SHORT-TERM SPECTRAL ANALYSIS OF HEART RATE VARIABILITY DURING SUPINE-STANDING-SUPINE TEST IN PATIENTS WITH PAROXYSMAL ATRIAL

The aim of the study was to assess the sympathovagal balance in group of 27 patients without significant structural heart disease after an attack of atrial fibrillation. The investigation was performed using spectral analysis of heart rate variability during examination under conditions of different orthostatic loads in single phases, called the supine-standing- supine test. The findings were compared with a group of healthy persons. These revealed a significantly decreased total spectral power (430.7 vs 1558.0 ms(2) supine1; 477.6 vs 1042,5 ms(2) standing; 567.5 vs 1948.5 ms(2) supine2), and spectral power of the high frequency spectral component (140.8 vs 619.3 ms(2) supine1; 96.2 vs 203.3 ms(2) standing; 186.3 vs 739.4 ms(2) supine2) in the studied group of patients in comparison with the control group.


INTRODUCTION
Atrial fi brillation (AF) is arrhythmia of increasing importance in ageing populations. 1,2 t has been characterized as a syndrome with complicated aetiopathogenetic mechanism involving the nervous system. 3The autonomic nervous system (ANS), in particular, is the homeostatic body system with infl uence on arrhythmogenesis as a modulating or even triggering factor. 4,5,6 Sectral analysis (SA) of heart rate variability (HRV) is a promising noninvasive method of ANS evaluation which off ers both qualitative and quantitative assessment of sympatho-vagal interactions. 7This method enables us in short-term recordings to distinguish three main components in the frequency domain: very low frequency VLF (10-50 mHz), low frequency LF (50-150 mHz) and high frequency HF (150-400 mHz). 7The majority of hitherto studies focused on ANS evaluation in paroxysmal AF have been based on long-term ECG recordings.In these tachograms, analysed were either total HRV during the whole period of the preceding sinus rhythm before AF attack 8 or short-term (20 minutes) time segments before 9 or after 10 attack.We examined the SA of short-term HRV during the supinestanding-supine test 11 in patients after AF attack.

MATERIAL AND METHODS
Inclusion criteria: ECG documented AF with a duration of less than 48 hours, normal left atrium diameter on echocardiography examination, 12 age under 70 years.
Exclusion criteria: signifi cant structural heart disease, acute coronary syndrome, valvular heart disease, other acute or severe diseases.
All the patients were examined under standard conditions between 8 and 11 a.m. at room temperature of 22 to 25 degrees Celsius after elimination of cigarettes, alcohol and coff ee consumption at least 12 hours beforehand.Medication potentially infl uencing HRV was stopped (52 %) or administered for the last time the day before examination (Propafenon 33 %, angiotensin converting enzyme inhibitors 29 %, Digoxin 11 %, betablockers 11 %).The patients remained in each position for 5 minutes during supine-standing-supine test to record at least 300 R-R intervals .The examinations were performed by means of the telemetric diagnostic system VARIA PULSE TF3 13 and VARIA CARDIO TF4. 14 The spectral analysis of HRV using fast Fourier transformation partially modifi ed according to Yamamoto (coarse-graining spectral analysis) was used for data processing.The following parameters were used for evaluation: total spectral power (TOT.PWR), spectral power of VLF, LF, HF components, LF/HF ratio which denominated as a marker of sympatho-vagal balance, 15 coeffi cient of component variance (CCV) to eliminate diff erences in mean R-R level and their impact on the amplitude of oscillations. 16The values of LF and HF were expressed in absolute (ms 2 ) and normalized units. 7he values of parameters obtained in the patient group were compared with the results of a corresponding (age and gender matched) heatlthy subject group (27 persons, mean age 51.1 years, range 35.1-68.8). 17,18

RESULTS
The results of the supine-standing-supine test are shown both in tabular and graphic formats, including 3-dimensional graphs.
The data are summarized in Tables 1-5.The mean values and standard deviation (SD) of the total spectral power and spectral powers of single frequency components in supine-standing-supine positions are shown in Table 1.
The mean values of CCV in the patient group (P) and in the healthy controls (C) in the course of the supinestanding-supine test are shown in Table 2.
The average values of total spectral power and spectral power and density of LF and HF components compared between the patient group (P) and the healthy controls (C) during the clinostatic phase of the supine-standingsupine test expressed as diff erences in values of standing phase minus supine 2 phase are shown in Table 3.
LF/HF spectral components ratios during the supine-standing-supine test in the patient group (P) and the healthy controls (C) are shown in Table 4.
Comparison of changes of LF and HF spectral components during the supine-standing-supine test between the patient group (P) and the healthy controls (C) was expressed in normalized units (Table 5).
The most important diff erences between mean values of each LF and HF components in normalized unit were found having compared supine 1 versus standing position, and standing versus supine 2 position.These diff erences have confi rmed the dynamics of the autonomic nervous system responses within the scope of this test when the values of the signifi cance of the changes equalled 0.006 and/or 0.001 in the patient group and 0.0003 for the healthy controls.
These diff erences demonstrate great sensitivity of the supine-standing-supine test for evaluation of autonomic regulation of heart rate.The results of spectral analysis of heart rate variability during the supine-standing-supine test in 3D graph show abnormal fi nding of striking attenuation of spectral component with the involvement of vagal and sympathetic component in a man 36 years old with a history of atrial fi brillation.
Legend: PSD -power spectral density, T1 -time of initial supine position, T2 -time of standing position, T3 -time of fi nal supine position.

