Abiological parameters display a certain variability.

Transcription

1 120 Effects of Aging on Blood Pressure Variability in Resting Conditions Derk P. Veerman, Ben P.M. mholz, Wouter Wieling, John M. Karemaker, Gert A. van Montfrans Abstract The objective of this study was to determine the effect of aging on beat-to-beat blood pressure and pulse interval variability in resting conditions and to determine the effect of aging on the sympathetic and vagal influence on the cardiovascular system by power spectral analysis of blood pressure and pulse interval. We studied three groups of healthy, normotensive subjects: young (10 to 15 years, n=16), adult (20 to 40 years, n=16), and elderly (70 to 90 years, n=25). Beat-to-beat blood pressure was measured by Finapres during 20 minutes supine and 10 minutes standing. Overall systolic and diastolic blood pressures and pulse interval variability were determined as SD and as coefficient of variation. Also, relative powers of the mid-frequency (0.08 to 0.12 Hz) and high-frequency bands (0.15 to 0.40 Hz) were determined by spectral analysis. n these subjects no differences in blood pressure variability (either as SD or coefficient of variation) were found between age groups, except for the coefficient of variation of standing diastolic blood pressure, which decreased with aging. Pulse interval variability decreased with aging. Power of the mid-frequency band of systolic and diastolic blood pressures was markedly decreased in the elderly, especially in the standing position. Power of the high-frequency band of pulse interval was also decreased in the elderly. Baroreflex sensitivity calculated by fast Fourier transformation spectral analysis was decreased in the elderly subjects compared with the younger groups. n conclusion, we found no change in the overall variability of blood pressure with aging. Mid-frequency spectral power of blood pressure and mid- and high-frequency spectral powers of pulse interval variability were decreased in the elderly. These results suggest that aging does not merely influence the magnitude of blood pressure and pulse interval variability but causes a complex rearrangement of the variability pattern by changes in neurocardiovascular regulation. (Hypertension. 1994;24: ) Key Words blood pressure autonomic nervous system aging Abiological parameters display a certain variability. Cardiovascular parameters such as blood pressure and heart rate distinguish themselves from many others because their variability has (patho)- physiological significance. 1 The study of beat-to-beat blood pressure and heart rate variabilities enhances the physiological understanding of neurocardiovascular regulation 2 " 6 ; from a clinical point of view, interest is currently focused on the possible role of blood pressure variability as a risk factor of hypertensive target-organ damage, independent of the average level of blood pressure. 7 Presently, the elderly constitute a large and fastgrowing proportion of the population that challenges cardiovascular research. Like hypertension, aging has been shown to increase blood pressure variability. 810 Heart rate variability, however, decreases with increasing age and blood pressure Because blood pressure rises with aging, 13 there is a substantial covariance, which makes the relative contribution of blood pressure and aging to variability difficult to disentangle. Other factors that may hamper proper understanding of studies on beat-to-beat blood pressure and heart rate variabilities in relation to age are Received August 10, 1993; accepted in revised form April 13, From the Departments of Medicine and Physiology (J.M.K.), Academic Medical Centre, Amsterdam, the Netherlands. Correspondence to G.A. van Montfrans, MD, Department of Medicine, Room B2-229, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands American Heart Association, nc. the lack of strict standardization of activities in ambulant subjects and the relatively small age range of the subjects. 810 Conventional descriptive techniques, such as frequency distribution, are used for the study of beat-tobeat blood pressure and heart rate variabilities. 8 " 10 Additional information can be obtained by power spectral analysis, which breaks down a signal to its constituent frequency components and quantifies the power of these components. 2 " Blood pressure and heart rate modulation by baroreflex and sympathetic and parasympathetic nervous activity has been shown to be involved in the genesis of the various frequency components of blood pressure and heart rate variabilities " 21 n the present work we determined the overall variability and frequency distribution of blood pressure and pulse interval in healthy, normotensive subjects, covering an age range of 10 to 90 years, with subjects in the supine and standing positions. We addressed two issues in this study: (1) the separated effects of aging and the rise in blood pressure with aging on overall blood pressure variability in healthy subjects covering a large age range during standardized resting conditions, and (2) the effect of aging on the influence of the baroreflex and sympathetic and parasympathetic nervous systems on the cardiovascular system as determined by power spectral analysis of blood pressure and pulse interval variabilities. Subjects were studied in both the supine and standing positions. These positions are of interest because the supine position increases the vagal drive to

2 Veerman et al Age and Blood Pressure Variability 121 the heart, and the standing position increases the sympathetic vasomotor traffic. 22 Methods Subjects The data of three studies were used to compose three different age groups: young, young adult, and elderly. Young (10 to 15 years) boys (n=10) and premenarcheal girls (n=10) and healthy elderly (70 to 90 years) male (n=20) and female (n=20) subjects were selected from a general practice. Healthy young adult (20 to 40 years) men (n=10) who were involved in a study on standardization of investigation of cardiovascular reflexes 25 comprised the male part of the adult group. This group was extended with seven young adult female subjects (20 to 40 years). Details of the selection procedure have been reported elsewhere. 23 " 25 All subjects were healthy, did not use any medication, and ate a normal diet without salt restriction. The elderly subjects were still physically and socially active. The protocol was approved by the hospital ethics committee, and all subjects gave their informed consent. Measurements Noninvasive continuous blood pressure was measured at the finger using a TNO FNAPRES model 5, which has been shown suitable for assessment of blood pressure variability. 