Épilepsie et coeur Épilepsies 2010 ; 22 (3) : 194-200

Heart rate variability as measurement of heart-brain interactions Mario Valderrama1, Vincent Navarro1,2, Michel Le Van Quyen1

Copyright © 2017 John Libbey Eurotext. Téléchargé par un robot venant de 37.44.207.187 le 29/01/2017.

1

Centre de recherche de l’Institut du cerveau et de la moelle épinière (CRICM) INSERM UMRS 975 - CNRS UMR 7225-UPMC, hôpital de la Pitié-Salpêtrière, 47, Bd de l’Hôpital, 75651 Paris Cedex 13, France 2 AP-HP, Epilepsy Unit, groupe hospitalier Pitié-Salpêtrière, 47, Bd de l’Hôpital, 75651 Paris Cedex 13, France

Abstract. Autonomic mediated heart-brain interactions are implicated in the regulation of different physiological and pathological situations. They reflect the influence of parasympathetic and sympathetic systems over the cardiac activity. In past years, a considerable number of methodological approaches applied to non-invasive electrocardiogram (ECG) recordings have been developed for the assessment of autonomic function. They are based on quantitative analyses of the heart rate variability (HRV), which, through different measurements in time and frequency domains or through non-linear estimations, have revealed useful information in diverse clinical contexts. In the present article, we provide a review of the general aspects related to the analysis of HRV including a brief summary of the physiological bases and methodological approaches together with some reported applications. Key words : autonomic modulation, heart-brain interactions, heart rate variability

Résumé.

Les variabilités du rythme cardiaque, témoins d’interactions coeur/cerveau

Les interactions cœur/cerveau sont régulées par le système nerveux autonome et sont impliquées dans différentes situations physiologiques ou pathologiques. Ces interactions sont médiées par l’influence des systèmes sympathiques et parasympathiques sur l’activité cardiaque. Durant ces dernières années, un nombre considérable d’approches méthodologiques, appliquées aux enregistrements non invasifs de l’électrocardiogramme (ECG), ont été développées pour caractériser la fonction du système autonome. Ces approches sont fondées sur des analyses quantitatives de la variabilité du rythme cardiaque (Heart rate Variability [HRV]), lesquelles, à travers différentes mesures dans le temps et la fréquence ou à travers des estimations non linéaires, ont permis d’extraire des informations utiles dans divers contextes cliniques. Dans cet article, nous présentons une revue des aspects généraux liés à l’analyse du HRV incluant un bref rappel des bases physiologiques et des approches méthodologiques ainsi que quelques applications. Mots clés: modulation du système autonome, interactions cœur-cerveau, variabilité du rythme cardiaque

Épilepsies, vol. 22, n° 3, juillet-août-septembre 2010

194

brain, the action of the second influences and/or controls the behavior of all same parts including the heart. Indeed, although the activity of the heart depends on the intrinsic rhythm of its peacemaker, the heart operation

doi: 10.1684/epi.2010.0323

Tirés à part : M. Le Van Quyen

The heart and the brain are two of the most important entities of the human body. While, in general terms, the action of the first sustains the supply of nutrients necessary for the functioning of all parts of the body including the

Copyright © 2017 John Libbey Eurotext. Téléchargé par un robot venant de 37.44.207.187 le 29/01/2017.

Heart rate variability as measurement of heart-brain interactions

Parasympathetic and sympathetic control centers in the brain are located principally in the brainstem, hypothalamus and spinal cord. Higher levels of the cortex can also activate the autonomous system via signals to lower centers. Similarly, visceral organs can induce autonomic reflex responses by a complex mechanism of afferent-efferent signals. Contrarily to sympathetic terminals that are distributed to all parts of the heart, parasympathetic ones innervate principally the sinoatrial (SA) and atrioventricular (AV) nodes (Guyton and Hall, 2006). Both systems are constituted by preganglionic and postganglionic neurons, however, conversely to preganglionic sympathetic fibers, preganglionic parasympathetic ones reach directly the wall of the heart where they make synaptic junctions with postganglionic neurons. Parasympathetic fibers are supplied to the heart by the vagus nerves. Cell bodies of vagal neurons are located predominantly in the nucleus ambiguous, the dorsal motor nucleus, and the intermediate zone between these brainstem’s regions (Taylor et al., 1999). The action of the parasympathetic cardiac fibers is mediated by the acetylcholine neurotransmitter which, when released, decreases, among other effects, the frequency of the heart peacemaker. Sympathetic cell bodies of neurons influencing the behavior of the heart are located mostly in the medulla, from where respective axons travel down the spinal cord, exit at specific thoracolumbar levels and terminate upon the postganglionic sympathetic neurons (Klabunde, 2004). The sympathetic action to the heart is mediated primarily by norepinephrine neurotransmitter. Sympathetic stimulation causes, among others effects, an increase in the discharge rate of SA node.

