P2.8 Action potential content in human muscle sympathetic nerve activity

76

Abstracts / Autonomic Neuroscience: Basic and Clinical 149 (2009) 1–126

P2.6 Comparison of baroreflex sensitivity determined by cross-spectral analysis at respiratory and 0.1 Hz frequencies in man N. Honzikova, B. Fiser, Z. Novakova, E. Zavodna (Department of Physiology, Faculty of Medicine, Masaryk University, Brno, Czech Republic) Non-invasive methods of determination of baroreflex sensitivity (BRS, ms/mmHg) are based on beat-to-beat systolic blood pressure and inter-beat interval recording. Sequential methods and spectral methods with spontaneous respiration include transient superposition of respiratory and 0.1 Hz rhythms. We use the cross-spectral method during breathing controlled by metronome at 0.33 Hz [1], which enables a separate determination of BRS at 0.1 Hz and respiratory frequencies. The aim of the present study was to evaluate the role of respiration in the spectral method of BRS determination. We recorded blood pressure by Finapres (5 min, metronome breathing 0.33 Hz) in 118 healthy subjects (age between 19 and 26 years) and 23 hypertensive patients (age between 24 and 60 years). We compared a modulus between the cross-spectral density of variation of pulse intervals and systolic blood pressure (ms⁎mmHg) and the spectral density of variation of systolic blood pressure (mmHg⁎mmHg) at 0.1 Hz (BRS 0.1 Hz), and at 0.33 Hz (BRS 0.33 Hz). A statistically significant correlation was found between BRS 0.1 Hz and BRS 0.33 Hz in controls (r = 0.52, p < 0.001), and in hypertensive patients (r = 0.52, p < 0.01) as well. The regression equations (controls: BRS 0.33 Hz = 2.63 + 1.14 ⁎ BRS 0.1 Hz; hypertensive patients: BRS 0.33 Hz = 2.43 + 1.38 ⁎ BRS 0.1 Hz, differences between the slopes and the identity line were insignificant) indicated the existence of a breathing-dependent BRS-non-related component. The ratio of BRS 0.1 Hz to BRS 0.33 Hz in both groups (controls: 0.876 ± 0.419; hypertensive patients: 0.579 ± 0.364) was significantly lower than 1 (p < 0.01). Thus, BRS evaluated at the breathing frequency overestimates the real baroreflex sensitivity. This is more pronounced at low values of BRS, which are more important for diagnostic purposes. In conclusion, determination of BRS by the spectral method at the breathing frequency overestimates the real BRS. For diagnostic purposes we recommend the evaluation of BRS at a frequency of 0.1 Hz using metronome-controlled breathing at a frequency which is substantially higher than 0.1 Hz. [1] Zavodna, E., Honzikova, N., Hrstkova, H., Novakova, Z., Moudr, J., Jira, M., Fiser, B., 2006. Can we detect the development of baroreflex sensitivity in humans between 11 and 20 years of age? Can. J. Physiol. Pharmacol. 84, 1275–1283. Acknowledgement: Supported by grant MSM 0021622402.

doi:10.1016/j.autneu.2009.05.135

P2.7 Cardiovascular effects of glossopharyngeal insufflation in divers K. Heusser (Institute of Clinical Pharmacology, Hannover Medical School, Hannover, Germany), G. Dzamonja (Department of Neurology, Clinical Hospital Split, Split, Croatia), J. Tank (Institute of Clinical Pharmacology, Hannover Medical School, Hannover, Germany), I. Palada (Department of Physiology, University of Split School of Medicine, Split, Croatia), Z. Valic (Department of Physiology, University of Split School of Medicine, Split, Croatia), D. Bakovic (Department of Physiology, University of Split School of Medicine, Split, Croatia), A. Obad (Department of Physiology, University of Split School of Medicine, Split, Croatia), V. Ivancev (Department of Physiology, University of Split School of Medicine, Split, Croatia), T. Breskovic (Department of Physiology, University of Split School of

Medicine, Split, Croatia), A. Diedrich (Autonomic Dysfunction Service, Vanderbilt University, Nashville, Tennessee, USA), M.J. Joyner (Department of Anesthesiology, Mayo Clinic College of Medicine, Rochester, Minnesota, USA), F.C. Luft (Experimental and Clinical Research Center, Charité, Max-Delbrueck-Centrum, HELIOS Klinikum, Berlin, Germany), J. Jordan (Institute of Clinical Pharmacology, Hannover Medical School, Hannover, Germany), Z. Dujic (Department of Physiology, University of Split School of Medicine, Split, Croatia) Trained apnea divers take a deeper breath before bouts of maximal apnea than control subjects. The greater reduction in cardiac output may cause a stronger increase in sympathetic vasoconstrictor activity resulting in expanded apnea time. We hypothesized that glossopharyngeal inhalation (GI) leads to even lower cardiac output and stronger sympathetic activation. In nine apnea divers (28.8 ± 4.5 years, 2 women) we monitored blood pressure, heart rate, thoracic impedance and cardiac output (impedance cardiography), arterial oxygen saturation, and muscle sympathetic nerve activity (MSNA) by microneurography. Recordings were taken during eight cumulative 4-seconds inhalation steps and during maximal end-inspiratory dry breath-holds with and without additional GI in randomized order. The experiments have been approved by the ethical committee of the University of Split, School of Medicine. We found a linear relationship between inhaled air volume and thoracic impedance allowing monitoring of GI effectiveness. GI elevated air volume within the lungs (p < 0.01) and prolonged apnea time (4.2 ± 0.9 vs 3.8 ± 1.1 min, p < 0.05) without changing desaturation level. The significant increases in arterial pressure and MSNA at the end of apnea were not affected by GI, whereas bradycardia was enhanced (−18.5 ± 18.3 vs − 11.2 ± 17.0 bpm, p < 0.05). Unexpectedly, the early reduction in cardiac output was not augmented by GI (−4.4 ± 2.0 vs −3.5 ± 1.3 l/min, p = ns). In trained apnea divers, GI prolongs apnea time. However, our data argue against the assumption that GI causes a greater reduction in cardiac output and, consequently, a larger sympathetic vasoconstriction, which in turn may conserve oxygen.

doi:10.1016/j.autneu.2009.05.136

P2.8 Action potential content in human muscle sympathetic nerve activity A. Salmanpour (Department of Electrical and Computer Engineering, the University of Western Ontario, London, Ontario, Canada N6A 5B9; The Neurovascular Research Laboratory, the School of Kinesiology, the University of Western Ontario, London, Ontario, Canada N6A 3K7), L.J. Brown (Department of Electrical and Computer Engineering, the University of Western Ontario, London, Ontario, Canada N6A 5B9), J.K. Shoemaker (The Neurovascular Research Laboratory, the School of Kinesiology, the University of Western Ontario, London, Ontario, Canada N6A 3K7; Department of Physiology and Pharmacology, the University of Western Ontario, London, Ontario, Canada N6A 5C1) Microneurographic studies in human sympathetic nerve recordings are traditionally interpreted through the integrated neurogram. This approach provides information on the summed magnitude of sympathetic outflow but loses considerable neurophysiologic information. Nonetheless, the integration process has been necessary because indepth interpretation of action potential discharge patterns in the raw multi-unit signal has not been possible due to signal-to-noise challenges. The aims of the present study were to a) establish a method of detecting sympathetic action potential (AP) content of single bursts in human muscle sympathetic nerve activity (MSNA), and b) use this approach to test the hypothesis that larger integrated bursts contain

Abstracts / Autonomic Neuroscience: Basic and Clinical 149 (2009) 1–126

new populations of axons that have a faster conduction velocity and higher recruitment threshold. Bursts of varying sizes were obtained from MSNA collected in the supine position after a stable baseline had been observed for at least 5 min. To obtain a suitable set of large and small bursts, the amplitude of the largest burst in the integrated MSNA was found and other large bursts were selected to have amplitudes between 75% and 98% of the amplitude of the largest burst. The small bursts were chosen to have amplitudes between 35% and 50% of the amplitude of the largest burst. Ten large and ten small bursts (both filtered and integrated MSNA) were selected from each of six healthy individuals. Filtered raw MSNA data were denoised with a stationary wavelet transform (SWT) and action potentials (APs) were detected with a peak detection algorithm. Detected APs were clustered using the k-means method and the cluster averages were calculated [1,2]. The large bursts had more APs (22 ± 3) than the small bursts (11 ± 2) (P < 0.001). Compared with small bursts (1.30 ± 0.044 s), the large bursts had a shorter reflex latency (1.23 ± 0.051 s; P < 0.001). Compared with small bursts, the large bursts also had greater average (27 ± 9 vs. 45 ± 10 Hz) (P < 0.001) and maximum (177 ± 33 vs. 251 ± 17 Hz) (P < 0.001) discharge frequencies. Also, there was more activity in the early part of the distribution of the APs in the large bursts compared with the small bursts. However, based on AP morphology, no evidence of a different population of neurons was present in the large versus small bursts. This study provides a methodological advance to study AP discharge patterns in human sympathetic neural recordings. The data indicate that larger bursts with shorter conduction delays are comprised of more APs, a higher discharge frequency and more activity in the front-end of the burst without differences in the AP morphology. Thus, the data suggest that, under baseline conditions, larger bursts reflect changes in central or ganglionic synaptic delays of existing neuronal populations rather than recruitment of a new population of faster-conducting neurons. [1] Salmanpour, A., Brown, L.J., Shoemaker, J.K., 2008. Detection and classification of raw action potential patterns in human muscle sympathetic nerve activity. Proc. IEEE Eng. Med. Biol. Soc. (IEEE EMBS), 2928–2931. [2] Salmanpour, A., Brown, L.J., Shoemaker, J.K., 2008. Performance analysis of stationary and discrete wavelet transform for action potential detection from sympathetic nerve recordings in humans. Proc. IEEE Eng. Med. Biol. Soc. (IEEE EMBS), 2932–2935. Acknowledgements: Supported by the Natural Sciences and Engineering Research Council of Canada and the Canadian Institutes for Health Research.

doi:10.1016/j.autneu.2009.05.137

P2.9 Baroreflex open-loop gain at rest and during exercise in man B. Fiser, N. Honzikova, J. Moudr, Z. Novakova, E. Zavodna (Department of Physiology, Faculty of Medicine, Masaryk University, Brno, Czech Republic) The evaluation of open-loop feedback gain (G) of the baroreflex in humans at rest and during exercise by a non-invasive method was the aim of the present study. G is the response of blood pressure on the baroreceptor stimulus of 1 mmHg described by the equation: G = (SV ⁎ HR ⁎ TPR) − (SV − BRSsv) ⁎ (HR − BRShr) ⁎ (TPR − BRStpr); (SV − stroke volume; HR − heart rate; TPR − total peripheral resistance; BRSsv, BRShr, BRStpr − baroreflex sensitivity, changes of SV, HR, and TPR respectively, elicited by stimulus 1 mmHg). We can drop second-order terms and BRSsv from the equation to simplify: G = Ghr + Gtpr; Ghr = MBP ⁎ (BRShr / HR); Gtpr = MBP ⁎

77

(BRStpr / TPR); (Ghr – gain of heart loop; Gtpr – gain of peripheral resistance loop; MBP – mean blood pressure). Ghr was determined from baroreflex sensitivity BRS (ms/mmHg) by spectral analysis of the Finapres blood pressure record. Gtpr was estimated from the resonance peak (between 0.05 and 0.15 Hz) by means of mathematical model stimulation. The model is a closed-loop system with two parallel branches. The cardiac branch is composed of time delay (0.2 s) and a low-pass filter with a time constant (T = 0.88 s). The peripheral resistance branch is composed of time delay (0.4 s) and two low-pass filters in series with a time constant (T = 1.6 s). We found Ghr (mean ± SD) 1.84 ± 0.27 and Gtpr 2.13 ± 0.23 at rest and Ghr 0.44 ± 0.20 and Gtpr 1.00 ± 0.15 during exercise (1W per kg body weight) in 12 young adult subjects. The difference in Ghr and Gtpr between the rest and the exercise is significant (p < 0.05). It is concluded that the open-loop gain of the baroreflex is decreased during exercise, which enables improved perfusion of the skeletal muscle. Acknowledgement: Supported by grant MSM 0021622402. doi:10.1016/j.autneu.2009.05.138

P2.10 Summation of afferent input affects sympathetic homeostasis: Mild skin tactile stimulation during painful isometric muscle contraction reduces perceived pain but augments muscle sympathoexcitation in man H. Krämer (Department of Clin Neurophysiology, Sahlgren Univ Hosp, Göteborg, Sweden; Department of Neurology, Mainz Univ Hosp, Mainz, Germany), Y.B. Sverrisdottir (Department of Clin Neurophysiology, Sahlgren Univ Hosp, Göteborg, Sweden), F. Birklein (Department of Neurology, Mainz Univ Hosp, Mainz, Germany), H. Olausson (Department of Clin Neurophysiology, Sahlgren Univ Hosp, Göteborg, Sweden), M. Elam (Department of Clin Neurophysiology, Sahlgren Univ Hosp, Göteborg, Sweden) Activation of myelinated and unmyelinated (C tactile) low threshold mechanoreceptor afferents via gentle stroking of the skin reduces the perception of pain induced by a simultaneously delivered noxious stimulus. We recorded pain perception (visual analogue scale; VAS) and muscle sympathetic nerve activity (MSNA) in 18 healthy subjects performing isometric handgrip to exhaustion (approx. 5 min; last 2 min rated as painful), and added vibration (activating myelinated afferents) or skin stroking (activating myelinated and C tactile afferents) of the contracting arm, to investigate if addition of a cutaneous stimulus affects the muscle metaboreflexinduced MSNA augmentation associated with isometric exercise. Compared to isometric handgrip alone, the addition of skin vibration resulted in a slight reduction of perceived pain whereas the contraction-induced MSNA increase remained unaffected. In contrast, adding gentle skin stroking resulted in a more marked pain reduction while the MSNA increase was markedly exaggerated. Based on the different effects of applying skin vibration or skin stroking during muscle contraction, our interpretation is that the augmented sympathoexcitation during skin stroking is driven by an additive effect of a different C afferent input, and not closely linked to the degree of pain induced by the muscle afferent input. Given that a generalized MSNA increase will favour a redistribution of blood flow to the active muscle, this added sympathoexcitation may contribute to the reduced pain perception. Supported by the Swedish MRC (proj no 12170), and approved by the Göteborg Ethics Com. doi:10.1016/j.autneu.2009.05.139