Experimental Physiology

Transcription

1 718 Exp Physiol 9.8 pp Research Paper α-adrenergic effects on low-frequency oscillations in blood pressure and R R intervals during sympathetic activation Antti M. Kiviniemi 1,, Maria F. Frances 3, Suvi Tiinanen, Rosemary Craen 5, Maxim Rachinsky 5, Robert J. Petrella 3,TapioSeppänen, Heikki V. Huikuri,MikkoP.Tulppo 1, and J. Kevin Shoemaker 3 1 Department of Exercise and Medical Physiology, Verve Research, Oulu, Finland Institute of Clinical Medicine and Department of Electrical and Information Engineering, The University of Oulu, Oulu, Finland 3 School of Kinesiology and 5 Department of Anesthesia and Perioperative Medicine, The University of Western Ontario, London, ON, Canada Experimental Physiology The present study was designed to address the contribution of α-adrenergic modulation to the genesis of low-frequency (LF;..15 Hz) oscillations in R R interval (RRi), blood pressure (BP) and muscle sympathetic nerve activity (MSNA) during different sympathetic stimuli. Blood pressure and RRi were measured continuously in 1 healthy subjects during 5 min periods each of lower body negative pressure (LBNP; mmhg), static handgrip exercise (HG; % of maximal force) and postexercise forearm circulatory occlusion (PECO) with and without α-adrenergic blockade by phentolamine. Muscle sympathetic nerve activity was recorded in five subjects during LBNP and in six subjects during HG and PECO. Low-frequency powers and median frequencies of BP, RRi and MSNA were calculated from power spectra. Low-frequency power during LBNP was lower with phentolamine versus without for both BP and RRi oscillations (1. ±. versus 1. ±.7 ln mmhg, P =.9; and.9 ±.8 versus 5. ±.9 ln ms, P =.1, respectively). In contrast, the LBNP with phentolamine increased the power of high-frequency oscillations (.15. Hz) in BP and MSNA (P <.1 for both), which was not observed during saline infusion. Phentolamine also blunted the increases in the LBNP-induced increase in frequency of LF oscillations in BP and RRi. Phentolamine decreased thelfpowerofrriduringhg(p =.15) but induced no other changes in LF powers or frequencies during HG. Phentolamine resulted in decreased frequency of LF oscillations in RRi (P =.) during PECO, and a similar tendency was observed in BP and MSNA. The power of LF oscillation in MSNA did not change during any intervention. We conclude that α-adrenergic modulation contributes to LF oscillations in BP and RRi during baroreceptor unloading (LBNP) but not during static exercise. Also, α-adrenergic modulation partly explains the shift to a higher frequency of LF oscillations during baroreceptor unloading and muscle metaboreflex activation. (Received 31 March 11; accepted after revision 18 May 11; first published online May 11) Corresponding author A. M. Kiviniemi: Verve Research, Kasarmintie 13, PO Box, FI-911 Oulu, Finland. antti.kiviniemi@verve.fi Spontaneous fluctuations in blood pressure (BP) and R R interval (RRi) at low frequency (LF;..15 Hz), also referred to as Mayer waves, have been investigated extensively. After utilization of power spectral methods (Akselrod et al. 1981), the power and the frequency of these M.P.T. and J.K.S. contributed equally to this work. cardiovascular oscillations have become routine research tools for the assessment of cardiovascular autonomic function. Several studies have shown that abnormal LF oscillations are related to increased risk for cardiovascular morbidities and severity of cardiovascular disease (van de Borne et al. 1997; La Rovere et al. 3; Wichterle et al. ; Kiviniemi et al. 7). The great interest in LF oscillations from cardiovascular signals is based on the need for a reliable noninvasive method to quantify sympathetic activity and DOI: /expphysiol C 11 The Authors. Journal compilation C 11 The Physiological Society

2 Exp Physiol 9.8 pp Cardiovascular variability and sympathetic activity 719 sympathovagal balance. These LF oscillations have been explained by central oscillations of sympathetic activity (Kaminski et al. 197; Preiss & Polosa, 197; Pagani et al. 1997) or by time delays within the baroreflex loop, resulting in resonance in BP and RRi (deboer et al. 1987; Bertram et al. 1998; Julien et al. 3). Although it is well established that LF oscillations in BP and RRi contain information on sympathetic modulation (Akselrod et al. 1981; Malliani et al. 1991; Pagani et al. 1997; Furlan et al. ), these oscillations have not shown a universal association with sympathetic activity. For instance, increases in the LF power of BP and RRi fluctuations, expressed in absolute or normalized units, have been observed during orthostatic stress (Montano et al. 199; Laitinen et al. 1999; Furlan et al. ) andpharmacologicaldecrementsinbp(vandeborne et al. 1), consistent with increased sympathetic activity. However, decreased power of these oscillations has been observed in some cardiovascular disease conditions (van de Borne et al. 1997; Mussalo et al. 3), despite the expected increases in sympathetic activity (Ng et al. 1993; van de Borne et al. 1997). Disparate findings on LF oscillations have also been observed during sympathetic pressor stimuli. Iellamo et al. (1999) observed both increased absolute and normalized LF power of RRi oscillations during static leg exercise and postexercise ischaemia. In contrast, our recent study showed no changes in absolute power of LF oscillation in RRi and BP during static handgrip and postexercise ischaemia, with changes observed only in normalized LF power of RRi oscillations during handgrip exercise and no changes seen during the ischaemic recovery (Kiviniemi et al. 1a). A possible explanation for these inconsistent findings during sympathetic activation is that the responses in the power and the frequency of LF oscillations may depend on the mode of sympathetic stimuli. While passive head-up tilt has resulted in dominant LF oscillations without changes in the frequency of these oscillations (Montano et al. 199; Cooke et al. 1999; Kiviniemi et al. 1a), sympathetic pressor stimuli have shown opposite effects (Kiviniemi et al. 1a). These differences may be related to the differences in proportional contribution of autonomic modulation of the sinoatrial node and vascular tone related to a given sympathetic stimulus. The pressor response during exercise is mainly produced by cardiac mechanisms (Shoemaker et al. 7), whereas peripheral mechanisms may be more involved in BP control during orthostatic stress and postexercise ischaemia (Ichinose et al. 7; Zamir et al. 1). While cardiovascular responses to a pressor stimulus and baroreceptor unloading are mediated by different pathways, the effects of peripheral sympathetic modulation on cardiovascular variability during these sympathetic stimuli are not fully understood. The present study was designed to address the contribution of α- adrenergic effects on LF oscillations with respect to RRi, BP and muscle sympathetic nerve activity (MSNA) during baroreceptor unloading, static exercise and postexercise ischaemia. We tested the hypothesis that α-adrenergic antagonism may decrease the power and the frequency of cardiovascular oscillations at LF during baroreceptor unloading and postexercise ischaemia but not during static exercise. Methods Ethical approval The study was performed according to the Declaration of Helsinki, and the Health Sciences Research Ethics Board of the University of Western Ontario approved the protocol. Written informed consent was received from each subject prior to the start of each experiment. Subjects The study population consisted of 13 healthy volunteers [nine men and four women, age 7 ± years, weight 75 ± 1 kg, height 17 ± 9 cm (mean ± SD)]. All subjects were non-smokers without diabetes mellitus, asthma or any cardiovascular disorders. The subjects abstained from food for 3 h, caffeinated drinks for 1 h, and alcohol or strenuous exercise for h prior to tests. Study protocol The study protocol included the baseline measurements in a supine position at rest, during a period of lower body negative pressure (LBNP), and during static handgrip exercise (HG) followed by postexercise circulatory occlusion (PECO; Fig. 1). The order of these manoeuvres was randomized within the protocol, and 15 min of recovery was completed prior to the start of baseline measurements for each subsequent manoeuvre. The protocols were performed during an intravenous infusion of saline (control), and then during infusion of phentolamine (PHE; non-selective α- adrenergic antagonist). Breathing was paced (.5 Hz) by metronome guidance throughout the protocols. The participants were sealed at the waist in a vacuum chamber to generate LBNP ( mmhg, 5 min) to mimic the effect of upright posture and induce baroreflexmediated vagal withdrawal with concurrent increases in sympathetic activity, secondary to decreased venous return (Kitano et al. 5). Maximal voluntary isometric handgripforcewasdeterminedasthehighestvalueofthree contractions. During the HG protocol, the participants maintained the force equal to % of their predetermined maximal voluntary contraction for 5 min. Feedback of C 11 The Authors. Journal compilation C 11 The Physiological Society

