Effects of aerobic training on heart rate dynamics in sedentary subjects

7 downloads 18137 Views 230KB Size Report
Mar 21, 2003 - obic training in sedentary subjects results in altered auto- nomic regulation of .... software (Heart Signal, Kempele, Finland). All R-R intervals.
J Appl Physiol 95: 364–372, 2003. First published March 21, 2003; 10.1152/japplphysiol.00751.2002.

Effects of aerobic training on heart rate dynamics in sedentary subjects Mikko P. Tulppo,1,2 Arto J. Hautala,1,2 Timo H. Ma¨kikallio,1,2 Raija T. Laukkanen,3 Seppo Nissila¨,3 Richard L. Hughson,4 and Heikki V. Huikuri2 1

Merikoski Rehabilitation and Research Center, 90100 Oulu; 2Department of Medicine, Division of Cardiology, University of Oulu, 90220 Oulu; 3Polar Electro, 90440 Kempele, Finland; and 4 Department of Kinesiology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1 Submitted 14 August 2002; accepted in final form 20 March 2003

Tulppo, Mikko P., Arto J. Hautala, Timo H. Ma¨kikallio, Raija T. Laukkanen, Seppo Nissila¨, Richard L. Hughson, and Heikki V. Huikuri. Effects of aerobic training on heart rate dynamics in sedentary subjects. J Appl Physiol 95: 364–372, 2003. First published March 21, 2003; 10.1152/japplphysiol.00751.2002.—This study was designed to assess the effects of moderate- and high-volume aerobic training on the time domain and on spectral and fractal heart rate (HR) variability indexes. Sedentary subjects were randomized into groups with moderate-volume training (n ⫽ 20), high-volume training (n ⫽ 20), and controls (n ⫽ 15). The training period was 8 wk, including 6 sessions/wk at an intensity of 70–80% of the maximum HR, lasting for 30 min/session in the moderate-volume group and 60 min/session in the high-volume group. Time domain, frequency domain, and short-term fractal scaling measures of HR variability were analyzed over a 24-h period. Mean HR decreased from 70 ⫾ 7 to 64 ⫾ 8 beats/min and from 67 ⫾ 5 to 60 ⫾ 6 beats/min (P ⬍ 0.001 for both) for the moderate- and highvolume training groups, respectively. The normalized highfrequency spectral component increased in both groups (P ⬍ 0.05). The normalized low-frequency component decreased significantly (P ⬍ 0.05), resulting in a marked decrease in low frequency-to-high frequency ratio in both groups. In addition, short-term scaling exponent decreased in both groups (P ⬍ 0.001). There were no significant differences in the changes of HR variability indexes between groups. Aerobic training in sedentary subjects results in altered autonomic regulation of HR toward vagal dominance. A moderate training volume is a sufficient intervention to induce these beneficial effects. heart rate variability; cardiovascular regulation; detrended fluctuation analysis; R-wave-R-wave interval

TIME AND FREQUENCY DOMAIN ANALYSES of heart rate (HR) variability are the noninvasive methods most commonly used to evaluate the autonomic regulation of HR. Long-term aerobic exercise results in changes in cardiac autonomic regulation or in intrinsic HR that are observed in the form of a decrease in the mean HR and an increase in the total variance of HR variability (1, 28, 42, 43, 46). The decreased HR level and consequent increased total variance of R-wave-R-wave (R-R) intervals, however, detract from the applicability of HR

Address for reprint requests and other correspondence: M. P. Tulppo, Merikoski Rehabilitation and Research Center, Kasarmintie 13, 90100 Oulu, Finland (E-mail: [email protected]). 364

variability methods on the basis of the magnitude of variance to the evaluation of altered cardiac autonomic regulation. Therefore, variables that are more independent of differences in HR levels, such as normalized spectral components and a coefficient of component variance (CCV%), might be more useful in detecting changes in autonomic regulation (16, 17, 34), but these indexes have not been studied after physical training in healthy sedentary subjects. Furthermore, new analysis techniques have also been developed to probe features in HR behavior that are not detectable by the traditional HR variability methods based on moment statistics. These include analysis of fractal scaling exponents by detrended fluctuation analysis (DFA), which reveals the scaling properties of R-R interval data (38). There are no controlled studies comparing the effects of moderate- and high-volume training on HR dynamics among sedentary subjects. The present study was designed to examine the effects of aerobic training and training volume on time domain and on spectral and fractal HR dynamics over a 24-h period in a population of healthy sedentary men. METHODS

Subjects. Subjects were recruited by advertising in a newspaper, which attracted 85 replies. Subjects were interviewed with a standardized scheme to ascertain their medical histories and levels of physical activity. All smokers, subjects with high body mass index (⬎30), subjects who did regular physical training more than twice a week, and those with diabetes mellitus, asthma, or cardiovascular disorders were excluded. We invited 60 male subjects to our laboratory (Merikoski, Oulu, Finland) for a more specific assessment of physical status and excluded two subjects on account of various relative contraindications for a maximal exercise test. We tested 58 subjects and excluded 3 from the final analysis because of the number of ectopic beats during data acquisition. Finally, 55 male subjects were included in the training study (Table 1). Subjects were randomized into a moderate-volume training group (n ⫽ 20), a high-volume training group (n ⫽ 20), and a control group (n ⫽ 15). The protocol was approved by the ethics committee of the Merikoski Rehabilitation and

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

8750-7587/03 $5.00 Copyright © 2003 the American Physiological Society

http://www.jap.org

365

AEROBIC TRAINING AND R-R INTERVAL DYNAMICS

˙ O2peak, HRmax, and Pacemax Table 1. Effects of training on diastolic and systolic blood pressure at rest, V Moderate Training (n ⫽ 19) Pre

Age, yr Min–max, yr Height, m Weight, kg BMI Min–max, yr Fat, % BPdia, mmHg BPsys, mmHg ˙ O2peak V l/min ml 䡠 kg⫺1 䡠 min⫺1 HRmax, beats/min Pacemax, km/h

High Training (n ⫽ 16)

Post

Pre

35 ⫾ 10 23–52 1.81 ⫾ 0.05 82 ⫾ 11 25 ⫾ 3 22–30 16 ⫾ 4 78 ⫾ 7 127 ⫾ 13 3.3 ⫾ 0.4 41 ⫾ 4 189 ⫾ 9 11.5 ⫾ 1.3

Control (n ⫽ 11)

Post

Pre

Post

81 ⫾ 11 25 ⫾ 3 21–30 16 ⫾ 3 79 ⫾ 6 124 ⫾ 9

35 ⫾ 10 24–50 1.80 ⫾ 0.05 79 ⫾ 9 25 ⫾ 2 20–28 17 ⫾ 4 77 ⫾ 8 126 ⫾ 9

79 ⫾ 8 24 ⫾ 2 20–27 16 ⫾ 4 82 ⫾ 5* 128 ⫾ 11

36 ⫾ 11 23–52 1.82 ⫾ 0.07 81 ⫾ 9 25 ⫾ 3 21–28 16 ⫾ 3 79 ⫾ 7 128 ⫾ 10

81 ⫾ 9 25 ⫾ 3 21–29 16 ⫾ 4 78 ⫾ 4 125 ⫾ 8

3.6 ⫾ 0.4‡ 45 ⫾ 4‡ 184 ⫾ 8† 12.9 ⫾ 1.1‡

3.4 ⫾ 0.4 42 ⫾ 5 187 ⫾ 8 12.3 ⫾ 1.4

3.7 ⫾ 0.5‡ 46 ⫾ 6‡ 180 ⫾ 10‡ 13.6 ⫾ 1.3‡

3.4 ⫾ 0.3 41 ⫾ 4 185 ⫾ 15 11.6 ⫾ 0.9

3.4 ⫾ 0.4 41 ⫾ 4 184 ⫾ 17 11.6 ⫾ 0.8

˙ O2peak, peak O2 consumption; HRmax, maximal heart rate; Pacemax, maximal pace in treadmill; min–max, Values are means ⫾ SD. V minimum to maximum; BMI, body mass index; BPdia, diastolic blood pressure; BPsys, systolic blood pressure. * P ⬍ 0.05; † P ⬍ 0.01; ‡ P ⬍ 0.001: comparisons are between pretraining (Pre) and posttraining (Post) values.

