Effect of diurnal variability of heart rate on development of arrhythmia in patients with chronic obstructive pulmonary disease

International Journal of Cardiology 88 (2003) 199–206 www.elsevier.com / locate / ijcard

Effect of diurnal variability of heart rate on development of arrhythmia in patients with chronic obstructive pulmonary disease a, *, Pınar Yildiz b , Dursun Atilgan a , Volkan Tuzcu a , Mehmet Eren a , Osman Erk c , ¨ Tufan Tukek S¸eref Demirel c , Vakur Akkaya c , Murat Dilmener c , Ferruh Korkut a a

Department of Cardiology, Istanbul Faculty of Medicine, Istanbul, Turkey b Yedikule Chest Diseases and Chest Surgery Hospital, Yedikule, Turkey c Department Internal Medicine, Istanbul Faculty of Medicine, Istanbul, Turkey Received 25 July 2001; received in revised form 25 May 2002; accepted 16 July 2002

Abstract We examined the possible effect of diurnal variability of heart rate on the development of arrhythmias in patients with chronic obstructive pulmonary disease (COPD). Forty-one COPD patients (M / F: 39 / 2, mean age: 5968.5 years) and 32 (M / F: 27 / 5, mean age: 57611 years) healthy controls were included. Twenty-four hour ECG recordings were analyzed for atrial fibrillation (AF) or ventricular premature beats (VPB), and circadian changes in heart rate variability (HRV) were assessed by dividing the 24-h period into day-time (08:00–24:00 h) and night-time (24:00–08:00 h) periods. Night-time total (TP), low frequency (LF) and high frequency (HF) powers were similarly lower from day-time parameters in AF(2) COPD patients (HF 3.9161 vs. 4.4361.04 ms 2 , P50.001) and controls (HF 3.9560.72 vs. 4.8260.66 ms 2 , P,0.001). The LF / HF ratios were also significantly reduced in the same patient groups (AF(2) COPD 1.3560.21 vs. 1.2760.19, P50.04, controls 1.4360.14 vs. 1.2460.09, P,0.001). Night-time TP and LF were increased, HF unchanged and LF / HF significantly increased (1.1160.25 vs. 1.1960.27, P,0.05) in AF(1) COPD patients. Frequency of VPB was correlated with corrected QT dispersion (QTc d ) (r50.52, P50.001) and the day-time / night-time HF ratio (r50.43, P50.02). Patients with QTc d $60 ms did not have the expected increase in night-time HF and had a statistically insignificant increase in LF / HF ratio. In COPD patients with QTc d ,60 ms, circadian changes in HRV parameters were parallel with the controls. We concluded that COPD patients with arrhythmia had circadian HRV disturbances such as unchanged night-time parasympathetic tone and disturbed sympatho-vagal balance in favor of the sympathetic system all day long, which may explain the increased frequency of arrhythmia.  2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Heart rate variability; COPD; Circadian rhythm; Arrhythmia

1. Introduction Supraventricular and ventricular rhythm disorders are common in chronic obstructive pulmonary disease (COPD) patients. There are reports of increased risk of sudden death in this group of patients [1,2]. ¨ ¨ *Corresponding author. Seyit Omer Mahallesi, Seyit Omer Cami Sokak, Armutcu Apt. No. 9, D:10 34280, S¸ehremini, Istanbul, Turkey. Fax: 190-212-531-4054. ¨ E-mail address: [email protected] (T. Tukek).

Though hypoxia, hypercapnia, acid–base disorders, pulmonary hypertension, and medications are implicated, none of the factors has been promoted as the most important [2,3]. It is known that disorders of cardiac autonomic function are important in the development of arrhythmias, and cardiac autonomic functions can be quantified by measuring beat-to-beat variability of the heart rate recorded during 24 h ECG monitoring. Though there are reports on changes in time-domain

0167-5273 / 02 / $ – see front matter  2002 Elsevier Science Ireland Ltd. All rights reserved. doi:10.1016/S0167-5273(02)00402-3

200

¨ et al. / International Journal of Cardiology 88 (2003) 199–206 T. Tukek

and frequency-domain heart rate variability (HRV) and prognosis in COPD patients, the relation of circadian changes in frequency-domain HRV parameters with arrhythmias was not adequately studied [4,5]. Analysis of hourly frequency domain HRV obtained from 24 h ECG monitoring demonstrated the presence of a circadian rhythm in parasympathetic and sympathetic tone and emphasis was given to its significance in the development of cardiac events [6,7]. The aim of our study is to examine the relation between the circadian rhythm of frequency-domain HRV parameters and dysrhythmias in COPD patients.