28 To avoid interference of hydrostatic errors, subjects were asked to keep the cuffed finger at heart level. The correct position of the finger, the fourth intercostal space in midaxillar line, was marked with a skin electrode. Continuous finger blood pressure and an event marker were recorded on magnetic tape using a cassette data recorder (TEAC R-61, TEAC Corp) for off-line data analysis. Maneuvers A 20-minute supine period and 10-minute standing period were used for analysis. The first 2 minutes of standing were omitted to allow for stabilization of blood pressure and pulse interval. The maneuvers were performed in the morning in a room with a constant temperature of 22 to 24 C. At least 2 hours before the experiment the subjects had eaten a light breakfast. All subjects abstained from coffee, tea, and smoking starting the evening before the experiments. Data Analysis Finger blood pressure and the event marker were replayed and analog-to-digital converted at 100 Hz (real time). By means of a signal-analysis program, the time of the systolic upstroke was identified for each beat, as were the actual systolic, mean, and diastolic pressures and the interbeat interval (resolution, 10 milliseconds). The automatic calibration of the Finapres (Physio-cal) was kept on during the supine and standing periods. The values of the missing beats were estimated by linear interpolation. The following calculations were made: Overall variability of systolic and diastolic blood pressures and pulse interval was expressed in absolute terms, as the within-person SD, and in relative terms, as the withinsubject coefficient of variation (CV). All beats of the supine period and beats of the period from 2 to 10 minutes of standing were used. The CV was calculated to remove the influence of the level of blood pressure or pulse interval on their variabilities. Before calculation of the SD and CV, the blood pressure and pulse interval tracings of the supine and standing periods of each subject were corrected for linear trends by the least-squares method. 29 Power spectra were estimated by means of fast Fourier transformation (FFT) of the above-mentioned time periods This technique breaks up the overall variability of a signal (in this case blood pressure and pulse interval) into its composing frequencies and reveals to what extent a given frequency contributes to the overall variability of a signal. TABLE 1. Age Group Young Adult Old Basic Subject Characteristics Male/Female 8/7 9/7 15/11 Mean Age, y 12±1 27±5 76±5 Height, m 160±9 175±6 171±10 Weight, kg 44±8 67±8 72±11 Values are mean±sd. Young, 10 to 15 years; adult, 20 to 40 years; and old, 70 to 90 years. ntegration of the curve for specific frequency bands yields variance for that particular frequency band. We calculated the power in two frequency bands: the mid-frequency (MF) band (0.08 to 0.12 Hz) and the high-frequency (HF) band (0.15 to 0.40 Hz). The percentage power of the MF and HF bands in relation to total power (total variability) in the supine and standing positions was also obtained in the three respective age groups. Spectral analysis by FFT not only gives variability as a function of frequency, it also quantifies covariation between signals in terms of phase (time shifts) and coherence (squared correlation in a specific frequency band) We computed phase and coherence between systolic blood pressure and pulse interval to test a baroreflex model for the genesis of variability in the MF and HF bands. Only when coherence exceeded.5 (which is comparable to a correlation coefficient of.7) did we compute the regression between systolic blood pressure and pulse interval in that band in milliseconds per millimeter of mercury, the usual measure for baroreflex sensitivity. Blood pressure recordings were screened for premature beats and movements artifacts, both known for their influence on the assessment of variability. 2 f these were present in a certain subject, that subject was omitted from the analysis. Consequently, 16 of 20 subjects remained in the young group, 16 of 17 subjects in the adult group, and 25 of 40 subjects in the elderly group. Averaged blood pressure in the supine and standing positions of the remaining subjects was comparable to the age group from which they originated. Table 1 lists basic subject characteristics. Statistics Statistical analysis was performed with BMDP statistical software (University of California, Los Angeles). Kruskal- Wallis and Mann-Whitney U tests were used to compare the different age groups. Paired comparisons for the effect of posture in the three age groups were performed by Wilcoxon's signed rank test. When no differences in variability (only SD) between age groups were found, the relation of the level of the tested parameter and SD was determined by a linear regression analysis on the data of all subjects, all age groups combined. Correlation coefficients, regression coefficients (re), and their standard errors (SEM) are given. Results are expressed as mean±sd. A value of f<.05 was considered statistically significant. Results Levels of Blood Pressure and Pulse nterval Table 2 presents the average blood pressure and pulse interval levels in the three age groups. Average systolic and diastolic blood pressures of young and adult subjects were significantly lower than those of the elderly subjects in both the supine and standing positions. The increase in diastolic pressure upon standing was significant ( in all three age groups). Average pulse interval in the young was shorter than in the elderly in both the supine and standing positions. pulse interval of the adult group was longer than

