|
|
||||||||||
|
J Am Coll Cardiol, 2002; 39:1524-1530 © 2002 by the American College of Cardiology Foundation |
* Laboratories for Cardiovascular Research, Hebrew Rehabilitation Center For Aged Research and Training Institute, and Harvard Medical School Division on Aging, Boston, MassachusettsUSA
Manuscript received May 31, 2001; revised manuscript received January 28, 2002, accepted February 6, 2002.
* Reprint requests and correspondence: J. Andrew Taylor, PhD, Director, Laboratories for Cardiovascular Research, HRCA Research and Training Institute, 1200 Centre Street, Boston, Massachusetts, USA 02131.
ataylor{at}mail.hrca.harvard.edu
| Abstract |
|---|
|
|
|---|
BACKGROUND: Increased CV risk is associated with lower carotid distensibility, impaired baroreflex function and reduced respiratory sinus arrhythmia (RSA), possibly representing a direct path between structural and neural CV deficits.
METHODS: We estimated the mechanostructural and neural components of baroreflex function and examined their relation to RSA in young (20 to 31 years) and older (59 to 71 years) subjects rigorously screened for CV and autonomic diseases.
RESULTS: In the older subjects, RSA was < 20% of that in the younger subjects. Moreover, mechanical transduction of pressure into barosensory vessel stretch was
40% lower (p < 0.05) and arterial baroreflex gain more than 60% lower (p < 0.05) in the older group. Although neural transduction of stretch into vagal outflow only tended to be less (p < 0.08), it was an important determinant of baroreflex function. A path analysis model showed comparable contributions of both the mechanical and neural components to baroreflex gain; however, lower overall baroreflex gain in the older group did not relate to lower RSA.
CONCLUSIONS: These data suggest that decreased carotid distensibility does reduce baroreflex function with age, but this does not lead to reduced resting vagal outflow.
| ||||||||
Decreased carotid distensibility may relate to CV risk by blunting arterial baroreflex function and thereby reducing cardiac vagal tone. Afferent nerves embedded in the carotid walls are stretch sensitive (11), and they determine baroreflex responses that buffer arterial pressure changes (12,13). Decreased distensibility would effectively desensitize these receptors and reduce baroreflex-mediated autonomic outflow in response to pressure changes. Decreased baroreflex sensitivity could lead directly to lower vagal tone as the arterial baroreflex may be a primary determinant of resting cardiac vagal outflow. Pharmacologically determined vagal tone is highly correlated to baroreflex sensitivity (13), and, in fact, RSA may derive directly from arterial baroreflex stimulation through waxing and waning arterial blood pressure at the respiratory cycle (14,15). Thus, increased CV risk associated with lower carotid distensibility (16), arterial baroreflex function (17), and RSA (18) may represent a direct path between structural and neural CV deficits.
We hypothesized that the age-related decline in carotid distensibility is related to decreased resting cardiac vagal tone through blunting of the arterial baroreflex. We used our recently developed technique (19) to estimate mechanostructural and neural afferentefferent components of arterial baroreflex function. If reduced vagal tone is related to decreased distensibility then there should be a direct path from the structural baroreflex component to RSA through arterial baroreflex gain. Therefore, we examined the relation of RSA to arterial baroreflex gain and its neural and structural components in both young and older individuals.
| Methods |
|---|
|
|
|---|
|
15 min recovery between trials. Respiration-related RR-interval fluctuations were used to assess basal cardiac vagal outflow. Phasic vagal modulation with respiration results in RSA, an oscillation proportional to the mean level of cardiac vagal outflow (20). The ECG was recorded during 5 min of paced frequency breathing (15 breaths/min). Paced breathing avoids breathing frequency changes, which influence RSA magnitude and alter its relation to vagal outflow (13).
After paced breathing, baroreflex testing was conducted using our recently developed approach (19). This technique provides robust linear gain estimates for arterial baroreflex gain and its mechanical and neural components. Arterial pressure is reduced below threshold and raised through linear and saturation regions, providing clear determination of linear relations. Moreover, acquisition of B-mode ultrasound common carotid images allows determination of carotid diameters on a near beat-by-beat basis. Briefly, we acquired the ECG, beat-by-beat arterial pressures (Finapres) and longitudinal common carotid artery images (Hewlett-Packard 7.5 MHz [Palo Alto, California]; Data Translations DT3152 Frame Grabber; Information Integrity CVI Acquisition software, Stowe, Massachusetts) immediately before and for 2 min after sequential boli of 100 µg nitroprusside followed in 60 s by 150 µg phenylephrine. This allowed estimation of mechanical transduction of pressure into barosensory vessel stretch (diameter/pressure), neural transduction of stretch into vagal outflow (RR/diameter) and conventional integrated cardiovagal baroreflex gain (RR/pressure).
