Baroreflex sensitivity and the blood pressure response to -blockade

Journal of Human Hypertension (1999) 13, 185 190 1999 Stockton Press. All rights reserved 0950-9240/99 $12.00 http://www.stockton-press.co.uk/jhh ORIGINAL ARTICLE Baroreflex sensitivity and the blood pressure response to -blockade X Chen, MO Hassan, JV Jones, P Sleight and JS Floras Department of Cardiovascular Medicine, University of Oxford, and Centre for Cardiovascular Research, University of Toronto, Canada The objective of this analysis was to determine whether changes in baroreflex sensitivity (BRS) within 35 hypertensive patients (25 M, 10 F, mean age 47 years) treated with -blockade as monotherapy relate to reductions in ambulatory blood pressure (BP) or its variability. BP was recorded intra-arterially directly from the brachial artery before and during submaximal exercise. BRS was determined by the phenylephrine injection technique. MAP and its variability were determined for the awake period of 24-h BP monitoring. Subjects were randomised to one of atenolol, metoprolol, pindolol, or propranolol, and restudied after a mean of 5 months. -blockade increased BRS in 24 patients and decreased BRS in 11. BRS increased from 6.53 4.94 to 9.40 8.62 ms/mm Hg (mean s.d.) (P 0.01). Waking ambulatory MAP decreased from 125.8 15.8 to 106.4 16.2 mm Hg (P 0.0001), but its variability did not change. Higher BRS after chronic -blockade was associated with a decrease in waking ambulatory MAP (r 0.55, P 0.001), but not with its variability (r 0.08). blockade attenuated the pressor response to exercise, but there was a positive relationship between the effect of -blockade on BRS, and on the rise in systolic BP during bicycling (r 0.63; P 0.001). Any dampening effect of -blockade on BP variability at rest in hypertensive patients with the greatest increase in BRS may be offset by increased pressor responses to physical activity such as exercise. Consequently, BP variability is unaffected, even though reductions in ambulatory BP during chronic -blockade are inversely related to changes in BRS. BP responses to -blockade may be a function of the action of this class of drugs on BRS. However, there is considerable variation, between subjects, in their effect on BRS. This may have implications for other conditions, such as dilated cardiomyopathy, or following myocardial infarction, in which improvement in BRS is one mechanism by which -adrenoceptor blockade could improve survival. Keywords: ambulatory blood pressure; baroreflex sensitivity for heart rate; -adrenoceptor blockade; blood pressure variability; exercise; hypertension Introduction There is considerable variation between hypertensive patients in the blood pressure (BP) response to -adrenoceptor blockade. Hypertensive patients also vary with respect to the effect of chronic -adrenoceptor blockade on baroreflex sensitivity (BRS) for heart rate. 1 6 The purpose of the present analysis was to determine if these two observations are related. We have previously reported that the arterial baroreflex regulation of heart rate is impaired in patients with essential hypertension, 7 and that this diminished arterial baroreceptor reflex is associated with increased variability of ambulatory BP and augmented BP responses to mental and physical activities. 8 After chronic -blockade using both cardioselective and non-selective drugs, mean arterial ambulatory BP was significantly decreased, 9 the mean baroreflex sensitivity was significantly increased, 1 and the pressor response to exercise was Correspondence: Dr John S Floras, Division of Cardiology, Room 1614, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario, Canada M5G 1X5 Received 11 November 1998; revised and accepted 17 December 1998 attenuated. 9 However, there was considerable difference between individuals in the effect of chronic blockade on BP and on baroreflex sensitivity, increasing in some, and decreasing in others. Moreover, the functional implications of these changes in individual subjects, with respect to ambulatory BP and its variability, have not been reported. In a recent publication, Vesalainen et al 2 contrasted the effects of 4 weeks of monotherapy with metoprolol and ramipril on heart rate variability, baroreflex sensitivity, and ambulatory BP. Both drugs caused comparable reductions in ambulatory BP, but only metoprolol increased baroreflex sensitivity, and cardiac vagal activity, as assessed by heart rate, heart rate variability, and the high frequency component of heart rate variability during supine rest. Reductions in ambulatory BP correlated significantly with increases in RR interval, its total variability, and high frequency variability. However, there was no significant relationship between changes in baroreflex sensitivity and ambulatory BP in these patients. We undertook the present re-analysis of our original data set to determine whether improvement in arterial baroreflex sensitivity for heart rate, with chronic -adrenoceptor blockade, will result in correspondingly lower ambulatory BP, or BP varia-