Research Center, and all subjects gave their written, informed consent. Protocol. Subjects were not allowed to eat or drink coffee for 2 h before the exercise test, and all physical exercise and alcohol were forbidden for 48 h before the day of testing. A 24-h ambulatory R-R interval recording was started 25 h before the exercise test. Before the exercise test, subjects remained in a supine position in a quiet room for 10 min, after which blood pressure was measured in a supine position by manual auscultation. Fat percentage was calculated from four skinfolds (sites: biceps, triceps, subscapular, and suprailiac) according to the equation described by Durnin and Womersley (11). All tests were performed at the same time of day before and after the training period. Measurement of peak O2 consumption. Subjects performed a graded maximal exercise test on a treadmill (Telineyhtyma¨ , Kotka, Finland), starting at 4.5 km/h followed by a work rate increase at a rate of 0.5 km/h every 2 min until voluntary exhaustion. Minute ventilation and gas exchange (M909 ergospirometer, Medikro, Kuopio, Finland) were monitored continuously during the ramp protocol. Minute ventilation and gas exchange were calculated on a breath-by-breath basis and reported as mean values for 1 min. The highest value of O2 consumption measured during the test was used as the peak O2 consumption. Measurement of R-R intervals. The R-R intervals were recorded over 24 h with Polar R-R Recorder (Polar Electro, Kempele, Finland) at an accuracy of 1 ms (41) and saved in a computer for further analysis of HR variability with Hearts software (Heart Signal, Kempele, Finland). All R-R intervals were edited by visual inspection based on electrocardiogram (ECG) portions, to exclude all the undesirable beats, which accounted for ⬍2% in every subject. The measures of R-R interval dynamics were calculated from the entire 24-h recording and also separately for the hours representing nighttime (midnight to 6 AM) and daytime (9 AM to 6 PM) hours to detect possible diurnal differences (21, 40). Subjects were asked to go to bed before midnight and to stay in bed until 6 AM on the R-R interval recording days. The R-R intervals were recorded during a nonexercise day before and after training intervention. At the end of training intervention, the R-R intervals were recorded after a 48-h nonexercise period. Time and frequency domain analysis of HR variability. The mean length of R-R intervals and the standard deviation of all R-R intervals were used as time-domain measures of J Appl Physiol • VOL

HR variability. An autoregressive model (order 20) was used to estimate the power spectrum densities of R-R interval variability. Ultralow-frequency power (⬍0.0033 Hz) and very low-frequency (LF) power (0.0033–0.04 Hz) were calculated from the entire 24-h segment. LF and high-frequency (HF) power values were calculated from the segments of 512 R-R intervals over 24-h recording (48). LF and HF were also calculated from 1-h segments of the 24-h recording (by using segments of 512 R-R intervals), and the mean values of these segments were used to detect day and night HR variability values (40). The spectral values are expressed as absolute values (ln; ms2/Hz) and in normalized units (nu), which were obtained by dividing the power of each component by total variance from which the very LF component had been subtracted and multiplying this value by 100 (34). The CCV% were calculated as in the following equation: CCV(%) ⫽ 100 䡠 (power of component)1/2/(mean R-R interval) (16). The frequency at the HF power peak was analyzed for every hour (from the segments of 512 R-R intervals), and the mean values were calculated separately for day and night to evaluate the effects of training on average breathing frequency. Fractal analysis of R-R intervals. DFA quantifies the fractal correlation properties of the R-R interval data (38). The root-mean-square fluctuations of the integrated and detrended data are measured in observation windows of different sizes and then plotted against the size of the window on a log-log scale (Fig. 1). The scaling exponent ␣ represents the slope of this line, which relates (log)fluctuation to (log)window size. In this study, the short-term (from 4 to 11 beats) scaling exponent (␣1) was used on the basis of previous experiments (32). The scaling exponent was calculated from 1-h segments of the 24-h recording, and the mean value of these segments was used (40). The power-law relationship of R-R interval variability (␤) was also calculated from the frequency range of 10⫺4 to 10⫺2 by a previously described method to assess the longer-term correlation of HR (4). Training program. The training period was 8 wk, including six 30-min sessions a week for the moderate-volume training group and six 60-min sessions a week for the highvolume training group at an intensity of 70–80% of maximum HR. The training mode was walking and jogging. All subjects wore a telemetric HR monitor (Polar Smart Edge, Polar Electro) during all training sessions to reach and stay at the correct training intensity. Subjects also recorded the

95 • JULY 2003 •

www.jap.org

366

AEROBIC TRAINING AND R-R INTERVAL DYNAMICS

Fig. 1. Representative examples of tachograms (A), power spectra analyzed in 4-h segments (B), and fractal scaling exponents (␣1) calculated by the detrended fluctuation analysis (DFA) method (C) before (left) and after (right) a moderate-volume training intervention. HR, heart rate; bpm, beats/ min; SDNN, standard deviation of all R-wave-R-wave (R-R) intervals; HFnu, normalized high frequency; LFnu, normalized low frequency; LF/HF, low frequency-to-high frequency ratio.