3 122 Hypertension Vol 24, No 1 July 1994 TABLE 2. Average Blood Pressure and Pulse nterval in the Three Age Groups SBP, mm Hg vs standing DBP, mm Hg vs standing Pulse interval, ms vs standing Young 111±6 115±9 58±5 70±7 861 ± ±134 Adult 112±18 118±14 58±9 72±9 979± ±72 Old 140±18 157± ±3 979± ±97 YvsO p YvsA A vs 0 Y indicates young (10 to 15 years); O, old (70 to 90 years); A, adult (20 to 40 years); SBP, systolic blood pressure; and DBP, diastolic blood pressure. Values are mean±sd. that of the young, whereas standing pulse interval of the adult group was shorter than that of the elderly. Overall Variability of Blood Pressure and Pulse nterval Table 3 and Figs 1 and 2 show the main results of overall blood pressure and pulse interval variabilities. Blood Pressure Variability Overall blood pressure variability, expressed as SD, did not differ among age groups for both systolic and diastolic pressures in neither the supine nor standing positions. from supine induced an increase in diastolic but not systolic SD for all three age groups. Expressed as the CV, there was no difference in systolic pressure variability between age groups in either the supine or standing positions. For diastolic pressure the variability (CV) in the standing position was even larger in the young group compared with the adult and elderly groups. n the supine position there were no differences among age groups. Using all subjects, regression analysis showed a significant relation between the levels of systolic and diastolic blood pressures and the overall variability (SD) of blood pressure in both the supine and standing positions. For systolic pressure the correlation coefficient was.37 (rc=0.038, SEM=0.013, P<.0\) in the supine position and.52 (rc=0.061, SEM=0.013, ) in the standing position. For diastolic pressure the correlation coefficient was.33 (rc=0.043, SEM=0.016, ) in the supine position and.29 (rc=0.033, SEM = 0.014, ) in the standing position. Regression of level versus variability of blood pressure for the three age groups separately did not show differences in slopes or intercepts among age groups. Scatterplots in Fig 3 illustrate the relation of height and variability (SD) of blood pressure for the three age groups. Pulse nterval Variability Overall pulse interval variability of the elderly, expressed as SD, was lower than that of the young and adults in the supine position. There was no significant difference in the SD of pulse interval in the standing position among the three age groups. Pulse interval variability expressed as CV of the young group was higher than that of the elderly in both the supine and standing positions. The pulse interval CV of the adult group was higher than the CV of the elderly group only in the supine position and not in the standing position. Spectral Analysis Fig 4 shows representative examples of blood pressure tracings of an elderly and adult subject in the supine and standing positions. Visually, there was a striking difference in the blood pressure variations in the standing position between the adult and elderly subject, the former apparently having the largest fluctuations. Nevertheless, the SDs of diastolic blood pressure were almost identical: 4.3 mm Hg in the elderly subject and 4.6 mm Hg in the adult female. n this case the basic difference was clarified by spectral analysis, which showed that in the adult subject 21% of variability was concentrated in the MF band, whereas in the elderly subject this was only 2%. Thus, the frequency distribution of the variability may disclose differences that, by using the SD, would remain hidden. Fig 5 (left and right) and Table 4 display the principal results of the spectral analysis. The relative power in the MF band in the supine position was larger in the young and adult subjects compared with the elderly for diastolic pressure but did not differ among age groups for systolic pressure. n the standing position both systolic and diastolic blood pressure variabilities in this band were larger in the young and adult subjects compared with the elderly. up from the supine position induced an increase in power of the MF band of systolic and diastolic blood pressures in all three age groups, but this increase in power was smaller in the elderly than in the young and adult subjects. The relative power of the MF band of both supine and standing pulse interval was significantly smaller in the elderly than in the young and

4 Veerman et al Age and Blood Pressure Variability 123 TABLE 3. Overall Variability of Blood Pressure and Pulse nterval Standard deviation SBP, mm Hg vs standing DBP, mm Hg vs standing Pulse interval, ms vs standing Coefficient of variation SBP, % vs standing DBP, % vs standing Pulse interval, % vs standing Young 5±1 6±1 4±1 5±1 74 ±28 61 ±26 5±1 6±1 6±1 7±1 9±3 9±2 Definitions are as in Table 2. Values are mean±sd. Adult 7±2 7±1 3±1 4±1 79±41 61 ±26 6±2 6±1 6±1 5±1 8±3 7±2 adult groups. up from the supine position induced an increase in the power of this band only in the young and adult groups. The relative power of the HF band of pulse interval in the supine position was larger in young and adult subjects than in the elderly. up from the supine position induced a decrease in the power of this band in all age groups, and in the standing position there was no longer a difference in the relative power of this band among the three age groups. Table 5 gives the results of the calculation of baroreflex sensitivity by FFT spectral analysis. n the acceptance test based on coherence we had to reject more than half the spectral outcomes in the elderly group. Closer inspection confirmed that most of the blood pressure and pulse interval variabilities in this group were in the low frequencies (<0.08 Hz). t seemed that the characteristic peak of the MF band was not so much absent but had shifted to a lower frequency in the elderly group (between 0.05 and 0.08 Hz), effectively shifting out of the frequency window that had been set at the start of this study. As Table 5 shows, baroreflex sensitivity decreased significantly Old 7±2 8±4 4±1 5±2 53 ±22 54 ±22 5±2 5±2 6±2 6±2 6±2 6±2 YvsO P YvsA A vs 0 from the supine to the standing posture in all age groups in both frequency bands. Values for baroreflex sensitivity were not significantly different between the young and the adult group but were significantly reduced in the elderly group. Discussion The main findings of this study on beat-to-beat blood pressure and pulse interval variabilities in stable conditions in subjects between 10 and 90 years of age were the following: (1) Absolute overall systolic and diastolic blood pressure variabilities did not change with aging; also, the relative overall systolic and supine diastolic blood pressure variabilities did not change with aging. relative diastolic blood pressure variability of the young subjects was larger than that of the adult or elderly subjects. (2) The power of the MF band of blood pressure and pulse interval variabilities decreased with aging. up from the supine position induced a rise in the power of this band; the magnitude of this rise was decreased in the elderly. (3) n the supine position the power of the HF band decreased with aging; in the standing position there was no age difference.