2. Materials and methods

2.1. Patients In this study, 41 patients with COPD (M / F: 39 / 2, with mean age: 5968.5 years) were included. Patients with hypertension, valvular heart disease, congenital heart disease, systolic left ventricular dysfunction (EF,50% and / or FS,25%), preexcitation syndrome, sick sinus syndrome, bundle branch blocks and chronic atrial fibrillation were excluded. Patients who had evidence of ischemic heart disease were also excluded. Patients were evaluated for ischemic heart disease by history, physical examination, ECG, echocardiography and 24 h ECG ST segment analysis. Cardio-active drugs like b 2 mimetics and xanthene derivatives were stopped 48 h before study. None of the patients were receiving anti-arrhythmic drugs. The control group comprised of a sex- and age-matched sample of 32 (M / F: 27 / 5, mean age: 57611years) healthy subjects. Echocardiography, 12-lead ECG, 24 h rhythm Holter, pulmonary function tests, arterial blood gas analysis and serum electrolytes were analyzed (Table 1).

2.2. Electrocardiographic data collection 2.2.1. The measurements of QT intervals Electrocardiograms were recorded by means of a 12-channel ECG recorder (Hewlett-Packard M 1709A) at a paper speed of 50 mm / s (gain 10 mm / mV). The QT interval was measured according to standard

methods. A minimum of nine leads were studied in each patient. Minimum (QT min ), maximum (QT max ) duration of QT intervals and their difference (QT dispersion; QT d ) were measured. Each QT interval was corrected for patient heart rate according to Bazett’s formula QTc 5 QT / œ(RR). One observer blindly performed all measures of QT intervals for each lead. Intra-observer coefficient of variation for QT dispersion measurements was 11%.

2.2.2. 24 h ECG recordings and heart rate variability analysis Twenty-four hour ambulatory ECG recordings were taken with a Marquette SEER solid-state recorder. The recordings were downloaded to a Marquette Laser SXP Holter system (Marquette Electronics, Milwaukee, WI, USA). The recordings were reviewed and the beat classifications were manually checked and corrected. Cardiac rhythms were screened for the presence of paroxysmal atrial fibrillation and ventricular premature beats. The diagnosis of paroxysmal atrial fibrillation was made if the patient had a history of documented-atrial fibrillation or atrial fibrillation was present in the Holter recording. Time-domain and long term frequency-domain HRV were calculated using the software present in the system (Marquette Electronics Series 8000 Holter Analysis System Version 5.8, 1-Sept-92). The software used fastFourier transform algorithm for the calculations. Hamming window was used for spectral smoothing. The standard output obtained included the following parameters. In time-domain analysis; SDNN (the standard deviation of all the RR intervals), SDANN (the standard deviation of the 5-min RR interval means), S.D. (the mean of the 5-min RR interval standard deviations), rMSSD (the square root of the mean of the squared differences of two consecutive RR intervals), pNN50 (the percentage of the beats with consecutive RR interval difference of more than 50 ms) were calculated. In frequency-domain analysis; total power (TP) (the area under the spectral curve from 0.01 to 1.0 Hz), low-frequency power (LF) (the area under the spectral curve from 0.04 to 0.15 Hz), high-frequency power (HF) (the area under

¨ et al. / International Journal of Cardiology 88 (2003) 199–206 T. Tukek

the spectral curve from 0.15 to 0.40 Hz), and the LF / HF ratio were calculated. For statistical analysis, 2 power spectral values (ln (ms )) were used. In time-domain analysis SDNN and S.D. were accepted as equivalent to TP, SDANN to ultra-low frequency not measured by our system, and rMSSD and pNN50 to HF. In frequency-domain analysis, LF was accepted as equivalent to the sympathetic plus parasympathetic components of autonomic function, HF was accepted as representing the parasympathetic component of autonomic function, and LF / HF as depicting the sympathovagal balance of the autonomic function [8]. On frequency-domain analysis, mean values for each 1-h period and for the total 24-h period were obtained as output. Circadian variation analysis for the 24-h period was done by dividing the recording into day-time (08:00 h to 12 midnight) and nighttime (12 midnight to 08:00 h) periods. The means of the hourly rates were used in the calculation of the representative values in these time intervals and were entered into the statistical analysis.

