Baroreflex sensitivity changes with calcium antagonist ... - Nature

Journal of Human Hypertension (1999) 13, 87–95  1999 Stockton Press. All rights reserved 0950-9240/99 $12.00 http://www.stockton-press.co.uk/jhh

REVIEW

Baroreflex sensitivity changes with calcium antagonist therapy in elderly subjects with isolated systolic hypertension MA James1, H Rakicka1, RB Panerai2 and JF Potter1 University Departments of 1Medicine for the Elderly, Glenfield General Hospital and 2Medical Physics, Leicester Royal Infirmary, Leicester, UK

In order to study the effects of calcium-blocking therapy on cardiovascular homeostasis in elderly subjects with isolated systolic hypertension, we performed a randomised double-blind placebo-controlled crossover study of 6 weeks therapy with modified-release nifedipine or placebo. Changes with calcium-blocker treatment in clinic and 24-h blood pressure (BP), heart rate, BP variability, baroreflex sensitivity (BRS) by three methods (Valsalva manoeuvre, phenylephrine and sodium nitroprusside injection), and in baroreflex- and non-baroreflexmediated reflexes (tilt and cold face stimulus) were studied in 14 elderly subjects (mean age [± SEM] 70 ± 1 years) with sustained isolated systolic hypertension (clinic BP 179 ± 3/85 ± 1 mm Hg). Clinic systolic BP, but not diastolic BP, was reduced with treatment (by 14 ± 6 mm Hg, P = 0.03, diastolic BP 4 ± 3 mm Hg, P = 0.16).

Twenty-four hour BP was also reduced by nifedipine treatment (by 18 ± 3/9 ± 2 mm Hg, both P ⬍ 0.001). Clinic and 24-h heart rate, and daytime BP variability, were unchanged with treatment. BRS was significantly increased during nifedipine therapy by all three measurement methods (all P ⬍ 0.05). With 60ⴗ tilt during active treatment, subjects exhibited a greater heart rate increase (P ⬍ 0.01), and a reduced fall in systolic (P ⬍ 0.05) and diastolic BP (P ⬍ 0.05). Thus despite the arteriosclerosis and reductions in large artery compliance described in elderly patients with isolated systolic hypertension, clinically important improvements in clinic and ambulatory BP and some aspects of cardiovascular homeostasis can be achieved with calciumchannel blocking therapy.

Keywords: elderly; baroreflex; nifedipine; phenylephrine; nitroprusside; Valsalva’s manoeuvre

Introduction Dihydropyridine calcium antagonists have become widely used in clinical hypertension practice and are proven effective antihypertensive agents in the elderly.1 Nonetheless, it is only recently that evidence has emerged from large-scale clinical trials regarding their long term efficacy in preventing adverse cerebrovascular and coronary outcomes.2,3 Furthermore, controversy has recently arisen concerning the safety of short-acting dihydropyridine calcium antagonists in the treatment of hypertension and their possible adverse effects on cardiovascular physiology.4,5 The arterial baroreflex is the most important physiological mechanism for the short-term regulation of blood pressure (BP). An improvement in the sensitivity of the baroreflex with the treatment of hypertension has been regarded as a useful secondary goal of therapy, conferring theoretical benefits in terms of improved short-term BP homeostasis Correspondence: Dr MA James, Royal Devon & Exeter Hospital, Barrack Road, Exeter EX2 5DW, UK Received 10 July 1998; revised 5 October 1998; accepted 21 November 1998

and reduced BP variability,6 itself independently related to target-organ damage.7,8 Several studies have addressed the issue of the modification of arterial baroreceptor-cardiac reflex sensitivity (BRS) with antihypertensive treatment in young and elderly subjects, predominantly with either angiotensin-converting enzyme inhibitors9–12 or with calcium channel blocking drugs.12–17 The basis for the observed increases in BRS with treatment, although not seen in every study,17 is thought to be an increase in large artery compliance permitting resetting of baroreflexes. On this basis it could be proposed that elderly subjects with isolated systolic hypertension (ISH), who represent over half of all elderly hypertensives18 and who have demonstrable reductions in large artery compliance when compared to their peers with combined hypertension19–21 (usually attributed to more advanced arteriosclerosis), may not respond with the same alterations in BRS with treatment as have been described in younger subjects.13,14 This may lead to impaired cardiovascular homeostasis and practical therapeutic problems such as orthostatic intolerance.22 Conversely, an increase in BRS may lead to improved cardiovascular reflex responses and reduced BP variability. The purpose of this study

Baroreflexes and nifedipine in the elderly MA James et al

88

therefore, was to examine if and to what extent BRS could be modified by long-acting calcium-channel blocking therapy in elderly subjects with sustained ISH.