5 124 Hypertension Vol 24, No 1 July SD PS 6 - (mm Hg) 3 - yy CV PS 7 - ok-p y/a 9 -i CD supine EZ3 standing 11 1 supine V7A standing SD PD 6 - (mm Hg) t- *-i -* CV PD 7 - Yvz/j- adult A 3 - r n young old 3-11? 100 n 11 n SD NT (msec) 50 - FG 1. Bar graphs show variability of systolic (PS) and diastolic (PD) blood pressures and pulse interval (NT) during supine and standing expressed by SD in three age groups (young: 10 to 15 years, n=15; adult: 20 to 40 years, n=16; old: 70 to 90 years, n=26). Results are mean±sd. *, **. CV NT Overall Blood Pressure and Pulse nterval Variabilities We found that absolute overall blood pressure variability of elderly subjects did not differ from that of younger subjects, whereas linear regression analysis revealed a significant relation between the height and the variability (expressed as SD) of blood pressure (Fig 3). Also, when we expressed blood pressure variability in relation to blood pressure level as the CV, no difference in blood pressure variability among age groups could be demonstrated, with the sole exception of standing diastolic blood pressure. n this case a decrease in variability with aging was shown. These results are in contrast with some earlier studies describing an increase in the variability of ambu FG 2. Bar graphs show variability of systolic (PS) and diastolic (PD) blood pressures and pulse interval (NT) during supine and standing expressed as coefficient of variation (CV) in three age groups (young: 10 to 15 years, n=15; adult: 20 to 40 years, n=16; old: 70 to 90 years, n=26). Results are mean±sd. *, **, ***. Note that the increased variability of systolic blood pressure when expressed by SD (Fig 1) in the elderly group compared with the other age groups disappears when variability is expressed as CV. latory recorded beat-to-beat blood pressure with aging independent of the height of blood pressure. 810 However, in one other study by Watson et al 31 no independent relation of blood pressure variability and age was found. n that study systolic variability increased with the height of pressure. To explain these differences in the relation between aging and blood pressure variability, consideration of the role of heart rate in blood pressure regulation may be helpful. The studies that show an increase in blood pressure variability with aging were performed by ambu-

6 Veerman et al Age and Blood Pressure Variability i y-0.03bi y-0.043x E 4 - A 8 5 a CO 2 A A A A a PS [mm HQ) 20 -i r-o.oeii-o.b PD (mm Hg) 10 -i T OO so E 10 - E «" a V) 2 - so PS (mm Hg) SO s so PD (mm Hg) s 110 FG 3. Scatterplots show systolic (PS) and diastolic (PD) blood pressure variability expressed as SD vs height of blood pressure of three age groups (young: 10 to 15 years, n=15; adult: 20 to 40 years, n=16; old: 70 to 90 years, n=26) n supine (top panels) and standing (bottom panels) positions. Regression lines for all subjects are displayed when the regression coefficient for the relation between height and variability of blood pressure was significant; see text. latory monitoring in freely moving subjects. 8 " 10 t has been shown that in resting conditions a marked reduction in heart rate variability through administration of atropine is not followed by a change in blood pressure variability. However, if during physical exercise heart rate variability is inhibited by atropine, this results in an increase in blood o m n oo m a oom mmho 200-, 0 J pressure variability. 32 Thus, it appears that during physical activity, which clearly dominates the blood pressure pattern during ambulatory blood pressure monitoring, 33 heart rate variations exert an antioscillatory influence on the variability of blood pressure. n elderly subjects heart rate variability was decreased (Figs 1 and 2), which has been 986 yrs 924 yrs FG 4. Tracings show typical example of finger blood pressure recordings during supine and standing positions in an elderly and an adult subject. Note pronounced variability n blood pressure that develops on standing n the adult subject, which is absent in the elderly subject. 0 J 1 mln

7 126 Hypertension Vol 24, No 1 July 1994 MF BAND HF BAND POWER PS young adult young adult old b'.-l supina V/A (landing 35 -**- 35 POWER PD NMH i POWER NT i FG 5. Bar graphs present results of spectral analysis, showing'percent power of mid-frequency (MF, 0.08 to 0.12 Hz) (left panels) and high-frequency (HF, 0.15 to 0.40 Hz) (right panels) bands of systolic (PS) and diastolic (PD) blood pressures and pulse interval (NT) during supine and standing positions of three age groups (young: 10 to 15 years, n=15; adult: 20 to 40 years, n=16; old: 70 to 90 years, n=26). Results are mean±sd. *, **, ***. n the MF bands (left), note the large increase upon standing in the young and adult groups, which is much smaller in the elderly subjects. attributed to diminished vagal activity in the elderly The decreased heart rate variability in elderly subjects thus may diminish the ability to limit the variability of ambulatory blood pressure. The baroreflex plays an important role in buffering the variations in blood pressure in ambulatory conditions, as shown by the large increase in blood pressure variability after surgical denervation of the baroreceptors t has also been shown that there is an inverse relation between baroreflex sensitivity and blood pressure variability in ambulatory conditions. 