duration and average training HR of each session. An extra resting day once a week was allowed if they felt exhausted. Statistical methods. Results are expressed as means ⫾ SD. The normal Gaussian distribution of the data was verified by the Kolmogorov-Smirnov goodness-of-fit test. The spectral values of 24-h HR variability were skewed. Therefore, these data were transformed by taking the natural logarithms of the absolute values. The differences within the groups after training were analyzed by a two-factor ANOVA with time and interventions followed by post hoc analysis (Student’s paired t-test). Pearson’s correlation analysis was performed on changes in the different HR variability parameters from baseline to the end of intervention. A P value of ⬍0.05 was considered significant. RESULTS

Training. One subject in the moderate-volume training group and four subjects in the high-volume group dropped out from the training because of a lack of motivation or leg problems. The final analysis included 19 subjects in the moderate-volume, 16 subjects in the high-volume, and 12 subjects in the control group. The average amount of training was 5.6 ⫾ 0.4 and 5.7 ⫾ 0.3 sessions/wk at an intensity of 76 ⫾ 2 and 75 ⫾ 1% of maximal HR for the moderate- and high-volume groups, respectively (P ⫽ not significant for both). The duration of training was 32 ⫾ 3 and 61 ⫾ 4 min/session for the moderate- and high-volume groups, respectively (P ⬍ 0.001). Eight weeks of both moderate- and high-volume training caused an increase in peak O2 consumption as well as a reduction in maximal HR. The diastolic blood pressure changed significantly in J Appl Physiol • VOL

the high-volume training group (Table 1). None of the measured variables changed within the control group during the study. Time and frequency domain analysis over a 24-h period. Representative examples of the effects of training on 24-h spectral HR variability are shown in Fig. 1B, and all mean values are shown in Tables 2 and 3. Mean HR decreased significantly for the moderate- and high-volume groups (P ⬍ 0.001; Fig. 2A). The mean value of the standard deviation of all R-R intervals increased for both training groups (P ⬍ 0.001). The absolute spectral values of HF, LF, and very LF power showed significant (P ⬍ 0.01) and similar increases for both training groups after the interventions. The mean value of ultralow-frequency power did not change after the training. HF power normalized by the average R-R interval (CCV%) increased in both training groups after the interventions (P ⬍ 0.01; Fig. 2B), but LF power (CCV%) only increased significantly in the moderatevolume training group (P ⬍ 0.05). The LF/HF ratio decreased in both training groups (P ⬍ 0.05; Fig. 2C). The mean value of normalized HF (HFnu) power increased in both training groups (P ⬍ 0.05). Similarly, the normalized LF (LFnu) power decreased significantly (P ⬍ 0.05). Scaling analysis of R-R intervals over a 24-h period. A representative example of the effects of training on 24-h short-term fractal HR behavior is shown in Fig. 1C, and all the mean values are shown in Table 3. The mean value of the short-term scaling exponent (␣1)

95 • JULY 2003 •

www.jap.org

367

AEROBIC TRAINING AND R-R INTERVAL DYNAMICS

Table 2. Effects of training on time and frequency domain HR variability measures over 24 h Moderate Training (n ⫽ 19) Pre

HR, beats/min 24 h Midnight–6 AM 9 AM–6 PM SDNN, ms 24 h HF, ln ms2 24 h Midnight–6 AM 9 AM–6 PM LF, ln ms2 24 h Midnight–6 AM 9 AM–6 PM VLF, ln ms2 24-h ULF, ln ms2 24 h LF/HF ratio 24 h Midnight–6 AM 9 AM–6 PM

High Training (n ⫽ 16)

Post

Pre

Control (n ⫽ 11)