3 7 A. M. Kiviniemi and others Exp Physiol 9.8 pp the magnitude of force reached was achieved through visual feedback. Postexercise circulatory occlusion was conducted by inflating a cuff, which was placed distal to the cubital fossa of the exercising arm, to a suprasystolic level (5 mmhg) 5 s before the end of HG. This pressure was maintained for 5 min. Rapid inflation prior to the end of exercise traps the metabolites produced during HG in the forearm, resulting in sustained sympathetic activation via the muscle metaboreflex (O Leary, 1993; Shoemaker et al. 7). Postexercise circulatory occlusion also produces rapid vagal reactivation due to the cessation of central command and baroreceptor loading by sustained high BP (O Leary, 1993). After the control protocol was completed, the effectiveness of the α-adrenergic blockade by PHE towards sympathetic vasoconstriction was assessed by the pressor response to an intravenous infusion of noradrenaline (NA; infusion rate 75 ng kg 1 min 1 for 3 min) via the antecubital catheter. Measurements were taken during 3 min periods each of baseline, infusion of NA and recovery. This was then repeated with the simultaneous infusion of PHE (infusion rate μgmin 1 )andna. Total peripheral resistance (TPR) was calculated from the last minute of each phase as the mean BP divided by cardiac output ( Q) obtained by modelflow from finger plethysmography (Finometer Pro; Finapres Medical Systems, Amsterdam, The Netherlands). The change in TPR induced by NA during PHE infusion was % of that without PHE. Phentolamine infusion (infusion rate μgmin 1 ) was continued, and the protocol with α- adrenergic blockade was started 5 min after cessation of NA infusion. Total dosage of PHE was 9. ±. mg. Lower body negative pressure data were analysed from 1 of the 13 subjects owing to symptoms of presyncope [nausea, light headedness, systolic blood pressure (SBP) <9 mmhg] from one subject during the PHE infusion. Additionally, HG and PECO data were analysed on 1 of the 13 subjects, because one participant experienced excessive discomfort and could not complete the PECO trial. Data acquisition Electrocardiogram (Pilot 9; Colin Medical Instruments, San Antonio, TX, USA), BP (Finometer Pro; Finapres Medical Systems), respiration (Pneumotrace; ADInstruments, Sydney, NSW, Australia) and MSNA were measured continuously using a PowerLab data acquisition system (PowerLab/1SP; ADInstruments) with a sampling frequency of 1 Hz. Blood pressure was also measured manually to validate the BP measured from finger plethysmography. Venous blood samples were drawn from the antecubital vein after min of each phase for subsequent analysis of circulating levels of NA (Musso et al. 1989). Multifibre recordings of MSNA were performed using the microneurography technique. A recording microelectrode was inserted transcutaneously into the peroneal nerve posterior to the fibular head and a reference electrode was inserted subcutaneously 1 3 cm from the recording site. Confirmation of an MSNA site was confirmed when pulse-synchronous burst activity was observed, in addition to increased activity during voluntary apnoeas but not during arousal to a loud SALINE protocol Randomization 15 min recovery Baseline LBNP Baseline HG PECO 5 min 5 min 5 min 5 min 5 min NA+PHE protocol NA Baseline Recovery 3 min PHE started PHE Baseline Baseline 3 min 3 min 3 min NA+PHE PHE Recovery 3 min 3 min 3 min 5 min PHE protocol Randomization 15 min recovery, PHE stopped PHE restarted 3 min before next baseline PHE Baseline PHE LBNP PHE Baseline PHE HG 5 min 5 min 5 min 5 min PHE PECO 5 min Figure 1. The study protocol Abbreviations: HG, static handgrip exercise; LBNP, lower body negative pressure; NA, noradrenaline infusion (infusion rate 75 ng kg 1 min 1 ); PECO, postexercise circulatory occlusion; PHE, phentolamine infusion (infusion rate μg min 1 ). C 11 The Authors. Journal compilation C 11 The Physiological Society

4 Exp Physiol 9.8 pp Cardiovascular variability and sympathetic activity 71 noise. The MSNA signal was amplified 1 times by a preamplifier and a further 75 times by a variablegain, isolated amplifier. Thereafter, the MSNA signal was bandpass filtered at a bandwidth of 7 Hz and then rectified and integrated (time constant.1 s) to obtain a mean voltage neurogram (model C-3; Iowa University Bioengineering) that was routed to the data acquisition system. Muscle sympathetic nerve activity was successfully recorded in six subjects during the HG and PECO protocol andinfivesubjectsduringlbnp. Data analysis Analyses were performed from the baseline recordings (5 min) and from the last min of each subsequent manoeuvre due to the time dependence of the responses to interventions. Mean values for diastolic (DBP), systolic (SBP) and mean blood pressure (MBP), mean heart rate (HR) and breathing frequency were calculated for each phase. Additionally, the means of Q and TPR were calculated by the modelflow method (Finometer Pro; Finapres Medical Systems). Bursts of MSNA activity with a :1 or greater signal-to-noise ratio were detected from a neurogram. The burst frequency (bursts min 1 ), burst incidence (bursts 1 RRi 1 ) and mean amplitude which was normalized to the highest individual burst observed across all protocols (arbitrary units, a.u.) were then calculated. Spectral characteristics of RRi, SBP and MSNA were analysed. Time series of RRi and SBP were extracted from the continuous recordings. The time series of MSNA, including the bursts only, was integrated over the time period between two consecutive diastolic values before resampling. These data of discrete event series were then resampled at Hz. Thereafter, very low-frequency components (<. Hz) were abolished by detrending with the Savitzky-Golay method (Orfanidis, 199). The power spectral analyses of RRi, SBP and MSNA variability were then performed using an autoregressive model (Burg s algorithm). Akaike s information criterion was used to determine the model order, validated by testing the whiteness of prediction error. The power spectrum densities of the LF (..15 Hz) and high-frequency (HF;.15. Hz) oscillations were calculated (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 199). The LF (Power LF RRi,Power LF SBP and Power LF MSNA )and HF (Power HF RRi,Power HF SBP and Power HF MSNA )ofrri, SBP and MSNA oscillations spectral powers are presented as absolute and normalized units (n.u.; Iellamo et al. 1999). Median frequencies of RRi, SBP and MSNA oscillations at LF (Med LF RRi,Med LF SBP and Med LF MSNA )andhf (Med HF RRi,Med HF SBP and Med HF MSNA ) were calculated as the frequency that divides the specific frequency band into two segments having equal power (Stulen & DeLuca, 1981). The cardiac baroreflex sensitivity was analysed using the α method separately for LF (BRS LF )andhf (BRS HF ) spectral bands (Pagani et al. 1988). Statistical analysis The data are presented as the means ± SD. The Gaussian distribution of the data was verified with the Shapiro Wilk goodness-of-fit test. The absolute spectral powers were not normally distributed and were transformed into natural logarithm (ln) before the statistical tests. The two-way ANOVA (PHE intervention, 3 for HG and for LBNP protocol) for repeated measures was used to assess the main effects. Bonferroni s test and Student s paired t test were used as post hoc tests accordingly. The data were analysed using IBM SPSS Statistics 19. (IBM Corporation, Somers, NY, USA). A P value <.5 was considered statistically significant. Results Lower body negative pressure Lower body negative pressure increased HR during both saline and PHE infusions (Table 1). Phentolamine resulted in higher HR at baseline and during LBNP compared with saline. Decreased SBP and unchanged DBP and MBP were observed during LBNP for both PHE and saline conditions. Additionally, SBP during LBNP tended to be lower with PHE compared with saline (P =.58). Compared with saline, NA plasma concentration was higher during LBNP with PHE (Table ). Muscle sympathetic nerve activity burst frequency increased during LBNP and was higher during LBNP with PHE than with saline. The MSNA burst incidence remained unchanged during LBNP. Burst amplitude increased during LBNP with PHE but not with saline (Table ). Blood pressure variability. Absolute Power LF-SBP during LBNP with saline did not change from baseline (Table 3 and Fig. A). However, both absolute (1. ±. versus 1. ±.7 ln mmhg, P =.9) and normalized Power LF-SBP (73 ± 1 versus 5 ± 19 n.u., P =.3) were lower during LBNP with PHE than with saline (Fig. A and B). The Med LF-SBP increased during LBNP (from.7 ±.5 to.83 ±.13 Hz, P =.), and this shift in frequency was blunted by PHE (from.71 ±.11 to.7 ±.11 Hz, P =.7; Fig. C). Lower body negative pressure with PHE resulted in a higher Power HF-SBP compared with LBNP without PHE (Table 3). Heart rate variability and baroreflex sensitivity. During PHE infusion, LBNP decreased absolute Power LF-RRi (from.7 ±.9 to 5. ±.9 ln ms, P =.) which C 11 The Authors. Journal compilation C 11 The Physiological Society

5 7 A. M. Kiviniemi and others Exp Physiol 9.8 pp Table 1. The effect of lower body negative pressure (LBNP) on heart rate, blood pressure, breathing frequency and central haemodynamics during saline and phentolamine (PHE) infusion P value for main effect Parameter Infusion Baseline LBNP PHE Intervention Interaction HR (beats min 1 ) Saline 5 ± 11 ± 1 <.1 <.1 <.1 PHE ± 1 89 ± SBP (mmhg) Saline 133 ± 15 1 ± 13.9 <.1.18 PHE 13 ± ± 1 DBP (mmhg) Saline 77 ± 78 ± PHE 75 ± 73 ± 9 MBP (mmhg) Saline 9 ± 9 9 ± PHE 9 ± ± 1 Breathing frequency (breaths min 1 ) Saline.3 ±.19.5 ± PHE.7 ±..5 ±.13 Cardiac output (L min 1 ) Saline 5.5 ± ± PHE.1 ± ±. TPR (mmhg L 1 min ) Saline 18.1 ± ± PHE 17.1 ± ±. Values are means ± SD. Abbreviations: DBP, diastolic blood pressure; HR, heart rate; MBP, mean blood pressure; SBP, systolic blood pressure; and TPR, total peripheral resistance. P <.5 and P <.1 between control and phentolamine infusion. P <.5 and P <.1 compared with baseline. Table. The effects of LBNP on plasma noradrenaline concentration (NA; n = 11), muscle sympathetic nerve activity (MSNA; n = 5) and oscillations of MSNA during saline and PHE infusion P value for main effect Parameter Infusion Baseline LBNP PHE Intervention Interaction NA (nmol L 1 ) Saline 1. ± ± PHE 1.51 ±.5.3 ±.91 Bursts min 1 Saline ± 1 8 ± PHE 7 ± 15 ± 1 Bursts 1 RRi 1 Saline 1 ± 1 3 ± PHE 3 ± ± 1 Burst amplitude (a.u.) Saline ± 7 ± PHE ± 31 ± 7 Low frequency Power (ln a.u. ) Saline.1 ±.8.3 ± PHE.1 ±.. ±.8 Power (n.u.) Saline 5 ± 15 3 ± PHE 9 ± 1 8 ± 33 Med (Hz) Saline.17 ±..19 ± PHE.18 ±..1 ±.7 High frequency Power (ln a.u. ) Saline. ±. 5.1 ± PHE 5. ±.8. ±. Power (n.u.) Saline 55 ± ± PHE 71 ± 1 7 ± 33 Med (Hz) Saline.53 ±.1.59 ± PHE.5 ±.7. ±.11 Values are means ± SD. Abbreviations: Med, median frequency; and RRi, R R interval. P <.1 between control and phentolamine infusion. P <.5 and P <.1 compared with baseline. was not observed with saline (from 7.1 ±.9 to.9 ±.8 ln ms, P =.339; Table 3 and Fig. D). Increased normalized Power LF-RRi was found during both LBNP with PHE (from ± 13 to 1± 15n.u., P =.) and saline (from 3 ± 1 to ± 18n.u., P =.13; Fig. E). The increment in Med LF-RRi during LBNP (from.71 ±.8 to.81 ±.13 Hz, P =.5) was inhibited by PHE (from.9 ±.7 C 11 The Authors. Journal compilation C 11 The Physiological Society

