INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE INTEREST IN BAROREFLEX SENSITIVITY (BRS), an indicator of cardiac autonomic regulation, has grown during the last decade. BRS becomes decreased in different cardiovascular diseases, including hypertension, coronary artery disease, and congestive heart failure (9). Perhaps most importantly, depressed BRS has been found to be a significant predictor of arrhythmic death in patients recovering from an acute myocardial infarction (10, 21, 22). In addition to cardiovascular diseases, physiological factors such as aging and gender influence BRS (1, 5, 6, 8, 14, 16, 17, 31, 32). Exercise capacity has also been found to have an impact on BRS (10, 16, 22), and there is evidence suggesting interaction between hormonal status and BRS (3, 29). However, in the previous studies the number of healthy subjects has been rather small (varying between 15 and 66) (5, 6, 8, 14, 16, 31, 32), in some cases only male subjects have been studied (8, 16, 31), and some studies have included also hypertensive subjects (14, 31). In addition, in earlier studies, the possible determinants of BRS have been studied separately, and no comprehensive studies that used a multivariate approach to assess the independent role of different BRS correlates have been published. Thus the aim of the present study was to evaluate the correlates of BRS and especially to investigate the relationships among BRS, age, and gender in a large population of healthy subjects.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Subjects. The study population consisted of 117 healthy subjects. They were divided according to age into three age groups representing young (23-39 yr, n = 44), middle-aged (40-59 yr, n = 38), and old (60-77 yr, n = 35) subjects. There was an approximately equal number of men and women in each group. The subjects were evaluated with a detailed history, physical examination, routine laboratory tests, and clinical exercise test. All subjects were free of hypertension and other systemic diseases. None of the subjects were on cardiovascular medication, and they were nonsmokers. Ten of the female subjects had postmenopausal hormone replacement therapy, and seven were taking oral contraceptives. All subjects were requested to abstain from beverages with caffeine for 12 h before the experiment. The subjects came to the laboratory between 7:30 and 9:00 AM, 45 min after ingestion of a light standard breakfast.

Electrocardiograph (ECG) and continuous blood pressure recordings. A Viggo-Spectramed (Helsingborg, Sweden) Teflon catheter was inserted into a large antecubital vein of left arm at least 1 h before the recordings. Electrodes were placed for the ECG, and a Finapres digital plethysmograph was installed around the middle finger of the right hand for continuous recording of arterial pressure (Ohmeda, Englewood, CO). ECG and blood pressure signals were recorded and simultaneously analog to digital converted with temporal resolution of 200 Hz/channel and amplitude resolution of 12 bits (35). All data acquisition, QRS detection (accuracy <2 ms), and analyses were performed with an IBM personal computer-compatible microcomputer with Cardiovascular Autonomic Function Test System (CAFTS) software (Medikro, Kuopio, Finland).

Assessment of BRS. Five minutes after the controlled breathing test, BRS was evaluated, following a modification of the method originally described by Smyth et al. (33). In brief, a bolus injection of 150 µg phenylephrine was administered into the antecubital vein to produce a rapid increase in blood pressure and a concomitant reduction in heart rate (increase of R-R interval). Beat-to-beat values of R-R intervals (RRIs) were plotted against the systolic arterial pressure (SAP) values of the preceding cardiac cycle [i.e., RRI(i) vs. SAP(i  1)] during a period in which blood pressure increased after phenylephrine injection. A linear regression analysis between RRI(i) and SAP(i  1) was performed and expressed as follows: RRI(i) (ms) = a + b · SAP(i  1) (mmHg), where i is one individual cardiac cycle and i  1 is the cardiac beat preceding the i beat. The slope of the regression line (b) represents BRS, and a and b are a constant and coefficient, respectively, that represent the linear regression (first-order equation). Only tests with correlation coefficients of r  0.80 or that were statistically significant (P < 0.05) were accepted. The phenylephrine test was repeated at 10-min intervals up to five times to get three acceptable measurements. The average of the three measurements was used for assessment of BRS. In two cases we were not able to get any acceptable BRS values because of technical problems or ectopic beats.

Assessment of heart rate variability. After subjects rested in the supine position for 1 h, heart rate variability was determined from 5-min ECG recordings. During the recording, subjects were asked to breathe according to a metronome set at 0.2 Hz for 5 min. Spectral estimates of R-R intervals were obtained from stationary regions free of ectopic beats. The length of these regions varied from 150 to 511 beats. After detrending (first degree), a modified covariance autoregressive model (fixed model order of 14) was used to obtain power spectral estimates of heart rate variability. Total power (variance) in the frequency range from 0 to 0.40 (Hz) was divided into three frequency bands: very-low-frequency band (0-0.07 Hz), low-frequency band (0.07-0.15 Hz), and high-frequency band (0.15-0.40 Hz). Signal powers of each band were calculated as integrals under the respective power spectral-density function and expressed in an absolute unit (ms²). In addition, time-domain analysis of heart rate variability was performed from the same regions by calculating SD of the R-R intervals and root mean square value of successive R-R interval differences.

Assessment of exercise capacity. Maximal exercise testing was performed by having the subjects pedal an electrically braked bicycle ergometer until exhaustion (Siemens Elema 380, Solna, Sweden) while they were in the sitting position. The initial workload was 20 W with subsequent increments of 20 W/min. During the test, perceived exertion, subjective symptoms, and blood pressure were recorded. In addition, heart rate, 12-lead ECG, and working time were recorded continuously by using a computerized system (Marquette Centra, Milwaukee, WI). Maximum oxygen uptake was measured as previously described (27).

Hormone measurements.. Blood samples for the assessment of plasma catecholamine, insulin, and arginine vasopressin concentrations and plasma renin activity were taken just before the phenylephrine tests. Within 1 h after the collection, the blood samples were centrifuged at 4°C for 15 min and plasma samples were aliquoted for different hormone measurements and stored immediately either at 20°C (insulin and renin activity) or 80°C (catecholamines and arginine vasopressin). Plasma catecholamine concentrations were measured by high-performance liquid chromatography with electrochemical detection (28). Plasma insulin and arginine vasopressin concentrations and plasma renin activity were measured by radioimmunoassay (RIA), insulin by using a commercial kit (Phasedeph, Pharmacia, Uppsala, Sweden), arginine vasopressin by using the RIA set in the laboratory with synthetic arginine vasopressin (375 IU/mg; gift from Ferring, Malmö, Sweden) and antiserum K9-IV (gift from Dr. L. C. Keil, National Aeronautics and Space Administration, Ames Research Center, Moffett Field, CA) (11), and renin activity by using a commercial kit (Medix angiotensin I test, Medix, Helsinki, Finland).

Statistical analyses. Because of skewed distribution, systolic blood pressure, exercise capacity, heart rate variability, BRS, and hormonal data were analyzed after logarithmic transformations. Student's t-test was used to test the significance of difference between genders. One-way analysis of variance with Duncan's multiple-range test was used to test the significance of differences between the age groups for continuous variables. Univariate correlations of BRS were calculated by using Pearson's correlation analysis. In addition, stepwise multiple-regression analysis was used to evaluate the independent correlates of BRS. A P value <0.05 was considered statistically significant. All values are presented as means ± SE. Statistical analyses were performed by using a SPSS/PC statistical package (version 5.0.1, SPSS, Chicago, IL).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Clinical neural and hormonal characteristics in relation to age and gender. Table 1 shows the clinical characteristics of the subjects in the three age groups studied. Systolic (P < 0.001) and diastolic blood pressures (P < 0.001) increased and maximum oxygen uptake (P < 0.001) decreased with aging. In addition, we found a significant decrease with aging in all time- and frequency-domain parameters of heart rate variability (Table 2). Baseline plasma norepinephrine concentration increased (P < 0.01), whereas baseline plasma renin activity decreased significantly with aging (P < 0.01) (Table 2).

Effects of age and gender on BRS. BRS correlated inversely with age in the pooled study population [ln BRS (ln ms/mmHg) = 3.77  0.032 × age (yr); r = 0.654, P < 0.0001]. BRS averaged 19.5 ± 1.4, 10.7 ± 1.2, and 6.0 ± 0.6 ms/mmHg in young, middle-aged, and old subjects, respectively.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Relationship between baroreflex sensitivity (BRS) and age in men (A) and in women (B). Regression slopes and 95% confidence intervals are shown.

Other univariate correlates of BRS. In addition to age and gender, systolic and diastolic blood pressures, heart rate, and heart rate variability indexes, maximum oxygen intake uptake during exercise test, plasma norepinephrine concentration, and plasma renin activity were significant univariate correlates of BRS (Table 6, Fig. 2, A-C).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Relationship between BRS and diastolic blood pressure (A), maximum O2 uptake (O2 max) during exercise (B), and high-frequency (HF) power of heart rate variability (HRV) (C). Regression slopes and 95% confidence intervals are shown.

Multivariate correlates of BRS. When the effect of age and gender on BRS was studied by using multivariate analysis, we found that these two factors together explained 52% (r² = 0.52) of the overall interindividual variation of BRS. When the significant univariate correlates of BRS were studied with stepwise multivariate analysis, BRS correlated independently only with age ( = 0.36, P < 0.001), gender ( = 0.29, P < 0.001), diastolic blood pressure ( = 0.22, P < 0.01), and high-frequency component of heart rate variability ( = 0.33, P < 0.001) (Table 7). All these factors together accounted for 63% (r² = 0.63) of the interindividual variation of BRS. The results of multiple regression analyses remained practically the same when diastolic blood pressure was substituted for systolic blood pressure or when men and women were analyzed separately (data not shown).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Assessment of BRS has become a promising tool in both experimental and clinical cardiology. Impaired BRS has been suggested to play a role in the development of hypertension (23, 30, 36). BRS has been found to be depressed in coronary artery disease (9), and in postinfarction patients BRS has been found to be a significant predictor of sudden death (10, 21, 22). The mechanisms leading to BRS depression in these cardiovascular diseases have, however, remained elusive. Clarification of these mechanisms and the use of BRS assessment in clinical practice require that the behavior of BRS in healthy population must be known. However, in earlier studies, the number of subjects has been limited, the age range has been narrow, and mainly male subjects have been studied. Thus these studies cannot be used to chacterize BRS distribution and determinants of BRS in a healthy population. Our study, which was based on healthy men and women of a wide age range, demonstrates that age and gender are the most important physiological correlates of BRS; they accounted for 52% of BRS variation. In addition, our study provides age- and gender-specific normal values for BRS.

BRS, age, and gender. Decrease of BRS with aging has been demonstrated in healthy subjects (6, 8, 14, 32) as well as in hypertensive patients (14) and in postinfarction patients (16). The mechanism responsible for decrease in BRS with aging is not known. Our findings support the concept of multifactorial mechanisms responsible for age-dependent BRS depression. In our study, the association between age and BRS probably reflects mechanisms related to changes in arterial compliance. Compliance of arterial walls decreases with aging (31), and, as a result, the stimulation of baroreceptors becomes reduced. In addition, age may contribute to BRS through changes in peripheral nervous pathways, central nervous control of the baroreflex system, and sinus node function. In our study, plasma norepinephrine concentration increased with aging and correlated inversely with BRS, whereas heart rate variabilty decreased with aging and correlated positively with BRS. This suggests that BRS decrease in elderly subjects reflects also changes in autonomic tone with aging.

BRS and blood pressure. In our study, low BRS was associated with elevated blood pressure levels corresponding with earlier studies (5, 14, 18, 20, 23, 30, 31, 32, 36). Systolic and diastolic blood pressures both correlated strongly with BRS. Reduced arterial compliance (18, 20, 31) and increased sympathetic activity (23, 36) have also been suggested to be responsible for the decreased BRS in hypertension. This is also in line with our results showing an inverse correlation between BRS and plasma norepinephrine concentration. There is also some evidence that impaired baroreflex regulation is not only a consequence of hypertension but also may contribute to the development of hypertension (23, 30, 36). Thus, besides being a consequence of high blood pressure, the decrease in BRS may be one of the mechanisms responsible for elevated blood pressure with aging.

BRS and autonomic nervous system. Cardiac vagal regulation consists of two components, tonic and dynamic. Heart rate variability represents the level of tonic vagal activity, whereas baroreflex sensitivity represents the dynamic component of cardiac vagal control. Previous studies suggest that preservation of both vagal tone and vagal reflexes protects against serious ventricular arrhythmias in postinfarction patients (10, 19, 21, 22). Although both are complementary components of vagal regulation, BRS has been found to provide somewhat different prognostic information than heart rate variability provides (10). In our study, BRS correlated positively with all frequency- and time-domain parameters of heart rate variability. In a previous study by Bigger et al. (2), BRS correlated only moderately with time- and frequency-domain variables of heart rate variability. The study of Bigger et al., however, was based on 24-h ECG recording, and our study was based on ECG recording during controlled conditions. The interaction between BRS and autonomic nervous tone is supported also by the findings that an increase in vagal tone by transdermal scopolamine increased both BRS and heart rate variability (4).

BRS and neurohormones. Circulatory vasoactive hormones such as arginine vasopressin and angiotensin influence cardiovascular regulation predominantly by controlling fluid balance and vascular smooth muscle tone. In addition, it has been suggested that they can modulate peripheral nerve activity through a baroreflex-dependent mechanism (3, 7, 24). Indeed, angiotensin II has been found to inhibit BRS in humans (24). However, it is probable that physiological arginine vasopressin levels do not have remarkable effects on BRS (12). These findings are in concordance with our results; we found only weak association between BRS and baseline plasma renin activity and no association between BRS and baseline arginine vasopressin. Also, other hormones have been suggested as influences on baroreflex function. An increase in plasma insulin elicits sympathetic activation (26), which might indirectly influence also BRS. However, there is only limited evidence about direct effects of insulin on BRS (29). Our findings do not exclude the possibility that, in situations also characterized by marked neurohumoral activation, neurohormones can have a more pronounced impact on BRS.

BRS and exercise capacity. Several studies have demonstrated a positive correlation between BRS and exercise capacity (10, 16, 22). In healthy subjects and borderline hypertensive patients, physical training has been found to be accompanied with augmentation of baroreflex control (13, 25, 34). In our study, exercise capacity correlated significantly with BRS in the univariate analysis. However, in the multivariate analysis, exercise capacity turned out not to be an independent correlate of BRS. This suggests that the effect of exercise capacity on BRS is not direct but is mediated through other factors e.g., sympathovagal balance.

Methodological considerations. In our study, the study population was large, equally distributed between genders, and had a wide age range. The subjects were well characterized and carefully evaluated for the exclusion of diseases and conditions known to influence BRS. This study was performed in a Caucasian population. Thus one has to be careful when applying our values in non-Caucasian populations.

Implications of the study. Impaired cardiac autonomic control has been associated with cardiac electrical instability and arrhythmias. A markedly depressed BRS (BRS value of 3 ms/mmHg) has been proposed to be useful in identifying patients at high risk of sudden death after myocardial infarction (10, 21, 22). In a large multicenter study [autonomic tone and reflexes after myocardial infarction (ATRAMI) study], the risk associated with markedly depressed BRS after myocardial infarction was higher in patients younger than 65 yr of age compared with older subjects (21). In addition, the mortality associated with depressed BRS was somewhat lower than expected based on earlier studies (10, 22). In the light of our results, the findings of ATRAMI study are not surprising. In our study 18% of healthy men >60 yr of age and particularly 24% of women >40 yr of age had markedly depressed BRS. Because of this, the positive predictive value of low BRS becomes reduced. In addition, because, in western countries, a great majority of myocardial infarction patients are old, the high prevalence of depressed BRS in healthy subjects may compromise the use of BRS in postinfarction risk stratification. These findings also raise the question of whether the absolute BRS value or the BRS decrease from the age- and gender-specific expected BRS should be used in risk assessment after myocardial infarction.