Post

Pre

Post

70 ⫾ 7 55 ⫾ 7 78 ⫾ 8

64 ⫾ 8‡ 52 ⫾ 6‡ 71 ⫾ 8‡

67 ⫾ 5 54 ⫾ 4 72 ⫾ 7

60 ⫾ 6‡ 50 ⫾ 4‡ 66 ⫾ 7†

67 ⫾ 8 53 ⫾ 5 76 ⫾ 8

68 ⫾ 8 53 ⫾ 6 75 ⫾ 8

183 ⫾ 44

202 ⫾ 46†

189 ⫾ 36

207 ⫾ 33*

194 ⫾ 38

195 ⫾ 35

6.19 ⫾ 1.02 6.68 ⫾ 1.20 5.77 ⫾ 0.95

6.76 ⫾ 0.96‡ 7.21 ⫾ 0.97‡ 6.35 ⫾ 0.93‡

6.61 ⫾ 1.01 6.96 ⫾ 1.22 6.18 ⫾ 1.01

7.12 ⫾ 0.92‡ 7.49 ⫾ 1.01‡ 6.69 ⫾ 0.95‡

6.46 ⫾ 0.83 6.88 ⫾ 0.86 5.92 ⫾ 1.07

6.56 ⫾ 1.00 7.09 ⫾ 1.14 5.96 ⫾ 1.07

7.42 ⫾ 0.56 7.61 ⫾ 0.67 7.30 ⫾ 0.52

7.75 ⫾ 0.59‡ 7.93 ⫾ 0.67† 7.60 ⫾ 0.51‡

7.62 ⫾ 0.57 7.78 ⫾ 0.79 7.48 ⫾ 0.61

7.89 ⫾ 0.63‡ 8.19 ⫾ 0.75† 7.63 ⫾ 0.57*

7.47 ⫾ 0.60 7.72 ⫾ 0.55 7.15 ⫾ 0.75

7.47 ⫾ 0.63 7.85 ⫾ 0.64 7.13 ⫾ 0.75

7.79 ⫾ 0.59

8.14 ⫾ 0.53‡

7.92 ⫾ 0.55

8.34 ⫾ 0.48‡

7.83 ⫾ 0.54

7.86 ⫾ 0.61

10.0 ⫾ 0.69

10.1 ⫾ 0.55

10.0 ⫾ 0.33

10.2 ⫾ 0.49

10.0 ⫾ 0.58

10.0 ⫾ 0.56

4.1 ⫾ 2.4 3.3 ⫾ 2.1 5.5 ⫾ 3.0

3.1 ⫾ 1.5† 2.5 ⫾ 1.2* 4.1 ⫾ 2.3‡

3.1 ⫾ 1.6 2.5 ⫾ 1.3 4.2 ⫾ 2.0

2.4 ⫾ 1.1† 2.2 ⫾ 1.2 2.9 ⫾ 1.6†

2.9 ⫾ 1.1 2.8 ⫾ 1.8 3.7 ⫾ 1.4

2.7 ⫾ 1.1 2.7 ⫾ 2.0 3.4 ⫾ 1.2

Values are means ⫾ SD. HR, heart rate; SDNN, standard deviation of R-wave-R-wave intervals; HF, high frequency; LF, low frequency; VLF, very LF; ULF, ultralow frequency. * P ⬍ 0.05; † P ⬍ 0.01; ‡ P ⬍ 0.001: comparisons are between Pre and Post values.

decreased for both training groups (P ⬍ 0.001; Fig. 2D). The mean value of ␤, the power law scaling exponent, did not change during the intervention. Effects of training on the HF spectral peak. The frequency at the HF spectral peak did not change after the intervention (moderate-volume group: from 0.25 ⫾ 0.02 to 0.25 ⫾ 0.02 Hz at night and from 0.32 ⫾ 0.03 to

0.31 ⫾ 0.03 Hz during daytime; high-volume group: from 0.23 ⫾ 0.02 to 0.23 ⫾ 0.03 Hz at night and from 0.30 ⫾ 0.03 to 0.30 ⫾ 0.03 Hz during daytime; P ⫽ not significant). Correlation between the ␣1 and normalized spectral components. The change in the scaling exponent ␣1 from the baseline to the end of intervention correlated

Table 3. Effects of training on normalized frequency domain HR variability measures and scaling exponents over 24 h Moderate Training (n ⫽ 19) Pre

HF, nu 24 h Midnight–6 AM 9 AM–6 PM LF, nu 24 h Midnight–6 AM 9 AM–6 PM HF, CCV% 24 h Midnight–6 AM 9 AM–6 PM LF, CCV% 24 h Midnight–6 AM 9 AM–6 PM ␣1 24 h Midnight–6 AM 9 AM–6 PM ␤ 24 h