6 Exp Physiol 9.8 pp Cardiovascular variability and sympathetic activity 73 Table 3. infusion The effects of LBNP on R R interval and systolic blood pressure variability and baroreflex sensitivity during saline and PHE P value for main effect Parameter Infusion Baseline LBNP PHE Intervention Interaction Low frequency Power SBP (ln mmhg ) Saline 1. ± ± PHE 1. ±.7 1. ±.7 Power SBP (n.u.) Saline 7 ± 1 73 ± PHE 9 ± 17 5 ± 19 Med SBP (Hz) Saline.7 ±.5.83 ± PHE PHE.71 ±.11 Power RRi (ln ms ) Saline 7.1 ±.9.9 ± PHE.7 ±.9 5. ±.9 Power RRi (n.u.) Saline 3 ± 1 ± PHE ± 13 1 ± 15 Med RRi (Hz) Saline.71 ±.8.81 ± PHE.9 ±.7.71 ±.9 BRS (ms mmhg 1 ) Saline.5 ± ± PHE 1. ± ± 5.8 High frequency Power SBP (ln mmhg ) Saline. ±..5 ± PHE.5 ± ±.8 Power SBP (n.u.) Saline 33 ± 1 7 ± PHE 31 ± 17 5 ± 19 Med SBP (Hz) Saline.8 ±.8.8 ± PHE.5 ±..51 ±.7 Power RRi (ln ms ) Saline 7.5 ±.7. ± <.1.1 PHE.9 ± 1..9 ± 1.3 Power RRi (n.u.) Saline 57 ± 1 38 ± PHE 5 ± ± 15 Med RRi (Hz) Saline. ±.8.5 ± PHE.8 ±..7 ±.7 BRS (ms mmhg 1 ) Saline 3.3 ± ± <.1.51 PHE 7.1 ± ± 11.5 Values are means ± SD. Abbreviations: BRS, baroreflex sensitivity; Med, median frequency; RRi, R R interval; and SBP, systolic blood pressure. P <.5 and P <.1 between control and phentolamine infusion. P <.5 and P <.1 compared with baseline. to.71 ±.9 Hz, P =.3; Fig. F). Lower body negative pressure with PHE resulted in a lower Power HF-RRi and BRS HF compared with LBNP with saline (Table 3). The BRS LF decreased during LBNP with saline and tended to decrease during LBNP with PHE as well (P =.51). Baseline BRS LF tendedtobelowerwithphecomparedwith saline (P =.). Muscle sympathetic nerve activity oscillations. Absolute and normalized Power LF-MSNA remained unchanged during the interventions (Table ). The Med LF-MSNA was found to be lower during LBNP with PHE than with saline (Table ). Absolute Power HF-MSNA also increased during LBNP with PHE but not with saline. Representative examples of raw data and corresponding spectra during LBNP with saline and PHE are presented in Fig. 3. Handgrip exercise and postexercise muscle metaboreflex During saline infusion, HR and BP increased during HG, followed by recovery of HR but a continued elevation in BP during PECO (Table ). Compared with saline, HR was constantly higher during HG and PECO with PHE, and this response did not subside to baseline levels during PECO. The BP increase during HG with PHE was similar to that observed with saline. However, DBP was lower during PECO with PHE compared with saline. Muscle sympathetic nerve activity burst frequency increased during HG and remained elevated during PECO with both saline and PHE (Table 5). Nonetheless, MSNA burst frequency was higher during HG and PECO when PHE was infused. Increased MSNA burst incidence was only observed during PECO, and this was not modified by PHE. C 11 The Authors. Journal compilation C 11 The Physiological Society

7 7 A. M. Kiviniemi and others Exp Physiol 9.8 pp Blood pressure variability. Absolute Power LF-SBP did not show any changes during HG or PECO, regardless of whether PHE was infused (Table and Fig. A). Normalized values were lower during PECO with PHE (8 ± 1 versus 5 ± 15 n.u., P =.19) and with saline (9 ± 9 versus 59 ± 11 n.u., P =.) compared with their respective baselines (Fig. B). The Med LF-SBP increased during HG with saline (from.7 ±.9 to.9 ±.1 Hz, P =.1) and PHE (from.9±.7 to.88 ±.13 Hz, P =.1) and was sustained at the higher level during PECO (.9 ±.1 and.8 ±.15 Hz, respectively, for saline and PHE, P <.1 for both; Fig. C). During PECO, however, Med LF-SBP tended to be lower with PHE compared with saline (P =. without significant main effects). The Power HF-SBP was increased during PECO with saline but not in any other conditions. A Power LF-SBP (ln, mmhg ) 3 1 Saline PHE Phentolamine.7 Intervention.73 Interaction.8 * D Power LF-RRi (ln, ms ) Phentolamine.3 Intervention.5 Interaction.19 ** B 1 E 1 8 C Med LF-SBP (Hz) Power LF-SBP (nu) Phentolamine.51 Intervention.78 Interaction.1 Phentolamine.7 Intervention.1 Interaction.1 ** Med LF-RRi (Hz) Power LF-RRi (nu) F Phentolamine.75 Intervention.3 Interaction.73 Phentolamine.7 Intervention. Interaction.8 *. Baseline LBNP. Baseline LBNP Figure. The absolute (A and D) and normalized powers (B and E), as well as the median frequencies (Med; C and F) of low-frequency (LF) oscillations in systolic blood pressure (SBP; A C) and R R interval (RRi; D F) at baseline and during lower body negative pressure (LBNP) with intravenous infusion of saline (filled diamonds) and phentolamine (open diamonds) P <.5 and P <.1 between control and phentolamine infusion. P <.5 and P <.1 compared with the baseline. C 11 The Authors. Journal compilation C 11 The Physiological Society