Fig. 2. (control)
The results of spectral analysis of heart rate variability during the supine-standing-supine test in 3D graph show normal fi nding with good expression of spectral component dynamics in a healthy, 42-years-old man.
Legend: PSD -power spectral density, T1-time of initial supine position, T2 -time of standing position, T3 -time of fi nal supine position.Short-term spectral analysis of heart rate variability during supine-standing-supine test in patients with paroxysmal atrial fi brillation Table 1.Spectral analysis of heart rate variability during the supine-standing-supine test -comparison of the mean spectral power values between the patient group (P) and healthy controls (C) expressed in spectral parameters of single frequency components and total power.), PSD power spectral density (ms 2 /Hz).There was a signifi cant increase in PWR.HF and PSD.HF components in the clinostase (negative value of diff erences) in both the groups under study, but much more pronounced in the control group.This change refl ecting vagal loading is also responsible for the signifi cant increase in total spectral power in both groups.

Spectral power (ms
Table 4. Spectral analysis of heart rate variability during the supine-standing-supine test; LF/HF ratio during the supine-standing-supine test in the patient group (P) and in the healthy controls (C) expressed as mean values and standard deviations.

Ratio
Patients Controls p LF/HF supine1 (SD) 1.6 (1.4) 1.5 (1.1) 0.71 LF/HF standing (SD) 5.4 (6.9) 5.5 (5.0) 0.28 LF/HF supine2 (SD) 1.5 (1.5) 1.5 (1.0) 0.382 Legend: Mann-Whitney U test, (P) Patients N=27, (C) Controls N=27.No signifi cant diff erences were found between patient and control groups.Changes in LF/HF balance were similar in both groups for single phases of the supine-standing-supine test.Legend: Normalized units are used to express the relationship between LF and HF components diff erently from the LF/HF ratio.A normalized unit is a relative value for each power component in propotion to the total minus the VLF component. 7These results show a similar sympathovagal balance between both the study groups during testing.
studied test but their regulatory mechanisms oscillated in a smaller range.It was not possible to withdraw medication before examination in all patients, but according to the references 7,19,20 we might to suppose their vagomimetic and/ or neutral infl uence on the automonic nervous system.Therefore, it was diffi cult to explain the total low power and especially low power of the HF component by the above medication.
Evaluation of ratio LF/HF > 2, which Lombardi 10 termed a predictor of early recurrence of AF revealed this value as 29 %.This percentage would probably be higher without the mentioned medication, in particular betablockers.
The supine-standing-supine test off ers an easy way for evaluating the integrity of the autonomic heart regulations.Our results confi rm signifi cant diff erences between both studied groups.The applied test was eff ective to reliably diff erentiate between the patient group and that of healthy individuals.The changes in the autonomic nervous system reactivity may be considered as a factor contributing to the origin of arrhythmias.One of the inclusion criteria for the patient group was a normal diameter of the left atrium, and hence it might be supposed that there was no atrial anatomic remodellation.More studies dealing with AF and focused on individual evaluation would be needed in the future because of the complexity of AF etiopathogenesis.The supine standing-supine test seems to be a suitable complement to standard examination of patients with AF.DISCUSSION Supine-standing-supine test measures a dynamic loading of the autonomic nervous system in the form of a sequentional increase in the sympathetic activity and concomitant decrease in the vagal activity in the standing position, and also decrease in the sympathetic and increase in the vagal activity in the supine position after lying back.
We found signifi cantly lower total spectral power as well as power of single spectral components in the patient group.There were no signifi cant diferences between the groups in sympathovagal balance represented by the LF/ HF ratio as a result of proportional attenuation of both the parameters.The patients had similar response to the

Fig. 1 .
Fig. 1. (patient)The results of spectral analysis of heart rate variability during the supine-standing-supine test in 3D graph show abnormal fi nding of striking attenuation of spectral component with the involvement of vagal and sympathetic component in a man 36 years old with a history of atrial fi brillation.Legend: PSD -power spectral density, T1 -time of initial supine position, T2 -time of standing position, T3 -time of fi nal supine position.

Table 2 .
Mann-Whitney U test, Patients N = 27, Controls N = 27.SD standard deviation.The most marked diff erences were in the HF component.The diff erences between the groups under study were more pronounced for all three spectral components in supine position, and sometimes they were even fourhold.Spectral analysis of heart rate variability during the supine-standing-supine test -comparison of coeffi cients of component variance mean values between the patient group (P) and the healthy controls (C) expressed in %.
Legend: Mann-Whitney U test, (P) Patients N = 27, (C) Controls N = 27.CCV coeffi cients of component variance.There were signifi cant diff erences between both the groups (P and C) in powers of all the three spectral components after calculations of heart rate infl uences expressed in CCV.

Table 3 .
Spectral analysis of heart rate variability during the clinostatic phase of the supine-standing-supine test -mean value diff erences in spectral power and spectral density of LF, HF components (standing value minus supine 2 value) in the patient group (P) and the healthy controls (C).

Table 5 .
Spectral analysis of heart rate variability during the supine-standing-supine test -comparison of mean values of spectral components LF, HF between the patient group (P) and the healthy controls (C) expressed in nu (normalized units).