Materials and methods Fourteen untreated elderly (aged ⭓60 years) subjects with sustained clinic ISH (systolic BP ⭓160 mm Hg with diastolic BP ⬍90 mm Hg) were studied. Subjects were recruited from among out-patients attending for assessment of their hypertension at two large teaching hospitals and through a liaison with several large local general practices. All subjects were active and living independently in the community, with a normal physical examination, on no treatment with cardiovascular or autonomic effects, and all were in sinus rhythm with a normal electrocardiogram. No subject had clinical evidence of secondary hypertension, ischaemic heart disease, cerebrovascular disease or renal disease, and none had previously received antihypertensive treatment. Clinic BP was measured using a standard mercury sphygmomanometer and a cuff of appropriate size, after 5 min of supine rest on three separate occasions at least 2 weeks apart before entry to the study. Subjects’ height, weight and body mass index (BMI; weight divided by square of height, kg/m2) were recorded. All subjects gave written informed consent to participate in the study which was approved by the local ethics committee. Study design Following entry to the study, subjects underwent 24 h ambulatory BP monitoring (ABPM) before returning the following day for the laboratory protocol for assessment of cardiovascular reflexes (described below). Subjects were then randomised to receive either nifedipine GITS (‘gastrointestinal system’; Adalat LA) 30 mg once daily or matching placebo for the first 6-week period of a double-blind randomised crossover study design.23 Subjects were reviewed after 3 weeks of active or placebo treatment at which visit specific enquiry was made regarding side effects. If target supine systolic BP had not been achieved (systolic BP ⬍160 mm Hg), therapy was doubled to nifedipine GITS 60 mg daily or two matching placebo tablets. Clinic BP (both supine and after 1 min of standing) were recorded again after 6 weeks, at which stage subjects underwent 24-h ABPM and assessment of cardiovascular reflexes using an identical laboratory protocol. Subjects then crossed over to the alternative treatment for the second 6-week period, with review of BP and side effects at the halfway stage in a similar manner. At the end of the second period the study was completed with a repeat 24-h ABPM and a further visit to the study laboratory for measurement of cardiovascular reflexes and arterial BRS. 24-hour ABPM Ambulatory BP monitoring was performed using the SpaceLabs 90207 system (SpaceLabs Inc, Redmond,

WA, USA). The monitor was programmed to take readings at 15-min intervals between 07.00 and 22.00, and at 30-min intervals between 22.00 and 07.00. This frequency of recording enabled an estimate to be made of daytime BP and heart rate variability in individual subjects.24 When subjects attended the research clinic to have the monitor fitted, using the large cuff if necessary, the accuracy of the monitor in each individual subject was verified by sequential same-arm comparison with a standard mercury sphygmomanometer.25 Monitor readings were acceptable if within 5 mm Hg of readings obtained by the standard method. Subjects were not given specific advice regarding physical activity apart from that necessary to preserve the accuracy of monitor readings. The monitor was removed 24 h later when subjects attended the cardiovascular laboratory. Data were downloaded to an IBM compatible computer without manual data editing, and only records with over 90% of possible readings were regarded as acceptable. Laboratory cardiovascular studies An identical protocol for estimation of cardiovascular reflexes and BRS was conducted at entry to the study (‘baseline’) and at the end of each 6-week active or placebo treatment period. Subjects attended the research laboratory for a morning session, having emptied the bladder and following a light breakfast, and having refrained from smoking, alcohol and caffeine for at least 12 h. The laboratory temperature was thermostatically controlled at 20°C. Subjects rested supine after the insertion of a cannula into a dorsal vein in the right hand, and were fitted with chest leads for recording of the continuous surface electrocardiogram (model CR7, Cardiac Recorders Ltd, London, UK). The appropriate-sized finger cuff of the Finapres 2300 non-invasive beatto-beat BP recording device (Ohmeda Monitoring Systems, Englewood, CL, USA) was fitted to the middle finger of the left hand, which rested throughout on an adjustable support at the level of the heart. The subject’s right forearm was placed in a thermostatically controlled warmer at 55°C, and after 30 min rest a 10 mL arterialised venous blood sample was drawn without occlusion, centrifuged and the plasma frozen at −70°C for later catecholamine analysis.26 Recordings were then made of the variation in RR interval with deep breathing at six breaths/minute for 1 min, and of the cardiovascular responses to the Valsalva manoeuvre. After several practices, subjects blew into a modified sphygmomanometer to maintain a pressure of 40 mm Hg for 15 sec whilst seated. The manometer contained a small air leak so that subjects had to maintain a constant respiratory effort. The manoeuvre was repeated three times with 5 min rest between each. Measurement of sino-aortic arterial BRS was then performed from the BP and pulse interval responses to phenylephrine injection.27,28 An initial bolus dose of 50 ␮g phenylephrine was progressively increased in 50 ␮g steps as necessary (up to a maximum of 200 ␮g) to achieve a peak systolic BP rise of 20– 40