9 Aging causes a decrease in baroreflex sensitivity, and thus in elderly subjects the baroreflex may be less able to limit variations in blood pressure in ambulatory conditions. 8 However, blood pressure variability during rest (night) is reduced, 8 and physical activity has been shown to be the main source of blood pressure variations during ambulatory conditions. 33 Thus, during resting conditions blood pressure is less variable, and consequently, the decrease in the efficacy of the baroreflex to inhibit blood pressure variations in the elderly may be of less importance in resting conditions. This can explain our finding that blood pressure variability does not increase with aging in resting conditions, whereas it has been shown to increase with aging in ambulatory conditions. 810 There is debate about the best way of expressing overall beat-to-beat blood pressure variability. 38 At

8 Veerman et al Age and Blood Pressure Variability 127 TABLE 4. Results of Spectral Analysis P Mid-frequency band SBP, % vs standing DBP, % vs standing Pulse interval, % vs standing High-frequency band SBP, % vs standing DBP, % vs standing Pulse interval, % vs standing Young 7.5± ± ± ± ± ± ± ± ±12.3 Adult 8.0± ± ± ± ± ± ± ±10.9 Old 5.2± ± ± ± ± ± ± ±9.4 YvsO <.01 YvsA A vs 0 Definitions are as in Table 2. The mid-frequency band was 0.08 to 0.12 Hz; the high-frequency band was 0.15 to 0.40 Hz. Values are mean±sd. least two approaches are possible: (1) the use of a measure of absolute variability (SD) alone or (2) the use of a relative measure of variability (CV), an index that expresses absolute variability as a percentage of baseline blood pressure. This choice will clearly affect the results and interpretation of studies on blood pressure variability. Some studies advocate the use of the SD because a relation between the variability and level of blood pressure cannot be demonstrated, 9-39 although in several other studies such a relation has been found. 8 ' 1031 Our data indicate that a relation between the level and variability of blood pressure exists, but we could not demonstrate such a relation between age and blood pressure variability, even in the wide age range we studied. Thus, because blood pressure variability increased with blood pressure level and blood pressure level also increases with aging, 13 we consider the CV more appropriate for studying the relation of blood pressure variability and aging. However, it can be argued that in studies on blood pressure variability one should use both the CV and SD to represent variability to facilitate comparison with other studies in which the <.01 <.01 SD was used for the calculation of blood pressure variability Another argument brought forward in favor of the use of the SD instead of the CV is that clearly visible differences in the variability of beat-to-beat blood pressure tracings are not supported by differences in CV. 39 But as demonstrated by the tracings in Fig 4, large differences in visible variability are not always represented in differences of either absolute or relative measures of overall variability of beat-to-beat blood pressure. n this case the use of only the SD is just as inappropriate as the use of only the CV; for a more detailed description of these differences in variability and for a better understanding of how the autonomic nervous system controls these beat-to-beat variations, a more sophisticated approach such as power spectral analysis may be appropriate. Spectral Analysis The present study confirmed for a large age range the observations of Pagani et al 2 and Lipsitz et al 4 that in healthy subjects pulse interval variability decreases in

9 128 Hypertension Vol 24, No 1 My 1994 TABLE 5. Baroreflex Sensitivity in thethree Age Groups Mid-frequency band n n vs standing High-frequency band n n vs standing Young 10 14± ± ±4 Adult Old YvsO 12 15± ± ± ±3 11 8±4 10 4± ±9 10 5±2 P YvsA A vs 0 Y indicates young (10 to 15 years); O, old (70 to 90 years); A, adult (20 to 40 years); and n, number of subjects with a coherence larger than.5 that were included in the calculation (see text). The mid-frequency band was 0.08 to 0.12 Hz; the high-frequency band was 0.15 to 0.40 Hz. Values are mean±sd. Baroreflex sensitivity is expressed in milliseconds per millimeter of mercury. the HF and MF bands with increasing age. n elderly subjects the power of the MF band of diastolic pressure and pulse interval was smaller than in young or adult subjects in both the standing and supine positions. For systolic pressure this was only true for the standing position. up from the supine position induced an increase in the power of the MF band of both blood pressure and pulse interval in all three age groups, with the exception of pulse interval MF power in the elderly, which did not increase. Also, in the elderly the increase in power of the MF band of systolic and diastolic blood pressures on arising was smaller than that of young and adult subjects. We did not calculate powers below 0.08 Hz; however, as the power above 0.08 Hz decreased in the elderly and overall blood pressure variability did not change, it is fair to assume that in the elderly the power of these very low frequencies increased. Several factors influence the blood pressure and pulse interval spectra. Neural mechanisms are considered to be the most important, and most of the spectral components are probably generated by sympathetic and parasympathetic nervous activity and the baroreflex. 