8 Exp Physiol 9.8 pp Cardiovascular variability and sympathetic activity 75 Heart rate variability and baroreflex sensitivity. Absolute Power LF-RRi decreased during HG with PHE (from.7 ±.7 to 5.7 ±. ln ms, P =.8; Table ) and was lower compared with HG during saline infusion (.3 ±.9 ln ms, P =.15). During both conditions, Power LF-RRi returned to baseline during PECO (Fig. D). Normalized Power LF-RRi during HG and PECO did not differ from the baseline in saline and PHE conditions (Fig. E). The Med LF-RRi increased from the baseline during HG and PECO with saline (from.73 ±.9 to.88±.1 and.99 ±.17 Hz, respectively, P <.1 for both) and with PHE (from.7 ±.9 to.85 ±.18 and.8 ±.19 Hz, respectively, P <.5 for both). The Med LF-RRi was significantly lower during PECO with PHE than without (P =.; Fig. F). The Power HF-RRi decreased during HG and returned to baseline during PECO, regardless of whether PHE was infused (Table ). However, PHE decreased Power HF-RRi throughout the protocol. Global decreases in BRS were observed during HG, whereas BRS returned to baseline during PECO. The BRS was also lower when PHE was infused compared with the saline infusion, during the baseline and HG conditions. However, BRS LF was unaffected by PHE during PECO. Muscle sympathetic nerve activity oscillations. Phentolamine tended to lower Med LF-MSNA during PECO (Table 5; P =.58, without significant main effects). The MSNA spectral powers did not change significantly during the interventions (Table ). Representative examples of raw data and corresponding spectra during HG and PECO with saline and PHE are presented in Figs 5 and, respectively. Figure 3. Representative example of data and corresponding spectra during lower-body negative pressure with saline (A) and with phentolamine (B) Spectra for muscle sympathetic nerve activity (MSNA) are normalized to baseline with saline infusion. C 11 The Authors. Journal compilation C 11 The Physiological Society

9 7 A. M. Kiviniemi and others Exp Physiol 9.8 pp Discussion The main finding of the present study was that the powers of LF oscillations in BP and RRi are affected by α- adrenergic blockade during baroreceptor unloading and to a lesser extent during static handgrip exercise and muscle metaboreflex activation. The power of LF BP oscillations did not increase with baroreceptor unloading, despite increased sympathetic activity; however, α-adrenergic blockade reduced the power of LF oscillations in BP during baroreceptor unloading. The frequency of LF fluctuations in BP and RRi increased during static exercise and muscle metaboreflex activation, regardless of the presence of an α-adrenergic blockade. However, α-adrenergic inhibition decreased the frequency of these oscillations in response to muscle metaboreflex activation (PECO). Slight increases in these frequencies during baroreceptor unloading were blunted by the α-adrenergic blockade. Thus, α-adrenergic activation appears to affect the frequency of LF fluctuation in BP and RRi. As the α-adrenergic effects on LF oscillatory patterns were less during exercise than during baroreceptor unloading or muscle metaboreflex activation (in the absence of volitional effort), the present findings support the hypothesis of neurogenic contributions to cardiac and peripheral mechanisms for BP regulation during sympathetic stimuli that affect LF oscillations. These contributions are complex and appear to depend on the mechanism of sympathetic activation and the outcome variable being assessed. Low-frequency oscillations in blood pressure and R R interval during baroreceptor unloading Baroreceptor unloading by passive head-up tilt or LBNP results in baroreflex-mediated increases in HR and Figure 3. Continued C 11 The Authors. Journal compilation C 11 The Physiological Society

10 Exp Physiol 9.8 pp Cardiovascular variability and sympathetic activity 77 Table. The effects of static handgrip exercise (HG) and subsequent postexercise forearm circulatory occlusion (PECO) on heart rate, blood pressure, breathing frequency and central haemodynamics during saline and PHE infusion P value for main effect Parameter Infusion Baseline HG PECO PHE Intervention Interaction HR (beats min 1 ) Saline 58 ± 1 7 ± 1 1 ± 9 <.1 <.1.1 PHE ± 1 9 ± 1 73 ± 1 SBP (mmhg) Saline 1 ± 9 15 ± 18 ±.3 <.1.31 PHE 133 ± ± ± 18 DBP (mmhg) Saline 7 ± 7 91 ± 9 88 ± <.1. PHE 77 ± 9 ± 8 83 ± 5 MBP (mmhg) Saline 91 ± 111 ± 1 18 ± 1. <.1.1 PHE 9 ± 113 ± 1 1 ± 8 Breathing frequency (breaths min 1 ) Saline. ±.9.5 ±..9 ± PHE.9 ±.11.1 ±.3.5 ±. Cardiac output (L min 1 ) Saline 5.5 ± ±.. ± PHE.1 ± ± ± 1.3 TPR (mmhg L 1 min ) Saline 17.9 ± ± ± PHE 1.8 ± ± ±.7 Values are means ± SD. Abbreviations: DBP, diastolic blood pressure; HR, heart rate; MBP, mean blood pressure; SBP, systolic blood pressure; and TPR, total peripheral resistance. P <.5 and P <.1 between control and phentolamine infusion. P <.5 and P <.1 compared with baseline. P <.5 and P <.1 compared with HG. Table 5. infusion The effects of HG and PECO on plasma noradrenaline concentration (NA; n = 11) and MSNA (n = ) during saline and PHE P value for main effect Parameter Infusion Baseline HG PECO PHE Intervention Interaction NA (nmol L 1 ) Saline 1.15 ±. 1.8 ± ±.5 <.1 <.1.9 PHE 1. ± ±..8 ±.7 Bursts min 1 Saline ± 7 33 ± ±.1 <.1.3 PHE 7 ± 1 5 ± 9 5 ± 1 Bursts 1 RRi 1 Saline 37 ± 11 ± 17 ± 1.73 < PHE ± 13 5 ± 8 5 ± 13 Burst amplitude (a.u.) Saline 5 ± 11 3 ± 9 9 ± PHE 3 ± 1 31 ± 13 8 ± 13 Low frequency Power (ln a.u. ) Saline. ± ± 1.5. ± PHE 3.8 ±.9.9 ± 1.7. ± 1. Power (n.u.) Saline 3 ± 7 3 ± 9 ± PHE ± 9 ± 31 ± 1 Med (Hz) Saline.13 ±.1.1 ±.1.11 ± PHE.98 ±..115 ±.1.11 ±.8 High frequency Power (ln a.u. ) Saline.8 ± ± ± PHE 5.1 ± 1.3. ± ± 1.7 Power (n.u.) Saline 7 ± 7 7 ± 71 ± PHE 78 ± 9 8 ± 9 ± 1 Med (Hz) Saline.5 ±.1.3 ±..59 ± PHE.1 ±.5.51 ±.9.5 ±.1 Values are means ± SD. Abbreviations: Med, median frequency; and RRi, R R interval. P <.5 and P <.1 between control and phentolamine infusion. P <.5 and P <.1 compared with baseline. systemic vascular resistance to maintain BP (Kitano et al. 5). This involves both vagal withdrawal and sympathetic excitation (Cooke et al. 1999; Furlan et al. ). Typically, baroreceptor unloading induces little change in the LF RRi power but decreases the power of HF oscillations in RRi (Cooke et al. 1999; Furlan et al. ). In contrast, the power of LF and HF BP oscillations increases during baroreceptor unloading (Cooke et al. C 11 The Authors. Journal compilation C 11 The Physiological Society

