Intraindividual Reproducibility of Heart Rate Variability STEFAN H. HOHNLOSER, THOMAS KLINGENHEBEN, MARKUS ZABEL, FRANK SCHRODER, and HANJORG JUST From the University Hospital, Department of Cardiology, Freiburg, Germany
HOHNLOSER, S.H., ET AL.: Intraindividual Reproducibility of Heart Rate Variability. Heart rate variabil-
ity was determined from three consecutive HoJter recordings performed on days 1, 7, and 28 in 17 normai subjects, in 13 patients with angiographicaJJy normal coronary arteries, and in 9 patients with remote myocardial infarctions. Group data of several time and frequency domain measures of heart rate variability were highly reproducible (correlation coefficients 0.629-0.894]. However, some individuals exhibited considerably larger day-to-day variations in heart rate variability. Single heart rate indices differed by up to 50% between two Holter recordings. Such potential di/ferences must be considered when repeated heart rate variability determinations are used to assess changes in neurocardiac re/lex regulation or effects of therapeutic interventions. (PACE, Vol. 15, November, Part II 1992] heart rate variability, Holter monitoring, reproducibility Introduction According to experimental as well as to clinical studies, heart rate variability represents a noninvasive assessment of autonomic tone, particularly cardiac vagal tone.^ During recent years, heart rate variability has been increasingly used to assess the risk of sudden cardiac death in patients after myocardial infarction.^"^ Several investigations have demonstrated that low heart rate variability constitutes an independent risk factor of mortality in these patient populations.^"^ Most measurments of heart rate variability have been computed from single 24-hour ambulatory ECG recordings and have then been compared with preselected cut-off values for purposes of individual risk stratification.^"'' In addition, some studies have attempted to evaluate the effects of interventions such as drug therapy^ on heart rate variability. These efforts have, however, been limited by the paucity of data available with respect to intraindividual stability and reproducibility of heart rate variability.^ This study was, therefore, designed to examine the short-term variations and
Address for reprints: Stefan H. Hohnloser, M.D., University Hospital, Dept. of Cardiology, Hugstetterstr. 55, 7800 Freiburg, Germany. Fax: 0761-72295.
PACE, Vol. 15
the reproducibility of frequently utilized heart rate variability measurements from three separate 24hour ambulatory ECG recordings in healthy subjects and in patients with cardiac diseases.
Methods Study Population
Three populations were examined: group I consisted of 17 healthy medical students (6 women, 11 men at a mean age of 24 ± 2.5 years) without cardiac disease, systemic hypertension, or diabetes mellitus. All subjects had a completely unremarkable medical history and a normal physical examination. None of these individuals were taking any kind of medication. Group II consisted of 13 patients with atypical chest pain undergoing left heart catheterization. Coronary angiography revealed normal coronary arteries in all patients. The mean age of these four women and nine men was 51 ± 10 years. Nine patients had systemic hypertension. Five of these nine patients were taking ace inhibitors, and four others were taking diltiazem. Group III consisted of two women and seven men at a mean age of 62 ± 8 years who had suffered a myocardial infarction at least 12 months before inclusion in the study. The extent and se-
November, Part n 1992
2211
HOHNLOSER, ET AL.
verity of coronary artery disease was establisbed in all by coronary angiography. No patient was taking beta adrenergic blocking drugs or digitalis glycosides, but all received nitrates, diltiazem, or nifedipine. In all patients in groups II and III, particular care was taken to ensure that medications and dosages utilized were held constant throughout the entire study period. Patients in whom medications had to be changed for clinical reasons were excluded from further analysis.
Statistics Values are given as mean ± SD. Results of the baseline and the two subsequent ambulatory ECG recordings were compared by Student's t-test for paired samples. Correlations between heart rate variability parameters, heart rate and clinical values were calculated by means of the SPSS statistical analysis program (version 3.1, SPSS Inc., Chicago, IL, USA). Statistical significance was considered as P < 0.05.