2 ' Blood pressure variations that cause baroreflex modulation of pulse interval by parasympathetic nerves affect both HF and MF powers, although the MF power of pulse interval is also influenced by sympathetic nervous activity due to the time delay of baroreflex action in the MF range There is much discussion on the origin of the blood pressure variability in the 0.1-Hz frequency range. Sympathetic vasomotor activity probably plays an important role as well as the baroreflex, as shown by the fact that blood pressure variability around this frequency decreases after sinoaortic denervation, whereas total blood pressure variability increases (mainly by an increase in lower-frequency variability). 19 Moreover, there are theoretical arguments for the existence of a 0.1-Hz resonance phenomenon in the feedback arterial baroreceptor-sympathetic vasomotor loop. 20 Much simplified, MF oscillations in blood pressure may be the result of the delayed sympathetic vasomotor reaction on blood pressure changes sensed by the baroreceptor and thus impose a specific rhythm on blood pressure Spontaneous peripheral vasomotor changes may also contribute to blood pressure variations around and below this frequency range The decrease in MF and HF power of pulse interval in the elderly may be related to a decrease of vagal modulation of heart rate with aging and the decrease in baroreflex sensitivity with aging (Table 5) ' 37 The decrease in MF pulse interval power in the elderly may also point to a decrease in cardiac sympathetic modulation. The decrease in power of the MF band of blood pressure with increasing age may be caused by a diminution of sympathetic vasomotor modulation or to the diminished effectiveness of the baroreflex in the elderly and thereby the weakening of the feedback loop responsible for the 0.1-Hz resonance. 12 ' Alternatively, slowing of sympathetic responses may have caused the 0.1-Hz band to shift outside the MF window used in this study. Various methods have been used to estimate the activity of the sympathetic nervous system. 42 Compared with measuring plasma levels of catecholamines or muscle sympathetic nerve activity, determining the power of the MF band by spectral analysis has an advantage: t does not determine sympathetic activity at some intermediate stage but may be related to the ultimate influence of the arterial baroreceptor-sympathetic vasomotor loop on the end organ. On the other hand, when changes occur in the power of the MF band, it is not possible to determine which site of the chain is responsible for these changes without information from other methods. n elderly subjects baseline sympathetic nerve activity per se seems to be increased, as shown by elevated levels of circulating catecholamines 43 and by a higher level of muscle sympathetic nerve activity recorded by microneurography The evidence concerning the increase in sympathetic nerve activity during orthostatic stress is conflicting. n an earlier study we found that in the same subjects after 10 minutes of standing, compared with

10 being supine, the increase in peripheral resistance was much larger in elderly than young subjects. 46 As the increase in peripheral resistance upon standing is mediated by the sympathetic nervous system, 22 it seems that in elderly subjects the sympathetic activation is larger than in young subjects. However, the increase in microneurographically recorded muscle sympathetic nerve activity induced by orthostatic stress was much smaller in elderly subjects than in young subjects, 44 but the increase in circulating catecholamines upon standing was larger in elderly than young subjects Thus, catecholamines and total peripheral resistance upon standing suggest an increase in sympathetic activity in the elderly. This makes it unlikely that the decreased MF power of blood pressure in the elderly is related to diminished sympathetic blood pressure modulation, and it is more probable that it is related to the diminished baroreflex sensitivity in the elderly. On the other hand, a decreased sensitivity of adrenergic receptors in the elderly may also partly explain the diminished MF power of blood pressure. Then, increased prereceptor nerve activity can result in less postreceptor effect in the end organ. Until now, a decrease in the sensitivity of these receptors in elderly subjects has been demonstrated unequivocally for /3-receptors. 49 For a-receptors the picture is less clear. The number of receptors probably decreases with aging, but the a-adrenergic receptor agonist response may still be preserved. 49 Structural changes in the vessel that occur with aging 50 may also play a role in the decrease of the MF variability of blood pressure and the smaller increase of this variability upon standing in the elderly. We already found that these vessel wall changes seemed to induce more vasoconstriction and thus a larger increase in blood pressure upon standing in elderly subjects than in younger subjects. 