rate or simply the heart rate (HR) is determined by a balance between influences coming from external sources located principally at the level of the brain. The external impulses innervating the heart correspond to two main branches of the autonomic nervous system: the parasympathetic and sympathetic systems. The action of these systems can induce modifications of the heart activity in different ways and circumstances. For example, situations of mental stress (Murata et al., 1999; Jouven et al., 2009), fear or anxiety (Baduí et al., 1996; Berntson et al., 1998) can rapidly increase the HR. Similarly, normal physiological states like rest, exercise (Perini and Veicsteinas, 2003) or sleep stages (Elsenbruch et al., 1999) are correlated to different changes in the autonomic modulation of the cardiac response. In addition, neuropathological conditions such as epilepsy or major depressive disorder that may be associated with autonomic dysfunctions can also chronologically modify the normal activity of the heart and increase the risk of sudden and unexpected death (Davis and Natelson, 1993; Carney et al., 2001; Nashef and Tomson, 2008). The regulatory function of the autonomic system can be quantitatively assessed by the analysis of the heart rate variability (HRV). HRV makes reference to the study of both the variability of intervals between consecutive heart beats and the variability between consecutive instantaneous heart rates (Task Force, 1996). Due to its relatively simple implementation methodology and the fact that it derived from a non-invasive technique, the analysis of different HRV measurements has became one of the most accepted and used procedures for the study of the fluctuating balance between the sympathetic and parasympathetic systems. In the following sections we will review the main physiological and analytical aspects related to the study of HRV. We will first shortly consider the physiological background of the autonomic modulation of the HR, we will continue next with the principal steps and methods for the quantitative assessment of the HRV, and finally, we will consider the relation of some HRV measurements and sleep and particularly the fluctuations of autonomic activity during specific sleep stages.

The chronotropic effects (i.e., those that change the heart rate) of sympathetic and vagus nerve electrical stimulation have been experimentally studied in animals (Henning and Khalil, 1989; Brack et al., 2004) and humans (Lewis et al., 2001). Vagal stimulation experiments have shown that the gradual increase of the stimulation frequency from ~ 0 – 20 Hz, while hyperbolically decreases the heart rate (Hondeghem et al., 1975; Parker et al., 1984), linearly increases the pulse interval between consecutive beats (Parker et al., 1984), suggesting that vagal activity instead of directly controlling the heart rate, rather influences it by the regulation of the time interval between successive beats (Hainsworth, 2004). Similarly, experiments where cardiac sympathetic nerves have been electrically stimulated have confirmed their positive chronotropic effect of increasing the heart rate but, conversely to vagal stimulation which can change the heart rate within one cardiac cycle (Levy et al., 1993), sympathetic stimulation entails a delay or latency before appreciating a change in the heart rate response (Ng et al., 2001). In addition, experiments where both autonomic branches have been simultaneously stimulated have indicated that while the positive chronotropic effect of sympathetic stimulation is reduced in the presence of background vagal stimulation, the negative chronotropic effect of vagal stimulation is on the contrary enhanced when

Autonomic regulation of heart rate As mentioned above, the heart is innervated with parasympathetic and sympathetic nerves. The effects of both autonomic branches on the HR are, in general terms, opposite: while the sympathetic stimulation tends to increase the HR, the parasympathetic one tends to decrease it. In absence of the external influence of any of these systems, the cardiac rhythm is marked, in normal cases, by the intrinsic rhythmicity of the sinoatrial node (the heart peacemaker), which varies from 100 to 120 beats per minute (bpm) (Hainsworth, 2004). In resting state, although both systems are activated, there is a predominance of parasympathetic influence which is more important in supine positions (Perini and Veicsteinas, 2003) and in trained individuals (Yamamoto et al., 2001). 195

Épilepsies, vol. 22, n° 3, juillet-août-septembre 2010

M. Valderrama, et al.

depolarization and repolarization of the atria and ventricles (figure 1A). Among these waves, those forming the so-called QRS complex, and more specifically the R wave, are usually selected for the extraction of beat-to-beat intervals or R-R signal. Within a single cardiac cycle, the QRS complex represents the depolarization of ventricles and so its detection can be useful for estimating the time interval between consecutive cycles. Due to its high amplitude and sharpness, the time of occurrence of each cardiac cycle (tR) is commonly associated with the maximal peak of the R wave and therefore a considerable number of online and offline methods have been developed for its detection (for review see Köhler et al., 2002). After having detected all successive tR, the R-R intervals signal, also known as R-R tachogram, is obtained by calculating the successive differences between consecutive tR (i.e., tR(i) - tR(i - 1)). An example of an ECG signal and the extracted R-R intervals is presented in figure 1. The R-R signal represents thus the evolution in time of the time duration between consecutive cardiac cycles. These successive intervals are usually measured in seconds or milliseconds. From this R-R signal, a corresponding instantaneous heart rate signal, with values in beats per minute, can be obtained through the following conversion:

Copyright © 2017 John Libbey Eurotext. Téléchargé par un robot venant de 37.44.207.187 le 29/01/2017.

background sympathetic stimulation is present, thus evidencing a dominant vagal role on chronotropic effects (Brack et al., 2004). As mentioned above, the activity of autonomic system is regulated by a complex mechanism of efferent reflexes in response to internal visceral afferent inputs. In the case of the heart, there are different types of reflexes that induce chronotropic effects as a consequence of an increment or decrement of the sympathetic or parasympathetic activity. Some of the reflexes accompanied by an increased parasympathetic activity are for example baroreceptors, which regulate the arterial blood pressure and are generated in response to an increased arterial pressure; cardiac reflexes are mediated by receptors in atria and ventricles and are generated under an increased chamber pressure which increases the parasympathetic activity and in consequence reduces the HR (Klabunde, 2004). Additional to these afferent-efferent system of visceral reflexes, other, diverse reflex mechanisms related to common activities such as breathing, physical exercise, changes in body position, etc., can also induce changes in HR (Hainsworth, 2004). In the case of breathing, the respiratory sinus arrhythmia is a normal arrhythmia associated with a synchrony between the respiration rate and the HR variability in which the beat-to-beat interval is shortened during inspiration and prolonged during expiration as a consequence of variations in cardiac vagal efferent activity (Yasuma and Hayano, 2004).

bpmðiÞ ¼ 60=t R ðiÞ−t R ði−1Þ where bpm(i) is an instantaneous estimation of the number of beats that would be contained in one minute (figure 1C). Although the instantaneous heart rate signal can be used, the R-R signal is usually utilized for the processing of different HRV measurements. Moreover, for the HRV estimation, it is recommended to select only normal beat-to-beat intervals, or the so-called normal-to-normal or NN intervals, and not those representing ectopic phenomena such as abnormal arrhythmias or their compensatory pauses (Task Force, 1996).

Higher cortical centers, especially those located in the limbic cortex, can also induce changes in HR. Some of changes can be associated with the occurrence of external stimuli. In the case of emotions (e.g., fear, anger, sadness) for instance, the implication of regions such as the amygdala has been correlated to chronotropic effects, with relative increases in HR associated with negative stimuli and relative decreases associated with positive ones (Critchley et al., 2005), or, in a general manner, with an increase in HR associated with a greater amygdala activation (Yang et al., 2007).

Time domain methods The most representative methods for the analysis of HRV in time domain can be divided in statistical and geometrical methods. Some of the most common statistical methods include: SDNN, the standard deviation of NN intervals; SDANN, the standard deviation of the average NN interval, usually calculated over short periods of 5 minutes; RMSSD, the root mean square estimated from the differences of successive NN intervals; NN50, the number of differences of successive NN intervals greater than 50 ms; pNNSO, the proportion calculated from the ration between NN50 and the total number of NN intervals. Geometrical methods are based on measurements that are estimated from the geometrical form of, for instance, the distribution of NN intervals or the distribution of differences between consecutive NN intervals. Sample density distributions can be approximated by the histogram of respective data series and from them, measurements like the HRV triangular index, which corresponds to the total number of data points (i.e., total number of NN or differences between consecutive NN intervals) divided by number of data points in the modal bin, can be calculated. Geometrical

Heart rate variability measurements The analysis of HRV measurements has been applied to the study of different physiological or pathological contexts where the assessment of the autonomic activation at particular times has been required (Kleiger et al., 2005; Lewis, 2005). Those analyses are derived from diverse methodological techniques that have been developed for the estimation of the autonomic function (parasympathetic and sympathetic) from the noninvasive electrocardiogram (ECG) signal (for a comprehensive review on HRV analysis see Task Force, 1996).

ECG preprocessing: R-R intervals signal Standard ECG recordings, for example those recorded from bipolar lead I, II or III derivations (Guyton and Hall, 2006), are shaped by particular waves that represent the sequence of

Épilepsies, vol. 22, n° 3, juillet-août-septembre 2010

196

Heart rate variability as measurement of heart-brain interactions

A

R

R−R

R

T P Q 1m V 0.5 sec

C

Instantaneous heart rate (bpm)

1

120

0.95 110 0.9 0.85 0.8 0.75 0.7 0.65 0.6

R − R interval (sec)

Copyright © 2017 John Libbey Eurotext. Téléchargé par un robot venant de 37.44.207.187 le 29/01/2017.

B

100

bpm =

90

60 R − R (sec)

80 70 60

0.55

0

100

200

300

400

500

600

700

800

0.5 900 1 000

0

100

200

300

400

500

600

700

800

50 900 1000

Time (sec)

Time (sec)

Figure 1. R-R intervals and instantaneous heart rate signals from the non-invasive ECG. (A) ECG recording with the main component waves. (B) R-R signal obtained from the time duration between successive cardiac cycles. (C) Instantaneous heart rate in beats per minute (bpm) determined from the R-R intervals.

distributed with frequency (Marple, 1987). Commonly, two different approaches have been implemented for the PSD estimation: non-parametric and parametric methods. While the first is usually based on the computation of the fast Fourier transform (FFT), which is computationally faster, the second requires the estimation of a parametric model (i.e., the estimation of the coefficients of the model), usually an autoregressive model, that depends on additional parameters such as the type and the order of the chosen model, but provides a smoother function from where is usually easier to identify the contribution of each one of the spectral bands of interest. Depending on the length of the R-R signal (preferentially the NN intervals), several spectral components have been defined which are supposed to be associated with individual or interactive activities of parasympathetic and/or sympathetic branches. For short-time or long-time recordings (~ 5 minutes), the main frequency bands have been defined as follows: high frequency component (HF), which refers to the spectral power between frequencies from ~ 0.15 to 0.4 Hz, is generally considered as a marker of a respiration-mediated vagal modulation, determined in consequence by the frequency of breathing (Sztajzel, 2004); low frequency component (LF), associated with the power between frequencies from ~ 0.04 to 0.15 Hz, reflects generally influences from both autonomic branches (Lewis, 2005); very low frequency component (VLF), referred to the

methods depend on the chosen size of the bin but are less sensible to the presence of not removed ectopic intervals than statistical ones. Different variations of HRV time-domain indexes have been correlated to diverse physiological or pathological contexts. In general terms, a reduced HRV, when compared to control sets, has been associated with pathology (Stein and Kleiger, 1999) or negative prognostic implications (Goldberger et al., 2001). This can be understood as the more healthy the heart, the more variable its rhythmic activity or the wider the range of possible reaction states. In postmyocardial infarction for instance, several studies have revealed that a low statistical and/or geometrical index values (e.g., SDNN or HRV triangular index) are associated with higher risks of mortality (Wolf et al., 1978; Kleiger et al., 1987; Farrell et al., 1992). This reduction on the variability has been suggested to reflect a modification of the normal sympatho-vagal balance consisting in a decreased parasympathetic tone and an increased sympathetic activity (Lombardi, 2004).

Frequency domain methods Frequency domain measurements constitute probably one of the most employed tools for the analysis of the HRV. They are based on the estimation of the power spectral density (PSD) function, which describes how the variance (i.e., power) is 197

Épilepsies, vol. 22, n° 3, juillet-août-septembre 2010

M. Valderrama, et al.

(e.g. sleep spindles) that alternate with a rather low-amplitude desynchronized background. During deep stages, the highamplitude slow activity becomes predominant and synchronized across large cortical extents. REM sleep is on the contrary characterized by a low-voltage desynchronized activity that is accompanied by an absence of muscular tone. When the autonomic regulation during these physiological states has been investigated, clear differences between different HRV quantifications have been documented. In the case of spectral indexes for instance, one study has shown an increase of the power in the HF component in light sleep accompanied by a simultaneous reduction of VLF and LF components, suggesting a concurrent increment of parasympathetic tone with a decrement of sympathetic activity during this stage of sleep (Vaughn et al., 1995). This same study reported a decrease in the power of all spectral components during deep sleep, which was associated with a reduction of both autonomic influences, and an increase of power in the VLF and LF components during REM sleep, explained by a possible increase in sympathetic influence. Although these results are consistent with those found by other studies, some of them have reported nevertheless an explicit increase of the power in the HR component during deep sleep, reflecting a preferential parasympathetic influence, as well as an explicit decrease of the same spectral component during REM, suggesting a supplementary parasympathetic withdraw during this stage (Baharav et al., 1995; Elsenbruch et al., 1996; Toscani et al., 1996). In two studies, the spectral power activity of the EEG has been correlated to the spectral components of the HRV (Ako et al., 2003; Miyashita et al., 2003). In both, the EEG spectral components related to the presence of absence of slow waves have been statistically significant correlated to the activity of LF and LF / HF components but not to the HF one, suggesting that the autonomic regulation during sleep is especially influenced by sympathetic or sympatho-vagal balance fluctuations rather than by a predominant vagal influence. In order to appreciate the cardiac autonomic fluctuations during sleep, figure 2 presents the simultaneous evolution of some HRV indexes in time and frequency domains during successive NREM and REM stages of a seizure-free night corresponding to an epileptic patient following an EEG-Video examination. The alternating sleep stages can be clearly appreciated in the scalogram (figure 2 Top), which represents the power content of different frequency bands as a function of time. In particular, during deep sleep periods of NREM sleep, characterized by a strong presence of slow waves (red boxes in figure 2) as can be confirmed by the high power in low frequency bands (~ 0.3 – 3 Hz), the instantaneous heart rate and the standard deviation of consecutive beat-to-beat intervals (SDNN) decrease, reflecting a slow and regular heart rhythm respectively, which, in turn, indicates a presumably preferential respiration-mediated vagal influence on the heart rate, manifested by a prominent increase of the HR spectral component at the same time of a notorious decrement of VLF and LF indexes (green ellipse in figure 2). On the contrary, during REM sleep (blue boxes in figure 2) the heart

Copyright © 2017 John Libbey Eurotext. Téléchargé par un robot venant de 37.44.207.187 le 29/01/2017.

power of frequencies below ~ 0.04 Hz, is a major determinant of physical activity and has been proposed to reflect a marked sympathetic activity (Sztajzel, 2004). When long-time recordings are available, one additional spectral component can be determined associated with frequencies lower than 0.003 Hz: ultra-low frequency component (ULF). Additionally, the ratio LF / HF is considered as an index of the global sympathovagal balance (Malliani et al., 1991). As time domain measurements, frequency domain ones have also been applied in different clinical and physiological contexts. In experiments where HRV has been investigated under diverse physiological conditions such as different corporal positions at rest, altitude, exercise, water immersion, light exposure, sleep and others, frequency domain indexes have revealed to be valuable markers of the autonomic interaction, with sensible variations of corresponding spectral components related to changes in levels of parasympathetic or sympathetic activations (Toscani et al., 1996; Tsunoda et al., 2001; Perini and Veicsteinas, 2003). Similarly, pathological conditions such as epilepsy (Harnod et al., 2009), depression (Carney et al., 2001), or different cardiovascular diseases (Sztajzel, 2004), have also appeared to present alterations of spectral components that have been related to an abnormal autonomic regulation (Malliani and Montano, 2004).

Complementary HRV analysis methods Besides time and frequency domain indexes, complementary approaches based on non-linear dynamics or timefrequency representations have also been applied to the analysis of HRV. Non-linear approaches attempt in general to extract information from the complex structure of the R-R signal, which is not readily available with other types of measurements. Among the most common non-linear methods, power-law slope, detrended fluctuation analysis (DFA) and approximate entropy have proved to be informative in clinical contexts (for a review see Mäkikallio et al., 2004).

Heart rate variability: an application to sleep During sleep, the evolution of different HRV indexes is marked by the presence of sleep stages. These changes reflect the fluctuation of the autonomic regulation during different physiological brain states. In general terms, numerous studies have shown that while a marked vagal modulation, mediated by the respiratory activity, is associated with non-rapid eye movement (NREM) sleep, an increased sympathetic activity is observed during rapid eye movement (REM) periods of sleep (Baharav et al., 1995; Vanoli et al., 1995; Vaughn et al., 1995; Toscani et al., 1996; Scholz et al., 1997; Elsenbruch and Harnish, 1999). NREM sleep can be subdivided in light and deep stages. During light stages, the electroencephalogram (EEG) presents a succession of short periods of synchronized highamplitude slow activity accompanied by faster oscillations

Épilepsies, vol. 22, n° 3, juillet-août-septembre 2010

198

Heart rate variability as measurement of heart-brain interactions

Slow waves 60

15 40 10

Power

Frequency (Hz)

Sleep spindles

20

5

0

Mean ins. heart rate (bpm) VLF

LF

HF 00

01

02

04

03

Frequency domain

Copyright © 2017 John Libbey Eurotext. Téléchargé par un robot venant de 37.44.207.187 le 29/01/2017.

SDNN

Time domain

Mean beat-tobeat (sec)

05

Time (hours)

Figure 2. Evolution of selected HRV indexes in time and frequency domain during sleep. HRV measurements were calculated for shortwindows of 5 minutes with a sliding step of 5 seconds. See explanations in the text.

the cardiac activity related to different neurological pathologies such as epilepsy where it is suspected to be one of the cause of sudden and unexpected deaths. However, despite these useful or potential applications, further studies are still required in order to better explore i) physiological significance of most of the HRV developed indexes, especially the complex ones, ii) new measurements of the simultaneous evolution of brain and heart activities which will permit to improve the knowledge of interactions between heart and brain.

rhythm becomes more irregular, as appreciated in the SDNN measurement, and the decrement of the HF index accompanied by a simultaneous increment of VLF and LF ones suggests a change in the sympatho-vagal balance, consisting in a presumably vagal withdraw with a concurrent increment of the sympathetic influence (purple ellipse in figure 2).

Conclusion

□

The analysis of HRV through the different methodological approaches presented provides valuable information about the cardiac autonomic modulation in diverse physiological and clinical situations. Additional to be quantitative indexes indicating fluctuations of normal or abnormal autonomic activities, HRV measurements are valuable markers of good prognostic of the outcome of particular cardiovascular-related diseases. Similarly, as being derived from non-invasive, routinely acquired ECG recordings, HRV analysis allows to study, over long-term, the altered autonomic regulation of

Conflict of interest: non declared.

References Ako M, Kawara T, Uchida S, et al. Correlation between electroencephalography and heart rate variability during sleep. Psychiatry Clin Neurosci 2003; 57: 59-65.

199

Épilepsies, vol. 22, n° 3, juillet-août-septembre 2010

M. Valderrama, et al.

Baduí E, Lepe L, Van Wielink G, Espinosa R. Paroxysmal atrial tachycardia recorded by Holter monitoring during an episode of fear. A case report. Angiology 1996; 47: 713-6.

Mäkikallio TH, Perkiömäki JS, Huikuri HV. Nonlinear Dynamics of RR intervals. In: Malik M, Camm J, eds. Dynamic Electrocardiography. Malden: Futura, Blackwell, 2004: 22-30.

Baharav A, Kotagal S, Gibbons V, et al. Fluctuations in autonomic nervous activity during sleep displayed by power spectrum analysis of heart rate variability. Neurology 1995; 45: 1183-7.

Malliani A, Pagani M, Lombardi F, Cerutti S. Cardiovascular neural regulation explored in the frequency domain. Circulation 1991; 84: 482-92. Malliani A, Montano N. Autonomic balance. In: Malik M, Camm J, eds. Dynamic Electrocardiography. Malden: Futura, Blackwell, 2004.

Berntson GG, Sarter M, Cacioppo JT. Anxiety and cardiovascular reactivity: the basal forebrain cholinergic link. Behav Brain Res 1998; 94: 225-48.

Marple Jr L. Digital spectral analysis. Upper Saddle River: Prentice-Hall Inc, 1987.

Brack KE, Coote JH, Ng GA. Interaction between direct sympathetic and vagus nerve stimulation on heart rate in the isolated rabbit heart. Exp Physiol 2004; 89: 128-39.

Miyashita T, Ogawa K, Itoh H, et al. Spectral analyses of electroencephalography and heart rate variability during sleep in normal subjects. Auton Neurosci 2003; 103: 114-20.

Copyright © 2017 John Libbey Eurotext. Téléchargé par un robot venant de 37.44.207.187 le 29/01/2017.

Carney RM, Blumenthal JA, Stein PK, et al. Depression, heart rate variability, and acute myocardial infarction. Circulation 2001; 104: 2024-8.

Murata J, Matsukawa K, Shimizu J, Matsumoto M, Wada T, Ninomiya I. Effects of mental stress on cardiac and motor rhythms. J Auton Nerv Syst 1999; 75: 32-7.

Critchley HD, Rotshtein P, Nagai Y, O’Doherty J, Mathias CJ, Dolan RJ. Activity in the human brain predicting differential heart rate responses to emotional facial expressions. Neuroimage 2005; 24: 751-62.

Nashef L, Tomson T. Sudden death in epilepsy. In: Engel Jr JJ, Pedley TA, eds. Epilepsy: A Comprehensive Textbook, Volume 2. Philadelphia: Lippincott Williams & Wilkins, 2008: 1991-8.

Elsenbruch S, Harnish MJ, Orr WC. Heart rate variability during waking and sleep in healthy males and females. Sleep 1999; 22: 1067-71. Farrell TG, Odemuyiwa O, Bashir Y, et al. Prognostic value of baroreflex sensitivity testing after acute myocardial infarction. Br Heart J 1992; 67: 129-37.

Ng GA, Brack KE, Coote JH. Effects of direct sympathetic and vagus nerve stimulation on the physiology of the whole heart--a novel model of isolated Langendorff perfused rabbit heart with intact dual autonomic innervation. Exp Physiol 2001; 86: 319-29.

Davis AM, Natelson BH. Brain-heart interactions. The neurocardiology of arrhythmia and sudden cardiac death. Tex Heart Inst J 1993; 20: 158-69.

Parker P, Celler BG, Potter EK, McCloskey DI. Vagal stimulation and cardiac slowing. J Auton Nerv Syst 1984; 11: 226-31.

Goldberger JJ, Challapalli S, Tung R, Parker MA, Kadish AH. Relationship of heart rate variability to parasympathetic effect. Circulation 2001; 103: 1977-83.

Perini R, Veicsteinas A. Heart rate variability and autonomic activity at rest and during exercise in various physiological conditions. Eur J Appl Physiol 2003; 903: 17-25.

Guyton AC, Hall JE. Textbook of Medical Physiology. Philadelphia: Elsevier Inc, USA, 2006.

Scholz UJ, Bianchi AM, Cerutti S, Kubicki S. Vegetative background of sleep: spectral analysis of the heart rate variability. Physiol Behav 1997; 62: 1037-43.

Hainsworth R. Physiological background of heart rate variability. In: Malik M, Camm J, eds. Dynamic Electrocardiography. Malden: Futura, Blackwell, 2004: 3-12.

Stein PK, Kleiger RE. Insights from the study of heart rate variability. Annu Rev Med 1999; 50: 249-61.

Harnod T, Yang CC, Hsin YL, Wang PJ, Shieh KR, Kuo TB. Heart rate variability in patients with frontal lobe epilepsy. Seizure 2009; 18: 21-5.

Sztajzel J. Heart rate variability: a noninvasive electrocardiographic method to measure the autonomic nervous system. Swiss Med Wkly 2004; 134: 514-22.

Henning RJ, Khalil I. Autonomic nervous stimulation affects left ventricular relaxation more than left ventricular contraction. J Auton Nerv Syst 1989; 28: 15-25. Hondeghem LM, Mouton E, Stassen T, De Geest H. Additive effects of acetylcholine released by vagal nerve stimulation on atrial rate. J Appl Physiol 1975; 38: 108-13.

Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Eur Heart J 1996; 17: 54-81.

Jouven X, Schwartz PJ, Escolano S, et al. Excessive heart rate increase during mild mental stress in preparation for exercise predicts sudden death in the general population. Eur Heart J 2009; 30: 1703-10.

Taylor EW, Jordan D, Coote JH. Central control of the cardiovascular and respiratory systems and their interactions in vertebrates. Physiol Rev 1999; 79: 855-916.

Klabunde RE. Cardiovascular Physiology Concepts. Philadelphia: Lippincott Williams & Wilkins, 2004.

Toscani L, Gangemi PF, Parigi A, et al. Human heart rate variability and sleep stages. Ital J Neurol Sci 1996; 17: 437-9.

Kleiger RE, Miller JP, Bigger Jr JT, Moss AJ. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 1987; 59: 256-62.

Tsunoda M, Endo T, Hashimoto S, Honma S, Honma KI. Effects of light and sleep stages on heart rate variability in humans. Psychiatry Clin Neurosci 2001: 55285-6.

Kleiger RE, Stein PK, Bigger Jr JT. Heart rate variability: measurement and clinical utility. Ann Noninvasive Electrocardiol 2005; 10: 88-101.

Vanoli E, Adamson PB. Ba-Lin, Pinna GD, Lazzara R, Orr WC. Heart rate variability during specific sleep stages. A comparison of healthy subjects with patients after myocardial infarction. Circulation 1995; 91: 1918-22.

Köhler BU, Hennig C, Orglmeister R. The principles of software QRS detection. IEEE Eng Med Biol Mag 2002; 21: 42-57.

Vaughn BV, Quint SR, Messenheimer JA, Robertson KR. Heart period variability in sleep. Electroencephalogr Clin Neurophysiol 1995; 94: 155-62.

Levy MN, Yang T, Wallick DW. Assessment of beat-by-beat control of heart rate by the autonomic nervous system: molecular biology technique are necessary, but not sufficient. J Cardiovasc Electrophysiol 1993; 4: 183-93.

Wolf MM, Varigos GA, Hunt D, Sloman JG. Sinus arrhythmia in acute myocardial infarction. Med J Aust 1978; 2: 52-3.

Lewis ME, Al-Khalidi AH, Bonser RS, et al. Vagus nerve stimulation decreases left ventricular contractility in vivo in the human and pig heart. J Physiol 2001; 534: 547-52.

Yamamoto K, Miyachi M, Saitoh T, Yoshioka A, Onodera S. Effects of endurance training on resting and post-exercise cardiac autonomic control. Med Sci Sports Exerc 2001; 33: 1496-502.

Lewis MJ. Heart rate variability analysis: a tool to assess cardiac autonomic function. Comput Inform Nurs 2005; 23: 335-41.

Yang TT, Simmons AN, Matthews SC, et al. Increased amygdala activation is related to heart rate during emotion processing in adolescent subjects. Neurosci Lett 2007; 428: 109-14.

Lombardi F. Physiological understanding of HRV components. In: Malik M, Camm J, eds. Dynamic Electrocardiography. Malden: Futura, Blackwell, 2004: 40-47.

Épilepsies, vol. 22, n° 3, juillet-août-septembre 2010

Yasuma F, Hayano J. Respiratory sinus arrhythmia: why does the heartbeat synchronize with respiratory rhythm? Chest 2004; 125: 683-90.

200