Analysis. All waveforms were stored to computer for signal-processing and subsequent analysis (Dataq Instruments WINDAQ software [Akron, Ohio]; DSP Development DADiSP software [Newton, Massachusetts]). Digitization at 500 Hz allowed accurate measurement of RR-interval to the nearest 2 ms. The RR-intervals were derived from time difference between successive R-wave peaks. Systolic and diastolic pressures were derived from maxima and minima of the pressure waveform.
Common carotid internal diameters were determined from digitized images by custom software (Information Integrity CVI Analysis software). Several points in proximity to near and far wall edges were selected, and the program fit an interpolated spline to each set of preliminary edge points and determined the best edge point from the interpolated line. Near and far wall edge points were modeled as parabolas constrained to have the same curvature, and diameter was estimated from the distance between parabolas. User-selected preliminary edge points were reused for successive images to reduce variance in the diameter time series.
Respiratory sinus arrhythmia was quantitated from power spectral analysis of each 300 s RR-interval time series during paced breathing. The time series was interpolated to 4 Hz; based on the Welch algorithm (21), seven overlapping periodograms were averaged to produce the spectrum estimate for the entire time series. Power within the respiratory frequency band, defined as 0.2 to 0.3 Hz, was summed to estimate RSA.
Arterial baroreflex gain and neural and mechanical components were derived from the associations among RR-interval, systolic pressure and systolic carotid diameter during the drug-induced rise in arterial pressure. Beat-by-beat values for each parameter were averaged across 3 mm Hg pressure increments to account for respiration-related variations and increase confidence in the relations among variables. Three linear relations were extracted from the sigmoid relations between: systolic carotid diameter and systolic pressure, representing the mechanical baroreflex component; RR-interval and systolic pressure, representing the integrated baroflex gain; RR-interval and carotid systolic diameter, representing the neural baroreflex component. Each linear relation was independently determined; that is, saturation and threshold regions were excluded from each relation without regard to where these regions might lie in the other two. (In addition, it should be noted that no a priori mathematical dependency exists among relations derived from three independent measures.)
Image consistency throughout pharmacologic interventions requires three baroreflex trials to obtain duplicate trials with adequate images for analysis on all subjects. Values were averaged across two trials or across the two best trials (highest r values) in subjects with three adequate trials. A minimum value of r = 0.65 identified significant relations among variables. Across all subjects and all trials, the three relations averaged r > 0.86. High reproducibility of these measures has been shown previously (19), and was comparable in these young and older subjects. Values for RSA were averaged from the two trials used for baroreflex assessment.
Statistics. Differences were assessed by Students t-test with p < 0.05. To determine whether the hypothesized indirect connection of vascular distensibility to vagal tone through the arterial baroreflex existed, we used an extension of multiple regression, structural equation modeling (22). For our application, we used a path analysis, which hypothesizes specific relations among variables and tests the model with a linear equation system. The a priori relationships among the arterial baroreflex, its components, and RSA were defined and a path diagram was constructed to represent direct and indirect effects. Path coefficients (beta weights) were calculated from structural equations reflecting defined effects. The goodness of overall model fit was determined by a nonsignificant chi-square and path coefficients were considered significant if r2 > 0.5 for the predicted variables (baroreflex gain and RSA).
| Results |
|---|
|
|
|---|
15% lower distensibility. Thus, the lesser sensitivity of the neural component in this older subject compared to the younger counterpart importantly contributed to the lower arterial baroreflex gain. Differences in these examples broadly applied to the groups as a whole (Fig. 2). Mechanical transduction of pressure into barosensory vessel stretch was
40% lower in the older subjects. However, arterial baroreflex gain was >60% lower; thus, even though differences in neural transduction of stretch into vagal outflow only approached significance, it likely played an important role in reducing baroreflex function in the older group.
|
|
|
|
| Discussion |
|---|
|
|
|---|
Effects of vascular structural deficits. Vascular changes associated with aging, especially in the aorta and carotid arteries, predict CV morbidity and mortality (16,23). Intuitively, a stiffer arterial vasculature may be related to greater CV morbidity and mortality due to increased systemic arterial pressures (24) and myocardial afterload stresses (25). However, across normotensive and hypertensive individuals, high resting heart rates are associated with greater carotid stiffness (26) and low baroreflex sensitivity relates to lower carotid compliance (27). Thus, it is reasonable to hypothesize that arterial distensibility could also exert effects on autonomic nervous control.