11 78 A. M. Kiviniemi and others Exp Physiol 9.8 pp Table. The effects of HG and subsequent PECO on R R interval and systolic blood pressure variability as well as baroreflex sensitivity during saline and PHE infusion P value for main effect Parameter Infusion Baseline HG PECO PHE Intervention Interaction Low frequency Power SBP (ln mmhg ) Saline 1. ±.7 1. ± ± PHE 1. ± ± ±.9 Power SBP (n.u.) Saline 9 ± 9 3 ± 1 59 ± PHE 8 ± 1 57 ± 17 5 ± 15 Med SBP (Hz) Saline.7 ±.9.9 ±.1.9 ±.1.17 < PHE.9 ±.7.88 ±.13.8 ±.15 Power RRi (ln ms ) Saline 7.1 ±.7.3 ±.9.9 ± PHE.7 ± ±..5 ±.8 Power RRi (n.u.) Saline ± 1 7 ± 1 33 ± PHE 5 ± ± 13 7 ± 15 Med RRi (Hz) Saline.73 ±.9.88 ±.1.99 ±.17. <.1.1 PHE.7 ±.9.85 ±.18.8 ±.19 BRS (ms mmhg 1 ) Saline 1. ± ±. 1.8 ± PHE 15. ± ± ± 8. High frequency Power SBP (ln mmhg ) Saline. ± ± ± PHE.7 ± ± ± 1.1 Power SBP (n.u.) Saline 31 ± 9 37 ± 1 1 ± PHE 3 ± 1 3 ± 17 ± 15 Med SBP (Hz) Saline.9 ±.5.53 ±.15.8 ± PHE.5 ±.. ±.17.5 ±. Power RRi (ln ms ) Saline 7. ±.8. ± ±.9 <.1 <.1.35 PHE.7 ± ± 1.. ± 1.3 Power RRi (n.u.) Saline 5 ± 1 53 ± 1 7 ± PHE 5 ± 15 7 ± ± 15 Med RRi (Hz) Saline.8 ±.3.51 ±.1.5 ± PHE.51 ±..39 ±.1.9 ±.18 BRS (ms mmhg 1 ) Saline 3. ± ± ± PHE 1. ± ± ± 11. Values are means ± SD. Abbreviations: BRS, baroreflex sensitivity; Med, median frequency; RRi, R R interval; and SBP, systolic blood pressure. P <.5 and P <.1 between control and phentolamine infusion. P <.5 and P <.1 compared with baseline. P <.5 and P <.1 compared with HG. 1999; Furlan et al. ; Kiviniemi et al. 1a). The frequency of LF oscillations in BP and RRi is suggested to be unaltered by passive head-up tilt (Cooke et al. 1999; Furlan et al. ; Kiviniemi et al. 1a). Owing to the tight coupling between oscillations in MSNA, BP and RRi at LF during orthostatic stimulus, it has been suggested that LF oscillations in BP and RRi provide pivotal information on sympathetic excitation (Montano et al. 199; Furlan et al. ). In the present study, LBNP without α-adrenergic blockade resulted in unchanged power of LF oscillations in RRi but only modest, non-significant increase in the power of BP oscillations at LF. Lower body negative pressure has shown similar responses in central haemodynamics and heart rate variability compared with passive head-up tilt (Butler et al. 1993; Kitano et al. 5). Therefore, this discrepancy remains unclear. Nonetheless, the main findings of the present study were observed during baroreceptor unloading with α-adrenergic blockade, suggesting that α-adrenergic activation has some impact on cardiovascular oscillations. First, α-adrenergic blockade during LBNP decreased both the absolute and the normalized power of BP oscillations at LF. Second, the inhibition of α- adrenergic vasoconstriction resulted in larger vagal withdrawal during baroreceptor unloading that most probably explains the lower power of LF oscillations in RRi, in addition to lower HF power of RRi oscillations. Interestingly, the responses of normalized LF power of RRi oscillations to LBNP were similar with and without α-adrenergic blockade. Third, α- adrenergic antagonism blunted the small increases in the frequency of LF oscillations in BP and RRi during LBNP. C 11 The Authors. Journal compilation C 11 The Physiological Society

12 Exp Physiol 9.8 pp Cardiovascular variability and sympathetic activity 79 The present results suggest that an α-adrenergic mechanism exerts an important impact on LF oscillations in BP during baroreceptor unloading by LBNP. Specifically, the normalized powers of BP oscillations showed that α-adrenergic inhibition resulted in pronounced BP oscillations at the breathing frequency as it blunted and slowed the LF oscillations in BP. This shift of spectral power was supported by the increased HF power of MSNA oscillations during LBNP with α- adrenergic inhibition. It seems that effective BP regulation during baroreceptor unloading results in pronounced or maintained LF oscillations in BP, whereas inhibition of α- adrenergic effects results in a shift in the oscillatory pattern of BP from the LF to the HF band. van de Borne et al. (1) observed this phenomenon at rest, while, in the present study, it was observed only during LBNP. It has been demonstrated that peripheral sympathetic activity oscillates with respiration (Eckberg et al. 1985). Therefore, inhibition of peripheral LF sympathetic effects appears to strengthen the coupling of MSNA and BP oscillations with respiration. Similar increases in normalized LF power of RRi oscillations during LBNP with and without α- adrenergic blockade support the hypothesis that these changes may be related to central increases in sympathetic activity and decreases in cardiac vagal modulation during baroreceptor unloading (Pagani et al. 1997; Furlan et al. A Power LF-SBP (ln, mmhg ) 3 1 Saline PHE Phentolamine.17 Intervention.5 Interaction. D Power LF-RRi (ln, ms ) Phentolamine. Intervention. Interaction.757 * B C Med LF-SBP (Hz) Power LF-SBP (n.u.) Phentolamine.3 Intervention.3 Interaction.3 Phentolamine.17 Intervention <.1 Interaction.181 E (n.u.) Med LF-RRi (Hz) Power LF-RRi F Phentolamine.7 Intervention. Interaction.31 Phentolamine. Intervention <.1 Interaction.1 **. Baseline HG PECO. Baseline HG PECO Figure. The absolute (A and D) and normalized powers (B and E) and the median frequencies (Med; C and F) of LF oscillations in SBP (A C) and RRi (D F) at baseline and during static handgrip exercise (HG) and subsequent postexercise forearm circulatory occlusion (PECO) with intravenous infusion of saline (filled diamonds) and phentolamine (open diamonds) P <.5 and P <.1 between control and phentolamine infusion. P <.5 and P <.1 compared with the baseline. C 11 The Authors. Journal compilation C 11 The Physiological Society

13 73 A. M. Kiviniemi and others Exp Physiol 9.8 pp ). However, the present data suggest that this hypothesis does not apply during the exercise pressor reflex. Low-frequency oscillations in blood pressure and R R interval during static exercise and muscle metaboreflex activation The pressor response to static exercise is related to increased central command and muscle metaboreflex activation that decrease cardiac vagal activity and increase sympathetic activity (Shoemaker et al. 7; Kiviniemi et al. 1a). During postexercise local ischaemia, cardiac vagal activity and HR return to baseline owing to the cessation of exercise-related central command, but BP remains high via sustained peripheral sympathetic activity by the muscle metaboreflex, potentially facilitating cardiac vagal activity via baroreflex (O Leary, 1993; Shoemaker et al. 7; Kiviniemi et al. 1a). As in the present study, earlier observations indicated that the absolute power of LF oscillations in BP did not change during static HG or PECO (Ichinose et al. 7; Kiviniemi et al. 1a). Also, we observed decreased normalized power of LF oscillations in BP during PECO. Together with the absence of α-adrenergic effects on LF power of BP oscillations during HG or PECO, these findings suggest that the LF power of BP oscillations may not serve as an indicator of sympathetic activity during the pressor response induced by static exercise and muscle metaboreflex activation. During static exercise, the power of LF oscillations in RRi has either remained A Blood pressure (mmhg) ECG / Heart rate (beats min -1 ) Power (mmhg. Hz -1 ) Power (ms Hz -1 ) x1-3 MSNA (mv) Power (au Hz -1 ). x1-3 Respiration Power (au Hz -1 ). seconds Frequency (Hz) Figure 5. Representative example of data and corresponding spectra during static handgrip exercise with saline (A) and with phentolamine (B) Spectra for muscle sympathetic nerve activity (MSNA) are normalized to baseline with saline infusion. C 11 The Authors. Journal compilation C 11 The Physiological Society

14 Exp Physiol 9.8 pp Cardiovascular variability and sympathetic activity 731 the same (Kiviniemi et al. 1a) or increased (Iellamo et al. 1999). Iellamo et al. (1999) conducted static leg exercise, which may partly explain this discrepancy. In the present study, HG did not significantly change the power of LF oscillations in RRi unless α-adrenergic blockade was present. As during LBNP, this is most probably due to the compensatory effect of larger vagal withdrawal during decreased vascular tone due to the α-adrenergic antagonism. This was observed more clearly in the HF power of RRi oscillations. The LF power of RRi oscillations returned to resting levels during PECO, regardless of α- adrenergic inhibition. Therefore, it seems that the LF power of RRi interval oscillations may follow changes in cardiac vagal activity rather than sympathetic effects during HG and PECO. In addition to the power of LF oscillations in RRi and BP, the frequency of these oscillations seems to carry significant clinical and physiological information (Takalo et al. 1997; Wichterle et al. ; Kiviniemi et al. 7, 1b). Reductions in the frequency of LF oscillations in RRi and BP are related to ageing and cardiovascular pathologies (Takalo et al. 1997; Kiviniemi et al. 1b). However, in a specific group of patients with recent myocardial infarction, the increased frequency of LF oscillations in RRi has been related to increased cardiac mortality (Wichterle et al. ; Kiviniemi et al. 7). It hasnot been establishedwhether these LF oscillations of RRi in high-risk patients are originating from baroreflex coupling between BP and RRi. In our recent study, the increased frequency of LF fluctuations in BP and RRi was related to the exercise pressor stimuli, whereas these frequencies were unaltered during baroreceptor unloading by passive head-up tilt B Blood pressure (mmhg) ECG / Heart rate (beats min -1 ) Power (mmhg. Hz -1 ) Power (ms Hz -1 ). 3 1 x1-8 x1-3 MSNA (mv) Power (au Hz -1 ). x1-3 Respiration Power (au Hz -1 ). seconds Frequency (Hz) Figure 5. Continued C 11 The Authors. Journal compilation C 11 The Physiological Society

15 73 A. M. Kiviniemi and others Exp Physiol 9.8 pp (Kiviniemi et al. 1a). The present results show that the frequencies of LF oscillations in BP and RRi are largely unrelated to α-adrenergic modulation during exercise, where cardiac autonomic regulation is the most potent determinant of the increased frequency of these LF oscillations. However, α-adrenergic blockade blunted, but did not completely abolish, the increase in the frequency of LF oscillations during PECO, whereas it completely removed the increases in these frequencies during LBNP. Thus, α-adrenergic effects are evidently involved in this respect during PECO and LBNP. Whether the α-adrenergic effects during PECO are related to changes in centrally driven oscillations of sympathetic activity or impaired coupling between MSNA and BP remains to be determined. Although the frequency of LF oscillations in MSNA was lower during PECO with A Blood pressure (mmhg) ECG / Heart rate (bpm) PHE compared with PECO without PHE, we cannot conclude which pathway was involved. Myogenic effects may also be present (Ichinose et al. 7; Zamir et al. 1), because BP was lower during PECO with PHE compared with without PHE. It is well known that changes in transmural pressure facilitate a vascular myogenic response, especially during increased sympathetic activity (Ping & Johnson, 199). In the present study, the blunted shift in the frequency of LF oscillations during LBNP with PHE may support the same hypothesis. Origin of low-frequency oscillations in blood pressure and R R interval There are two theories for the genesis of LF oscillations in BP. One theory suggests the presence of a central Power (mmhg. Hz -1 ) Power (ms Hz -1 ). 3 1 x1-8 x1-3 MSNA (mv) Power (au Hz -1 ). x1-3 Respiration Power (au Hz -1 ). seconds Frequency (Hz) Figure. Representative example of data and corresponding spectra during postexercise forearm circulatory occlusion with saline (A) and with phentolamine (B) Spectra for muscle sympathetic nerve activity (MSNA) are normalized to baseline with saline infusion. C 11 The Authors. Journal compilation C 11 The Physiological Society

16 Exp Physiol 9.8 pp Cardiovascular variability and sympathetic activity 733 oscillator in the autonomic nervous system which operates at.1 Hz. This idea is based on observations from animal studies, where persistent LF oscillations in BP have been observed after baroreceptor denervation (Kaminski et al. 197; Preiss & Polosa, 197). The other theory proposes that LF oscillations in arterial pressure are resonance that is caused by time delays and low-pass filtering properties of the baroreflex-mediated modulation of BP (Bertram et al. 1998; Julien et al. 3). Regardless, it is still not clear whether these oscillations are governed by the changes in systemic vascular resistance and peripheral sympathetic nerve activity (Preiss & Polosa, 197; O Leary & Woodbury, 199; Furlan et al. ; Liu et al. ) or autonomic effects on cardiac output (Myers et al. 1). Although it is known that different pathways of sympathetic activation are involved during the exercise pressor stimulus and baroreceptor unloading, the present data provide some new insight into the control of LF oscillations. They suggest that the contributions of cardiac and α-adrenergic modulation vary as a function of both the mode of sympathetic stimulus and the measured outcome. Different responses in LF oscillations were observed during exercise, PECO and LBNP, despite the similar amount of sympathetic activity during these interventions, as measured by circulating levels of NA and MSNA. Taken together, the power of LF oscillations in BP and RRi is largely affected by α-adrenergic antagonism during baroreceptor unloading, but to a lesser extent during static exercise and PECO. These findings may support the hypothesis that the origin of LF oscillatory patterns in BP may be dependent on the mode of sympathetic stimulus, with oscillations driven by cortical or metaboreceptor-based pathways being more prominent B Blood pressure (mmhg) Power (mmhg. Hz -1 ) 3 1 x1 - ECG / Heart rate (bpm) 8.7 Power (ms. Hz -1 ) 8 x1-3 MSNA (mv) Power (au. Hz -1 ) x1-3 Respiration Power (au. Hz -1 ) seconds Frequency (Hz) Figure. Continued C 11 The Authors. Journal compilation C 11 The Physiological Society

17 73 A. M. Kiviniemi and others Exp Physiol 9.8 pp during static exercise and delays in the baroreflex loop during orthostatic stimulus. Study limitations The present study is limited by the lack of successful MSNA recording from all subjects. Another limitation concerns the α-adrenergic blockade. As PHE is a competitive blocker,the complete blockade of α-adrenergic inputs is not possible. The interventions were conducted in a single session, and we were restricted to a relatively low dose of PHE. Nevertheless, presyncope symptoms were observed during LBNP in one participant, whose data were excluded from the analysis. The breathing frequency was guided throughout the interventions because of its considerable variations during spontaneous breathing (Kiviniemi et al. 1b). Although the breathing frequency was well maintained, it is an intervention on the cardiovascular system (Morris et al. 1), which may alter the oscillatory patterns studied here (Montano et al. 9). Conclusions These findings indicate that α-adrenergic modulation of BP has an important contribution to the power and the frequency of LF oscillations in BP and RRi during baroreceptor unloading. This α-adrenergic effect partly explains the increased frequency of these oscillations during muscle metaboreflex activation. The lack of α-adrenergic effects during static handgrip exercise may indicate that cardiac or other peripheral regulatory mechanisms may be more important in determining LF oscillations in BP and RRi during exercise. References Akselrod S, Gordon D, Ubel FA, Shannon DC, Berger AC & Cohen RJ (1981). Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat-to-beat cardiovascular control. Science 13,. Bertram D, Barrès C, Cuisinaud G & Julien C (1998). The arterial baroreceptor reflex of the rat exhibits positive feedback properties at the frequency of Mayer waves. J Physiol 513, Butler GC, Yamamoto Y, Xing HC, Northey DR & Hughson RL (1993). Heart rate variability and fractal dimension during orthostatic challenges. JApplPhysiol75, 1. Cooke WH, Hoag JB, Crossman AA, Kuusela TA, Tahvanainen KU & Eckberg DL (1999). Human responses to upright tilt: a window on central autonomic integration. JPhysiol517, deboer RW, Karemaker JM & Strackee J (1987). Hemodynamic fluctuations and baroreflex sensitivity in humans: a beat-to-beat model. Am J Physiol Heart Circ Physiol 53, H8 H89. Eckberg DL, Nerhed C & Wallin BG (1985). Respiratory modulation of muscle sympathetic and vagal cardiac outflow in man. JPhysiol35, Furlan R, Porta A, Costa F, Tank J, Baker L, Schiavi R, Robertson D, Malliani A & Mosqueda-Garcia R (). Oscillatory patterns in sympathetic neural discharge and cardiovascular variables during orthostatic stimulus. Circulation 11, Ichinose M, Koga S, Fujii N, Kondo N & Nishiyasu T (7). Modulation of the spontaneous beat-to-beat fluctuations in peripheral vascular resistance during activation of muscle metaboreflex. Am J Physiol Heart Circ Physiol 93, H1 H. Iellamo F, Pizzinelli P, Massaro M, Raimondi G, Peruzzi G & Legramante JM (1999). Muscle metaboreflex contribution to sinus node regulation during static exercise: insights from spectral analysis of heart rate variability. Circulation 1, 7 3. Julien C, Chapuis B, Cheng Y & Barres C (3). Dynamic interactions between arterial pressure and sympathetic nerve activity: role of arterial baroreceptors. Am J Physiol Regul Integr Comp Physiol 85, R83 R81. Kaminski RJ, Meyer GA & Winter DL (197). Sympathetic unit activity associated with Mayer waves in the spinal dog. Am J Physiol 19, Kitano A, Shoemaker JK, Ichinose M, Wada H & Nishiyasu T (5). Comparison of cardiovascular responses between lower body negative pressure and head-up tilt. JApplPhysiol 98, Kiviniemi AM, Tiinanen S, Hautala AJ, Seppanen T, Makikallio TH, Huikuri HV & Tulppo MP (1a). Frequency of slow oscillations in arterial pressure and R-R intervals during muscle metaboreflex activation. Auton Neurosci 15, Kiviniemi AM, Tiinanen S, Hautala AJ, Seppanen T, Norton KN,FrancesMF,NolanRP,HuikuriHV,TulppoMP& ShoemakerJK (1b). Low-frequency oscillations in R-R interval and blood pressure across the continuum of cardiovascular risk. Auton Neurosci 158, Kiviniemi AM, Tulppo MP, Wichterle D, Hautala AJ, Tiinanen S, Seppanen T, Makikallio TH & Huikuri HV (7). Novel spectral indexes of heart rate variability as predictors of sudden and non-sudden cardiac death after an acute myocardial infarction. Ann Med 39, 5. LaitinenT,HartikainenJ,NiskanenL,GeelenG&LansimiesE (1999). Sympathovagal balance is major determinant of short-term blood pressure variability in healthy subjects. Am J Physiol Heart Circ Physiol 7, H15 H15. La Rovere MT, Pinna GD, Maestri R, Mortara A, Capomolla S, Febo O, Ferrari R, Franchini M, Gnemmi M, Opasich C, Riccardi PG, Traversi E & Cobelli F (3). Short-term heart rate variability strongly predicts sudden cardiac death in chronic heart failure patients. Circulation 17, Liu HK, Guild SJ, Ringwood JV, Barrett CJ, Leonard BL, Nguang SK, Navakatikyan MA & Malpas SC (). Dynamic baroreflex control of blood pressure: influence of the heart vs. peripheral resistance. Am J Physiol Regul Integr Comp Physiol 83, R533 R5. Malliani A, Pagani M, Lombardi F & Cerutti S (1991). Cardiovascular neural regulation explored in the frequency domain. Circulation 8, 8 9. C 11 The Authors. Journal compilation C 11 The Physiological Society