Holter Monitoring and Measurement of Heart Rate Variability
Ambulatory monitoring was performed utilizing a two-channel bipolar recorder and evaluated semiautomatically after digitization by an AMmodulated analysis system (Marquette 8000, Marquette Electronics, Inc., Milwaukee, WI, USA). For each hour, the following data were computed and tabulated on a printout: heart rate, total ventricular premature beats, couplets, ventricular tachycardia runs and beats, and time analyzed. The Marquette 8000 computerized Holter system (version 5.7 software) provides R wave detection via precise fiducial point markings. Only RR cycles in which beats had a normal morphology and whose cycle length duration was within 20% of the preceding cycle length were measured to assure the rejection of ectopic beats. In instances in which sinus rhythm was interrupted by a premature atrial or ventricular beat, one RR interval preceding and one following the nonsinus beat were rejected. This beat classification was verified, manually overread, and corrected where appropriate by two independent investigators. Each RR interval during the respective recording time was determined and further analyzed. For each recording, the following time domain measurements of heart rate variability were obtained: SD (the standard deviation of the mean RR interval); pNN50 (proportion of adjacent RR intervals > 50 msec different); rMSSD (root mean square of difference of successive RR intervals). In addition, the software provides three frequency domain measurements of heart rate variability utilizing fast Fourier transform, namely the low frequency (0.04—0.15 Hz), the high frequency (0.15-0.40 Hz), and the total frequency (0.01-1.00 Hz) components.
2212
Results Heart Rate Variability in Normal Volunteers and in Patients witb Cardiac Disease
Analysis of the first 24-hour ambulatory ECG revealed comparable mean heart rates in normal individuals versus patients from groups II and III (Table I). However, time and frequency domain measurements of heart rate variability were significantly lower in subjects with heart diseases as compared to normals. Particularly, measurements of parasympathetic nervous system activity such as rMSSD, pNN50, or high frequency were depressed in patients with cardiac diseases during this inital recording.
Table I. Measures of Heart Rate Variability in Normal Volunteers and in Patients with Cardiac Disease
Mean RR (msec) SD (msec) rMSSD (msec) pNN50(%) HF (msec)
Normai Volunteers (N = 17)
Group It and III Patients (N = 22)
P Value
843 ± 55
829 ± 123
NS
185 ± 39 68 ± 34
146 ± 58 46 ± 39
50 msec different; rMMSO = root mean square of difference of successive RR intervals; SD = standard deviation.
November, Part II 1992
2213
HOHNLOSER, ET AL.
well as by time domain measures sucb as rMSSD or pNN50.'^ Since tbe initial study by Kleiger and co-workers in 1987,^ beart rate variability has been sbown to predict mortality in postinfarction patients independently of other major risk factors.^-* Bigger and co-workers^ have recently demonstrated that, in postinfarction patients, heart rate variability remained stable when assessed by means of ambulatory monitoring on two consecutive days. Similar results were reported for patients with congestive heart failure.^ The present study adds to these observations by extending the time period over which heart rate variability was repeatedly determined. Gomparing group data from 24-hour ambulatory EGGs recorded 1 to 4 weeks apart revealed high correlation coefficients for frequently utilized indices of heart rate variability. As expected, patients with heart disease had lower heart rate variability than normal subjects. However, wben data from individual subjects were compared, larger day-to-day variations of the various heart rate variability measures were found. Variations ranged from 0% to approximately 50% for normal individuals and patients with cardiac diseases. Similar observations have been reported by van Hoogenhuyze and co-workers^ who also examined the heart rate variability of normal individuals and of patients witb congestive heart failure. The exact mechanisms causing these varia-
tions are unknown. Since age and gender have been demonstrated to influence heart rate variability, these factors may also contribute to its reproducibility. Heart rate variability has been shown to be higher in the young as opposed to middleaged or elderly subjects;^ variations in heart rate variability over time are thus expected to be larger in the young. Others have demonstrated that heart rate variability is depressed in patients with congestive heart failure compared to healthy individuals; this implies that variations in heart rate variability would tend to be lower in such patients.^ Finally, it has been demonstrated tliat beart rate variability is dependent on heart rate.® Though heart rate variability reflects predominantly vagal activity, it could nevertheless be influenced by sympatbetically mediated changes in heart rate as well. In conclusion, this study demonstrates that group measurements of heart rate variability indices remain stable over a short period of time. The larger day-to-day variation in individual patients, however, must be considered when heart rate variability determinations are used to assess cbanges in neurocardiac reflex regulation or therapeutic interventions. Furthermore, the data emphasize the need to establish normal limits of heart rate variability with respect to clinical variables such as age, gender, or presence or absence of heart disease.