186 -blockade, baroreflex sensitivity, and BP bility, or attenuated BP responses to submaximal exercise in hypertensive patients. Because it is now recognised that 24-h ambulatory BP and BP variability are more closely related to end-organ damage in hypertension than is the office BP measurement, 10 12 and that increased BP variability is an independent adverse risk factor for greater vascular and target organ damage, 10,13 this question now has clinical relevance. Subjects and methods Details of the study protocol have been published. 1,7 9 Observations in 35 subjects (25 males, 10 females) referred for the assessment of newly diagnosed, untreated hypertension are reported. Subjects ranged from 16 to 69 years old, with a mean age of 47 (s.d. 12) years. Blood pressure was measured on three or more clinic visits using a standard mercury sphygmomanometer. For the purpose of this study, patients were considered hypertensive when their clinic BP was 140/90 mm Hg or greater, if less than 40 years old, or at or above 160/95 mm Hg, if 40 years or older. Subjects with secondary hypertension were excluded. The study protocol was approved by the Hospital Ethics Committee and informed, written consent was obtained from each subject. Protocol Subjects were studied in the morning after a light breakfast at home, avoiding tea, coffee, or cigarettes. Arterial BP was recorded from a left brachial artery cannula, and an adjacent antecubital vein was also cannulated for injection of phenylephrine. ECG (lead II) and intra-arterial BP were recorded continuously for 24 h 7,8 and digitised for computer analysis. Baroreflex sensitivity was measured during supine rest, using the phenylephrine method. 7,8 Subjects performed a submaximal bicycle exercise test, 8,9 then left hospital to resume routine activities, and kept a diary noting, in particular, times of sleep and waking. Because the frequency response of the entire ambulatory recording and replay system falls off rapidly above 10 Hz, 14 we used mean, rather than systolic, BP in these calculations. Average awake mean arterial pressure (MAP) was computed using all valid beats over this period. Blood pressure variability was defined as the standard deviation of the waking mean arterial pressures. 7,8 After the first study, subjects were randomised to one of four adrenergic receptor blocking drugs: atenolol (n=9), metoprolol (n = 9), pindolol (n = 9), or slow-release propranolol (n=8). All drugs were taken once daily, in the morning, for 3 to 8 months (mean: 5 months). 1,9 Subjects were evaluated at monthly intervals at the same time of the day, and the medication dose adjusted if a cuff BP of 140/90 mm Hg or less was not achieved. The study protocol was replicated while on chronic treatment, with the last tablet taken at 08.00 am on the morning of the second study. Statistics Data are presented as mean ± standard deviation. Values for baroreflex sensitivity were logarithmically transformed to approximate a normal distribution. Student s t-test was used for paired data when comparing control versus treatment days. P values 0.05 were considered statistically significant. Results Chronic -adrenoceptor blockade increased baroreflex sensitivity in 24 subjects, and decreased it in 11. Patients whose BRS increased tended to be younger (45.6 ± 12.4 vs 50.6 ± 10.9 years), and comprised eight of the 10 women, but these age and sex distributions were not significantly different between the two groups. The average ambulatory mean arterial pressures (126 mm Hg) and heart rates (89 bpm) were identical in the two groups, and the number of patients in whom BRS increased was similar for each drug. Mean values were significantly augmented, from 6.53 ± 4.94 msec/mm Hg before, to 9.40 ± 8.62 msec/mm Hg during -adrenoceptor blockade (P 0.01). Corresponding values for the Log transformation of baroreflex sensitivity were 0.725 ± 0.276 and 0.831 ± 0.347 respectively (P 0.02). Paired data were available for analysis of ambulatory BP in 34 patients. As demonstrated in Table 1, adrenoceptor blockade reduced significantly MAP, from 125.8 ± 15.8 to 106.4 ± 16.2 mm Hg (P 0.0001). -blockade had no effect on the variability of MAP (14.8 ± 4.1 mm Hg untreated vs 15.0 ± 3.2 mm Hg on treatment). The coefficient of variation increased, from 11.9 ± 2.9% to 14.3 ± 3.2% (P 0.002). Increases in Log (BRS) were inversely related to reductions in ambulatory mean arterial pressure (r = 0.55, P 0.001) (Figure 1). As illustrated in Figure 2, there was no relationship between these changes in Log BRS and changes in the variability of mean arterial pressure, as a result of treatment (r = 0.08). Paired data were available for analysis of BP during exercise in 33 patients. Chronic -adrenoceptor blockade attenuated significantly the increase in BP during bicycle exercise. Systolic BP increased by 59.6 ± 18.5 mm Hg before, and 42.5 ± 20.2 mm Hg after, -blockade (P 0.05), while diastolic BP rose by 17.7 ± 9.4 mm Hg before and by 14.7 ± 9.1 mm Hg after -blockade (P 0.05). There was a positive linear relationship between increases in Log (BRS) on chronic -adrenoceptor blockade and changes in the systolic BP response to bicycle exercise (Figure 3, r = 0.63, P 0.001). Discussion The purpose of this re-analysis was to determine if increases in arterial baroreflex sensitivity for heart rate with chronic -blockade relate to reductions in ambulatory BP, BP variability, or blunted BP responses to bicycle exercise. There are several

-blockade, baroreflex sensitivity, and BP Table 1 Effects of chronic -adrenergic blockade on arterial baroreflex sensitivity, ambulatory blood pressure and its variability 187 Variable Pre -blockade -blockade Difference Ambulatory blood pressure MAP (mm Hg) 125.8 ± 15.8 106.4 ± 16.2 19.4 ± 14.8*** Variability (mm Hg) 14.8 ± 4.1 15.0 ± 3.2 0.14 ± 4.7 CV (%) 11.9 ± 2.9 14.3 ± 3.2 2.4 ± 4.0** Baroreflex sensitivity BRS (ms/mm Hg) 6.53 ± 4.94 9.40 ± 8.62 2.87 ± 5.88* Log (BRS) 0.725 ± 0.276 0.831 ± 0.347 0.106 ± 0.244+ Mean ± s.d. MAP; mean arterial blood pressure, CV; coefficient of variance, BRS; baroreflex sensitivity. + P 0.02, *P 0.01, **P 0.002, ***P 0.0001. Figure 1 Relationship between changes in arterial baroreflex sensitivity for heart rate (Log (BRS)) and changes in ambulatory awake mean arterial blood pressure after chronic -adrenoceptor blockade (n = 34, r = 0.55, P 0.001). unique aspects to this analysis that distinguish these findings from other reports. Blood pressure was measured directly, using intra-arterial ambulatory monitoring, and baroreflex sensitivity measured directly, using the phenylephrine technique. Only those BP data obtained when subjects were awake and active were included. Patients were studied after long-term treatment (on average 5 months) and, as the recordings were done approximately 5 h after taking the last dose of the drug, the effects of chronic, rather than the acute-on-chronic effect, of -blockade were assessed. Increased baroreflex sensitivity for heart rate following chronic -blockade in the present study was associated with reduced ambulatory BP. This finding contrasts with the smaller series of Vesalainen et al, 2 who restricted their study to metoprolol. Those authors demonstrated an inverse relationship between changes in ambulatory BP, and indices of cardiac vagal activity based on heart rate variability, but did not detect an inverse relationship between changes in baroreflex sensitivity and ambulatory BP. Despite the fall in average ambulatory BP, increased baroreflex sensitivity was not accompanied by attenuation of blood pressure variability, or by a blunting of the pressor response to exercise. Our principal observation cannot be explained as a non-specific response to reductions in BP with treatment. In a few of our patients (Figure 1), a reduction in awake ambulatory BP was observed in conjunction with a decrease in baroreflex sensitivity. Diuretics, angiotensin-converting enzyme inhibitors and calcium channel antagonists lower BP equally well, but they do not appear to increase baroreflex sensitivity consistently. 2,15 22 For example, in previous reports, captopril increased

-blockade, baroreflex sensitivity, and BP 188 Figure 2 Absence of relationship between changes in arterial baroreflex sensitivity for heart rate (Log (BRS)) and changes in the variability of ambulatory awake arterial blood pressure (n = 34, r = 0.08, P 0.6). Figure 3 Relationship between changes in arterial baroreflex sensitivity for heart rate (Log (BRS)) and the effect of chronic -adrenoceptor blockade on the rise in systolic blood pressure during bicycle exercise (n = 33, r =+0.63, P 0.001).