2.2.3. Pulmonary function tests Pulmonary function tests were performed with a Sensor Medics 2400 (Serial [ 1677, Holland). Vital capacity, FEV1 , FVC, PEF, FEF25-75 were measured and compared to the predicted value. Arterial blood gases were also obtained simultaneously. All the patients were examined with standard medical and laboratory practice according to the recommendations set forth by the Declaration of Helsinki on Biomedical Research involving Human Subjects [9]. Informed consent was taken from all patients. 2.3. Statistical analysis Statistical analysis was carried out with Statistical Package for Social Sciences for Windows version 10.0 (SPSS, Chicago, USA). All non-numeric variables are reported as frequency and percentage, and all numeric variables are reported as mean6S.D. The comparison of the means of two numerical parameters was done with Student’s t-test. When three groups, controls, COPD without atrial fibrillation and COPD with atrial fibrillation groups, and QTc d lower and longer than 60 ms with COPD patients and control groups, were compared, one-way ANOVA

201

with post hoc Tukey’s honestly significant difference was used. The HRV frequency-domain and timedomain parameters were not normally distributed; therefore the statistical comparisons of these values were done on power-spectral data. Correlation between two numerical variables with non-normal distribution was done with Spearman’s bivariate nonparametric correlation test. P-value less than 0.05 was accepted as significant.

3. Results Chronic obstructive pulmonary disease patients had increased interventricular wall thickness, right atrial diameter, right ventricular internal dimension, and right ventricular free wall thickness. They also had lower ejection fraction (EF) and fractional shortening (FS) compared to controls (Table 1). However, none of the COPD patients had systolic dysfunction (EF, 50% and / or FS,25%). Left ventricular posterior wall thickness, internal diameter (diastolic), left ventricular mass index, left atrial diameter, and heart rate were similar between COPD patients and controls. Heart rate variability analysis revealed that COPD patients had decreased SDANN, SDNN, S.D. in timedomain, and decreased LF in frequency-domain parameters. However, no significant difference in rMSSD, pNN50, total power and HF was found between study groups. When the circadian rhythm of frequency-domain power spectral HRV parameters in COPD patients were studied, night-time TP (6.0860.83 vs. 6.5360.91 ms 2 , P,0.0001), LF (4.8661.1 vs. 5.2461.2 ms 2 , P,0.0001), and HF (3.9460.95 vs. 4.2961.1 ms 2 , P50.003) were significantly increased but LF / HF ratio (1.2660.25 vs. 1.2560.22, P50.6) was unchanged from day-time parameters. In the controls, HF (3.9560.72 vs. 4.8260.66 ms 2 , P,0.0001) increased more than TP (6.4360.7 vs. 6.9960.6 ms 2 , P,0.001) and LF (5.5960.76 vs. 6.9660.6 ms 2 , P50.02) which significantly decreased the LF / HF ratio (1.4360.14 vs. 1.2460.1, P,0.0001).

3.1. Supraventricular arrhythmia Atrial fibrillation was present in 13 (31%) of the

202

¨ et al. / International Journal of Cardiology 88 (2003) 199–206 T. Tukek

Table 1 Results of respiratory function tests, arterial blood gases and echocardiographic examinations COPD n541

Control n532

Statistics

Pulmonary function tests

VC (%) FVC (%) FEV1 (%) FEV1 / FVC

60615 57616 40616 54612

100612 102614 104615 8967

,0.001 ,0.001 ,0.001 ,0.001

Arterial blood gases

pH PO 2 (mmHg) PCO 2 (mmHg) HCO 3 (mmol / l)