Post

High Training (n ⫽ 16) Pre

Post

Control (n ⫽ 11) Pre

Post

24 ⫾ 12 30 ⫾ 15 19 ⫾ 10

28 ⫾ 11† 33 ⫾ 12 24 ⫾ 11†

28 ⫾ 11 31 ⫾ 10 23 ⫾ 10

32 ⫾ 10† 34 ⫾ 10 29 ⫾ 10‡

27 ⫾ 8 31 ⫾ 13 23 ⫾ 7

29 ⫾ 10 33 ⫾ 14 24 ⫾ 8

76 ⫾ 12 70 ⫾ 15 81 ⫾ 10

72 ⫾ 11† 67 ⫾ 12 76 ⫾ 11†

72 ⫾ 11 69 ⫾ 10 77 ⫾ 10

68 ⫾ 10† 66 ⫾ 10 71 ⫾ 10‡

73 ⫾ 8 69 ⫾ 12 77 ⫾ 7

71 ⫾ 10 67 ⫾ 13 75 ⫾ 8

2.82 ⫾ 1.17 2.81 ⫾ 1.31 2.44 ⫾ 1.00

3.40 ⫾ 1.40† 3.40 ⫾ 1.41† 3.00 ⫾ 1.18†

3.31 ⫾ 1.70 3.32 ⫾ 1.88 2.85 ⫾ 1.36

3.88 ⫾ 1.81‡ 3.91 ⫾ 1.92† 3.38 ⫾ 1.50†

3.06 ⫾ 1.15 2.97 ⫾ 1.09 2.68 ⫾ 1.21

3.32 ⫾ 1.56 3.45 ⫾ 1.70 2.78 ⫾ 1.35

4.87 ⫾ 1.00 4.18 ⫾ 0.97 5.05 ⫾ 1.08

5.25 ⫾ 1.11† 4.73 ⫾ 1.27* 5.30 ⫾ 0.97

5.10 ⫾ 1.48 4.61 ⫾ 1.78 5.16 ⫾ 1.41

5.41 ⫾ 1.67 5.28 ⫾ 1.98† 5.10 ⫾ 1.28

4.87 ⫾ 1.22 4.28 ⫾ 0.93 4.71 ⫾ 1.40

4.90 ⫾ 1.37 4.66 ⫾ 1.36 4.67 ⫾ 1.49

1.36 ⫾ 0.16 1.25 ⫾ 0.20 1.44 ⫾ 0.14

1.27 ⫾ 0.15‡ 1.18 ⫾ 0.17* 1.33 ⫾ 0.17‡

1.31 ⫾ 0.15 1.22 ⫾ 0.17 1.36 ⫾ 0.15

1.21 ⫾ 0.14‡ 1.14 ⫾ 0.15‡ 1.25 ⫾ 0.14‡

1.31 ⫾ 0.10 1.23 ⫾ 0.16 1.36 ⫾ 0.09

1.29 ⫾ 0.13 1.20 ⫾ 0.20 1.34 ⫾ 0.11

⫺1.15 ⫾ 0.14

⫺1.17 ⫾ 0.12

⫺1.13 ⫾ 0.11

⫺1.12 ⫾ 0.13

⫺1.14 ⫾ 0.16

⫺1.11 ⫾ 0.13

Values are means ⫾ SD. nu, Normalized units; CCV%, coefficient of component variance; ␣1, fractal scaling exponent; ␤, power law scaling exponent. * P ⬍ 0.05; † P ⬍ 0.01; ‡ P ⬍ 0.001: comparisons are between Pre and Post values. J Appl Physiol • VOL

95 • JULY 2003 •

www.jap.org

368

AEROBIC TRAINING AND R-R INTERVAL DYNAMICS

Fig. 2. Mean ⫾ SD and individual values analyzed over 24 h for HR (A), high-frequency (HF) power normalized by the average R-R interval [coefficient of component variance (CCV%)] (B), LF/HF (C), and ␣1 (D). Pre, pretraining; Post, posttraining. The differences within the groups after training were evaluated by Student’s paired t-test. NS, not significant.

moderately with the changes in HFnu (r ⫽ ⫺0.76, P ⬍ 0.001), LFnu (r ⫽ 0.76, P ⬍ 0.001), and the LF/HF ratio (r ⫽ 0.69, P ⬍ 0.001) as analyzed over the 24-h recording. The correlation between the change in ␣1 and the change in the LF/HF ratio became weaker during the daytime hours (r ⫽ 0.57, P ⬍ 0.001) compared with the nighttime correlation (r ⫽ 0.68, P ⬍ 0.001). DISCUSSION

