apex, over a distance of 4 cm. Pulmonary vein flow was also measured from the four-chamber view.
Statistical Methods
First, simple linear regression analysis (Statview, SAS Institute, Cary, NC) of diastolic parameters against age was performed. Second, the factors independently associated with aspects of the LV filling pattern known to accompany aging (E, IVRT, and A) were assessed through forward stepwise regression using models that incorporated the potentially important physiological determinants. An index of LA pressure (the pulmonary vein systolic-to-total flow velocity ratio) was then added to these models, and the effects of that addition to the determinants of the filling pattern were assessed. Data are expressed as means ± SD. A two-tailed P value of <0.05 was required for statistical significance.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The sample included 63 women and 59 men aged 54 ± 20 yr, ranging from 21 to 92 yr, with a median age of 56.3 yr for women and 52.2 yr for men.
LV Filling Patterns
Results are detailed in Table 1 and selected regression plots are shown in Fig. 1. Linear regression analysis demonstrated the expected reduction in peak E and E/A and an increase in peak A, DT, and IVRT with age. There was an increase in the duration of systole (Q-S2) with advancing age. There was no relation between age and resting heart rate.
|
|
LV Pressure Decline
LV pressure decline, assessed both by color M-mode (Vp) and tagged MRI (recoil rate), did not correlate with age.
LA Pressure
The systolic fraction of pulmonary vein flow, DT of the pulmonary vein diastolic flow, and E/Em increased with age, all suggesting reduced early diastolic LA pressure. In addition, E/Vp was reduced with advancing age, also indicating reduced early diastolic LA pressure.
Multivariate Models
E as the dependent variable. In a model including age, Vp, heart rate, and end-systolic BP (ESBP; estimated as diastolic pressure plus two-thirds of the pulse pressure) as independent variables to predict peak E velocity, 15% of the variance of E was attributable to age and an additional 13% was explained by the rate of pressure decline (Table 2). There was no independent relation of E to heart rate or ESBP. However, when an index of LA pressure, the pulmonary vein systolic fraction, was added to the model, age became nonsignificant; 15% of the variance was then explained by the rate of pressure decline and 19% by LA pressure. This is consistent with the notion that age exerts its effects on the E wave through age changes in LA pressure.
|
IVRT as the dependent variable. In a similar model, including age, Vp, heart rate, and ESBP as independent variables to predict IVRT, 31% of the variance of IVRT was attributable to age and an additional 11% was explained by heart rate (Table 3). There was no independent relation of IVRT to LV pressure decline or ESBP. However, when an index of LA pressure, the pulmonary vein systolic fraction, was added to the model, age became nonsignificant; 53% of the variance was explained by LA pressure. Again, this is consistent with the notion that age exerts its effects on IVRT through LA pressure.
|
A as the dependent variable. In a similar model, 45% of the variance in peak A velocity was determined by age and 8% by the rate of pressure decline (Table 4). The importance of age was not reduced by including an index of LA pressure (pulmonary vein systolic fraction) in the model.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Prior Studies
The present findings run counter to the generally accepted notion, derived largely from isolated papillary muscle preparations, that an impaired rate of LV pressure decline is the primary age-associated change in LV diastolic function. However, scant data supporting an important age dependence of pressure decline are available in normal humans. Although Hirota et al. (14) showed a direct age relationship of in 18 patients, that group included athletes and patients with murmurs, edema, pulmonary stenosis, and dextrocardia, and the age relationship may have reflected changes due to disease or lifestyle. Furthermore, the inclusion of five children in the that study may have skewed the results such that they did not properly reflect trends in adult aging (44).
A number of recent observations obtained using invasive hemodynamic methods cast doubt on the notion that an alteration in pressure decline fully explains the profound changes in filling pattern due to aging in humans. First, Yamakado et al. (44) reported measures of by micromanometer in 55 normal subjects aged 20–77 yr; was calculated by multiple methods and was not related to age. The patient group studied appeared to represent a typical adult population, demonstrating a small but significant rise in systolic BP with advancing age.