baroreflex sensitivity in normotensive and hypertensive humans, 15,16 and lisinopril in elderly hypertensives, 17 whereas ramipril, in a dose therapeutically equal to metoprolol, had no such effect. 2 There are conflicting reports on the effects of calcium channel antagonists on baroreflex sensitivity. 17 20 In contrast to the present analysis, neither of those studies, nor other studies of short-term treatment with -adrenoceptor antagonists 3,4 addressed the functional implications of these observations with respect to changes in ambulatory BP or its variability within individual subjects. When we first reported group mean data from this cohort, 1,7 we attempted to relate changes in baroreflex sensitivity to changes in systolic BP at the time of the phenylephrine injection, 1 but we did not specifically address the question as to whether changes in baroreflex sensitivity within subjects might determine the ambulatory BP response to blockade. However, recent developments prompted the analysis of these data, in the present form, with the goal of stimulating hypotheses for testing in future studies. For example, there has been increasing recognition of the importance of ambulatory BP monitoring and BP variability, as risk factors for end-organ damage in hypertension, 10 12 and increased BP variability is now recognised as an independent adverse risk factor for greater vascular and target organ damage. 10 12 Reduced baroreflex sensitivity following myocardial infarction is now known to be an important risk factor for premature mortality. 23 Whether increasing baroreflex sensitivity pharmacologically will improve prognosis has not been determined, but the benefits of chronic adrenoceptor blockade, when given following myocardial infarction, have been well established 24 and there is increasing interest in the administration of -adrenoceptor blockade to patients with left ventricular dysfunction. 25 There is considerable variation between patients in the BP response to -blockade, and the mechanism by which -adrenoceptor antagonism reduces BP remains obscure. Our data, which are derived from the entire ambulatory recording during the awake period, argue for a potential relationship between this hypotensive action, and augmented baroreflex sensitivity for heart rate. If so, this may be mediated at the level of the arterial baroreceptor afferents. 26 Long-term -adrenoceptor blockade also reduces central sympathetic outflow. 27 Thus, there may be two mechanisms by which the -adrenoceptor blockade could cause the vagal control of heart rate to predominate: a reduction in sympathetic traffic to cardiac nerves, and reduction in the normal sympathetic antagonism of vagal effects at the sinoatrial node. 28 If the BP response to -adrenoceptor blockade is indeed a function of changes in baroreflex sensitivity, future studies might focus on the reasons for the variability between subjects of the effects of blockade on the baroreflex control of heart rate. One factor may be the duration of hypertension that has elapsed prior to the initiation of treatment. When given to rats early in the course of spontaneous hypertension, atenolol increased the gain of the -blockade, baroreflex sensitivity, and BP baroreceptor reflex control of renal sympathetic nerve activity, and of heart rate, whereas, when given late in the course of spontaneous hypertension, these effects of atenolol were lost. 29 The authors of this experimental study suggested that there may be a critical phase representing the optimal time to initiate antihypertensive treatment. 29 A second factor may be patient age. It has been proposed that increases in BRS with chronic -adrenoceptor blockade occur primarily in younger hypertensives, 5,6 possibly due to alterations in compliance of the arterial conduits that house arterial baroreceptor afferents, or due to an age-related decrease in cardiac muscarinic receptor responsiveness. 30 Subjects whose BRS decreased tended to be older, in the present study, but this difference was not significant. We did not address the issue of how these changes in baroreflex sensitivity relate to changes in central and peripheral haemodynamics at rest, and during bicycle exercise, or the influence of these drugs on sympathetic nerve traffic to the heart and periphery. In the absence of such information, we can only speculate as to why there was a direct relationship between changes in baroreflex sensitivity and changes in systolic BP during bicycle exercise. For example, pulse pressure may have increased to compensate for a blunted heart rate response to exercise. These novel observations suggest that the antihypertensive action of -blockade may be a function of the effect of this class of drugs on the arterial baroreflex sensitivity for heart rate. However, our observations would indicate that -adrenoceptor antagonists are unlikely to attenuate any vascular and target organ damage in hypertensive patients arising from their increased BP variability. An effect of -adrenoceptor blockade on baroreflex sensitivity may be fundamental to the cardioprotective action of this class of drugs following myocardial infarction, or in mediating the medium- to long-term benefits of blockade in heart failure. If so, those patients whose baroreflex sensitivity does not improve following chronic -adrenoceptor blockade may be at highest risk for adverse outcome in these two conditions. This hypothesis merits prospective evaluation. References 1 Floras JS, Jones JV, Hassan MO, Sleight P. Effects of acute and chronic -adrenoceptor blockade on baroreflex sensitivity in humans. J Autonomic Nervous System 1988; 25: 87 94. 