There were two main findings in this study. First, moderate-volume training results in similar changes in the average HR and HR variability indexes as in highvolume training. Second, the autonomic regulation, analyzed as nu by spectral analysis of HR variability, changed toward increased vagal dominance after aerobic training in both high- and moderate-volume training groups. Similarly, aerobic training changed the J Appl Physiol • VOL

short-term HR behavior (⬍11 beats) toward less correlated short-term of R-R interval dynamics. Methodological considerations for HR variability analysis. A decreased mean HR is normally associated with an increased total variance of R-R interval fluctuation and increases the absolute HF and LF spectral components (16). Therefore, the time domain HR variability indexes and the absolute spectral components may not be ideal methods for detecting changes in autonomic regulation after physical training interventions, where a decrease in basal HR occurs. In the present study, too, the time domain and absolute spectral calculations of HR variability are influenced by decreased HR because of the long-term aerobic training. The absolute spectral components are not able to indicate which physiological mechanisms, the enhancement in vagal tone and/or the decreased sympathetic activity, are responsible for the altered HR after training. The normalized spectral values and LH/HF ratio represent the relative changes in the characteristics of the HR fluctuation pattern rather than the magnitude of HR variability (34). Therefore, the nu were used in the present study to assess the effects of training on the autonomic regulation of HR. Spectral components normalized by the average R-R interval length (CCV%) have also been used to evaluate the altered autonomic regulation in various pharmacological and clinical settings, where differences in the HR levels disturb the comparison of the absolute spectral values (16–18). Specifically, the HF power (CCV%) showed a good linear correlation with pharmacologically assessed cardiac vagal tone (16). The DFA technique is a method to detect qualitative rather than quantitative changes in HR dynamics. From the mathematical point of view, spectral measures are associated with fractal scaling measures when analyzed as nu during strictly controlled external conditions, because both describe relative changes in the characteristics of HR fluctuations rather than the magnitude of HR variability, and, therefore, these indexes are not related to changes in the average HR. Indeed, our laboratory has recently reported (49, 52) a relationship between ␣1 values and the LF/HF ratio in controlled external situations with a fixed respiratory rate. However, the relationship is weak during “freerunning” ambulatory ECG recordings (31, 40, 49) because measurement of scaling exponents by the DFA method provide information on the scaling properties of HR fluctuations over several segmented time windows, whereas conventionally computed spectral ratios describe HR fluctuations only in two predetermined time windows. Therefore, both normalized spectral units of HR variability and fractal scaling exponent were used in this study to describe changes in HR dynamics caused by exercise training. From the physiological point of view, we have described the altered short-term fractal HR behavior in various pharmacological and physiological interventions (49, 52). Briefly, a change in the R-R interval dynamics from fractal behavior (␣1 of ⬃1.0) toward a stronger

95 • JULY 2003 •

www.jap.org

AEROBIC TRAINING AND R-R INTERVAL DYNAMICS

correlation in short-term R-R interval dynamics (␣1 of ⬃1.5) was observed during a withdrawal of vagal activity by graded atropine infusion and dynamic low-intensity exercise (49, 52). Similarly, in the present study, the reduced ␣1 values correspond to increased vagal activity of the heart after aerobic training. Effects of training on HR dynamics. Bradycardia is a well-known consequence of aerobic training and has been attributed to changes in the autonomic nervous system, i.e., either an increase in parasympathetic activity or a decrease in sympathetic activity, or else to a reduced intrinsic HR. Good aerobic fitness has been suggested to be associated with increased HR variability, especially with vagally mediated respiratory sinus arrhythmia, which is expressed as various time domain indexes or as absolute spectral HF power (8–10, 13, 14). However, some studies have failed to show any association between good aerobic fitness and enhanced HR variability (7, 26). The results of prospective, wellcontrolled interventions are also controversial. All of the studies have indicated decreased HR, and most have revealed increased vagal activity measured by absolute HF power after aerobic training (1, 28, 42–44, 46). Contrariwise, some controlled studies have failed to show any association between aerobic training and HR variability (5, 29, 30). The results of these previous contradictory studies are partly limited by either the small sample size (26, 30) and subject characteristics, e.g., a narrow range in aerobic fitness (7). Also, the duration of interventions (from 6 to 36 wk), training frequency (from 3 to 7 sessions/wk), and the intensity of exercise (from ⬃60 to 90% of maximal HR) varies between the previous studies, and thus the results of different studies are difficult to compare. However, long-duration and/or intense training may not necessarily lead to greater enhancement in HR variability as prolonged (12 mo) and intense training (from walking to running) may return these changes in HR variability back to the control level (22). The training sessions performed during the previous days also affect HR variability and autonomic function (2, 12, 15). In our laboratory’s recent study (15), the autonomic function recovered back to the baseline level at the second day after strenuous aerobic exercise. Therefore, 2 days of nonexercise were used in the present study before R-R intervals were recorded. Several other methodological issues may also explain the controversial results. First, in our populationbased cross-sectional study, good aerobic fitness was associated with an augmentation of vagally mediated beat-to-beat R-R interval fluctuation during a controlled exercise but not at controlled rest (51). This suggests that short-term HR variability recording at rest (⬃10 min) is not an ideal method for evaluating changes in autonomic regulation. Although the spectral HR variability methods with short-term recording at rest may not be sensitive enough to detect subtle changes in autonomic regulation after training, standard spectral methods over 24 h also failed to show any association between aerobic training and spectral HR variability indexes (26, 29). Second, sampling freJ Appl Physiol • VOL