Second, Lenihan et al. (26) studied the effect of load and contractility alteration on the early diastolic AV gradient and on LV filling in humans with diastolic dysfunction. Changes in the AV gradient were determined primarily by loading conditions and were not importantly influenced by active relaxation. For example, dobutamine reduced early filling velocities despite a marked augmentation of relaxation, because small declines in LA pressure dwarfed the importance of the rate of the LV pressure fall.
Third, Fagard et al. (7) measured pulmonary capillary wedge pressure in a group of 110 untreated mild hypertensives and divided the group into those aged <30, 30–44, and 
45 yr. There were no differences in systolic BP among the groups. The wedge pressure was 6.7 mmHg in the young subjects, 5.6 mmHg in the middle-aged subjects, and 5.3 mmHg in the older subjects (P < 0.05). Although this difference seems small, the impedence of the normal mitral apparatus is very low. Therefore, small changes in the AV pressure gradient produce large changes in flow patterns (44). Similar finding were reported by Granath et al. (13).
Finally, Hoit et al. (16) determined that large changes in peak filling rates occur with small changes in LA pressure at low, but not at high, LV end-diastolic pressure. That is, early filling is most dependent on preload in subjects with normal LV function. Our results, obtained in normal subjects using noninvasive methods, are concordant with the above conclusions.
Possible Mechanisms
The echocardiographic indexes that suggest reduced LA pressure in older versus younger adults are all measured during rapid LV filling, probably reflecting pressure in early, rather than late, diastole. Therefore, these findings do not imply that mean LA pressure is necessarily lower with advanced age; LA pressure might start at a lower baseline after mitral valve opening and rise more steeply throughout diastole. We found no relationship of age to an index of late diastolic LA pressure, the difference between the duration of the mitral and pulmonary vein A wave. Several hypotheses may be advanced to explain these findings.
LA passive compliance might be altered with advancing age.
Tissue properties of the atrial wall determine the LA pressure at any given volume (17, 18). Morphologically, the fibrous endocardial layer of the atrial wall is known to more than double in thickness and to increase in density between the third and ninth decades (30). At advanced ages, this connective tissue layer occupies nearly a third of the total wall thickness (30) and is likely to influence its physical properties. Likewise, aging is associated with a progressive fragmentation and rupture of elastic fibers, which contribute to tissue changes in the arterial system, skin, and other organs, and might also influence atrial properties. We hypothesize that the aging LA becomes both flaccid due to loss of elasticity and stiff due to endocardial fibrosis. At low volumes, the LA pressure might be low due to flaccidity, but at high volumes the pressure might be high due to increased stiffness. These changes in the LA wall might be considered to be analogous to those of aging skin: wrinkled and redundant, with reduced tone, causing distensibility that is high upon initial stretch but then low with continued stretch (34). A primary enlargement of the LA might contribute to this effect by leaving the LA wall largely unstressed during its reservoir phase. This hypothesis would explain both the reduced peak E velocity in normal older adults and also the increased propensity to "pseudonormalize," that is, to develop congestive heart failure, in elderly patients with volume overload.
Atrial contractility might be increased with advancing age. If increasing atrial contractility were the primary change, perhaps due to LA hypertrophy, effective atrial emptying to a lower post-A wave volume might leave the atrium underfilled and at lower pressure during the next rapid filling phase. In this construct, atrial compliance would not change, but the LA would operate on the lower portion of its Starling curve. Multiple studies have shown that the extent of LA ejection does increase with advancing age (28, 37). However, the atria are subject to the Frank-Starling mechanism (31), and atrial ejection is generally assumed to increase on the basis of augmented preload rather than true contractility in the settings of aging and hypertension (43).