46 The decrease of MF blood pressure variations with aging may also be explained by vessel wall changes. The fact that the end organ (the peripheral vessels) now reacted by a diminished instead of an enhanced response to sympathetic activation may be related to the increased stiffness of the vessel wall in the elderly, 50 which may decrease and slow down its ability to react 51 ' 52 to changes in neurosympathetic drive. Veerman et al Age and Blood Pressure Variability 129 Conclusions Overall blood pressure variability at rest does not increase with aging. n other studies it has been found that ambulatory blood pressure variability is increased in the elderly. The difference between our findings and the literature can probably be attributed to the fact that the antioscillatory influence of heart rate, which decreases with aging, is less important at rest than in ambulatory conditions. The power in the MF band of blood pressure variability is decreased in healthy elderly subjects. This may be related to the decreased baroreflex sensitivity in the elderly or to a decrease in end-organ responsiveness to sympathetic nerve activity due to adrenergic receptor sensitivity decrease or to structural changes of the vessel wall with aging. Overall blood pressure variability does not change in the elderly, and as MF blood pressure variability decreases, it seems that aging does not merely influence the magnitude of blood pressure variability but causes a rearrangement of the variability pattern by changes in neurocardiovascular regulation. References 1. Mancia G, Zanchetti A. Blood pressure variability. n: Zanchctti A, Tarazi RC, eds. Handbook Of Hypertension, Volume 7: Pathophysiology of Hypertension Cardiovascular Aspects. Amsterdam, the Netherlands: Elsevier Science Publishers (Biomedical Division); 1986: Pagani M, Lombardi R, Guzetti S, Rimoldi O, Furlan R, Pizzinelli P, Sandrone G, Malfatto G, Dell'Orto S, Piccaluga E, Turiel M, Baselli G, Serutti S, Malliani A. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res. 1986;59: Weise F, Heydenreich F, Runge U. Contributions of sympathetic and vagal mechanisms to the genesis of heart rate fluctuations during orthostatic load: a spectral analysis. J Auton Nerv Syst. 1987;21: Lipsitz LA, Mietus J, Moody GB, Goldberger AL. Spectral characteristics of heart rate variability before and during postural tilt: relations to aging and risk of syncope. Circulation. 1990;81: Lipsitz LA, Storch HA, Minaker KL, Rowe JW. ntra-individual variability in postural blood pressure in the elderly. Clin Sci. 1985; 69: Turjanmaa V, Kalli S, Sydanmaa M, Uusitalo A. Short-term variability of systolic blood pressure and heart rate in normotensive subjects. Clin Physiol. 1990;10: Parati G, Pomidossi G, Albini F, Malaspina D, Mancia G. Relationship of 24-hour blood pressure mean and variability to severity of target-organ damage in hypertension. J Hypertens. 1987;5: Mancia G, Ferrari A, Gregorini L, Parati G, Pomidossi G, Bertinieri G, Grassi G, di Rienzo M, Pedotti A, Zanchetti A. Blood pressure and heart rate variabilities in normotensive and hypertensive human beings. Circ Res. 1983;53: Floras JS, Hassan MO, Vann Jones J, Osikowska BA, Sever PS, Sleight P. Factors influencing blood pressure and heart rate variability in hypertensive humans. Hypertension. 1988; 11: Mann S, Millar Craig MW, Altman DG, Raftery EB, Hunyor SN. Blood pressure variability in health, hypertension and autonomic failure. Clin Exp Hypertens A. 1985;7: Wieling W, van Brederode JFM, de Rijk LG, Borst C, Dunning AJ. Reflex control of heart rate in normal subjects in relation to age: a data base for cardiac vagal neuropathy. Diabetologia. 1982;22: Kawamota A, Shimada K, Matsubayashi K, Chikamori T, Kuzumc O, Ogura H, Ozawa T. Cardiovascular regulatory functions in elderly patients with hypertension. Hypertension. 1989; 13: Kannel WB. Host and environmental determinants of hypertension: perspective from the Framingham study. n: Kesteloot H, Joossens JV, eds. Epidemiology of Arterial Blood Pressure. The Hague, the Netherlands: Martinus Nijhoff; 1980: Akselrod S, Gordon D, Ubel FA, Sahnnon DC, Barger AC, Cohen RJ. Power spectrum analysis of heart rate fluctuations: a quantitative probe of beat-to-beat cardiovascular control. Science. 1981; 213: de Boer RW, Karemaker JM, Strackee J. Hemodynamic fluctuations and baroreflex sensitivity in humans: a beat-to-beat model. Am J Physiol. 1987;253:H680-H Karemaker JM. Analysis of blood pressure and heart rate variability: theoretical considerations and clinical applicability. n: Low PA, ed. Clinical Autonomic Disorders. Boston, Mass: Little, Brown & Co; 1993: Rimoldi O, Pierini S, Ferrari A, Cerutti S, Paganni M, Malliani A. Analysis of short-term oscillations of R-R and arterial pressure in conscious dogs. Am J Physiol. 1990;258:H967-H Saul JP, Rea RF, Eckberg DL, Berger RD, Cohen RJ. Heart rate and muscle sympathetic nerve variability during reflex changes of autonomic activity. Am J Physiol. 1990;258:H713-H Wesseling KH, Settels JJ, Walstra HG, van Esch HJ, Donders JJH. Baromodulation as the cause of short term blood pressure variability? n: Alberi G, Bajzer Z, Baxa P, eds. Application of Physics to Medicine and Biology. Singapore: World Scientific; 1983: Di Rienzo M, Parati G, Castiglioni P, Omboni S, Ferrari AU, Ramirez AJ, Pedotti A, Mancia G. Role of sinoaortic afferents in modulating BP and pulse-interval spectral characteristics in unanesthetized cats. Am J Physiol. 1991;261:H1911-H1918.