mm Hg, and the effective dose was repeated to obtain a minimum of three adequate responses. Bolus injections of 0.9% saline were interspersed at random between the drug injections and the subject was blinded to the nature of each injection. BRS was similarly assessed from the BP and pulse interval responses to a graded sodium nitroprusside infusion.29,30 The infusion was commenced at 0.25 ␮g/kg/min and increased (by 0.25 ␮g/kg/min each minute) until a fall in systolic BP of 20– 40 mm Hg was observed. The cardiovascular responses to head-up tilt were then recorded. After familiarising subjects with the tilting procedure, they were tilted to 60° while lightly strapped to a hydraulic tilt table fitted with foot support (Akron Medical Products, Ipswich, UK) and then held in that position for 3 min. The move from the supine to the tilted position took about 5 sec. During tilt BP and pulse interval were recorded from the Finapres, and forearm blood flow (FBF) in the right arm was measured with a mercury-in-silastic strain guage plethysmograph (QMC Medical Physics, Nottingham, UK). The hand was excluded from FBF estimations by the inflation of a wrist cuff to 40 mm Hg above systolic pressure. FBF was recorded for 1 min at baseline and then during the 3 min of tilt. This manoeuvre was repeated a further two times with at least 5 min in between and the average of the three responses was taken. After returning to the supine position and a further period of supine rest, subjects underwent the cold face stimulus, where the BP, heart rate and FBF responses to the application of a refrigerated gel pack (0– 4°C) to the subject’s forehead for 45 sec were recorded.31 Subjects wore protective glasses to avoid the stimulation of the oculo-cardiac reflex. The average of two responses was taken. A suitable interval (at least 5 min) was observed between all the above tests to permit the recovery of stable baseline values for BP and heart rate. During all the above manoeuvres the hand and forearm attached to the Finapres was supported on an adjustable shelf to maintain it at heart level and eliminate hydrostatic artefacts, and the contralateral arm attached to the FBF equipment was also raised on an adjustable support to allow passive venous emptying and to retain the angle to the horizontal during tilt. The entire duration of the laboratory protocol was approximately 3 h. Data and chemical analyses The analogue outputs from the Finapres and the simultaneous ECG were routed to a dedicated personal computer fitted with an analogue-to-digital converter sampling at 200 Hz per channel. A third channel recorded pressure from a transducer and amplifier connected to the sphygmomanometer used for the Valsalva manoeuvre. Purpose-written software allowed the recording, calibration and editing of the digitised signal and the derivation for later off-line analysis of beat-to-beat data for systolic, mean arterial and diastolic BP, together with the pulse interval from the Finapres and ECG signals and the Valsalva pressure. The data from the three

Valsalva manoeuvres were analysed to derive an index for BRS from Phase 4 of the manoeuvre according to the method of Smith et al.32 ValsalvaBRS thus described the BRS from the linear regression of pulse interval on systolic BP for the overshoot portion of Phase 4 (from the first beat where systolic BP exceeded that prevailing before the manoeuvre to the peak value observed several seconds later). BRS was derived from the phenylephrine ‘ramp’ technique by the method of Smyth et al,27 but including both inspiratory and expiratory beats. A value for phenylephrine-BRS for each individual was obtained from the average of the (minimum three) effective phenylephrine responses. Nitroprusside-BRS was derived from the sodium nitroprusside depressor response by the same ‘ramp’ method. FBF was expressed as ml/min/100 mls of forearm tissue, and the ratio of FBF to the mean arterial pressure prevailing during that cycle (taken from the simultaneous Finapres recording) was calculated and expressed in arbitrary units representing forearm vascular resistance (FVR). Baseline values for FBF prior to tilting and the cold face stimulus were obtained from the mean of the four curves in the minute immediately before the stimulus. FBF and FVR values at 15, 30, 60, 90, 120 and 180 seconds following tilt were considered. The mean of the three values obtained in each subject for each of those time points was taken as the final value. Similarly the two values were averaged for 15, 30, and 45 sec following application of the cold face stimulus. Plasma adrenaline and noradrenaline samples were stored at −70°C and duplicate samples were analysed using high pressure liquid chromatography with electrochemical detection. The coefficient of variation of this technique has previously been shown to be 8% for noradrenaline and 13% for adrenaline.26 Statistical analysis Data are presented as mean ± SEM or as median (interquartile range). After testing for period effects and treatment-period interactions,33 analysis of treatment effects was performed by comparison of changes from baseline after the treatment and placebo phases of the study. Paired t-tests were used after confirmation of normality using the ShapiroFrancia W-test, but for the BRS and catecholamine data non-parametric tests were used (Wilcoxon rank sum test). For the tilt and cold face stimulus data, statistical analysis was performed using two-way repeated measures analysis of variance, with group (nifedipine or placebo) and time point as factors. Summary statistics (systolic and diastolic BP and heart rate change, and time to maximum change) were also examined for the tilt and cold face stimulus tests. A probability of ⬍0.05 was regarded as indicating statistical significance.

Results Thirteen subjects successfully completed the study, one subject having withdrawn within the first

89


90

period because of side effects on placebo. Other side effects during the active treatment period were minor and consisted of slight ankle swelling in three patients and mild headache on initiating therapy in two subjects. None required dose adjustment or stopping therapy. Seven women and six men of mean age 70 ± 1 years (range 60–77 years), weight 75.9 ± 2.5 kg and BMI 27.2 ± 0.7 kg/m2 completed the study. Group mean supine clinic systolic BP at entry to the study was 179 ± 3 mm Hg, range 161– 197; diastolic BP was 86 ± 1 mm Hg, range 75–89. After 6 weeks of nifedipine therapy, clinic supine systolic BP was significantly lower by 14 ± 6 mm Hg compared to the corresponding placebo period (P = 0.028; Figure 1 and Table 1). Clinic supine diastolic BP was not significantly affected by nifedipine treatment, being reduced by only 4 ± 3 mm Hg compared to placebo (P = 0.16). Similarly, clinic heart rate was not affected by treatment (difference nifedipine vs placebo −2.5 ± 2.0 bpm, P = 0.26, Table 1). Active treatment did not affect the changes in BP or heart rate after 1 min of standing in the clinic. At baseline, the change in BP on standing was −6 ± 3/+9 ± 2 mm Hg. After 6 weeks’ nifedipine therapy, the clinic BP change at 1 min of standing was −6 ± 3/+4 ± 2 mm Hg; similarly after 6 weeks of placebo the clinic BP change on standing was −9 ± 4/+7 ± 2 mm Hg (P = 0.39 for systolic BP, P = 0.20 for diastolic BP). The heart rate response to active standing was similarly unaltered by active treatment (Table 1). Data were available for 24-h ABPM recordings at

Table 1 Clinic and ambulatory BP and heart rate at baseline and after 6 weeks’ active (nifedipine) or placebo therapy

Figure 1 Clinic and 24-h blood pressure and heart rate at baseline (entry to the study; BL) and after 6 weeks therapy with placebo (P) or nifedipine (N). BPM: beats per minute.

Figure 2 24-h ambulatory blood pressure and heart rate at baseline (BL; solid lines) and after 6 weeks therapy with placebo (dotted lines) or nifedipine (dashed lines).

Clinic Supine

SBP DBP HR Standing SBP DBP HR Ambulatory* 24-h SBP DBP HR Daytime SBP var DBP var HR var

Baseline

Placebo

Nifedipine

P

179 ± 3 86 ± 1 66 ± 2 173 ± 3 94 ± 3 70 ± 3

172 ± 6 80 ± 3 66 ± 3 162 ± 6 87 ± 3 71 ± 3

157 ± 5 75 ± 3 64 ± 2 152 ± 6 80 ± 3 69 ± 2

0.028 0.16 0.26 0.12 0.047 0.19

151 ± 2 151 ± 3 133 ± 2 83 ± 3 84 ± 2 75 ± 2 72 ± 2 70 ± 2 70 ± 2 13.5 ± 0.5 13.1 ± 1.0 13.4 ± 0.9 10.7 ± 0.6 9.7 ± 0.9 9.4 ± 0.7 9.3 ± 1.2 9.0 ± 1.0 9.6 ± 0.9

0.0002 0.0009 0.88 0.71 0.63 0.50

*n = 11. Data are given as mean ± SEM. Daytime: 07.00 to 22.00. Var: variability.

the end of both the nifedipine and placebo periods in 11 subjects. Both 24-h systolic and diastolic BP were significantly lower after nifedipine treatment (by 18 ± 3/9 ± 2 mm Hg; Table 1 and Figure 1) with 24-h heart rate unaffected by therapy (Figure 2). Daytime BP and heart rate variability were calculated from the standard deviation of all daytime reading taken at 15-min intervals between 07.00 and 22.00 for each individual.24 The ‘longer-term’ variability of BP and heart rate was unchanged from baseline after both the nifedipine and placebo treatment periods (Table 1).


Figure 3 The systolic blood pressure and pulse interval responses to bolus phenylephrine injections in a typical study subject.

The systolic BP and pulse interval responses to bolus phenylephrine injections in a typical study subject are shown in Figure 3. The results of cardiovascular reflex and arterial BRS testing for the laboratory visits at baseline and at the end of the nifedipine and placebo periods in all 13 subjects are shown in Table 2. Heart rate variability with regular deep breathing was unaffected by nifedipine treatment. By the Valsalva method, 12 subjects demonstrated an increase in their BRS with 6 weeks’ active treatment and one a decrease compared to placebo. After nifedipine treatment, Valsalva-BRS was a median 1.6 msec/mm Hg higher (95% confidence interval (CI) 0.10–3.20) than after placebo. By the phenylephrine pressor method, 11 subjects showed an increase in BRS following the nifedipine phase, and two a decrease compared to placebo. Phenylephrine-BRS was a median 1.8 msec/mm Hg higher with treatment (95% CI 0.30–3.20) compared to after the corresponding placebo period. With the sodium nitroprusside depressor method, again 11 subjects manifested an increase in nitroprusside-BRS with active treatment, and two a decrease. NitroprussideBRS was higher by a median 2.3 msec/mm Hg after nifedipine therapy (95% CI 0.23– 4.35) compared to after placebo. Overall eight subjects demonstrated an increase relative to placebo in all measures of

BRS with active treatment and although the absolute treatment effects on BRS were small, these represented relative increases with nifedipine of 46%, 78% and 190% in Valsalva, phenylephrine and sodium nitroprusside measures of BRS respectively. Changes in BRS after 6 weeks placebo or nifedipine therapy are indicated in Figure 4. There were no significant differences between the different methods for the assessment of BRS (Valsalva, phenylephrine or nitroprusside) at baseline or during either the nifedipine or placebo periods. Changes in plasma adrenaline and noradrenaline levels, and baseline FBF and FVR were no different between the nifedipine and placebo phases of the study (Table 2). In examining the overall response of FBF and FVR to tilt, there was no significant difference in the proportional fall in FBF between nifedipine and placebo (Figure 5; effect of group: F = 2.91, P = 0.09; effect of time: F = 0.55, P = 0.74). Similarly, the overall increase in FVR with tilt was not different between nifedipine and placebo (Figure 5; effect of group: F = 1.26, P = 0.26; effect of time: F = 0.37, P = 0.86). The maximum BP and heart rate responses to tilt and the cold face stimulus at baseline and after each of the study phases are given in Table 3. The maximum heart rate increase with

Figure 4 Individual changes in baroreceptor sensitivity by the three methods with placebo (P) and nifedipine (N) treatment. *P ⬍ 0.05; **P ⬍ 0.01 for treatment effect. Horizontal bars represent group medians.

Table 2 Cardiovascular reflexes, forearm blood flow and plasma catecholamines at baseline and after 6 weeks’ active (nifedipine) or placebo therapy

Heart rate variability (bpm) BRS Valsalva PE SNP Forearm blood flow (ml/min/100 ml) Forearm vascular resistance (units) Plasma noradrenaline (nmol/L) Plasma adrenaline (pmol/L)

Baseline

Placebo

Nifedipine

P (treatment effect)

8.8 ± 1.0 4.3 (2.5–6.4) 3.5 (1.2–6.1) 1.8 (0.9– 4.6) 2.3 ± 0.2 51.4 ± 5.8 1.29 (1.02–2.69) 165 (92–307)

10.4 ± 1.4 3.5 (2.2– 4.7) 2.3 (1.4 –6.0) 1.2 (0.4 –4.3) 2.3 ± 0.3 54.2 ± 6.0 1.03 (0.49–1.43) 121 (82–150)

9.1 ± 0.8 4.6 (3.9–7.6) 3.7 (2.5–8.5) 3.5 (2.4 –5.8) 2.7 ± 0.4 42.7 ± 5.7 1.28 (0.44 –2.05) 123 (75–218)

0.48 0.019 0.006 0.03 0.34 0.17 0.55 0.39

Data are given as mean ± SEM or median (interquartile range). Bpm: beats per minute; BRS: baroreflex sensitivity; PE: phenylephrine; SNP: sodium nitroprusside. All baroreflex sensitivities given in msec/mmHg.

91


92

Figure 5 Changes in forearm blood flow and vascular resistance with 60° passive tilt at baseline (entry to the study; solid line, open circles) and after 6 weeks therapy with placebo (dotted line, closed circles) or nifedipine (dashed line, open circles). FBF: forearm blood flow; FVR: forearm vascular resistance; some error bars omitted for clarity.

tilt was significantly greater (P = 0.039), and the maximum falls in systolic and diastolic BP were significantly less (P = 0.03 and P = 0.047 respectively), following the nifedipine phase of the study, although when the entire response was examined the overall heart rate and systolic BP responses to tilt were not different between nifedipine and placebo (for heart rate; effect of group: F = 3.27, P = 0.07; effect of time: F = 0.73, P = 0.63. For systolic BP; effect of group: F = 0.02, P = 0.89; effect of time: F = 0.84, P = 0.53, Figure 6). In general there was a fall of approximately 20% in FBF with a rise in FVR of about 25% following

Figure 6 Changes in heart rate and systolic blood pressure with 60° passive tilt at baseline (entry to the study; solid line, open circles) and after 6 weeks therapy with placebo (dotted line, closed circles) or nifedipine (dashed line, open circles). BPM: beats per minute; some error bars omitted for clarity.

application of the cold face stimulus. However, the changes in both FBF and FVR with the cold face stimulus were not different between the nifedipine and placebo phases of the study (For FBF; effect of group: F = 3.19, P = 0.08; effect of time: F = 0.04, P = 0.97. For FVR; effect of group: F = 1.49, P = 0.28; effect of time: F = 0.04, P = 0.97). Similarly, although systolic BP rose by on average 10–25 mm Hg with the cold face stimulus, the changes in systolic and diastolic BP and heart rate were not significantly different between the two phases of the study (Table 3).

Table 3 Maximum observed changes in blood pressure and heart rate in response to 60° passive tilt and the cold face stimulus at baseline and after 6 weeks active (nifedipine) or placebo therapy Baseline

Placebo

Nifedipine

Time 0

Maximum/ minimum

Change

Time 0

Maximum/ minimum

Change

Time 0

Maximum/ minimum

Change

Tilt SBP DBP HR

169 ± 7 77 ± 3 61 ± 2

121 ± 8 59 ± 4 77 ± 3

48 ± 6 18 ± 2 16 ± 2

170 ± 8 79 ± 4 62 ± 3

124 ± 6 64 ± 3 76 ± 3

46 ± 6 16 ± 3 14 ± 2

154 ± 5 72 ± 3 61 ± 2

124 ± 4 62 ± 2 82 ± 3

30 ± 4* 10 ± 2* 21 ± 2**

Cold face stimulus SBP DBP HR

163 ± 8 77 ± 3 61 ± 3

182 ± 7 84 ± 4 63 ± 3

18 ± 4 7±2 2±1

166 ± 7 79 ± 3 62 ± 2

185 ± 11 89 ± 6 63 ± 2

19 ± 4 10 ± 3 1±1

156 ± 6 73 ± 3 63 ± 2

169 ± 7 79 ± 3 67 ± 3

14 ± 3 6±1 5±2

Data are given as mean ± SEM. SBP: systolic blood pressure; DBP: diastolic blood pressure (both in mm Hg); HR: heart rate (beats per minute). *P ⬍ 0.05, **P ⬍ 0.01.


No significant period effects or treatment-period interactions were found in the analysis of clinic or ambulatory BP, or arterial BRS.

Discussion This study of the effect of 6 weeks of therapy with a long-acting calcium-channel blocking agent in elderly subjects with isolated systolic hypertension has demonstrated a significant reduction in both clinic and ambulatory systolic BP with a smaller fall in 24h ambulatory diastolic BP and no change in heart rate. The study has also shown a modest but statistically significant increase in baroreceptor-cardiac reflex sensitivity when measured by three comparable methods,34 which appears to translate into an improved reflex response to orthostatic stress with a greater increase in maximum heart rate and a consequent smaller fall in systolic and diastolic BP with passive tilt. Notably, both the vagal efferent function and the sympathetically-mediated cold face stimulus were unaffected by nifedipine treatment. The reliability of the Finapres beat-to-beat BP monitoring device in cardiovascular studies of this kind has previously been established during pressor and depressor drug injection, Valsalva manoeuvre and vasomotor stimuli.35–37 Previous studies have provided us with varying evidence of the effects of antihypertensive treatment in general, and calcium-channel blocking therapy in particular, on the function of the arterial baroreflex in human hypertension.17 Few of these studies have gone on to include an assessment of the implications of changes in the baroreflex on the physiological response to orthostatic stress. No published studies have looked at the specific case of elderly subjects with isolated systolic hypertension, despite the prevalence of this condition and the different pathophysiology that characterises it.20,21,34,38,39 This study is thus the first to indicate that some of the adverse pathophysiology of ISH can be reversed by treatment with calcium-blocker therapy with few problems in terms of tolerability. The mechanism by which these beneficial effects are seen cannot be asserted with certainty from this study. In particular, the study does not permit the assertion that the observed effects are independent of the BP reduction achieved with active treatment, or specific to calcium-channel blocking drugs. However, the study protocol enables some scrutiny of the various components of cardiovascular reflex homeostasis. The heart rate variation with controlled deep breathing acts as a test of the vagal reflex loop,40 and this was unchanged during the active leg of the study, suggesting that the alterations seen in BRS with nifedipine therapy were not mediated either through a direct effect on the vagus or through any effect on the sinus node producing alterations in heart rate, which was unchanged with chronic dosing. This would also suggest that the increase in BRS occurred through either an increase in the sensitivity of the baroreceptors themselves, or through an effect on the central integration of baroreflex control. The unchanged plasma catecholamine levels with the chronic administration of nifedipine also miti-

gate against any prolonged state of sympathetic activation with active treatment with the sustainedrelease preparation used for this study, although such plasma levels have their limitations as indicators of sympathetic activity. A direct effect of nifedipine on the afferent baroreflex is suggested by the findings of other studies in human and experimental hypertension. Lowering of extracellular calcium has been observed to reduce the threshold and increase the sensitivity of baroreceptors in a feline model,41 and nifedipine has been shown to enhance baroreceptor sensitivity in the isolated canine carotid sinus.42 A direct action in relaxing the smooth muscle in the region of the baroreceptors themselves and reducing the ‘splinting’ effect may be responsible for any enhancement of sensitivity, although traditionally the diminished BRS observed in ISH has been attributed to arteriosclerosis and arterial rigidity not readily reversed by short-term antihypertensive therapy, rather than any modifiable functional alteration in the vessel wall.19 Recent work has shown a direct relation between baroreceptor-sympathetic nerve activity and arterial compliance in patients with congestive heart failure,43 but no similar data is available to confirm this relation in systolic hypertension. We have made no study of the independent effect of calcium-blocker therapy on arterial compliance in the region of the aortic arch and carotid sinus in the present study, although other studies of older hypertensives have not identified an independent effect of nifedipine on large artery compliance.12 A recent study using the dihydropyridine calcium antagonist lacidipine in younger mild-to-moderate combined hypertensives observed resetting of the arterial baroreflex independent of alterations in the elastic properties of the arterial wall induced by acute dosing.44 Significantly, the present study would suggest that when BP is lowered with chronic calcium-channel blocking therapy, at least in elderly patients with isolated systolic hypertension, BRS resetting is not the principal outcome, but rather increases in baroreceptorcardiac and possibly also baroreceptor-mediated vascular reflex responses. Were this to lead to increased background sympathetic activation this might be regarded as an unfavourable consequence of the baroreflex alterations, but this is not supported by the unchanged FVR and plasma catecholamines with nifedipine treatment, although these measures alone give only a limited assessment of longer-term sympathetic changes. The study has not examined other possible influences that may be partly responsible for the observed alterations in BRS. We have not measured left ventricular mass, which is known in animal studies to modify cardiac baroreflexes.45 Although it is unlikely that left ventricular mass would be altered to any great extent by 6 weeks of antihypertensive treatment, previous studies in younger subjects with moderate combined hypertension have found left ventricular mass to be significantly reduced after only 16 weeks of nifedipine therapy, so some alteration cannot be entirely discounted.13 The tilt test examines the integrated response of the cardiovascular system to orthostatic stress and does not

93


94

discriminate between arterial high-pressure baroreflex responses and those evoked by the cardiopulmonary low-pressure receptors that are stimulated by pooling of blood in the lower limbs. Although the two responses can be separated to an extent by use of low and high-level lower body negative pressure,46 passive tilt has the advantage of simulating the stresses to which the cardiovascular system is subjected during normal activities and when the drug is used in clinical practice. The present study does not however offer proof that the described changes on the arterial side of the baroreflex are exclusively responsible for the alterations seen in the response to tilt with nifedipine treatment. Nonetheless studies in younger normotensive and hypertensive subjects have not demonstrated a significant effect of dihydropyridine calcium antagonists on cardiopulmonary baroreflex function.46,47 Active standing in the clinic did not demonstrate any adverse effect of nifedipine therapy on the BP and heart rate responses under those circumstances, when some clinicians might caution against antihypertensive treatment because of the recognised association between ISH and orthostatic hypotension in the elderly.48 We have not shown in the present study that the observed increase in arterial BRS with short-term chronic dosing with a long-acting nifedipine preparation leads to reductions in BP variability measured by 24-h ABPM. The estimation of BP and heart rate variability with intermittent ambulatory recorders does have its limitations24 and provides only an incomplete assessment of overall short-and long-term variability, with additional information being available in recent studies using new techniques such as wide-band power spectral analysis49 which have not been used in the current study. BP variability from ambulatory recorders has been independently related to adverse outcomes,7,8 and theoretically improvements in baroreflex homeostatic mechanisms should translate into reduced BP variability and increased heart rate variability. However, other workers have similarly not found that increases in BRS with the treatment of hypertension necessarily lead to reduced longer-term BP variability.50 Whether this mitigates some of the apparent benefits of antihypertensive treatment in such cases is not known. The results of the present study would indicate that far from simply reflecting a rigid and irreversibly damaged arterial tree, some of the adverse pathophysiology of ISH is potentially reversible with simple antihypertensive treatment that is well tolerated. Furthermore, the present study has indicated that the lowering of BP and increased BRS seen with sustained-release nifedipine treatment leads to improved orthostatic tolerance. Thus the study we report here has shown that despite the observed reductions in large artery compliance and reported arteriosclerosis in elderly patients with ISH, modest but clinically significant improvements in some aspects of cardiovascular reflex homeostasis can be achieved with calcium antagonist therapy that may relate to improved management and even outcomes

in this important group of elderly hypertensive patients.

Acknowledgements MAJ was supported by a project grant from the Sir Jules Thorn Trust. The authors acknowledge the support provided by Bayer UK for this study. The authors are also grateful to Professor Ian MacDonald for the catecholamine measurements made in the study.

References 1 Massie BM. Demographic considerations in the selection of antihypertensive therapy. Am J Cardiol 1987; 60: 121I–126I. 2 Gong L et al. Shanghai Trial of Nifedipine in the Elderly (STONE). J Hypertens 1996; 14: 1237–1245. 3 Staessen JA et al. Randomised double-blind comparison of placebo and active treatment for older patients with isolated systolic hypertension. Lancet 1997; 350: 757–764. 4 Furberg CD, Psaty BM. Calcium antagonists: not appropriate as first-line antihypertensive agents. Am J Hypertens 1996; 9: 122–125. 5 Epstein M. Calcium antagonists: still appropriate as first-line antihypertensive agents. Am J Hypertens 1996; 9: 110–121. 6 Siche´ JP et al. Non-invasive ambulatory blood pressure variability and cardiac baroreflex sensitivity. J Hypertens 1995; 13 (Pt 2): 1654 –1659. 7 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. 8 Pickering TG. Ambulatory monitoring and blood pressure variability. Science Press: London, 1990. 9 Mancia G et al. Modification of arterial baroreflexes by captopril in essential hypertension. Am J Cardiol 1982; 49: 1415–1419. 10 West JNW, Smith SA, Stallard TJ, Littler WA. Effects of perindopril on ambulatory intra-arterial blood pressure, cardiovascular reflexes and forearm blood flow in essential hypertension. J Hypertens 1989; 7: 97–104. 11 West JNW, Champion de Crespigny PC, Stallard TJ, Littler WA. Effects of angiotensin converting enzyme inhibitor, benazepril, on the sino-aortic baroreceptor heart rate reflex. Cardiovasc Drugs Ther 1991; 5: 747–752. 12 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. 13 McLeay RAB, Stallard TJ, Watson RDS, Littler WA. The effect of nifedipine on arterial pressure and reflex cardiac control. Circulation 1983; 67: 1084 –1090. 14 Littler WA, Young MA. The effect of nicardipine on blood pressure, its variability and reflex cardiac control. Br J Clin Pharm 1985; 20: 115S–119S. 15 Smith SA, Mace PJE, Littler WA. Felodipine, blood pressure, and cardiovascular reflexes in hypertensive humans. Hypertension 1986; 8: 1172–1178. 16 Goldsmith SR. Effect of amlodipine and felodipine on sympathetic activity and baroreflex function in normal humans. Am J Hypertens 1995; 8: 902–908. 17 Harron DWG. Antihypertensive drugs and baroreflex sensitivity. Am J Med 1989; 87 (Suppl 3C): 57S–62S. 18 Wilking SB et al. Determinants of isolated systolic hypertension. JAMA 1988; 260: 3451–3455.


19 Sumimoto T, Mukai M, Murakami E, Hiwada K. Haemodynamic characteristics in elderly patients with isolated systolic hypertension. J Hum Hypertens 1990; 4: 521–526. 20 O’Rourke MF. Arterial stiffness, systolic blood pressure and logical treatment of arterial hypertension. Hypertension 1990; 15: 339–347. 21 Weber MA, Neutel JM, Cheung DG. Hypertension in the aged: a pathophysiologic basis for treatment. Am J Cardiol 1989; 63: 25H–32H. 22 James MA, Robinson TG, Panerai RB, Potter JF. Orthostatic blood pressure changes and arterial baroreflex sensitivity in elderly subjects. Age Ageing (in press) 23 Sever PS, Poulter NR, Bulpitt CS. Double-blind crossover versus parallel groups in hypertension. Am Heart J 1989; 117: 735–739. 24 Di Rienzo M, Grassi G, Pedotti A, Mancia G. Continuous vs intermittent blood pressure measurements in estimating 24-hour average blood pressure. Hypertension 1983; 5: 264 –269. 25 Atkins N, Mee F, O’Malley K, O’Brien E. The relative accuracy of simultaneous same arm, simultaneous opposite arm and sequential same arm measurements in the validation of automated blood pressure measuring devices. J Hum Hypertens 1990; 4: 647–649. 26 Potter JF, Fotherby MD, Macdonald IA. Repeatability and relationship between arterialized catecholamines and blood pressure in elderly subjects. Age Ageing 1993; 222: 404 –410. 27 Smyth HS, Sleight P, Pickering GW. Reflex regulation of arterial pressure during sleep in man: a quantitative method of assessing baroreflex sensitivity. Circ Res 1969; 24: 109–121. 28 Sleight P. Reflex control of the heart. Am J Cardiol 1979; 44: 889–894. 29 McGarry K et al. Baroreflex function in elderly hypertensives. Hypertension 1983; 5: 763–766. 30 Lage SG, Polak JF, O’Leary DH, Creager MA. Relationship of arterial compliance to baroreflex function in hypertensive patients. Am J Physiol 1993; 265: H232–H237. 31 Anderson NB et al. Racial differences in blood pressure and forearm vascular responses to the cold face stimulus. Psychosom Med 1988; 50: 57–63. 32 Smith SA, Stallard TJ, Salih MM, Littler WA. Can sinoaortic baroreceptor heart rate reflex sensitivity be determined from phase IV of the Valsalva manoeuvre? Cardiovasc Res 1987; 21: 422– 427. 33 Altman DG. Practical statistics for medical research. Chapman & Hall, London, 1991. 34 James MA, Robinson TG, Panerai RB, Potter JF. Arterial baroreceptor-cardiac reflex sensitivity in the elderly. Hypertension 1996; 28: 953–960. 35 Parati G et al. Comparison of finger and intra-arterial

36

37

38

39 40 41 42

43 44

45

46 47

48 49

50

blood pressure monitoring at rest and during laboratory testing. Hypertension 1989; 13: 647–655. Imholz BPM et al. Continuous non-invasive blood pressure monitoring: reliability of Finapres device during the Valsalva manoeuvre. Cardiovas Res 1988; 22: 390–397. Imholz BPM et al. Non-invasive continuous finger blood pressure monitoring during orthostatic stress compared to intra-arterial pressure. Cardiovasc Res 1990; 24: 214 –221. James MA, Robinson TG, Panerai RB, Potter JF. Abnormal cardiovascular reflexes in isolated systolic hypertension (abstract). J Hypertens 1996; 14 (Suppl 1): S108. Koch-Weser J. Correlation of pathophysiology and pharmacotherapy in primary hypertension. Am J Cardiol 1973; 32: 499–510. Ewing DJ, Clarke BF. Diagnosis and management of diabetic autonomic neuropathy. Br Med J 1982; 285: 916–918. Kunze DL. Calcium and magnesium sensitivity of the carotid baroreceptor reflex in cats. Circ Res 1979; 45: 815–821. Heesch CM, Miller BM, Thames MD, Abboud FM. Effects of calcium antagonists on carotid sinus baroreceptors in the anesthetized dog. Am J Physiol 1983; 245: H653–H661. Grassi G et al. Sympathetic modulation of radial artery compliance in congestive heart failure. Hypertension 1995; 26: 348–354. Caretta R et al. Relationship between mechanical properties of the carotid artery wall and baroreflex function in acutely treated hypertensive patients. J Hypertens 1996; 14: 1105–1110. Minami N, Head GA. Relationship between cardiovascular hypertrophy and cardiac baroreflex function in spontaneously hypertensive and stroke-prone rats. J Hypertens 1993; 11: 523–533. Ferguson DW, Dorsey JK. Effects of nifedipine on baroreflex modulation of vascular resistance in man. Am Heart J 1985; 109: 55–62. Hirsch AT et al. Effect of isradipine on cardiopulmonary baroreflex function, regional blood flow and vascular responsiveness in hypertensive patients. J Cardiovasc Pharmacol 1992; 19: 272–281. Lipsitz LA, Storch HA, Minaker KL, Rowe JW. Intraindividual variability in postural blood pressure in the elderly. Clin Sci 1985; 69: 337–341. Di Rienzo M et al. Effects of sino-aortic denervation on spectral characteristics of blood pressure and pulse interval variability: a wide-band approach. Med Biol Eng Comput 1996; 34: 133–141. Mancia G et al. Structural cardiovascular alterations and blood pressure variability in human hypertension. J Hypertens 1995; 13 (Suppl 2): S7–S14.

95