Mild hypertension causes changes in diastolic filling, adrenergic responsiveness, and even gene expression (25) that are almost identical to those seen in aging. Therefore, previous studies of mild hypertension are relevant to this hypothesis. Matsuzaki et al. (31) recorded pressure-dimension loops in normal subjects and patients with healed myocardial infarction (MI) and hypertension. Both patients with hypertension and MI performed more atrial work than normals, but the relation between atrial shortening and tension was not different between normals and patients with MI. However, the relation was shifted upward and rightward for the hypertensives, suggesting an enhanced atrial inotropic state. A primary atrial hypercontractility in mild hypertension was also suggested by Phillips et al. (34) based on the fact that the A wave increases occurred early, before E wave decreases were seen. Similar mechanisms may be operative in normal aging.
Limitations
The noninvasive methods utilized here may lack the sensitivity to detect small changes in the LV pressure decline. However, the decline in peak E wave velocity and increase in IVRT are 
40–50% between the ages of 20 and 90 yr. If this difference were solely due to changes in , which follows an exponential course, this would require about a 25% decline in LV pressure decline, which should have been detectible using our methods (5) or the invasive approach used by Yamakado et al. (44). Further studies utilizing simultaneous LA and LV high-fidelity pressure recordings and pressure-volume loops would be useful but are difficult to obtain in normal subjects. In addition, caution should be applied in the interpretation of multiple regression analyses, especially when applied to small data sets such as this.
Implications
In a recent commentary summarizing the current concepts of diastolic function, Yellin (45) concludes by lamenting that "it is unfortunate that the role of left atrial properties has not received the attention required, considering that it is more than 10 years since Ishida et al. (19) demonstrated the importance of LA pressure in creating the early filling wave." Indeed, the present data suggest that the marked decline in early diastolic filling known to occur with age may be more dependent on changes in LA characteristics than on changes in the rate of LV pressure decline. Active and passive properties of the LA should be studied further, probably by invasive measurement of pressure-volume relationships. Drugs that affect tissue compliance, such as angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or advanced glycation end-product cross-link breakers (22), might influence these properties.
|
ACKNOWLEDGMENTS |
|---|
This study was presented in part at the 49th Annual Scientific Session of the American College of Cardiology, Anaheim, CA, in March 2000.
GRANTS
This study was supported by National Heart, Lung, and Blood Institute Grant RO1 HL-46223.
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Am J Physiol Heart Circ Physiol 281: H2002–H2009, 2001.This article has been cited by other articles:
|
|
 |
|
  J. Wang, D. S. Khoury, Y. Yue, G. Torre-Amione, and S. F. Nagueh Left Ventricular Untwisting Rate by Speckle Tracking Echocardiography Circulation, November 27, 2007; 116(22): 2580 - 2586. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Melenovsky, B. A. Borlaug, B. Rosen, I. Hay, L. Ferruci, C. H. Morell, E. G. Lakatta, S. S. Najjar, and D. A. Kass Cardiovascular Features of Heart Failure With Preserved Ejection Fraction Versus Nonfailing Hypertensive Left Ventricular Hypertrophy in the Urban Baltimore Community: The Role of Atrial Remodeling/Dysfunction J. Am. Coll. Cardiol., January 16, 2007; 49(2): 198 - 207. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Foster and K. E. Lease New Untwist on Diastole: What Goes Around Comes Back Circulation, May 30, 2006; 113(21): 2477 - 2479. [Full Text] [PDF]
|
|
|
|
|
|
Y. Notomi, G. Srinath, T. Shiota, M. G. Martin-Miklovic, L. Beachler, K. Howell, S. J. Oryszak, D. G. Deserranno, A. D. Freed, N. L. Greenberg, et al. Maturational and Adaptive Modulation of Left Ventricular Torsional Biomechanics: Doppler Tissue Imaging Observation From Infancy to Adulthood Circulation, May 30, 2006; 113(21): 2534 - 2541. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. de Simone, D. W. Kitzman, M. Chinali, A. Oberman, P. N. Hopkins, D. C. Rao, D. K. Arnett, and R. B. Devereux Left ventricular concentric geometry is associated with impaired relaxation in hypertension: the HyperGEN study Eur. Heart J., May 2, 2005; 26(10): 1039 - 1045. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. E. Weyman The year in echocardiography J. Am. Coll. Cardiol., February 1, 2005; 45(3): 448 - 455. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