References 1. Bigger JT Jr, Kleiger RE, Fleiss JL, et al. Components of heart rate veiriability measured during healing of acute myocardial infarction. Am J Cardiol 1988; 61:208-215. 2. Kleiger RE, Miller JP, Bigger JT Jr, et al. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 1987; 59:256-262. 3. Bigger JT Jr, Fleiss JL, Steinman RC, et al. Frequency domain measures of heart period variability and mortality after myocardial infarction. Circulation 1992; 85:164-171. 4. Farrell TG, Bashir Y, Cripps T, et al. Risk stratification for arrhythmic events in postinfarction patients based on heart rate variability, ambulatory electrocardiographic variables and the signal-averaged electrocarcliogram. J Am Coll Cardiol 1991; 18:687-697. 5. Zuanetti G, Latini R, Neilson JMM, et al. Heart rate variability in patients with ventricular arrhyth-
2214
6.
7.
8.
9.
mias: Effect of antiarrhythmic drugs. J Am Coll Cardiol 1991; 17:604-612. van Hoogenhuyze D, Weinstein N, Martin GJ, et al. Reproducibility and relation to mean heart rate of heart rate variability in normal subjects and in patients with congestive heart failure secondary to coronary artery disease. Am J Cardiol 1991; 68: 1668-1676. Bigger JT Jr, Fleiss JL, Steinman RC, et al. Correlations among time and frequency domain measures of heart period variability two weeks after acute myocardial infarction. Am J Cardiol 1992; 69: 891-898. Bigger JT Jr, Fleiss JL, Rolnitzky LM, et al. Stability over time of heart period variability in patients with previous myocardied infarction and ventricular arrhythmias. Am J Cardiol 1992; 69:718-723. Masaoka S, Lev-Ran A, Hill LR, et al. Heart rate variability in diabetes: Relationship to age and duration of the disease. Diabetes Care 1985; 8:64-69.
November, Part II 1992
PACE, Vol. 15
HOHNLOSER, S.H., ET AL.: Intraindividual Reproducibility of Heart Rate Variability. Heart rate variabil-
ity was determined from three consecutive HoJter recordings performed on days 1, 7, and 28 in 17 normai subjects, in 13 patients with angiographicaJJy normal coronary arteries, and in 9 patients with remote myocardial infarctions. Group data of several time and frequency domain measures of heart rate variability were highly reproducible (correlation coefficients 0.629-0.894]. However, some individuals exhibited considerably larger day-to-day variations in heart rate variability. Single heart rate indices differed by up to 50% between two Holter recordings. Such potential di/ferences must be considered when repeated heart rate variability determinations are used to assess changes in neurocardiac re/lex regulation or effects of therapeutic interventions. (PACE, Vol. 15, November, Part II 1992] heart rate variability, Holter monitoring, reproducibility Introduction According to experimental as well as to clinical studies, heart rate variability represents a noninvasive assessment of autonomic tone, particularly cardiac vagal tone.^ During recent years, heart rate variability has been increasingly used to assess the risk of sudden cardiac death in patients after myocardial infarction.^"^ Several investigations have demonstrated that low heart rate variability constitutes an independent risk factor of mortality in these patient populations.^"^ Most measurments of heart rate variability have been computed from single 24-hour ambulatory ECG recordings and have then been compared with preselected cut-off values for purposes of individual risk stratification.^"'' In addition, some studies have attempted to evaluate the effects of interventions such as drug therapy^ on heart rate variability. These efforts have, however, been limited by the paucity of data available with respect to intraindividual stability and reproducibility of heart rate variability.^ This study was, therefore, designed to examine the short-term variations and
Address for reprints: Stefan H. Hohnloser, M.D., University Hospital, Dept. of Cardiology, Hugstetterstr. 55, 7800 Freiburg, Germany. Fax: 0761-72295.
PACE, Vol. 15
the reproducibility of frequently utilized heart rate variability measurements from three separate 24hour ambulatory ECG recordings in healthy subjects and in patients with cardiac diseases.
Methods Study Population
Three populations were examined: group I consisted of 17 healthy medical students (6 women, 11 men at a mean age of 24 ± 2.5 years) without cardiac disease, systemic hypertension, or diabetes mellitus. All subjects had a completely unremarkable medical history and a normal physical examination. None of these individuals were taking any kind of medication. Group II consisted of 13 patients with atypical chest pain undergoing left heart catheterization. Coronary angiography revealed normal coronary arteries in all patients. The mean age of these four women and nine men was 51 ± 10 years. Nine patients had systemic hypertension. Five of these nine patients were taking ace inhibitors, and four others were taking diltiazem. Group III consisted of two women and seven men at a mean age of 62 ± 8 years who had suffered a myocardial infarction at least 12 months before inclusion in the study. The extent and se-
November, Part n 1992
2211
HOHNLOSER, ET AL.
verity of coronary artery disease was establisbed in all by coronary angiography. No patient was taking beta adrenergic blocking drugs or digitalis glycosides, but all received nitrates, diltiazem, or nifedipine. In all patients in groups II and III, particular care was taken to ensure that medications and dosages utilized were held constant throughout the entire study period. Patients in whom medications had to be changed for clinical reasons were excluded from further analysis.
Statistics Values are given as mean ± SD. Results of the baseline and the two subsequent ambulatory ECG recordings were compared by Student's t-test for paired samples. Correlations between heart rate variability parameters, heart rate and clinical values were calculated by means of the SPSS statistical analysis program (version 3.1, SPSS Inc., Chicago, IL, USA). Statistical significance was considered as P < 0.05.
Holter Monitoring and Measurement of Heart Rate Variability
Ambulatory monitoring was performed utilizing a two-channel bipolar recorder and evaluated semiautomatically after digitization by an AMmodulated analysis system (Marquette 8000, Marquette Electronics, Inc., Milwaukee, WI, USA). For each hour, the following data were computed and tabulated on a printout: heart rate, total ventricular premature beats, couplets, ventricular tachycardia runs and beats, and time analyzed. The Marquette 8000 computerized Holter system (version 5.7 software) provides R wave detection via precise fiducial point markings. Only RR cycles in which beats had a normal morphology and whose cycle length duration was within 20% of the preceding cycle length were measured to assure the rejection of ectopic beats. In instances in which sinus rhythm was interrupted by a premature atrial or ventricular beat, one RR interval preceding and one following the nonsinus beat were rejected. This beat classification was verified, manually overread, and corrected where appropriate by two independent investigators. Each RR interval during the respective recording time was determined and further analyzed. For each recording, the following time domain measurements of heart rate variability were obtained: SD (the standard deviation of the mean RR interval); pNN50 (proportion of adjacent RR intervals > 50 msec different); rMSSD (root mean square of difference of successive RR intervals). In addition, the software provides three frequency domain measurements of heart rate variability utilizing fast Fourier transform, namely the low frequency (0.04—0.15 Hz), the high frequency (0.15-0.40 Hz), and the total frequency (0.01-1.00 Hz) components.
2212
Results Heart Rate Variability in Normal Volunteers and in Patients witb Cardiac Disease
Analysis of the first 24-hour ambulatory ECG revealed comparable mean heart rates in normal individuals versus patients from groups II and III (Table I). However, time and frequency domain measurements of heart rate variability were significantly lower in subjects with heart diseases as compared to normals. Particularly, measurements of parasympathetic nervous system activity such as rMSSD, pNN50, or high frequency were depressed in patients with cardiac diseases during this inital recording.
Table I. Measures of Heart Rate Variability in Normal Volunteers and in Patients with Cardiac Disease
Mean RR (msec) SD (msec) rMSSD (msec) pNN50(%) HF (msec)
Normai Volunteers (N = 17)
Group It and III Patients (N = 22)
P Value
843 ± 55
829 ± 123
NS
185 ± 39 68 ± 34
146 ± 58 46 ± 39
50 msec different; rMMSO = root mean square of difference of successive RR intervals; SD = standard deviation.
November, Part II 1992
2213
HOHNLOSER, ET AL.
well as by time domain measures sucb as rMSSD or pNN50.'^ Since tbe initial study by Kleiger and co-workers in 1987,^ beart rate variability has been sbown to predict mortality in postinfarction patients independently of other major risk factors.^-* Bigger and co-workers^ have recently demonstrated that, in postinfarction patients, heart rate variability remained stable when assessed by means of ambulatory monitoring on two consecutive days. Similar results were reported for patients with congestive heart failure.^ The present study adds to these observations by extending the time period over which heart rate variability was repeatedly determined. Gomparing group data from 24-hour ambulatory EGGs recorded 1 to 4 weeks apart revealed high correlation coefficients for frequently utilized indices of heart rate variability. As expected, patients with heart disease had lower heart rate variability than normal subjects. However, wben data from individual subjects were compared, larger day-to-day variations of the various heart rate variability measures were found. Variations ranged from 0% to approximately 50% for normal individuals and patients with cardiac diseases. Similar observations have been reported by van Hoogenhuyze and co-workers^ who also examined the heart rate variability of normal individuals and of patients witb congestive heart failure. The exact mechanisms causing these varia-
tions are unknown. Since age and gender have been demonstrated to influence heart rate variability, these factors may also contribute to its reproducibility. Heart rate variability has been shown to be higher in the young as opposed to middleaged or elderly subjects;^ variations in heart rate variability over time are thus expected to be larger in the young. Others have demonstrated that heart rate variability is depressed in patients with congestive heart failure compared to healthy individuals; this implies that variations in heart rate variability would tend to be lower in such patients.^ Finally, it has been demonstrated tliat beart rate variability is dependent on heart rate.® Though heart rate variability reflects predominantly vagal activity, it could nevertheless be influenced by sympatbetically mediated changes in heart rate as well. In conclusion, this study demonstrates that group measurements of heart rate variability indices remain stable over a short period of time. The larger day-to-day variation in individual patients, however, must be considered when heart rate variability determinations are used to assess cbanges in neurocardiac reflex regulation or therapeutic interventions. Furthermore, the data emphasize the need to establish normal limits of heart rate variability with respect to clinical variables such as age, gender, or presence or absence of heart disease.
References 1. Bigger JT Jr, Kleiger RE, Fleiss JL, et al. Components of heart rate veiriability measured during healing of acute myocardial infarction. Am J Cardiol 1988; 61:208-215. 2. Kleiger RE, Miller JP, Bigger JT Jr, et al. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 1987; 59:256-262. 3. Bigger JT Jr, Fleiss JL, Steinman RC, et al. Frequency domain measures of heart period variability and mortality after myocardial infarction. Circulation 1992; 85:164-171. 4. Farrell TG, Bashir Y, Cripps T, et al. Risk stratification for arrhythmic events in postinfarction patients based on heart rate variability, ambulatory electrocardiographic variables and the signal-averaged electrocarcliogram. J Am Coll Cardiol 1991; 18:687-697. 5. Zuanetti G, Latini R, Neilson JMM, et al. Heart rate variability in patients with ventricular arrhyth-
2214
6.
7.
8.
9.
mias: Effect of antiarrhythmic drugs. J Am Coll Cardiol 1991; 17:604-612. van Hoogenhuyze D, Weinstein N, Martin GJ, et al. Reproducibility and relation to mean heart rate of heart rate variability in normal subjects and in patients with congestive heart failure secondary to coronary artery disease. Am J Cardiol 1991; 68: 1668-1676. Bigger JT Jr, Fleiss JL, Steinman RC, et al. Correlations among time and frequency domain measures of heart period variability two weeks after acute myocardial infarction. Am J Cardiol 1992; 69: 891-898. Bigger JT Jr, Fleiss JL, Rolnitzky LM, et al. Stability over time of heart period variability in patients with previous myocardied infarction and ventricular arrhythmias. Am J Cardiol 1992; 69:718-723. Masaoka S, Lev-Ran A, Hill LR, et al. Heart rate variability in diabetes: Relationship to age and duration of the disease. Diabetes Care 1985; 8:64-69.
November, Part II 1992
PACE, Vol. 15
