ATRIAL FIBRILLATION
The assessment of autonomic function in chronic atrial fibrillation: description of a non-invasive technique based on circadian rhythm of atrioventricular nodal functional refractory periods
1 Academic Unit of Cardiology, University of Hull, Hull, UK; 2 Department of Medical Cardiology, University of Glasgow, Glasgow Royal Infirmary, Glasgow G31 2ER, Glasgow, UK
Manuscript submitted 1 January 2006. Accepted after revision 1 July 2006.
* Corresponding author. Tel/fax: +44 1412114409. E-mail address: akhand31{at}aol.com
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionImplications and conclusionAppendixReferences |
|---|
Aims Heart rate variability (HRV) parameters can be used to assess autonomic function and to predict outcome, but this has been done exclusively in patients with sinus rhythm. Atrial fibrillation (AF) is the commonest sustained arrhythmia and is particularly prevalent in heart failure. We have developed a simple index to assess autonomic function in patients with chronic AF.
Methods and results Forty patients with chronic AF (>1 month) and symptoms of heart failure underwent ambulatory 24 h electrocardiography recording as well as evaluation of symptoms, exercise capacity (6 min walk distance), ventricular function (echocardiography and radionuclide ventriculography), and neuroendocrine activation. A number of standard HRV parameters shown to have prognostic significance in sinus rhythm were also determined. A modified in-house HRV statistical programme was used to filter labelled QRS intervals and to compute the 5th percentile RR interval in each hour. This parameter has been shown to approximate the functional refractory period (FRP) of the atrioventricular node (AVN). A cosine curve was fitted to hourly 5th percentile RR intervals for each patient and from this was estimated the diurnal change in hourly 5th percentile RR interval (approximating 
FRP of the AVN) and, by inference, diurnal variation in sympathovagal input to the AVN. Digoxin was the sole agent permitted for control of ventricular rate. FRP of the AVN varied and revealed a significant correlation, on multivariate analysis, with mean RR interval (P<0.001), SDARR (SD of 5-min average RR intervals during 24 h, P<0.001), and NYHA class of heart failure (classes III and IV heart failure vs. classes I and II, P=0.02). SDARR has previously been shown independently to predict mortality in patients with chronic AF and heart failure.
Conclusion This analysis describes a novel non-invasive method for assessing autonomic function in chronic AF. Whether FRP in chronic AF patients can independently predict adverse prognosis or sudden death requires further study.
Key Words: Autonomic function, Atrial fibrillation, Heart failure, Heart rate variability
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionImplications and conclusionAppendixReferences |
|---|
Heart rate variability (HRV) reflects the influence of the autonomic nervous system on the cardiovascular system and has been shown to be a reliable and reproducible technique for assessing autonomic function.1
,2 Time domain and spectral HRV analyses have indicated that there is sympathetic activation3 and a reduction in parasympathetic activity4 in patients with heart failure. Direct recordings of muscle sympathetic nerve activity5 and studies investigating sympathetic neurotransmitter release and clearance have strengthened these findings.6 The effect of heart failure on autonomic dysfunction is expressed, in basic terms, as a profound reduction in RR interval variance, and the degree of reduction in this measure has been demonstrated to be independently associated with mortality.7–9 Measures of HRV have also been shown to correlate with the severity of symptoms in patients with heart failure.10
Atrial fibrillation (AF) is the commonest sustained arrhythmia in patients with heart failure.11 Although autonomic dysfunction is likely to confer the same risks for cardiac death or electrical stability in patients with AF as in sinus rhythm, the relevance or significance of ‘traditional’ time domain or spectral HRV parameters to assess autonomic function in chronic AF is uncertain. However, although RR intervals are intrinsically irregular in AF, this irregularity is not random; it is complex and is dependent on a number of factors: the refractory period and conductivity of the atrioventricular (AV) node; the degree of concealed conduction; and the irregularity, frequency, and direction of atrial wavefronts impacting on the AV node.12–16 The autonomic nervous system modifies ventricular response by influencing all these factors. In particular, vagal and sympathetic efferents to the compact region of the AV node have a profound influence on ventricular response,3,13,14 and there is a clear pattern of sympathovagal balance during a 24 h cycle, with vagal effects dominating at night and the sympathetic nervous system activity dominating during the day.3 It may therefore be possible to assess autonomic function in AF by measuring the change in electrophysiological parameters of the AV node over a 24 h period.