7.3960.04 67614 4669 2663

– – – –

– – – –

Echocardiographic measurements

Heart rate (bpm) IVSd thickness (cm) LVIDd (cm) LVPW (cm) LVMI (g / m 2 ) EF (%) FS (%) Left atrial diameter (cm) Right atrial diameter (cm) RV wall thickness (cm) RV diameter (cm)

82612 1.1360.2 4.860.6 1.0560.15 124630 6369 3567 3.9360.41 4.160.5 1.260.17 3.560.6

7668 0.9860.09 4.660.4 1.0660.19 118625 7164 4163 3.9660.39 3.7060.45 0.8060.25 2.2860.54

NS 0.003 NS NS NS ,0.001 ,0.001 NS ,0.01 ,0.001 ,0.001

EF, ejection fraction; FEV1 , force expiratory volume 1 s; FS, fractional shortening; FVC, force vital capacity; IVSd, interventricular septum diastolic; LVID, left ventricular internal diameter; LVMI, left ventricular mass index; LVPW, left ventricular posterior wall; RV, right ventricular; VC, vital capacity.

COPD patients. COPD patients with atrial fibrillation were older (6465 vs. 5769 years, P50.003), SDANN and S.D. were significantly lower, LF and LF / HF ratio were decreased without reaching statistical significance compared to patients without atrial fibrillation. When the circadian changes in frequencydomain HRV were examined, patients without atrial fibrillation had significantly increased night-time TP,

LF, HF and significantly decreased LF / HF ratio. In patients with atrial fibrillation, night-time TP and LF showed slighter increase, night-time HF showed no change and night-time LF / HF ratio was significantly increased (Table 2). When day-time night-time ratios of LF and HF were compared in patients with or without atrial fibrillation, LF ratios were not different (0.9360.02 vs. 0.9360.1, P50.9) but HF ratios were

Table 2 Day-time night-time frequency domain HRV parameters in COPD patients with or without AF and controls Atrial fibrillation (1) COPD n513 TP (ln (ms 2 )) Day Night LF (ln (ms 2 )) Day Night HF (ln (ms 2 )) Day Night LF / HF Day Night a

Controls

5.8460.6 6.2360.82

6.2460.92 6.7460.94

6.4360.7 6.9960.58

4.3560.97 4.7661.3

5.1761.08 5.5361.1

5.5960.76 5.9660.6

460.88 4.0661.16

3.9161 4.4361.04

3.9560.72 4.8260.65

1.1160.25 1.1960.27

1.3560.21 1.2760.19

1.4360.14 1.2460.09

Atrial fibrillation (1) COPD day–night. Atrial fibrillation (2) COPD day–night. c Control day–night. b

Atrial fibrillation (2) COPD n528

P value

P50.01 a P,0.001 b P,0.001 c P50.02 a P,0.001 b P50.01 c P50.62 a P50.001 b P,0.001 c P,0.05 a P50.04 b P,0.001 c

¨ et al. / International Journal of Cardiology 88 (2003) 199–206 T. Tukek

increased in patients with atrial fibrillation (0.8860.12 vs. 1.060.15, P50.02). The same relation was also present when the comparison was carried out between patients with atrial fibrillation and controls (0.9460.1 vs. 0.9360.1, P50.8 for LF, 0.8260.13 vs. 1.060.15, P50.001 for HF). The findings were confirmed when one-way ANOVA analysis of controls, COPD without atrial fibrillation, and COPD with atrial fibrillation was done.

3.2. Ventricular arrhythmia When the recordings were examined for the frequency of ventricular premature beat, COPD patients had more ventricular premature beats compared to controls (89962925 vs. 366112, P,0.0001). On bivariate correlation, log transformed ventricular premature beat frequency was significantly correlated to QT d (r50.52, P50.001) and day-time night-time ratio of HF (r50.43, P50.02). Arterial blood gases, respiratory function tests or the remaining HRV parameters were not correlated. When QT parameters were examined, QTc max (438630 vs. 424634 ms, P50.05) and QTc d (57610 vs. 4068 ms, P,0.0001) were longer and QTc min (382626 vs. 384638 ms, P50.6) unchanged in COPD patients compared to controls. There were 18 (44%) patients with QTc d longer than 60 ms and 23 (56%) patients with QTc d shorter than 60 ms. Patients with QTc d longer than 60 ms had similar age (6167.7 vs. 5869 years P50.2), HRV parameters, respiratory function test parameters, arterial blood gases compared to the rest of the COPD patients. However day-time night-time ratio of HF (1.0060.13 vs. 0.8560.12, P50.003) was significantly increased in COPD patients with QTc d longer than 60 ms (Table 3). The findings were confirmed when oneway ANOVA analysis of controls, QTc d longer than 60 ms and QTc d lower than 60 ms with COPD patients was done (one-way ANOVA for HF ratio P,0.001, on post hoc Tukey’s HSD analysis P value 0.008 between QTc d longer than 60 ms and QTc d lower than 60 ms, P value,0.001 between controls and QTc d longer than 60 ms, P value 0.5 between controls and QTc d lower than 60 ms with COPD patients). When the other circadian rhythm parameters were

203

Table 3 The comparison of COPD patients with QTc d under or over 60 ms

Age (years) SDANN (ms) S.D. (ms) RMSSD (ms) PNN50 (%) SDNN (ms) TP (ln (ms 2 )) LF (ln (ms 2 )) HF (ln (ms 2 )) LF / HF LF / HF day-time LF / HF night-time LF day / LF night ratio HF day / HF night ratio VC (l) FVC (l) FEV1 (l) Pa O 2 (mmHg) Pa CO 2 (mmHg)

QTc d ,60 ms n523

QTc d $60 ms n518

P value

5869 89637 36615 2469 5.465 97638 6.160.9 4.861.1 3.861 1.360.23 1.3460.25 1.2460.21 0.9260.07 0.8560.1 2.360.8 2.1560.8 1.1960.68 64613 4769

6168 76621 39614 28610 7.166.4 88623 6.460.7 5.261.1 4.460.9 1.2160.23 1.1760.23 1.2460.24 0.9560.07 160.1 2.461 2.1860.9 1.2260.66 69614 4468

P50.2 P50.4 P50.4 P50.2 P50.4 P50.4 P50.2 P50.3 P50.09 P50.3 P50.07 P50.9 P50.3 P50.003 P50.8 P50.9 P50.8 P50.3 P50.2

examined in patients with QTc d longer than 60 ms, there was a trend for night-time increase in TP, LF and LF / HF ratio without reaching statistical significance. The circadian parameters in patients with QTc d less than 60 ms were similar to controls (Table 4). When bivariate correlation of QTc d with other parameters were investigated, day-time night-time HF ratio (r50.47, P50.009) (Fig. 1) was correlated. Day-time LF / HF ratio was significantly correlated with FEV1 (r50.52, P50.004) and PEF (r50.57, Table 4 The comparison of the circadian changes in frequency domain HRV parameters in COPD patients with QTc d over or under 60 ms and controls QTc d $60 ms n518 TP (ln (ms 2 )) Day 6.3160.7 Night 6.6460.88 LF (ln (ms 2 )) Day 5.0761.1 Night 5.461.25 HF (ln (ms 2 )) Day 4.3560.8 Night 4.4261.1 LF / HF Day 1.1860.24 Night 1.2460.24 a b

QTc d ,60 ms n523

Controls

P value

5.8560.9 6.4360.96

6.4360.7 6.9960.58

P50.004 a P,0.001 b

4.6561.1 5.0861.24

5.5960.76 5.9660.6

P50.006 a P50.002 b

3.5460.95 4.1661.1

3.9560.72 4.8260.65

P50.5 a P50.001 b

1.3460.25 1.2460.21

1.4360.14 1.2460.09

P50.08 a P50.02 b

QTc d $60 ms COPD day–night. QTc d ,60 ms COPD day–night.

204

¨ et al. / International Journal of Cardiology 88 (2003) 199–206 T. Tukek

Fig. 1. Correlation of QT d with HF day / HF night ratio.

P50.001) among the respiratory function test parameters.

4. Discussion In our study, COPD patients had increased interventricular septum diastolic diameter and decreased left ventricular EF without causing overt systolic left ventricular dysfunction. These findings were consistent with previous reports [10–12]. The increase in interventricular septum diastolic diameter and right ventricular free wall thickness without concomitant increase in left ventricular mass and left ventricular diastolic internal diameter could be related to increased pulmonary pressure in the COPD patients. Ventricular and supraventricular arrhythmias are common in COPD patients [1,13–15]. The putative factors of arrhythmia include concomitant diseases, age, medications, hypoxia, hypercarbia, acid–base disorders, autonomic dysfunction, but no one factor was purported to be dominant [2,4,14,15]. Correlation between arrhythmia and arterial blood gases were only shown in a few patient groups [2,15]. Incalzi et

al. studying atrial fibrillation and ventricular arrhythmias in COPD patients reported that rhythm disorders were more frequent at night. They reported that while ventricular arrhythmias were related to left ventricular diastolic dysfunction, atrial premature beats were related to Pa O 2 and Pa CO 2 changes. However, they believed that other factors may be as important in the development of arrhythmia [3]. In our analysis we were unable to confirm these findings. Autonomic dysfunction and QTc d are important factors in the development of arrhythmia and are known to be disturbed in COPD patients. There are reports of their recovery after 24 h oxygen therapy and correction of hypoxia [16,17]. However in these studies the relation of arrhythmia with autonomic dysfunction was not examined. Additionally, Stein et al. [5] reported no correlation between autonomic function parameters and atrial or ventricular arrhythmia in COPD patients. In our study we found an important positive correlation between VPB and prolonged QTc d . Additionally, ventricular premature beat frequency and QT d were correlated with daytime night-time variability of HF. These findings imply that suppression in the normal increase in

¨ et al. / International Journal of Cardiology 88 (2003) 199–206 T. Tukek

night-time HF predispose to arrhythmia. However presence of hypoxia ( pO 2 ,60 mmHg) or hypercarbia ( pCO 2 .45 mmHg) did not cause a significant increase in ventricular premature beat frequency or prolongation of QTc d . When COPD patients were examined, factors for development of paroxysmal atrial fibrillation were found to be age, unchanged HF hence increased night-time LF / HF, and decreased HF variability. In a recent study, a relation between age and HF was reported. Crasset et al. [18] found aging was linked to decrease in deep sleep and was related to lower night-time HF. They related these findings to impaired deep sleep pattern with aging which disturbed the rate and rhythm of the respiration. These findings may also be valid in our patients with atrial fibrillation who may have disturbed sleep and breathing pattern at night which may impair circadian changes in autonomic function. However, age was not retained as significant in the multivariate analysis. There are two studies which examine autonomic function assessed by HRV in COPD patients. Volteranni et al. [19] reported increased HF in COPD patients and put forth that cardiac autonomic control works in tandem with pulmonary parasympathetic control. Stein et al. [5] examined young alpha 1 PiZ antitrypsine deficient COPD patients and reported a decrease in all HRV parameters. Our results resemble the findings of Volteranni et al. [19] in that we found no change in HF and other parameters were decreased consistent with the findings of Stein et al. This discrepant finding on HF may partially be explained by the continued use of medications, which may modify autonomic tone in the Stein et al. study. We found that night-time TP and HF increased and LF / HF decreased in controls consistent with previous findings [6,20]. We found that night-time LF and HF were increased and LF / HF ratio was unchanged in COPD patients. This was also consistent with a previous report which did not report on their relation with arrhythmia [5]. In our study we did not observe the night-time rise in HF in COPD patients with atrial fibrillation or QTc d $60 ms. Day-time and night-time ratios of HF were significantly correlated with ventricular premature beat frequency and QTc d . It seems that the lack of circadian rhythm in HF is closely associated with supraventricular or ventricular arrhythmia. This lack of rise in night-time HF may

205

represent unopposed sympathetic activity during night-time. The rise in LF / HF ratio in patients with arrhythmia may substantiate this idea. We also observed that circadian rhythm was intact in COPD patients without arrhythmia regardless of their arterial blood gases and respiratory function test parameters. It was concluded that circadian changes like nighttime lack of HF increase may result in increased arrhythmogenicity due to relatively unopposed sympathetic stimulation. This may be related to impaired deep sleep and breathing patterns.

Acknowledgements ¨ and Hulya ¨ We gratefully thank Zahide Ac¸ıkgoz Bedır for their excellent technical aid in data collection.

References [1] Kleiger RE, Senior RM. Long-term electrocardiographic monitoring of ambulatory patients with chronic airway obstruction. Chest 1974;65:483–7. [2] Shih HT, Webb CR, Conway WA, Peterson E, Tilley B, Goldstein S. Frequency and significance of cardiac arrhythmias in chronic obstructive lung disease. Chest 1988;94:44–8. [3] Incalzi RA, Pistelli R, Fuso L, Cocchi A, Bonetti MG, Giordano A. Cardiac arrhythmias and left ventricular function in respiratory failure from chronic obstructive pulmonary disease. Chest 1990;97:1092–7. [4] Stewart AG, Waterhouse JC, Howard P. The QTc interval, autonomic neuropathy and mortality in hypoxaemic COPD. Respir Med 1995;89:79–84. [5] Stein PK, Nelson P, Rottman JN et al. Heart rate variability reflects severity of COPD in PiZ alpha-1-antitrypsin deficiency. Chest 1998;113:327–33. [6] Burger AJ, Kamalesh M. Effect of beta-adrenergic blocker therapy on the circadian rhythm of heart rate variability in patients with chronic stable angina pectoris. Am J Cardiol 1999;83:596–8. [7] Lombardi F, Sandrone G, Mortara A et al. Circadian variation of spectral indices of heart rate variability after myocardial infarction. Am Heart J 1992;123:1521–9. [8] 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. Circulation 1996;93(5), Special report. [9] Declaration of Helsinki. Recommendation guiding physicians in biomedical research involving human subjects. In: 41st World Medical Assembly, Hong-Kong, September, 1989. [10] Incalzi RA, Pistelli R, Fuso L et al. Cardiac arrhythmias and left ventricular function in respiratory failure from chronic obstructive pulmonary disease. Chest 1990;97:1092–7.

206

¨ et al. / International Journal of Cardiology 88 (2003) 199–206 T. Tukek

[11] Vizza CD, Lynch JP, Ochoa LL et al. Right and left ventricular dysfunction in patients with severe pulmonary disease. Chest 1998;113:576–83. [12] Jardin F, Gueret P, Prost JF et al. Two-dimensional echocardiographic assessment of left ventricular function in chronic obstructive pulmonary disease. Am Rev Respir Dis 1984;129:135–42. [13] Gorecka D. Cardiac arrhythmias in chronic obstructive pulmonary disease. Monaldi Arch Chest Dis 1997;52:278–81. [14] Flick MR, Block AJ. Nocturnal vs. diurnal cardiac arrhythmias in patients with chronic obstructive pulmonary disease. Chest 1979;75:8–11. [15] Shepard Jr. JW, Garrison MW, Grither DA, Evans R, Schweitzer PK. Relationship of ventricular ectopy to nocturnal oxygen desaturation in patients with chronic obstructive pulmonary disease. Am J Med 1985;78:28–34. [16] Sarubbi B, Esposito V, Ducceschi V et al. Effect of blood gas derangement on QTc dispersion in severe chronic obstructive pulmonary disease: evidence of an electropathy? Int J Cardiol 1997;58:287–92.

[17] Scalvini S, Porta R, Zanelli E et al. Effects of oxygen on autonomic nervous system dysfunction in patients with chronic obstructive pulmonary disease. Eur Respir J 1999;13:119–24. [18] Crasset V, Mezzetti S, Antoine M, Linkowski P, Degaute JP, van de Borne P. Effects of aging and cardiac denervation on heart rate variability during sleep. Circulation 2001;103:84–8. [19] Volterrani M, Scalvini S, Mazzuero G et al. Decreased heart rate variability in patients with chronic obstructive pulmonary disease. Chest 1994;106:1432–7. [20] Huikuri HV, Kessler KM, Terracall E, Castellanos A, Linnaluoto MK, Myerburg RJ. Reproducibility and circadian rhythm of heart rate variability in healthy subjects. Am J Cardiol 1990;65:391–3.