2 Vesalainen RK et al. Vagal cardiac activity in essential hypertension: the effects of metoprolol and ramipril. Am J Hypertens 1998; 11: 649 658. 3 Parati G et al. -adrenergic blocking treatment and 24- hour baroreflex sensitivity in essential hypertensive patients. Hypertension 1994; 23[part 2]: 992 996. 4 Lucini D, Pagani M, Malliani A. Improved baroreflex control of the heart rate with chronic beta-adrenergic blockade in mild hypertension. J Hypertens 1993; 11 (Suppl 5): S156 S157. 5 Simon G, Kiowski W, Julius S. Effect of beta adrenoceptor antagonists on baroreceptor reflex sensitivity in hypertension. Clin Pharmacol 1997; 22: 293 298. 6 Watson RDS, Stallard TJ, Littler WA. Effects of -adrenoceptor antagonists on sino-aortic baroreflex sensi- 189

190 -blockade, baroreflex sensitivity, and BP tivity and blood pressure in hypertensive man. Clin Sci 1979; 57: 241 247. 7 Floras JS et al. Factors influencing blood pressure and heart rate variability in hypertensive humans. Hypertension 1988; 11: 273 281. 8 Floras JS et al. Consequences of impaired arterial baroreflexes in essential hypertension: effects on pressor responses, plasma noradrenaline and blood pressure variability. J Hypertens 1988; 6: 525 535. 9 Floras JS, Hassan MO, Jones JV, Sleight P. Cardioselective and nonselective beta-adrenoceptor blocking drugs in hypertension: a comparison of their effect on blood pressure during mental and physical activity. J Am Coll Cardiol 1986; 6: 186 195. 10 Parati G et al. Relationship of 24-hour blood pressure mean and variability to severity of target-organ damage in hypertension. J Hypertens 1987; 5: 93 98. 11 Verdecchia P et al. Ambulatory blood pressure. An independent predictor of prognosis in essential hypertension. Hypertension 1994; 24: 793 801. 12 Mancia G et al. Ambulatory blood pressure is superior to clinic blood pressure in predicting treatmentinduced regression of left ventricular hypertrophy. SAMPLE Study Group. Study on Ambulatory Monitoring of Blood Pressure and Lisinopril. Circulation 1997; 95: 1460 1470. 13 Frattola A et al. Prognostic value of 24-hour blood pressure variability. J Hypertens 1993; 11: 1133 1137. 14 Stott FD. The Oxford portable blood pressure transducer. In: Clement DL, ed. Blood pressure variability. MTP Press: Lancaster, UK, 1979, pp 55 60. 15 Ebert TJ. Captopril potentiates chronotropic baroreflex responses to carotid stimuli in humans. Hypertension 1985; 7: 602 606. 16 Giannattasio G et al. Investigation of reflexes from volume and baroreceptors during converting enzyme inhibition in humans. Am Heart J 1989; 117: 740 745. 17 Egan BM et al. Improved baroreflex sensitivity in elderly hypertensives on lisinopril is not explained by blood pressure reduction alone. J Hypertens 1993; 11: 1113 1120. 18 Tomiyana H et al. Effects of an ACE inhibitor and a calcium channel blocker on cardiovascular autonomic nervous system and carotid distensibility in patients with mild to moderate hypertension. Am J Hypertens 1998; 11: 682 689. 19 McLeay RAB, Stallard TJ, Watson RDS, Littler WA. The effect of nifedipine on arterial pressure and reflex cardiac control. Circulation 1983; 67: 1084 1090. 20 Young MA, Watson RDS, Littler WA. Baroreflex setting and sensitivity after acute and chronic nicardipine therapy. Clin Sci 1984; 66: 233 235. 21 Young MA, Watson RDS, Littler WA. Acute and chronic effects of the loop diuretic, piretanide, on baroreflex set point and sensitivity in hypertensive man. J Hypertens 1985; 3: 481 483. 22 Kailasam MT et al. Divergent effects of dihydropyridine and phenylalkylamine calcium channel antagonist classes on autonomic function in human hypertension. Hypertension 1995; 26: 143 149. 23 La Rovere et al, for the ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. Lancet 1998; 351: 478 484. 24 Yusuf S et al. Beta blockade during and after myocardial infarction: An overview of the randomized trials. Progress in Cardiovasc Dis 1985; 27: 335 371. 25 Lechat P et al. Clinical effects of -adrenergic blockade in chronic heart failure. A meta-analysis of doubleblind, placebo-controlled, randomized trials. Circulation 1998; 98: 1184 1191. 26 Ichikawa M et al. Differential modulation of baroreceptor sensitivity by long-term antihypertensive treatment. Hypertension 1995; 26: 425 431. 27 Wallin BG, Sundlof G, Stromgren E, Aberg H. Sympathetic outflow to muscles during treatment of hypertension with metoprolol. Hypertension 1984; 6: 557 562. 28 Levy MN. Sympathetic-parasympathetic interactions in the heart. Circ Res 1971; 29: 437 445. 29 Kumagai K et al. Comparison of early and late start of antihypertensive agents and baroreceptor reflexes. Hypertension 1996; 27: 209 218. 30 Poller U et al. Age-dependent changes in cardiac muscarinic receptor function in healthy volunteers. JAm Coll Cardiol 1997; 29: 187 193.