This study investigates autonomic function in patients with chronic AF by studying the circadian rhythm of the AV nodal functional refractory period (FRP) as estimated by the minimum RR interval. This is compared with the severity of heart failure and standard HRV parameters.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionImplications and conclusionAppendixReferences |
|---|
The minimal hourly RR interval in chronic AF was assumed to approximate the FRP of the AV node for that hour. The evidence for this came principally from two sources.
Billete et al.12 undertook electrophysiological studies of isolated, perfused dog hearts. They described the 5th percentile RR interval (lowest 5% RR interval) during induction of AF to be equal to the FRP of the AV node when the FRP was short (as determined by the shortest H1H2 that could be achieved) and approximated (slightly overestimated) the FRP when the FRP was long.
Toivonen et al.17 undertook electrophysiological studies of AV nodal conduction properties in humans and assessed the relation of these to atrial pacing and RR intervals after induction of AF. The minimum RR interval in AF was found to approximate the FRP of the AV node (r=0.92). As in the study by Billete et al.,12 a short FRP, induced by an infusion of the ß-agonist isoprenaline, caused almost equivalence between the measures, whereas a longer FRP, after ß-blockade, weakened the association.
Patient population
Patients with chronic AF (>1 month) and a clinical diagnosis of heart failure were studied. The criteria for the definition of heart failure included both: symptoms consistent with heart failure for >2 months and echocardiographic evidence of cardiac dysfunction [systolic dysfunction or evidence for diastolic dysfunction (left ventricular hypertrophy in association with dilated left atrium) in the absence of significant valvular heart disease]. Patients with paroxysmal AF (on 24 h tape) were excluded, as were patients with alternating periods of flutter and fibrillation, detected by analysis of full disclosure rhythm strips and characteristic patterns in Lorenz plots.
To avoid the confounding effects of medication, patients on current treatment with a ß-blocker or heart-rate lowering calcium channel antagonist were excluded. Also excluded were patients with evidence of native conduction disease, pacemaker implantation, or recent major cardiovascular event. All patients were on digoxin alone for ventricular rate control.
The main purpose of the analysis was to investigate the circadian rhythm of FRP as an index of autonomic function in patients with chronic AF and varying severity of heart failure. Therefore, patients with concomitant conditions known to have an independent effect on autonomic activity [diabetes mellitus, chronic renal failure, a history of alcohol abuse, clinical evidence of autonomic neuropathy, or a recent (<6 months) myocardial infarction] were excluded. Also excluded were patients with severe chronic obstructive pulmonary heart disease and periodic or Cheyne–Stokes (patients with FEV 1 <50% predicted were excluded) breathing, as marked or erratic respiratory efforts could confound assessment of autonomic tone.
Baseline data collection
A number of baseline variables were collected: left and right ventricular ejection fraction was determined by electrocardiography (ECG)-gated radionuclide ventriculography, left atrial size and left ventricular dimensions were determined by echocardiography (parasternal long axis view). Plasma concentrations of brain natriuretic peptide (C-terminal—Peninsula) and digoxin were measured. Patients were grouped into New York Heart Association (NYHA) classes according to standard definition. A more detailed examination of symptoms was undertaken by a symptom questionnaire designed and previously tested in the heart failure population. From this was derived a score that assessed the symptom burden of the arrhythmia (palpitation) as well as functional status (symptoms related to heart failure). The maximum score indicating severe symptoms was 33, whereas the minimum score (0) indicated no symptoms. All patients also underwent a 6 min walk test.
This study was approved by the local Ethics Committee and all subjects gave informed consent.
Twenty-four-hour ambulatory ECG
Twenty-four hour ambulatory ECG was undertaken in all subjects during normal, unrestricted out-of-hospital activity using a miniature tape recorder with a crystal time-generated reference track that allowed correction for recording and replay speed errors to within 0.5%. For a tape to be eligible, it had to have AF as the basic rhythm. The ECG recordings were processed with standard precision on MEDILOG Excel 2 system (Oxford instruments, Abingdon, UK). All QRS complexes were detected and labelled automatically. The results of this analysis were completely reviewed, and any errors in R wave detection or QRS labelling (dominant or non-dominant beat) were edited manually by experienced technicians. Patients with ventricular ectopics >20% of entire RR intervals were excluded. The edited RR intervals were downloaded to discs. The discs were run through a statistical programme, modified for use in AF, which was used to eliminate outlying RR intervals that may not have been correctly detected by the initial scanning of the Holter recordings. Normally, RR interval variation in successive beats is limited to 20% for HRV analysis in sinus rhythm. This part of the programme was deleted and minimum and maximum RR intervals were arbitrarily set at 80 and 6400 ms, respectively.
Subsequently, RR intervals (aligned chronologically) were sliced into hourly segments. For each hourly segment, an RR interval histogram was plotted (Figure 1). The 5th percentile RR interval was determined for each hour. This and other common HRV parameters are summarized in Table 1. The 5th percentile RR interval, by definition, is an RR interval on the hourly RR interval histogram to the right of which lie 95% of all remaining intervals (Figure 1). The 5th percentile was used for analysis rather than the absolute minimum RR interval for two main reasons: to provide a greater check against rogue or artefact values and because this was the parameter used by Billete et al.12 in his work on dog hearts to validate the relationship between minimum RR interval and the AV nodal FRP.
|
|
Circadian rhythms were plotted using cosine curves as described in Figure 2 for hourly 5th percentile RR interval. The parameter used for assessing circadian rhythm was the normalized amplitude. This was the amplitude (difference between the acrophase or maximum FRP and the mean value) divided by the mean (or mesor) for a particular patient. This parameter gives an indication of the relative change in sympathovagal balance. Normalizing corrects for individual differences in electrophysiological parameters.
|
Statistical analysis
Spearman's correlation coefficient was used to assess the degree of association between normalized amplitude of 5th percentile hourly RR interval and a range of other variables. Variables that demonstrated significant correlation were then entered into a multivariate regression model to determine those that were independently associated with the normalized 5th percentile RR interval.
To ensure robustness of findings, the analysis of differences, between groups of patients, in the parameters from fitted curves was undertaken in two ways. The first analysis involved a comparison of parameters derived from curves fitted to each patient. The values for a group were then compared with the other group (single-patient method). The second analysis involved fitting a cosine graph to the data as a whole for a group (grouped method) and then comparing the parameters derived from the fit between groups (Figure 2).
MATLAB statistical software (version 5.0, Natick, MA, USA) was used to retrieve the hourly histogram and values, whereas SPSS version 9.0 (SPSS Inc., Chicago, IL, USA) was used to undertake cosinor and other statistical analyses.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionImplications and conclusionAppendixReferences |
|---|
Forty-one patients were initially included in this analysis, but one patient was subsequently excluded because of frequent (>100 per hour), complex multi-focal ventricular extrasystoles on the 24 h ambulatory ECG.
Table 2 describes the baseline characteristics of all patients and compares the population with NHYA classes I and II heart failure symptoms and those in classes III and IV. There was no significant difference between patients with NYHA classes I and II and NYHA classes III and IV heart failure for a host of variables, such as left atrial diameter, ACE-inhibitor use, LVEF, and mean 24 h heart rate. As expected, symptom burden was greater in the severe heart failure group and the 6 min walk distance was shorter.
|
Relation of normalized amplitude of 5th percentile RR interval to clinical, haemodynamic, and humoral factors
Table 3 presents correlation coefficients of normalized amplitude of 5th percentile RR intervals with a range of variables, including those that are known to be independently associated with reduced life expectancy in patients with heart failure in sinus rhythm.18
There was a positive correlation with normalized amplitude of mean hourly RR interval, SDARR, and NYHA Class. On multivariate analysis, after exclusion of normalized amplitude of mean hourly RR interval (which is predicted by FRP rather than predicting FRP), SDARR (P<0.001) and NYHA classes, I and II vs. III and IV (P=0.02) remained independently associated with normalized amplitude of 5th percentile RR intervals.
|
Circadian rhythms of 5th percentile RR interval in relation to severity of heart failure
Figure 3 describes the profiles of two patients: one in NYHA class I and the second in NYHA class IV heart failure. The RR interval histogram at 2—3 a.m. is broader in the patient with NYHA class I, with evidence of greater variability in RR intervals (Figure 3A). During the day (11 a.m to 12 mid-day), the difference is less striking. Circadian rhythm of hourly 5th percentile RR interval and hourly mean RR interval is blunted in NYHA class IV heart failure patient, compared with patient in NYHA class I heart failure (Figure 3B).
|
Table 4 describes the group values for circadian rhythms of 5th percentile RR interval and mean RR interval in patients with NYHA classes I and II compared with patients in NYHA classes III and IV heart failure.
|
All patients, except two in the severe heart failure group, had a significant circadian rhythm of hourly 5th percentile RR interval. No patient had evidence of two peaks on visual inspection of observed points.
There was a significantly greater circadian variation in normalized 5th percentile RR intervals in the mild heart failure group compared with the severe heart failure group (Figure 4; Table 4), but the mesor and acrophase (mean and time of peak respectively) did not differ between groups. Similar results were discovered for hourly mean RR interval curves. The results of the above analyses were consistent using either single patient curves or cosine graphs fitted to the data of an entire group.
|
The patients were also analysed according to supra- and infra-median SDARR values, as this was demonstrated to have the greatest independent contribution to the normalized amplitude of 5th percentile RR intervals. (Figure 5; Table 5). There was an even greater difference in the degree of circadian rhythm for hourly 5th percentile RR intervals between patients above and below the median for SDARR compared with the dichotomy by NYHA class.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionImplications and conclusionAppendixReferences |
|---|
This analysis describes a simple technique for estimating the FRP of the AV node and quantifying the circadian variation of this parameter and, thereby, the circadian sympathovagal interaction at the level of the AV node. Circadian variation of the mean RR interval mirrored the circadian variation in 5th percentile RR intervals (FRP), confirming an important role of AV nodal conduction properties in determining the mean ventricular rate in AF.17
,19 This analysis further describes a blunting of circadian sympathovagal interaction, signifying autonomic dysfunction in patients with severe heart failure (NYHA classes III and IV) (Figure 4; Table 4) and/or low SDARR (Figure 5; Table 5).
Autonomic modulation of AV nodal conduction in AF
Although the AV node, rather than the sinus node, regulates ventricular rate in AF, mean hourly ventricular rate has been shown to exhibit a circadian rhythm similar to that observed in sinus rhythm.20 Cinca et al.,21 by direct electrophysiological measurement, discovered a circadian rhythm of the AV nodal FRP, with an acrophase during the night. This analysis is consistent with the pattern of the above results and indicates the influence of the AV nodal FRP in determining mean ventricular rate.
A blunting of the circadian rhythm of the FRP (reflected by 5th percentile RR interval), corrected for individual electrophysiological characteristics, may imply either sympathetic predominance or parasympathetic withdrawal. Both these phenomena have been demonstrated in heart failure and sinus rhythm.4,7,22 However, spectral analysis of short-term recordings in sinus rhythm has indicated a more complex relationship between arms of the autonomic nervous system in heart failure as well as in the setting of myocardial infarction, although sympathovagal balance as assessed by LF/HF ratios is consistently altered.23,24
A second or additional mechanism of blunting of circadian rhythm of FRP in patients with severe heart failure could be a diminished responsiveness of the AV node to neural inputs, which might be contributed to by high levels of circulating catecholamines and changes in ß-receptor sensitivity.25
Follow-up in a large cohort will be required to assess whether blunting of circadian rhythm of FRP of the AV node, as assessed non-invasively by this method, is a valuable tool in stratifying patients for the risk of cardiac events, including sudden death in patients with chronic AF. A second potential value of this assessment may be in determining the likelihood of maintenance of sinus rhythm post-cardioversion. Autonomic dysfunction and specifically sympathetic activation, as discerned post-cardioversion, are known to be associated with early reversion to AF.26
Previous literature
Only one other study has assessed circadian changes in AV nodal conduction properties in patients with chronic AF. This study used more complex methods compared with our study, but came to similar conclusions.27 The normalized amplitude of the FRP, reflecting the change in sympathovagal balance is easily derived by a programme which first filters and time sequences the hourly RR intervals followed by cosinor analysis of the data points. It could be easily incorporated into a Holter arrhythmia analysis profile if proven to reflect autonomic function, predict cardiac events or the success of cardioversion, and long-term maintenance of sinus rhythm.
Frey et al.16 computed HRV parameters and determined invasive haemodynamic indices in patients with chronic AF and severe heart failure. SD of RR intervals and SDARR were compared with 13 pre-selected clinical and haemodynamic variables for prediction of outcome. Patients were followed-up for 12 months. Only SDARR was independently associated with survival on multivariate analysis, and no HRV parameter correlated with any haemodynamic measure. SDARR, out of other HRV markers and haemodynamic and neuroendocrine variables, proved to correlate most strongly with FRP in our study. This suggests that each may measure some component of autonomic function in AF.
Recently, Yamada et al.28 showed that patients with AF and reduced ventricular response irregularity, as determined by novel non-linear dynamic measures, had a worse prognosis. The authors speculated that reduction in vagal activity could provide a link between increased mortality and reduced ventricular response irregularity. Reduced vagal activity would result in an increase in atrial refractoriness (thereby reducing the number of re-entrant wavelets) and reduce the filtering properties of the AV node. However, conclusive proof of this is lacking. SDARR describes variability of response rather than irregularity or unpredictably of ventricular response. It is possible that non-linear dynamics may provide a potent marker of autonomic dysfunction, which could be complementary to circadian rhythms.
Digoxin is known to improve autonomic function in patients with heart failure.29 This will influence AV nodal conductivity, but there was no difference in serum digoxin concentrations between those with mild as opposed to severe heart failure (Table 2). Digoxin is, therefore, unlikely to have altered the relative differences in FRP between the two groups.
Limitations
Although autonomic dysfunction is a plausible explanation for the reduced circadian variation in FRP there may be confounding factors. The effect of change from standing to a supine position and abnormal respiratory patterns may be two such factors. The former is unlikely to be an adequate explanation as all patients went to bed at roughly the same time (according to diaries). The second factor may be a valid explanation for a reduction in vagal activity due to apnoeic spells, which are well documented in patients in severe heart failure. This does not cast doubt on the validity of the results, but implies that there is no autonomic dysfunction per se, only a reduction in vagal efferent discharge because of reduction in central respiratory drive.
Respiratory variations that could induce more short-term high frequency changes in HRV measures were not examined in this study. The interpretation of short periods of recordings in AF to estimate spectral analysis is poorly validated, and important methodological aspects of interpolation of RR intervals for ventricular ectopics are crucial, but difficult if not impossible to solve in the context of AF. Therefore, the intention in this study was to examine long-term circadian trends.
Although age, gender, and anatomy of the AV node (slow and fast pathways) will have an influence on AV-conducting properties, this study investigated a relative change from baseline of 5th percentile RR beat interval over 24 h and is more likely, therefore, to overcome individual differences in baseline AV-conducting properties.
|
Implications and conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionImplications and conclusionAppendixReferences |
|---|
This study describes a non-invasive marker for quantifying the dynamic interplay between sympathetic and parasympathetic inputs to the AV nodal compact region. The index described will require further validation by way of electrophysiological studies, but could prove to be a valuable tool for assessing autonomic function in chronic AF. This may ultimately help in the risk stratification of patients with AF for sudden cardiac death or progression of heart failure. It may also help in predicting the utility of DC cardioversion in patients with chronic AF.
|
Appendix |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionImplications and conclusionAppendixReferences |
|---|
Symptom assessment in atrial fibrillation and heart failure
| |||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionImplications and conclusionAppendixReferences |
|---|
[1] Batin PD and Nolan J. Assessment of autonomic function: reflex testing or variability analysis? J Amb Mon 1996; 9: 255–73.
[2] Nolan J, Flapan AD, Goodfield NE, Prescott RJ, Bloomfield P, Neilson JM, et al. Measurement of parasympathetic activity from 24-hour ambulatory electrocardiograms and its reproducibility and sensitivity in normal subjects, patients with symptomatic myocardial ischaemia, and patients with diabetes mellitus. Am J Cardiol 1996; 77: 154–8.[CrossRef][Web of Science][Medline]
[3] Kienzle MG, Ferguson DW, Birkett CL, Myers GA, Berg WJ, Mariano DJ. Clinical haemodynamic and sympathetic neural correlates of heart rate variability in congestive heart failure. Am J Cardiol 1992; 69: 482–5.[CrossRef][Web of Science][Medline]
[4] Casolo G, Balli E, Taddei T, Amuhasi J, Gori C. Decreased spontaneous heart rate variability in congestive heart rate failure. Am J Cardiol 1989; 64: 1162–7.[CrossRef][Web of Science][Medline]
[5] Leimbach WN, Wallin BG, Victor RG, Aylward PE, Sundlöf G, Mark AL. Direct evidence from intraneural recordings for increased central sympathetic outflow in patients with heart failure. Circulation 1986; 73: 913–9.
[6] Hasking GJ, Esler MD, Jennings GL, Burton D, Johns JA, Korner PI. Norepinephrine spillover to plasma in patients with congestive heart failure: evidence of increased overall and cardiorenal sympathetic nervous activity. Circulation 1986; 73: 615–21.
[7] Binkley PF, Nunziata E, Haas GJ, Nelson SD, Cody RJ. Parasympathetic withdrawal is an integral component of autonomic imbalance in congestive heart failure: demonstration in human subjects and verification in a paced canine model of ventricular failure. J Am Coll Cardiol 1991; 18: 464–72.[Abstract]
[8] Szabo BM, Van Velduisen DJ, Van der Weer N, Brouwer J, De Graeff PA, Crijns HJGM. Prognostic value of heart rate variability in chronic congestive heart failure secondary to idiopathic or ischaemic dilated cariomyopathy. Am J Cardiol 1997; 79: 978–80.[CrossRef][Web of Science][Medline]
[9] Brouwer J, Van Veldhuisen DJ, Man in't Veld AJ, Haaksma J, Dijk WA, Visser KR, et al. Prognostic value of heart rate variability during long-term follow-up in patients with mild to moderate heart failure. J Am Coll Cardiol 1996; 28: 1183–9.[Abstract]
[10] Guzzeti S, Signorini MG, Cogliati C, Mezzetti S, Porta A, Cerutti S, et al. Non linear dynamics and chaotic indices in heart rate variability of normal subjects and heart transplanted patients. Cardiovasc Res 1996; 31: 441–6.[CrossRef][Web of Science][Medline]
[11] Khand AU, Rankin AC, Kaye G, Cleland JGF. Systematic review of the management of atrial fibrillation in patients with heart failure. Eur Heart J 2000; 21: 614–32.
[12] Billete J, Nadeau RA, Roberge F. Relation between the minimum RR interval during atrial fibrillation and the functional refractory period of the AV junction. Cardiovasc Res 1974; 8: 347–51.[Web of Science][Medline]
[13] Mazgalev TN, Garrigue S, Mowrey KA, Yamanouchi Y, Tchou PJ. Autonomic modification of the atrioventricular node during atrial fibrillation. Role in the slowing of ventricular rate. Circulation 1999; 99: 2806–14.
[14] Van Den Berg MP, Haaksma J, Brouwer J, Tieleman RG, Mulder G, Crijns HJGM. Heart rate variability in patients with atrial fibrillation is related to vagal tone. Circulation 1997; 96: 1209–16.
[15] Stein K, Borer J, Hochreiter C, Devereux RB, Kligfield P. Variability of the ventricular response in atrial fibrillation and prognosis in chronic nonischemic mitral regurgitation. Am J Cardiol 1994; 74: 906–11.[CrossRef][Web of Science][Medline]
[16] Frey B, Heinz G, Binder T, Wutte M, Schneider B, Schmidinger H, et al. Diurnal variation of ventricular response to atrial fibrillation in patients with advanced heart failure. Am Heart J 1995; 129: 58–65.[CrossRef][Web of Science][Medline]
[17] Toivonen L, Kadish A, Kow W, Morady F. Determinants of the ventricular rate during atrial fibrillation. J Am Coll Cardiol 1990; 16: 1194–2000.[Abstract]
[18] Cowburn PJ, Cleland JGF, Coats AJS, Komajda M. Risk stratification in chronic heart failure. Eur Heart J 1998; 19: 696–710.
[19] Fujiki A, Tani M, Mizumaki K, Yoshida S, Sasayama S. Quantification of human concealed atrioventricular nodal conduction: relation to ventricular response during atrial fibrillation. Am Heart J 1990; 120: 598–603.[CrossRef][Web of Science][Medline]
[20] Raeder EA. Circadian fluctuations in ventricular response to atrial fibrillation. Am J Cardiol 1990; 66: 1013–6.[CrossRef][Web of Science][Medline]
[21] Cinca J, Moya A, Figueras J, Roma F, Rius J. Circadian variations in the electrical properties of the human heart assessed by sequential bedside electrophysiologic testing. Am Heart J 1986; 112: 315–21.[CrossRef][Web of Science][Medline]
[22] Saul JP, Arai Y, Berger RD, Lilly LS, Colucci WS, Cohen RJ. Assessment of autonomic regulation in chronic congestive heart failure by heart rate spectral analysis. Am J Cardiol 1988; 61: 1292–9.[CrossRef][Web of Science][Medline]
[23] Lombardi F, Malliani A, Pagani M, Cerruti S. Heart rate variability and its sympatho-vagal modulation. Cardiovasc Res 1996; 32: 208–16.
[24] Lombardi F, Sandrone G, Mortara A, La Rovere MT, Colombo E, Guzzetti S, et al. Circadian variation of spectral indices of heart rate variability after myocardial infarction. Am Heart J 1992; 123: 1521–9.[CrossRef][Web of Science][Medline]
[25] Bristow MR. Beta-adrenergic receptor blockade in chronic heart failure. Circulation 2000; 101: 558–69.
[26] Lombardi F, Columbo A, Basilico B, Ravaglia R, Garbin M, Vergani D, et al. Heart rate variability and early recurrence of atrial fibrillation after electrical cardioversion. J Am Coll Cardiol 2001; 37: 157–62.
[27] Hayano J, Sakata S, Okada A, Mukai S, Fujinami T. Circadian rhythms of atrioventricular conduction properties in chronic atrial fibrillation with and without heart failure. J Am Coll Cardiol 1998; 31: 158–66.
[28] Yamada A, Hayano J, Sakata S, Okada A, Mukai S, Ohte N, et al. Reduced ventricular response irregularity is associated with increased mortality in patients with chronic atrial fibrillation. Circulation 2000; 102: 300–6.
[29] Krum H, Bigger T, Goldsmith RL, Packer M. Effect of long-term digoxin therapy on autonomic function in patients with chronic heart failure. J m Coll Cardiol 1995; 25: 289–94.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||