Indeed, we found that the age-related decrease in mechanical transduction of arterial pressure into carotid stretch significantly reduced baroreflex sensitivity. However, lesser carotid distensibility was not linked to reduced RSA. Thus, the autonomic effect of reduced distensibility in our older subjects was confined to blunting baroreflex sensitivity. Although some data suggest that the baroreflex is the primary driving force for cardiac vagal tone (13) and for vagally mediated heart period fluctuations at the respiratory frequency (14), the baroreflex role in generating RSA may be minimal in supine subjects (28) and sympathetic effects on RSA can be significant (29). Therefore, it is not surprising that our analysis did not suggest a direct path from lesser distensibility to reduced RSA through lower cardiovagal baroreflex gain.
Effects of vagal neural deficits. Cardiac vagal outflow is crucial for both baroreflex-mediated bradycardic responses (30) and nearly all frequencies of heart rate variability (31). Both have significant prognostic value for cardiac mortality (18,32) because of the cardioprotective effects of this common vagal effector. We found that neural control of bradycardic responses was low and that it strongly determined baroreflex gain in older individuals. Moreover, we found a link between age-related reductions in our neural baroreflex component and RSA. Thus, a generalized vagal deficit may result in both lesser baroreflex-mediated bradycardia and lower RSA.
Our neural baroreflex component encompasses various aspects of cardiovagal function. However, the link between this aspect of baroreflex control and RSA may indicate that significant age-related neural deficits exist beyond the afferent baroreceptive nerves. For example, there is evidence of altered central autonomic integration (33), reduced vagal outflow (34) and lower muscarinic sinoatrial node receptor density (6) with advancing age. Our findings suggest that these declines may play a direct role in age-related reductions of both resting vagal tone and cardiovagal baroreflex gain.
Clinical implications. Although carotid distensibility correlates to baroreflex sensitivity in heart disease patients (35), Eckberg et al. (30) concluded, over 30 years ago, that reduced compliance could not fully explain the baroreflex derangement they observed. In their examination of patients with a range of heart diseases not known to involve autonomic nerves directly, they found blunted tachycardiac responses to vagal blockade and a level of baroreflex impairment related directly to the severity of cardiac symptoms. Subsequent work has made use of RSA to characterize these vagal deficits (7,9) and to assess CV risk (8,32). However, our data suggest that baroreflex gain and not RSA crucially depends upon vascular function. This could explain why baroreflex function provides a better index of risk for malignant arrhythmias (17,36) and overall cardiac mortality (18) than heart rate variability. Arterial stiffness may carry risk independent of its autonomic effects, and low baroreflex gain would encompass this risk as well as that associated with compromised vagal function. Despite the fact that both RSA and baroreflex gain are useful indices, our results suggest the information contained within them differ.
Conclusions. These data provide insight to carotid arterial distensibility, arterial baroreflex sensitivity and RSA as indices for CV risk. Vascular structural declines may not impact resting vagal tone, but they do reduce baroreflex sensitivity. Moreover, because baroreflex gain encompasses both vascular stiffness and vagal function, arterial baroreflex gain may provide greater insight to CV risk than heart rate variability indices. At the least, our data suggest that baroreflex sensitivity and RSA are independent markers of autonomic function.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
X. Shi, F. A. Schaller, N. Tierney, P. Chanthavong, S. Chen, P. B. Raven, and M. L. Smith Physically Active Lifestyle Enhances Vagal-Cardiac Function but Not Central Autonomic Neural Interaction in Elderly Humans Experimental Biology and Medicine, February 1, 2008; 233(2): 209 - 218. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Studinger, R. Goldstein, and J. A. Taylor Mechanical and neural contributions to hysteresis in the cardiac vagal limb of the arterial baroreflex J. Physiol., September 15, 2007; 583(3): 1041 - 1048. [Abstract] [Full Text] [PDF] |
||||
![]() |
Z. Lenard, P. Studinger, B. Mersich, L. Kocsis, and M. Kollai Maturation of Cardiovagal Autonomic Function From Childhood to Young Adult Age Circulation, October 19, 2004; 110(16): 2307 - 2312. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. D. Lipman, J. K. Salisbury, and J. A. Taylor Spontaneous Indices Are Inconsistent With Arterial Baroreflex Gain Hypertension, October 1, 2003; 42(4): 481 - 487. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |