Effect of adenosine on heart rate variability in ... - Semantic Scholar

Clinical Science (1999) 96, 597ñ604 (Printed in Great Britain)

Effect of adenosine on heart rate variability in humans Gerard A. RONGEN, Steven C. BROOKS, Michael J. POLLARD, Shin-ichi ANDO, Hilmi R. DAJANI, Catherine F. NOTARIUS and John S. FLORAS Division of Cardiology, Mount Sinai and Toronto Hospitals, and the Centre for Cardiovascular Research, University of Toronto, Toronto, Ontario, Canada

A

B

S

T

R

A

C

T

By stimulating afferent nerve endings in skeletal muscle, heart, kidney and the carotid body, adenosine infusion evokes a receptor-specific sympatho-excitatory reflex in humans that overrides its direct negative chronotropic effect. We tested the hypothesis that adenosine increases heart rate by suppressing parasympathetic and augmenting sympathetic components of heart rate variability. High-frequency (PH ; 0.15–0.50 Hz) and low-frequency (PL ; 0.05–0.15 Hz) components of heart rate variability total power (PT) were determined by spectral analysis. The ratios PH/PT and PL/PH respectively were used to estimate parasympathetic and sympathetic input to the sino–atrial node. Heart rate was recorded before and during a 5 min intravenous infusion of adenosine (140 µg:min−1:kg−1) in seven healthy men. Adenosine did not affect blood pressure, but increased heart rate by 33p6 beats/min, and reduced PT, PH, PL and PH/PT. In contrast, there was an increase in PL/PH. In a second experiment in nine men, brachial artery infusion of adenosine (15 µg:min−1:100 ml−1 forearm tissue) increased heart rate by 3 beats/min, had no effect on PT, PH, PL or PH/PT, yet increased PL/PH. Intra-arterial adenosine exerts a modest effect on heart rate by modulating cardiac sympathetic indices, without affecting parasympathetic indices, of heart rate variability, whereas intravenous infusion of adenosine reduces heart rate variability and raises heart rate by decreasing parasympathetic and increasing cardiac sympathetic tone. These reflex effects may become clinically relevant during adenosine stress testing, or when endogenous adenosine is increased, such as during ischaemia, exercise or vasodepressor reactions, or in heart failure.

INTRODUCTION The cardiovascular actions of adenosine are a topic of current interest [1]. Adenosine can act directly on the sinus node to reduce heart rate and slow conduction within the atrial–ventricular node [2]. These properties are the basis for one clinical application of adenosine, namely the interruption of supraventricular tachycardia by bolus injection. However, when infused intravenously

into conscious humans, to achieve lower plasma concentrations, adenosine increases both heart rate and sympathetic nerve traffic to muscle [3,4]. There are two stimuli to this activation of the sympathetic nervous system : peripheral vasodilation, which reduces baroreceptor afferent nerve discharge, and a receptor-specific sympatho-excitatory reflex elicited by stimulating afferent nerve endings in the carotid body, skeletal muscle, heart and kidney [4–10].

Key words : adenosine, blood pressure, heart rate variability, parasympathetic nervous system, sympathetic nervous system. Abbreviations : PT, total spectral power ; PH, high-frequency power ; PL, low-frequency power ; R–R interval, interval between successive R waves on ECG. Correspondence : Dr John S. Floras, Suite 1614, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario, Canada M5G 1X5.

# 1999 The Biochemical Society and the Medical Research Society

597

598

G. A. Rongen and others

Under resting conditions, there is a good correlation between sympathetic nerve traffic to skeletal muscle and biochemical estimates of sympathetic outflow to the human heart [11]. Whether this relationship persists when sympathetic nerve traffic to skeletal muscle is activated reflexively by adenosine is unknown. If so, the tachycardia observed during adenosine infusion could arise from a concordant increase in sympathetic neural drive to the heart, which would override the direct negative chronotropic effects of this purine. On the other hand, decreased parasympathetic tone [12] could also contribute to this increase in heart rate. Indeed, in a study involving six young healthy subjects given intravenous propranolol and atropine, Conradson et al. [13] concluded that the heart rate response to an infusion of 120 µg of adenosine:min−":kg−" was due to a reduction in vagal tone. However, the chronotropic response to adenosine tended to be less after β-adrenoceptor blockade alone. Thus the relative contributions of sympathetic nervous system activation and parasympathetic nervous system withdrawal to the increase in heart rate observed when adenosine is infused in normal humans for clinical purposes is unclear. Spectral analysis can provide quantitative information on the relative balance between the parasympathetic neural and sympathetic neural contributions to heart rate variability [14–16]. Dipyridamole inhibits the cellular uptake of adenosine, leading to its accumulation [2]. The effects of dipyridamole on frequency-domain indices of heart rate variability have been studied in both normal subjects and patients with ischaemia [17], but the effects of the purine itself on heart rate variability have not been reported. Our objective in the present study was to test the hypothesis that adenosine reduces the parasympathetic and augments the sympathetic components of heart rate variability in healthy subjects. Two series of experiments were performed. In the first series, we determined the effect of a systemic infusion of adenosine, as given commonly during myocardial perfusion imaging [18], on heart rate variability. In the second series, we determined the effect on heart rate variability of brachial artery infusion of adenosine at a dose that is without systemic haemodynamic effect, but which has a demonstrated reflex sympatho-excitatory action [9].

(meanspS.D.)], known to consume at least two cups of coffee per day, were enrolled in this series [19]. All subjects provided written informed consent for this study, and the study was approved by the Ethics Committee of our Institute.

Protocol Because caffeine is a competitive antagonist that inhibits virtually all actions of adenosine studied, adenosine was infused after 30 h of abstinence [19,20]. All studies were performed in the same quiet temperature-controlled room at the same time of day (afternoon), at least 2 h after subjects’ last food intake. An antecubital vein of the left arm was cannulated, and a blood sample was collected for measurement of total (protein-bound plus free) serum caffeine concentration by a homogeneous enzyme immunoassay technique using a commercially available kit (Emit Caffeine Assay ; Syva Canada, Kanata, Ontario, Canada). The limit of detection of this assay is 5 µg\ml. In our clinical laboratory, the batch-to-batch coefficient of variation is 4.2 % at a concentration of 77.2 µmol\l. Lead II of the ECG was recorded continuously during spontaneous respiration, inscribed on to paper by an ink recorder and digitized at 1000 Hz into a personal computer for data analysis. Blood pressure was measured at 1 min intervals, using a non-invasive oscillometric technique with an inflatable cuff applied on each occasion to the right arm (Dinamap ; Kritikon, Tampa, FL, U.S.A.). At 20 min after the intravenous cannulation, a 5 min baseline was recorded. Adenosine was infused to a maximum concentration of 140 µg:min−":kg−", which was sustained for 5 min (this being the dose most commonly administered during myocardial perfusion imaging [18]).

Series 2 : effect of intra-arterial adenosine infusion on heart rate variability Subjects Nine healthy, non-smoking, male volunteers (age 33p9 years ; weight 79p6 kg ; height 178p5 cm) were enrolled in this series, which was approved by our institutional Ethics Committee [9]. All subjects provided written informed consent prior to their participation.

Protocol

METHODS Series 1 : effect of intravenous adenosine infusion on heart rate variability Subjects Seven healthy, non-smoking, male volunteers [age 37p10 years ; weight 82p10 kg ; height 178p5 cm # 1999 The Biochemical Society and the Medical Research Society

All studies were performed in the same quiet temperature-controlled laboratory, at the same time of day, after 24 h of caffeine abstinence, with subjects supine. Lead II of the ECG was recorded continuously, inscribed on to paper by an ink recorder and digitized at 1000 Hz into a personal computer for data analysis. After local anaesthesia, the brachial artery of the non-dominant arm (mean volume 1156p151 ml) was cannulated (5 cm ;

Adenosine and heart rate variability

20 G catheter) for the continuous recording of intraarterial blood pressure, and for intra-arterial infusions. Forearm blood flow was measured by venous occlusion strain gauge plethysmography. After a rest period of at least 30 min, saline (as vehicle) was infused into the brachial artery for 5 min, followed by adenosine, which was infused to a maximum concentration of 15 µg:min−":100 ml−" forearm tissue, which was sustained for 5 min. After a further rest period of at least 30 min, glucose (5 %, w\v ; as vehicle) was infused into the brachial artery for 5 min, followed by sodium nitroprusside, which was infused as a time and vasodilator control to a maximum concentration of 0.6 µg:min−":100 ml−" forearm tissue, which was sustained for 5 min. In five subjects (randomly allocated), adenosine was infused before nitroprusside. In the remainder, the order of infusions was reversed.

Spectral analysis The ECG was sampled at 1000 Hz to determine the interval between successive R waves (R–R interval). Data over the last 3.5 min of the baseline period and for each infusion were subjected to spectral analysis by fast Fourier transformation, as described in [20]. Extra or missing beats were edited and replaced with substitute R–R intervals calculated by linear interpolation from adjacent cycles. The unequal R–R intervals were aligned sequentially to obtain equally spaced samples with spectra that do not differ significantly from heart rate variability spectra for the interpolated time series [21]. The 3.5 min data set was subdivided into five overlapping segments. The mean and linear trend were subtracted from each segment. A cosine tapered window was applied to each segment to reduce spectral leakage. Power spectra were obtained and ensemble-averaged over the five segments using the Coolman–Tukey algorithm. Total spectral power (PT) and integrated power within the lowfrequency (0.05–0.15 Hz ; PL) and high-frequency (0.15– 0.50 Hz ; PH) regions were then computed. The ratios PH\PT and PL\PH were used as conventional indices of parasympathetic nervous system and sympathetic nervous system modulation respectively of sino–atrial node discharge [14–16].

Statistics Baseline values and responses to adenosine (140 µg:min−":kg−" intravenous for Series 1 ; 15 µg:min−":100 ml−" forearm tissue intra-arterial for Series 2) were compared. Mean values for blood pressure and heart rate, acquired over the 5 min baseline periods and during the last 3.5 min of adenosine infusions, were calculated. Student’s t-test was used for paired comparisons of heart rate and blood pressure, whereas the Wilcoxon signed rank test was used to assess changes in frequency-domain indices of heart rate variability from

baseline values. P values of 0.05 were considered statistically significant. Results are presented as meanspS.D.

RESULTS Series 1 : effect of intravenous adenosine infusion on heart rate variability All subjects reported complete abstinence from any caffeine intake during the preceding 30 h, and plasma caffeine concentrations in all seven subjects were below the limits of detection. Adenosine (140 µg:min−":kg−", intravenous) increased heart rate by 33.3p6.2 beats\min (P 0.001). Systolic (j4.2p8.2 mmHg) and diastolic (j0.5p5.4 mmHg) blood pressure did not change significantly (Table 1 ; Figure 1). Adenosine suppressed PT (P 0.02), PH (P 0.02) and PL (P 0.02), as well as the parasympathetic nervous system indicator PH\PT (P 0.02). In contrast, there was an increase in the sympathetic nervous system indicator PL\PH (Table 1 ; Figure 2).

Series 2 : effect of intra-arterial adenosine infusion on heart rate variability Adenosine (15 µg:min−":100 ml−" forearm tissue, intraarterial) increased forearm blood flow from 3.3p0.7 to 20.1p7.1 ml:min−":100 ml−" forearm tissue (P 0.01). There was no significant effect on systolic blood pressure, but diastolic blood pressure was increased by 2.7p2.4 mmHg (P 0.01). Mean heart rate rose from 58 to 61 beats\min (P 0.02) (Table 2). Adenosine had no effect on PT, PH or PL, or on the parasympathetic nervous system indicator PH\PT. In contrast, there was a significant increase in the sympathetic nervous system indicator PL\PH (P l 0.04) (Table 2 ; Figure 2). Sodium nitroprusside (0.6 µg:min−":100 ml−" forearm tissue, intra-arterial) had no significant effect on Table 1 Influence of intravenous adenosine infusion on blood pressure, heart rate and its variability

Adenosine was infused intravenously at a concentration of 140 µg:min−1:kg−1. Results are meanspS.D. from seven healthy subjects. BP, blood pressure ; n.s., not significant (P 0.05). Variable

Baseline

jAdenosine

P

Systolic BP (mmHg) Diastolic BP (mmHg) Heart rate (min−1) PT (ms2) PH (ms2) PL (ms2) PH/PT PL/PH

122p6 68p8 67p4 1696p1829 440p421 437p385 0.28p0.11 1.15p0.36

126p12 68p9 100p6 280p410 32p56 43p41 0.08p0.05 3.00p1.54

n.s. n.s. 0.001 0.02 0.02 0.02 0.02 0.03


599

600


Figure 1

Effects of intravenous adenosine infusion on heart rate variability

Shown are representative tracings in a single subject of pulse (RR) intervals over 5 min (A) and corresponding coarse graining power spectra across the frequency range 0ñ0.5 Hz (B) under supine rest (baseline ; left panels) and during intravenous infusion of adenosine at 140 µg:min−1:kg−1 (right panels).

Figure 2

Effects of intravenous and intra-arterial adenosine infusion on heart rate variability

Shown are group mean values for heart rate, PT, PH/PT as an index of heart rate modulation by the parasympathetic nervous system, and PL/PH as an index of heart rate modulation by the sympathetic nervous system, at baseline and during intravenous (i.v.) infusion of adenosine (140 µg:min−1:kg−1 ; left panels) (n l 7 ; meanpS.D.), and at baseline and during intra-arterial (i.a.) infusion of adenosine (15 µg:min−1:100 ml−1 forearm tissue ; right panels) (n l 9 ; meanpS.D.). Significance of differences : *P 0.05 ; **P 0.02 ; ***P 0.001. systolic or diastolic blood pressure, or on heart rate. Forearm blood flow rose from 3.2p0.7 to 14.0p3.5 ml:min−":100 ml−" forearm tissue (P 0.01). # 1999 The Biochemical Society and the Medical Research Society

This increase in forearm blood flow (j10.9p 8.8 ml:min−":100 ml−" forearm tissue) was not significantly different from that caused by adenosine


Table 2 Influence of intra-arterial adenosine infusion on blood pressure, heart rate and its variability

Adenosine was infused intra-arterially at a concentration of 15 µg:min-1:100 ml−1 forearm tissue. Results are meanspS.D. from nine healthy subjects. BP, blood pressure ; n.s., not significant (P 0.05). Variable

Baseline

Systolic BP (mmHg) 119p20 Diastolic BP (mmHg) 67p11 Heart rate (min−1) 58p9 PT (ms2) 4668p6424 PH (ms2) 1200p178 PL (ms2) 978p990 PH/PT 0.22p0.10 PL/PH 1.65p0.92

jAdenosine

P

120p22 69p11 61p9 3833p3051 839p1050 1042p859 0.18p0.12 2.24p1.29

n.s. 0.01 0.02 n.s n.s n.s. n.s. 0.04

(j16.8p19.5 ml:min−":100 ml−" forearm tissue ; P l 0.15). Nitroprusside had no effect on PT, PH or PL, or on the ratios PH\PT and PL\PH. The latter was 1.69p1.24 during dextrose infusion and 1.81p1.44 during sodium nitroprusside infusion.

DISCUSSION In conscious humans, adenosine evokes a receptorspecific sympatho-excitatory reflex by stimulating afferent nerve endings in the carotid body, skeletal muscle, heart and kidney [3–8]. Several independent, yet complementary, methods have been used to characterize the reflex adrenergic response to adenosine infusion : peroneal muscle sympathetic nerve traffic [3,4], plasma noradrenaline concentration [22] and total body noradrenaline spillover [23]. However, the effect of exogenous adenosine on cardiac sympathetic neural drive, as assessed by heart rate variability, has not been described. From experiments in six healthy volunteers in which intravenous propranolol plus atropine abolished the increase in heart rate elicited by adenosine, but propranolol alone did not, Conradson et al. [13] concluded that vagal inhibition is the primary mechanism by which adenosine infusion increases heart rate. However, these authors did not report on the interaction of adenosine with atropine alone. Moreover, the combination of propranolol and atropine increased baseline heart rate significantly, complicating the interpretation of the results. Systemic infusion of adenosine could affect heart rate variability through a number of pathways : by a direct negative chronotropic action on the sino–atrial node ; or indirectly, either by a central neural action or reflexively by unloading arterial baroreceptors or by stimulating adenosine-sensitive afferent nerve endings in the carotid body, myocardium, kidney and skeletal muscle. Previously we reported that intra-arterial infusion of

adenosine increased total body noradrenaline spillover, presumably by stimulating selectively the latter set of afferents [9]. In the present series, we determined whether selective stimulation of skeletal muscle afferent nerve endings by local infusion of adenosine has similar reflex effects on the sympathetic neural modulation of heart rate variability. If so, this mechanism could assume importance during local ischaemia or exercise. In the first series of experiments, systemic infusion of adenosine caused a 50 % rise in heart rate and increased the PL\PH ratio, a frequency-domain representation of heart rate modulation by the sympathetic nervous system. Because diastolic blood pressure did not fall significantly, arterial baroreflex unloading [23] is probably less important than the stimulation of adenosine-receptor sympathetic neural afferents in mediating these sympathoneural responses. In the second series, intraarterial infusion of adenosine also increased the sympathetic nervous system indicator PL\PH ; this increase occurred in the setting of a much more modest (3 beats\min) increase in heart rate. This was not a time effect, or a non-specific response to forearm vasodilation, since nitroprusside infusion did not alter either variable. High local concentrations of adenosine, as may be achieved clinically by bolus adminstration, have a direct negative chronotropic effect and a direct inhibitory action on cardiac noradrenergic neurotransmission [24,25], yet heart rate increased during the intravenous and the intraarterial infusions. Thus observations with both protocols are consistent with the concept that the cardiac adrenergic nervous system participates in the efferent reflex sympatho-excitatory response to infused adenosine. Systemic infusion of adenosine reduced spectral power across all frequency bands. This effect, not observed during local infusion, can be attributed to the additional stimulation of adenosine-sensitive neural afferents in other skeletal muscle vascular beds, and in the heart, kidneys and carotid bodies. Several mechanisms could account for this broad-band decrease in spectral power. First, by acting directly on the sinus node [2], adenosine could exert a damping effect that would render the heart rate less responsive to neural modulation. Secondly, adenosine reduced two generally accepted indices of the parasympathetic modulation of heart rate : PH and the ratio PH\PT. These actions were not observed during intra-arterial infusion, indicating that the reflex stimulated by adenosine-receptor-specific neural afferents in the forearm does not include the parasympathetic nervous system in its efferent limb. Observations during systemic infusion indicate that sympathetic activation and vagal withdrawal are important effectors of the marked chronotropic response to adenosine. Such parasympathetic withdrawal could occur through several mechanisms, or their interactions : as an appropriate arterial baroreflex-mediated efferent response to peripheral vasodilation, via central inhibition of acetyl# 1999 The Biochemical Society and the Medical Research Society

601

602


choline release [26], or as a direct inhibitory effect of adenosine on vagally induced bradycardia [12]. Chemoreceptor reflex stimulation is an important component of the sympatho-excitatory response to the systemic infusion of adenosine in humans [5,27]. Adenosine infusion at 140 µg:min−":kg−" increases tidal volume and causes respiratory alkalosis, but does not change breathing frequency [5]. Increasing the depth of respiration will increase vagal drive reflexively (which should augment power within the high-frequency spectral band) and decrease sympathetic outflow [28], i.e. effects opposite to those observed. Because tidal volume (and hence these additional respiratory influences) was not controlled in the present experiments, our results might have underestimated both the potential inhibitory effect of exogenous adenosine on parasympathetic tone and its excitatory influence on cardiac sympathetic neural drive. The effects of adenosine infusion on heart rate variability in normal subjects have not been reported, but Petrucci et al. [17] assessed heart rate variability changes, using a time variant spectral algorithm, during the infusion of dipyridamole, which inhibits the cellular uptake of adenosine, leading to its accumulation [2]. In control subjects without objective evidence for coronary artery disease, dipyridamole infusion increased heart rate, and decreased PT, PH and PL [17]. The last parameter decreased to 31 % of its baseline value. The sympathetic nervous system indicator PL\PH tended to increase, but this was not significant. These results indicate that our observations during adenosine infusion are qualitatively similar to the actions of endogenous adenosine on heart rate variability in healthy subjects. In contrast, during ischaemia, when adenosine accumulates, there was a 15fold increase in PL in their experimental group of patients with functionally important coronary artery disease [17], consistent with an enhanced, adenosine-induced, sympatho-excitatory response under these conditions [10]. The sympathetic nervous system responds to some internal or external challenges in a highly selective and organ-specific manner, whereas, under other circumstances, such stimuli elicit a generalized sympathoexcitatory response [29–31]. When considered in the context of previous reports that adenosine, infused at the dose used in the present study, increases muscle sympathetic nerve activity and heart rate [3,4], these observations with respect to heart rate variability indicate that the systemic infusion of adenosine evokes a generalized reflex sympatho-excitatory response involving the heart, in addition to other vascular beds. The capacity to vary the heart rate appropriately allows the cardiovascular system a broad repertoire of responses to maintain stability when faced by acute pathology or external challenges. A reduction in the variability or complexity of heart rate is a risk factor for early mortality # 1999 The Biochemical Society and the Medical Research Society

following myocardial infarction [32]. The increases in heart rate, reductions in heart rate variability and augmented adrenergic drive to the heart observed during adenosine infusion may be clinically relevant in the setting of adenosine-radiotracer stress testing, and may contribute to its pro-arrhythmic effects, as observed in clinical practice [33]. These actions may also have relevance for cardioprotective interventions involving adenosine or molecules designed to replicate or potentiate its local actions [34]. In conditions such as congestive heart failure or coronary artery disease, tissue ischaemia or relative hypoperfusion of skeletal or cardiac muscle could stimulate the local production of endogenous adenosine and cause reflex increases in sympathetic outflow to the heart and other vascular beds [9]. In a recent report, 10–20-fold increases in plasma adenosine concentrations were described in patients with NYHA (New York Heart Association) Class III and Class IV heart failure, as compared with healthy control subjects [35]. Decreases in both the variability and the complexity of the heart rate in heart failure [36] have been attributed to impaired neural circulatory control, and to decreased sino–atrial responsiveness to autonomic input in this condition. Increased interstitial and plasma adenosine concentrations might underlie both alterations. In summary, exogenous adenosine reduces heart rate variability by decreasing parasympathetic and increasing cardiac sympathetic nervous system tone. These actions overwhelm adenosine’s direct negative chronotropic effects, resulting in a marked increase in heart rate. Brachial artery infusion of adenosine has a more modest, positive, effect on heart rate, and increases its modulation by the cardiac sympathetic nervous system without affecting parasympathetic tone. These findings indicate concordance between this purine’s reflex actions on the heart and on sympathetic discharge to other vascular beds. These reflex effects override its direct negative chronotropic and sympatho-inhibitory actions, and may be clinically relevant in the setting of exercise, vasodepressor syncope [37], congestive heart failure or myocardial ischaemia, or during adenosine-radiotracer stress testing.

ACKNOWLEDGMENTS The expert technical assistance of Beverley L. Senn is gratefully acknowledged. This work was supported by an Operating Grant from the Medical Research Council of Canada (MT9721). G. A. R. was the recipient of a Research Fellowship from the Department of Medicine Research Fund of the Mount Sinai Hospital, a grant from Janssen-Cilag B.V. (Tilburg, The Netherlands), and travel grant R95114 from the Dutch Heart Foundation (‘ De Gelderfonds ’ ; The Hague, The Netherlands). S. C. B. received a John D. Schultz Science Student Scholarship


from the Heart and Stroke Foundation of Ontario. S-i.A. was the recipient of a Canadian Hypertension Society\ Merck Frosst Fellowship. C. F. N. was supported by a Research Fellowship from the Department of Medicine, Mount Sinai Hospital, and by Fellowship Awards from the Department of Medicine and the Centre for Cardiovascular Research of the University of Toronto. J. S. F. is a Career Investigator of the Heart and Stroke Foundation of Ontario.

REFERENCES 1 Rongen, G. A., Floras, J. S., Lenders, J. W. M., Thein, T. and Smits, P. (1997) Cardiovascular pharmacology of purines. Clin. Sci. 92, 13–24 2 Belardinelli, L., Linden, J. and Berne, R. M. (1989) The cardiac effects of adenosine. Prog. Cardiovasc. Dis. 32, 73–97 3 Biaggioni, I., Killian, T. J., Mosqueda Garcia, R., Robertson, R. M. and Robertson, D. (1991) Adenosine increases sympathetic nerve traffic in humans. Circulation 83, 1668–1675 4 Rongen, G. A., Senn, B. L., Ando, S., Notarius, C. F., Stone, J. A. and Floras, J. S. (1997) Comparison of hemodynamic and sympathoneural responses to adenosine and lower body negative pressure in man. Can. J. Physiol. Pharmacol. 75, 128–134 5 Biaggioni, I., Olafsson, B., Robertson, R. M., Hollister, A. S. and Robertson, D. (1987) Cardiovascular and respiratory effects of adenosine in conscious man. Evidence for chemoreceptor activation. Circ. Res. 61, 779–786 6 Costa, F. and Biaggioni, I. (1993) Adenosine activates afferent fibers in the forearm, producing sympathetic stimulation in humans. J. Pharmacol. Exp. Ther. 267, 1369–1374 7 Cox, D. A., Vita, J. A., Treasure, C. B., Fish, R. D., Selwyn, A. P. and Ganz, P. (1989) Reflex increase in blood pressure during intracoronary administration of adenosine in man. J. Clin. Invest. 84, 592–596 8 Edlund, A., Ohlsen, H. and Sollevi, A. (1994) Renal effects of local infusion of adenosine in man. Clin. Sci. 87, 143–149 9 Rongen, G. A., Brooks, S. C., Ando, S., Abramson, B. L. and Floras, J. S. (1998) Angiotensin AT receptor blockade abolishes the reflex sympatho-excitatory" response to adenosine. J. Clin. Invest. 101, 769–776 10 Thames, M. D., Kinugawa, T. and Dibner-Dunlap, M. E. (1993) Reflex sympathoexcitation by cardiac sympathetic afferents during myocardial ischemia. Role of adenosine. Circulation 87, 1698–1704 11 Wallin, B. G., Esler, M., Dorward, P., Eisenhofer, G., Ferrier, C., Westerman, R. and Jennings, G. (1992) Simultaneous measurements of cardiac noradrenaline spillover and sympathetic outflow to skeletal muscle in humans. J. Physiol. (London) 453, 45–58 12 Monteiro, E. C. and Ribeiro, J. A. (1991) Adenosine and the bradycardic response to vagus nerve stimulation in rats. Eur. J. Pharmacol. 204, 193–202 13 Conradson, T. B., Clarke, B., Dixon, C. M., Dalton, R. N. and Barnes, P. J. (1987) Effects of adenosine on autonomic control of heart rate in man. Acta Physiol. Scand. 131, 525–531 14 Butler, G. C., Yamamoto, Y., Xing, H. C., Northey, D. R. and Hughson, R. L. (1993) Heart rate variability and fractal dimension during orthostatic challenges. J. Appl. Physiol. 75, 2602–2612 15 Malliani, A., Pagani, A., Lombardi, F. and Cerutti, S. (1991) Cardiovascular and neural regulation explored in the frequency domain. Circulation 84, 482–492

16 Pomeranz, B., Macauley, R. J. B., Caudill, M. A. et al. (1985) Assessment of autonomic function in humans by heart rate spectral analysis. Am. J. Physiol. 248, H151–H153 17 Petrucci, E., Mainardi, L. T., Balian, V. et al. (1996) Assessment of heart rate variability changes during dipyridamole infusion and dipyridamole-induced myocardial ischemia : a time variant spectral approach. J. Am. Coll. Cardiol. 28, 924–934 18 Cerquiera, M. D., Verani, M. S., Schwaiger, M., Heo, J., Iskandrian, A. S. and the Investigators of the Multicenter Adenoscan Trial (1994) Safety profile of adenosine stress perfusion imaging : results from the Adenoscan Multicenter Trial Registry. J. Am. Coll. Cardiol. 23, 384–389 19 Rongen, G. A., Brooks, S. C., Ando, S., Notarius, C. F. and Floras, J. S. (1998) Caffeine abstinence augments the systolic blood pressure response to adenosine in humans. Am. J. Cardiol. 81, 1382–1385 20 Yamamoto, Y. and Hughson, R. L. (1993) Extracting fractal components from time series. Physica D 68, 250–264 21 Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology (1996) Heart rate variability. Standards of measurement, physiological interpretation and clinical use. Circulation 93, 1043–1065 22 Smits, P., Boekema, P., Thien, T., De Abreu, R. and van’t Laar, A. (1987) Evidence for an antagonism between caffeine and adenosine in the cardiovascular system in man. J. Cardiovasc. Pharmacol. 10, 136–143 23 Sanders, J. S. and Ferguson, D. W. (1989) Diastolic pressure determines autonomic responses to pressure perturbations in humans. J. Appl. Physiol. 66, 800–807 24 Nakatsuka, H., Nagano, O., Foldes, F. F., Nagashima, H. and Vizi, E. S. (1995) Effects of adenosine on norepinephrine and acetylcholine release from guinea-pig right atrium : role of A -receptors. Neurochem. Int. 27, " 345–353 25 Richardt, G., Waas, W., Kranzhofer, R., Cheng, B., Lohse, M. J. and Schomig, A. (1989) Interaction between the release of adenosine and noradrenaline during sympathetic stimulation : a feed-back mechanism in rat heart. J. Mol. Cell. Cardiol. 21, 269–277 26 Richardson, P. J., Brown, S. J., Bailyes, E. M. and Luzio, J. P. (1987) Ectoenzymes control adenosine modulation of immunoisolated cholinergic synapses. Nature (London) 327, 232–234 27 Engelstein, E. D., Lerman, B. B., Somers, V. K. and Rea, R. F. (1994) Role of arterial chemoreceptors in mediating the effects of endogenous adenosine on sympathetic nerve activity. Circulation 90, 2919–2926 28 Eckberg, D. L., Nerhed, C. and Wallin, B. G. (1985) Respiratory modulation of muscle sympathetic and vagal cardiac outflow in man. J. Physiol. (London) 365, 181–196 29 Delius, W., Hagbarth, K.-E., Hongell, A. and Wallin, B. G. (1972) Manoeuvres affecting sympathetic outflow in human muscle nerves. Acta. Physiol. Scand. 84, 82–94 30 Esler, M., Jennings, G., Korner, P. et al. (1988) Assessment of human sympathetic nervous system activity from measurements of norepinephrine turnover. Hypertension 11, 3–20 31 Floras, J., Jones, J. V., Hassan, M. O., Osikowska, B. A., Sever, P. S. and Sleight, P. (1986) Failure of plasma norepinephrine to consistently reflect sympathetic activity in humans. Hypertension 8, 641–649 32 La Rovere, M. T., Bigger, J. T., Marcus, F. I., Mortara, A. and Schwartz, P. J., for the ATRAMI Investigators (1998) Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. Lancet 351, 478–484 33 Strickberger, S. A., Man, K. C., Daoud, E. G. et al. (1997) Adenosine-induced atrial arrhythmia : a prospective analysis. Ann. Intern. Med. 127, 417–422 34 Mangano, D. T., for The Multicenter Study of Perioperative Ischemia (McSPI) Research Group (1996) Effects of acadesine on myocardial infarction, stroke, and


603

604


death following surgery. A Meta-analysis of the 5 International Randomized Trials. J. Am. Med. Assoc. 277, 325–332 35 Funaya, H., Kitakaze, M., Node, K., Minamino, T., Komamura, K. and Hori, M. (1997) Plasma adenosine levels increase in patients with chronic heart failure. Circulation 95, 1363–1365

36 Butler, G. C., Ando, S-i. and Floras, J. S. (1997) Fractal component of heart rate and systolic blood pressure in congestive heart failure. Clin. Sci. 92, 543–550 37 Waxman, M. B. and Asta, J. A. (1998) Role of adenosine receptors in the paradoxic bradycardia response of rats to inferior vena cava occlusion during an infusion of isoproterenol. Circulation 98, 1228–1235 Received 12 October 1998/4 January 1999; accepted 12 March 1999