11 130 Hypertension Vol 24, No 1 July Saul JP, Berger RD, Albrecht P, Stein SP, Chen MH, Cohen RJ. Transfer function analysis of the circulation: unique insights into cardiovascular regulation. Am J Physiol. 1991;261:H1213-H Rowell LB. Human Cardiovascular Control. Oxford, England: Oxford University Press; 1993: Dambrink JHA, mholz BPM, Karemaker JM, Wieling W. Circulatory adaptation to orthostatic stress in healthy year old children investigated in a general practice. Clin Sci. 1991;81: mholz BPM, Dambrink JHA, Karemaker JM, Wieling W. Orthostatic circulatory control in the elderly evaluated by non-invasive continuous blood pressure measurement. Clin Sci. 1990;79: Ten Harkel ADJ, van Lieshout JJ, van Lieshout EJ, Wieling W. The assessment of cardiovascular reflexes: influence of posture and period of preceding rest. J Appl Physiol. 1990;68: mholz BPM, Wieling W, Langewouters GJ, van Montfrans GA. Continuous finger arterial pressure: utility in the cardiovascular laboratory. Clin Auton Res. 1991;l: Parati G, Casadei R, Gropelli A, Di Rienzo M, Mancia G. Comparison of finger and intra-arterial blood pressure monitoring at rest and during laboratory testing. Hypertension. 1989;13: Omboni S, Parati G, Frattola A, Mutti E, Di Rienzo M, Castiglioni P, Mancia G. Spectral and sequence analysis of finger blood pressure variability: comparison with analysis of intra-arterial recording. Hypertension. 1993;22: Armitage P, Berry G. Statistical Methods in Medical Research. Oxford, England: Blackwell Scientific Publications; 1987: Robbe HW, Mulder U, Ruddel H, Langewitz WA, Veldman JB, Mulder C. Assessment of baroreceptor reflex sensitivity by means of spectral analysis. Hypertension. 1987;10: Watson RDS, Stallard TJ, Flinn RM, Littler WA. Factors determining direct arterial pressure and its variability in hypertensive man. Hypertension. 1980;2: Parati G, Pomidossi G, Casadei R, Gropelli A, Trazzi S, di Rienzo M, Mancia G. Role of heart rate variability in the production of blood pressure variability in man. J Hypertens. 1987;5: Clark LA, Denby L, Pregibon D, Harshfield GA, Pickering TG, Blank S, Laragh JH. A quantitative analysis of the effects of activity and time of day on the diurnal variations of blood pressure. J Chronic Dis. 1987;40: Cowley AC, Liard JF, Guyton AC. Role of baroreceptor reflex in daily control of arterial blood pressure and other variables in dogs. Am J Physiol. 1973;22: Shade RE, Bishop VS, Haywood JR, Hamm CK. Cardiovascular and neuroendocrine response to baroreceptor denervation in baboons. Am J Physiol. 1990;258:R930-R Karemaker JM, Wieling W, Dunning AJ. Aging and the baroreflex. n: Amery A, Staessen J, eds. Handbook of Hypertension, Volume 12: Hypertension in the Elderly. Amsterdam, the Netherlands: Elsevier Science Publishers (Biomedical Division); 1989: Gribbin B, Pickering TG, Sleight P, Peto R. Effect of age and high blood pressure on baroreflex sensitivity in man. Circ Res. 1971;29: Pickering TG. Ambulatory Monitoring and Blood Pressure Variability. London, England: Science Press; 1990: Jacob HJ, Alper RH, Brody MJ. Lability of arterial pressure after baroreceptor denervation is not pressure dependent. Hypertension. 1989;14: Hayoz D, Tardy Y, Rutschmann B, Mignot JP, Achakri H, Feihl F, Meister JJ, Waeber B, Brunner HR. Spontaneous diameter oscillations of the radial artery in humans. Am J Physiol. 1993;264: H2080-H Siegel G. Principles of vascular rhythmogenesis. Prog Appl Microcirc. 1983;3: Eckberg DL, Sleight P. Human Baroreflexes in Health and Disease. Oxford, England: Clarendon Press; 1992: Esler M, Lambert G, Jennings G. The influence of aging on catecholamine metabolism. n: Amery A, Staessen J, eds. Handbook of Hypertension, Volume 12: Hypertension in the Elderly. Amsterdam, the Netherlands: Elsevier Science Publishers (Biomedical Division); 1989: wase S, Mano T, Watanabe T, Saito M, Kobayashi F. Age-related changes of sympathetic outflow to muscles in humans. J Gerontol. 1991;46:M1-M Ng AV, Callister R, Johnson DG, Seals DR. Age and gender influence muscle sympathetic nerve activity at rest in healthy humans. Hypertension. 1993;21: Wieling W, Veerman DP, Dambrink JHA, mholz BPM. Disparities in circulatory adaptation upon standing between young and old subjects explained by pulse contour analysis. Clin Sci. 1992;83: Ziegler MG, Lake CR, Kopin J. Plasma noradrenaline increases with age. Nature. 1976;261: Young J, Rowe J, Pallota J, Sparrow D, Landsberg L. Enhanced plasma norepinephrine response to upright posture and oral glucose administration in elderly human subjects. Metabolism. 1980;29: Nielson CP, Vestal RE. a-adrenoreceptors, /3-adrenoreceptors and aging. n: Amery A, Staessen J, eds. Handbook of Hypertension, Volume 12: Hypertension in the Elderly. Amsterdam, the Netherlands: Elsevier Science Publishers (Biomedical Division); 1989: Nichols WW, O'Rourke MF. McDonald's Blood Flow in Arteries. Theoretical, Experimental and Clinical Principles. London, England: Edward Arnold; 1990: Vanhoutte PM. Aging and vascular responsiveness. J Cardiovasc Pharmacol. 1988;12(suppl 8):S11-S Fleg JL, Tzankoff SP, Lakatta EG. Age-related augmentation of plasma catecholamines during dynamic exercise in healthy humans. J Appl Physiol. 1985;59: