INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

WE RECENTLY IDENTIFIED a greater decrease in arterial blood pressure (ABP) associated with a diminished tachycardiac response at the onset of 40 Torr lower body negative pressure (LBNP) in elderly compared with young adults (21). This attenuated increase in heart rate (HR) at the onset of orthostatic challenge results primarily from an age-related diminution in vagal cardiac function (21). When the muscarinic cholinergic (MC) antagonists atropine or glycopyrrolate were administered to young subjects to block cardiac parasympathetic efferent influences, their tachycardiac response to the onset of 40 Torr LBNP was abolished. The attenuated tachycardia was associated with a significant hypotension, a response similar to that seen in the elderly individual (21). When the ₁-adrenoceptor antagonist metoprolol was administrated to the same young subjects to block the cardiac sympathetic nerve efferent influence, neither the ABP nor the HR response was affected at the onset of 40 Torr LBNP (21). However, the impact of vagal cardiac function on ABP variability during graded sustained orthostatic challenge has yet to be established.

One of the major purposes of this study was to investigate the role of the parasympathetic efferent arm of the baroreflex in maintaining ABP stability. Previously, Saul et al. (19) observed that both systolic (SBP) and diastolic blood pressure (DBP) variance increased significantly during a body postural change from supine to standing with or without cardiac autonomic blockade. However, they did not describe the changes in SBP and DBP spectral density elicited by standing (19). Taylor et al. (24) demonstrated that the low-frequency (LF) power of ABP oscillation was significantly augmented during 40° head-up tilt. However, the augmentation in the LF power of ABP oscillation during tilt combined with atropine did not reach significance (23), probably due to the lack of significant hypotensive challenge. This investigation was designed to further consider the relationship between hypotension and LF ABP variability during graded and sustained hypotensive stimuli before and after vagal cardiac blockade.

Atropine is a common antimuscarinic agent that has been applied to study vagal cardiac function under the influence of heart disease (7, 13, 17) and during orthostatic challenge (19, 20, 22, 24). It has been established that baroreflex function is impaired after atropine treatment. Because atropine penetrates the blood-brain barrier, it blocks both central and peripheral MC receptors (1, 3) and alters central autonomic nervous system activity (10, 14). The antimuscarinic agent glycopyrrolate has little influence on the central nervous system (3). Equipotent doses of atropine and glycopyrrolate (2, 5) were thus applied to establish the vagal cardiac contribution to ABP stability during dynamic hypertensive and hypotensive stimuli. Although previous studies have compared atropine and glycopyrrolate differences in HR response (5) and HR variability (2), we believe that this is the first study to investigate the effects of these antimuscarinic agents on the dynamic control and regulation of ABP. We hypothesized that use of antimuscarinic agents would simulate vagal dysfunction and compromise ABP regulation but that the degree of change in ABP regulation with atropine blockade would differ from that during glycopyrrolate blockade of the MC receptors.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. Ten disease-free and drug-free volunteers (age, 22 ± 1.8 yr; n = 7 men and 3 women) were recruited as subjects. On a preliminary screening day, all subjects completed a medical history questionnaire and provided a resting 12-lead electrocardiogram (ECG) and arterial blood pressure (ABP). All female subjects who participated in the study had a negative pregnancy test before the initiation of the study. Eligible subjects received both oral and written explanation of the experimental protocol and signed a written consent form that was approved by the Institutional Review Board for the Protection of Human Subjects at the University of North Texas Health Science Center (Fort Worth, TX).

Measurements. During each experiment, beat-to-beat HR, SBP, DBP, and mean ABP (MAP) signals were digitized (Vetter Digital 3000) and collected using an on-line personal computer (Gateway 2000, 486DX) interfaced by an analog-to-digital board (Data Translation DT2801; Marlboro, MA) with a software sampling rate at 150 Hz. The lead II ECG signal was used to monitor HR on an electrocardiograph (Hewlett-Packard 78342A). On experimental day 1, an ABP signal was established using intraradial arterial catheterization. A physician, using local anesthesia to minimize subject discomfort, performed the catheter insertion. During experimental day 2, indirect radial tonometry (Colin 7000) was utilized. Previous data (11) indicated a close correlation between tonographic and direct ABP measurements. Our laboratory observations indicated that carefully monitored indirect measurements of radial arterial pressure by tonography were highly correlated with intraradial arterial catheter measurements (Fig. 1). Although radial pressure signals have a tendency to amplify systolic ABP values compared with brachial or aortic pressure measurements, this effect is insignificant when the comparisons are made before and after atropine or glycopyrrolate in the same subjects. A venous catheter was inserted into the subject's median antecubital vein for intravenous drug injections and blood samples. Venous blood was collected during baseline and 40 Torr LBNP before and after drug administration for measurement of the plasma catecholamines. Each 5-ml sample was centrifuged at 4°C for 15 min and frozen at 90°C, and the supernatant was analyzed using HPLC within 2 wk. Stroke volume (SV) and cardiac output (CO) values were collected by impedance plethysmography (Minnesota Impedance Cardiograph). Minute SV was averaged from the data collected over 50 s using commercially available software (Bio-Impedance Technology; Chapel Hill, NC). Peripheral vascular resistance (PVR) was then calculated from the equation PVR = MAP/CO. A 2.8% coefficient of variation was found for SV repeatedly measured in one subject at rest on three different days. Because thoracic impedance and tonographic ABP measurements shared the same computer channel, SV and CO were only determined during the experiments in which direct ABP was monitored.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   A summary of the relationship between beat-to-beat systolic (SBP) and diastolic blood pressure (DBP) measured by radial tonography and an intra-arterial catheter during 2-min baseline and 50 Torr lower body negative pressure (LBNP) conditions in one subject. Solid line, identity line.

Protocol. Before each experiment, subjects were familiarized with the procedures and measurements to be used during the test. All experiments were carried out with the subject supine in the LBNP chamber, with an ambient temperature of 24-26°C. The subjects were encouraged to stay relaxed throughout the testing. After 30 min of rest, baseline values for 10 min of HR, ABP, SV, and CO were recorded. At the end of baseline data collection, bilateral thigh cuffs were inflated to 200 mmHg for 3 min to create regional ischemia. The tachycardiac response to a systemic hypotension after deflation of the bilateral thigh cuffs was monitored. Phenylephrine was then injected at 1.0 µg/kg body wt to create transient hypertension. In addition, the Valsalva maneuver was performed in some subjects before and after parasympathetic blockade to compare the difference in hypertensive effect during the release phase (Fig. 2).

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Mean arterial pressure (MAP), R-R interval (RRI), and thoracic impedance during a Valsalva maneuver before (left) and after (right) vagal blockade with atropine in one representative subject. Ten-beat data were collected and calculated after the release of the Valsalva strain (dotted lines). The increase in thoracic impedance indicates the Valsalva strain induced a decrease in thoracic blood volume.

Data management. Each 10-min data set was plotted using SigmaPlot software and examined for variance, beat-to-beat pulse interval [R-R interval (RRI)], HR, SBP, and DBP from the selected 6-min segment with the least variation. The data from the 6-min "steady-state" segment were then linearly interpolated and resampled at 2 Hz. After resampling, the data were detrended with a third-order polynomial fitting and divided into 128-point segments with 50% overlap for fast Fourier transform analysis using the templates created with the data analysis and display software (DADiSP/ AdvDSP, DSP Development; Cambridge, MA). Group power spectral density curves for RRI, SBP, and DBP variability are shown in Fig. 3. Harmonic power in the high frequency (HF; 0.20-0.28 Hz) and LF (0.04-0.12 Hz) was extracted. In addition, HF and LF of RRI variability was normalized from the ratio of HF to LF + HF [HF/(LF + HF)] and the ratio of LF to LF + HF [LF/(LF + HF)]. The magnitude at LF and HF spectra was compared between the baseline and during LBNP before and after parasympathetic blockade.

View larger version (63K): [in this window] [in a new window]   Fig. 3.   Group RRI, SBP, and DBP variability during graded LBNP with and without parasympathetic blockade. There is a 10-fold reduction in the RRI variability (y-axis) after atropine and glycopyrrolate. Insets: amplified high-frequency RRI spectral density from 0.15 Hz.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline HR or pulse interval (RRI), SBP, DBP, MAP, pulse pressure (PP), RRI variability, and ABP variability were not different in the two experiments before MC blockade using atropine or glycopyrrolate. Thus the baseline data were combined into one baseline. Both atropine and glycopyrrolate increased the baseline HR without a significant effect on ABP (Table 1). Previously, studies (19, 23) have shown a significant increase in ABP after atropine, whereas others (17, 22) found no significant ABP augmentation. The cause for the inconsistency is unknown. The tachycardiac effect was not different between the two antimuscarinic agents. LBNP elicited significant tachycardia associated with significant decreases in SBP and PP. However, MAP and DBP were not significantly altered by LBNP. The magnitude and the rate of decrease in PP or SBP during LBNP were greater after MC blockade with atropine or glycopyrrolate. However, this change in PP or SBP was not statistically different between atropine and glycopyrrolate (Table 1).

Atropine and glycopyrrolate administration produced similar reductions (P < 0.05) in the baroreflex slope (calculated from the changes in HR to SBP) during systemic hypotension induced by the bilateral thigh cuff deflation after 3-min suprasystolic occlusion and systemic hypertension induced by a phenylephrine injection of 1 µg/kg body wt (Table 2). However, there was no drug-related difference in the baroreflex slopes between the two antimuscarinic agents. After atropine or glycopyrrolate blockade, the increase in ABP to the same dose of phenylephrine and the decrease in ABP to the same cuff deflation protocol were significantly augmented (Table 2).

Figure 2 illustrates individual changes in RRI and MAP after release of the Valsalva maneuver (a nonpharmacological intervention) before and after atropine administration. The group SBP and DBP (n = 5) were not different before the Valsalva maneuver with (134 ± 8 and 73 ± 9 mmHg) or without MC blockade (136 ± 5 and 72 ± 6 mmHg). After the release of the Valsalva maneuver, increases in SBP and DBP were substantially (P < 0.05) augmented with MC blockade (+51 ± 12 and +40 ± 7 mmHg) compared with control (+26 ± 9 and +13 ± 4 mmHg). The duration and intensity of the Valsalva strain, indicated by the increases in thoracic impedance, were similar before (15.1 ± 1.8 s and +1.75 ± 0.22 ) and after (14.4 ± 1.4 s and +1.49 ± 0.15 ) MC blockade (Fig. 2). This augmented ABP overshoot was associated with a significant diminution of the bradycardiac response (before vs. after parasympathetic blockade: RRI/t, +40.6 ± 7.0 vs. +6.8 ± 1.3 ms/s; or HR/t, 3.8 ± 0.8 vs. 1.9 ± 0.4 beats · min¹ · s¹; Fig. 2).

Parasympathetic blockade using atropine or glycopyrrolate significantly diminished RRI variability (Fig. 3). The density of the LF spectrum of the baseline RRI variability was significantly less after glycopyrrolate than after atropine blockade (Table 3). After normalization with the control (before the antimuscarinic agents), this difference between atropine (410 ± 121 ms²/Hz) and glycopyrrolate (346 ± 123 ms²/Hz) blockade was not statistically significant. The control values before atropine (416 ± 120 ms²/Hz) and glycopyrrolate (349 ± 123 ms²/Hz) blockade were not statistically different. During LBNP, RRI variability tended to increase in the LF spectrum and to decrease in the HF spectrum with or without normalization (Table 3). The LF spectrum of RRI variability during LBNP tended to be greater with atropine than glycopyrrolate (P = 0.08). However, the changes in RRI variability during LBNP were not significantly different between atropine and glycopyrrolate. The RRI variability LF-to-HF ratio was not affected by MC blockade; however, it was increased with LBNP (Table 3). Parasympathetic blockade decreased (P < 0.05) the baseline LF ABP variability without affecting the HF ABP variability (Table 4). Baseline LF ABP variability was significantly less after glycopyrrolate than after atropine blockade, and the effect was associated with the baseline LF RRI variability (Fig. 4). Application of LBNP augmented both LF SBP variability (P = 0.002) and LF DBP variability (P = 0.001; Fig. 3 and Table 4). The augmented LF ABP variability during LBNP was greater after parasympathetic blockade, and the interaction of LBNP factor and drug factor (parasympathetic blockade) on both SBP variability (P = 0.038) and DBP variability (P = 0.020) was significant (Table 4). The augmented LF ABP variability was significantly associated with the LBNP-induced systemic hypotension (Fig. 5). Figure 5 illustrates the group increases in LF SBP variability associated with decreases in SBP before and after parasympathetic blockade during graded LBNP (r = 0.79, P = 0.037), which was consistent with the correlation when calculated using individual data (r = 0.46, P = 0.001).

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Association of low-frequency (LF; 0.04-0.12 Hz) spectral density of SBP and log RRI variability at rest before (open symbols) and after (closed symbols) parasympathetic blockade using atropine (circles) and glycopyrrolate (squares). Shaded diamonds, group means ± SE. Before parasympathetic blockade, baseline group LF SBP and RRI variability was not statistically different between atropine and glycopyrrolate. Although group LF SBP and RRI variability was significantly smaller after glycopyrrolate than atropine, the changes from the baseline at rest were not statistically different between the two drugs. Dashed line, regression line for individual data (r = 0.68, P = 0.001).

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Association of increases in SBP variability at the LF spectrum and decreases in SBP induced by graded LBNP with and without parasympathetic blockade (r = 0.79, P = 0.037).

Atropine tended to have a greater effect on LBNP-induced augmentation in LF ABP variability compared with the glycopyrrolate effect. However, the difference between the two drugs did not reach a significance level in SBP variability (P = 0.09) or DBP variability (P = 0.16). Parasympathetic blockade with atropine or glycopyrrolate significantly decreased the magnitude of the transfer function of SBP variability to RRI variability (Table 4 and Fig. 6). There was no difference in the abolished transfer function between the two drugs. The coherence of the transfer function at the HF spectrum disappeared after atropine or glycopyrrolate treatment. LBNP significantly decreased the HF magnitude of the transfer function in the control condition (without parasympathetic blockade) but had no apparent effect on the LF transfer function. The coherence of the transfer function during LBNP tended to increase at the LF spectrum and decrease at the HF spectrum (Table 4).

View larger version (72K): [in this window] [in a new window]   Fig. 6.   Group data of the transfer function during graded LBNP with and without parasympathetic blockade.

Baseline plasma epinephrine and norepinephrine concentrations (0.43 ± 0.18 and 1.42 ± 0.21 pmol/ml) were not affected by atropine (0.22 ± 0.07 and 1.18 ± 0.33 pmol/ml) or glycopyrrolate (0.38 ± 0.12 and 1.21 ± 0.11 pmol/ml) blockade. LBNP (40 Torr) did not change epinephrine (0.39 ± 0.08, 0.18 ± 0.05, and 0.30 ± 0.01 pmol/ml in control, atropine, and glycopyrrolate, respectively) but significantly increased norepinephrine in the control (2.05 ± 0.27 pmol/ml) and MC blockade conditions (atropine and glycopyrrolate: 2.44 ± 0.78 and 1.63 ± 0.28 pmol/ml). The relative increase in norepinephrine during LBNP appeared to be greater (P = 0.18) after atropine (+108 ± 25%) than glycopyrrolate (+49 ± 26%) compared with the responses before MC blockade (+61 ± 31% vs. +61 ± 15%, P = 0.99).

There was no significant difference in CO after either atropine (n = 5) or glycopyrrolate (n = 3) administration. Because there was no observed drug difference and the sample size was small, CO and PVR data were merged and classified into "before" and "after MC blockade" groups. Drug administration did not significantly change CO (7.0 ± 0.3 vs. 6.1 ± 0.2 l/min), although SV decreased significantly from a baseline of 113 ± 8.5 to 58.3 ± 2.8 ml/beat after drug administration. PVR tended to increase after MC blockade (12.8 ± 0.7 to 14.9 ± 0.7 peripheral resistance units, P = 0.06). Figure 7 shows the decrease in CO in all groups during minute 2 (before vs. after blockade: 17.5 ± 2.2% vs. 29.6 ± 6.4%), minute 5 (20.3 ± 2.8% vs. 34.8 ± 4.5%), and minute 8 (22.4 ± 3.0% vs. 28.8 ± 3.9%) of 40 Torr LBNP and minute 2 (24.8 ± 3.3% vs. 29.2 ± 6.2%), minute 5 (27.5 ± 2.5% vs. 34.4 ± 4.0%), and minute 8 (28.4 ± 2.4% vs. 35.4 ± 5.1%) of 50 Torr LBNP, associated with a decrease in MAP (Fig. 7, left). The changes in CO and MAP were consistently greater after MC blockade (P = 0.001 from ANOVA for drug factor on %CO, P = 0.003 from ANOVA for drug factor on %MAP), although post hoc analysis with stratification of time or LBNP did not reach significance. However, the rate of decrease in MAP per unit decrease in CO was similar to the control condition (slope = 0.43 for both groups, P = 0.66 from analysis of covariance). In addition, a greater relative decrease in CO was explained by a reduced tachycardiac compensation after MC blockade (Fig. 7, right), because the decrease in CO in terms of per unit decrease in SV was greater (P = 0.016 from analysis of covariance) after antimuscarinic agents.

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Group changes in mean intraradial arterial pressure (MAP) associated with cardiac output (CO) (left) and group changes in CO associated with stroke volume (SV) during minutes 2, 5, and 8 of 40 and 50 Torr LBNP (right) before and after parasympathetic blockade. Left: slope 0 indicates perfect vasoconstrictive compensation, i.e., decrease in CO without decrease in MAP. Before CO decreases by 20%, there is no hypotension. Slope 1 represents no vasoconstritive compensation, i.e., a unit decrease in MAP explained by unit decrease in CO. The rate of decrease in MAP per unit decrease in CO was similar before and after vagal blockade (slope = 0.43). However, the decreases in CO and MAP are consistently greater (ANOVA, P < 0.05) after vagal blockade (left). Right: dashed line, no tachycardiac compensation, i.e., unit decreases in CO explained by unit decreases in SV. Tachycardiac compensation is seen at 20% CO before versus 30% CO after vagal blockade. The slope of %CO/%SV before parasympathetic blockade is 0.45, suggesting that 55% of the decreased SV is compensated by the tachycardiac response, which is diminished to 28% after vagal blockade (right).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study provided three new findings that extend previous observations on the importance of vagal cardiac function in maintaining ABP. First, our data manifested that MC blockade using atropine or glycopyrrolate was associated with an augmented amplitude of ABP fluctuations during dynamic hypertensive and hypotensive stimuli. These augmented ABP fluctuations were related to diminished baroreflex sensitivity. Second, our data indicated that sustained LBNP increased the LF spectral density of ABP variability and that the increased LF ABP oscillation was associated with systemic hypotension, which was augmented after MC blockade. Third, the difference in the arterial baroreflex function and ABP variability elicited by atropine and glycopyrrolate was not statistically significant. These data demonstrated that the integrity of vagal cardiac function was essential for the maintenance of ABP stability during various hemodynamic challenges.

Mechanism of the augmented amplitude of ABP variability. A previous study (8) reported that the pressor response to phenylephrine injection in tranquilized dogs was substantially augmented after atropine treatment combined with the -adrenoceptor antagonist propranolol. These authors postulated that the augmented pressor response to phenylephrine after propranolol and atropine was related to the blockade of a peripheral vasodilation mechanism modulated by the -adrenoceptor and the cholinergic fiber (8). Our data indicated that an augmented pressor response to phenylephrine was present after MC blockade alone with either atropine or glycopyrrolate (Table 2), suggesting that the -adrenoceptor mechanism of peripheral vasodilation described by Evans et al. (8) had little effect on ABP changes elicited by phenylephrine. Sanders et al. (18) reported a significant increase in forearm blood flow during contralateral arm exercise, which was attenuated after atropine. However, the overall response to exercise was an increase in MAP, indicating a limited functional significance for the cholinergic vasodilatory mechanism. Conversely, administration of atropine in conscious rats has demonstrated no effect on the resting ABP or blood flow in various tissues and organs (4). Although species and protocol differences must be considered when interpreting these results, we believe that these previous studies and our current data suggest a minimal physiological contribution from the -adrenoceptor or cholinergic fiber elicited vasodilation at rest. Furthermore, neither the previous study (8) nor the current data support the theory of a sensitized -adrenoceptor mechanism in the augmented pressor response to phenylephrine. If the -adrenoceptor sensitivity were enhanced by MC blockade, a hypotension or depressor response after bilateral thigh cuff inflation/deflation would be diminished after atropine or glycopyrrolate. However, MC blockade significantly exacerbated the systemic hypotension after the cuff deflation (Table 2).

LF ABP oscillations after antimuscarinic drugs. The LF spectral density of ABP variability tended to decrease after MC blockade. These data were consistent with a previous study (23) describing a significant decrease in the LF power of ABP variability after atropine with or without atenolol. This diminished LF ABP oscillation was more significant after glycopyrrolate than atropine administration (Table 4). The mechanism underlying this difference in the LF ABP oscillation before and after the MC blockade or between the two antimuscarinic agents was associated with a low LF RRI variability (Fig. 4).

LF ABP oscillation during LBNP. Previously, Taylor et al. (24) demonstrated that the LF power of SBP and DBP variability were significantly increased following postural change from supine to 40° head-up tilt. When the HR response was fixed by paced rhythm, the increased LF DBP power during head-up tilt was significantly greater than that during sinus rhythm (24). It has been demonstrated that the ABP variability around 0.1 Hz is significantly and positively correlated with the LF power density of MSNA over the variety of ABP range induced by nitroprusside and phenylephrine infusion (16). Furthermore, it has been observed that prazosin, an -adrenoceptor antagonist, reduces hemorrhage-augmented LF ABP variability (9, 15). Oz et al. (15) compared the LF ABP variability between Wistar-Kyoto normotensive (WKY) rats and spontaneously hypertensive rats (SHR) before and after hemorrhage with and without prazosin or the angiotensin enzyme inhibitor captopril. They demonstrated that hemorrhage increased the ABP variability in both groups of rats (15). Although the baseline ABP variability was significantly less in SHR than WKY rats, the increase in ABP variability after hemorrhage was greater in SHR than WKY rats. This hemorrhage-induced augmentation in the ABP variability was diminished to a similar degree in SHR and WKY groups after prazosin or captopril administration, suggesting that activation of both the sympathetic nervous system and the renin-angiotensin system was required for the response to hemorrhage (15). However, LF ABP oscillation is not significantly altered by blocking angiotensin-converting enzyme using enalaprilat in humans during a posture change from supine to 40° head-up tilt (23) or captopril in rats during hypotensive hemorrhage (9).

Implication of the difference in LF ABP oscillation. Application of LBNP has been commonly applied to simulate hemorrhage or orthostatic challenge. In this study, the augmented LF ABP oscillation during LBNP was attributed to changes in arterial vasomotion, modulated by the activation of the sympathetic nervous system. This enhanced vasomotion appeared to be related to systemic hypotension during LBNP, which in turn appeared to be an indication of ABP instability (Fig. 5). Our data indicated that this ABP instability, evidenced by the augmented LF ABP oscillation during LBNP, was exacerbated after vagal cardiac blockade.

Comparison of atropine and glycopyrrolate blockade. The present study suggested that after high-dose intravenous injection of atropine and glycopyrrolate, changes in HR and ABP during pressor and depressor response were not statistically different in young healthy humans. A relatively greater increase in plasma norepinephrine levels during LBNP occurred after atropine than after glycopyrrolate blockade. However, this difference did not reach statistical significance. The baseline LF ABP and RRI variabilities were lower after glycopyrrolate than atropine blockade. This antimuscarinic agent-related difference in the baseline LF ABP and RRI variability was related to the subject's individual control values, because the changes in baseline LF ABP and RRI variability (normalized with the values before the blockade) were not statistically different. The present study confirmed that the atropine inhibitory effect on the vagal cardiac function was dominant and that the efficacy of parasympathetic blockade using 40 µg/kg atropine and 16 µg/kg glycopyrrolate was equivalent.