369

quency is important for an accurate detection of beatto-beat fluctuations in R-R intervals, as the precision of sampling of 8 vs. 1 ms resulted in a significant difference in the values of HR variability indexes, e.g., the difference was 14% for the LF/HF ratio in a 24-h recording (47). In the present study, the R-R intervals were recorded with a real-time microprocessor QRS detector system with 1-ms timing accuracy (41). In the present study, the increased HFnu power and the decreased LFnu power as well as the decreased LF/HF ratio showed an alteration in the autonomic regulation of HR toward increased vagal dominance after aerobic training. The LF/HF ratio decreased also during the night hours after moderate-volume training intervention. This is an important finding because the night hours may reflect a more standardized condition and a subject’s behavior pattern does not disturb the R-R interval recordings. Concurrently, the present data also showed a decreased short-term scaling exponent ␣1 in R-R interval dynamics during the freerunning daytime hours as well as during the night hours after aerobic training. The normalized spectral components remained unaltered during the night hours after a high-volume training period, suggesting that the scaling indexes detect subtle alterations in HR more effectively than simple spectral ratios. In the previous pharmacological studies, both HFnu power indexes (CCV% and nu) are associated with vagal activity (16, 33). Also in the present study, both HFnu power indexes increased similarly as evidence of increased vagal activity after aerobic training. The opposite changes occur in the LF power normalized by average R-R interval (CCV%) and LFnu after aerobic training. These changes are concurrent with pharmacological studies where LF (CCV%) power decreased significantly during atropine infusion, revealing that this index is partly related to vagal activity (16). On the

Fig. 3. Mean ⫾ SD and individual values for systolic (A) and for diastolic blood pressure (BP) (B). The differences within the groups after training were evaluated by Student’s paired t-test.

95 • JULY 2003 •

www.jap.org

370

AEROBIC TRAINING AND R-R INTERVAL DYNAMICS

contrary, LFnu is related to sympathetic activity studied by muscle sympathetic activity recordings from the peroneal nerve (33, 37), and decreased LFnu in the present study may therefore indicate a decreased sympathetic activity after aerobic training. Although not a focus of this study, it was observed that both moderate- and high-volume training groups had reductions of 5–7 beats/min in maximal HR. This could have been a consequence of altered autonomic balance, although vagal activity is elevated only at lower HR. Alternatively, a reduction in intrinsic HR and altered sinoatrial node responses to norepinephrine have been identified as potential contributors to reduced maximal HR after training (19). This might suggest that altered autonomic balance and reduced intrinsic HR contribute to the changes that we observed in HR after training. Another interesting finding is an increased diastolic blood pressure after intervention for the high-volume training group (Fig. 3). This may be caused by an early stage of over-training usually referred to as “overreaching” syndrome. However, specific and sensitive parameters are not available to identify overreaching syndrome, and we did not focus on that issue in the present study (25). Study limitations. Endurance training produces well-known functional improvements in cardiac performance. An enlarged blood volume, together with greater ventricular compliance and distensibility (27), may influence HR variability with changes of blood pressure variability as the input signal of the baroreflex system. In conscious dogs, changes in cardiac mechanics affect blood pressure variability, especially in HF by influencing stroke volume (36). Furthermore, the tidal volume becomes larger, and breathing frequency is considerably reduced after training interventions, at least in submaximal exercise intensity levels (23, 53). The reduced breathing frequency is known to increase the absolute spectral values of HF power of R-R intervals (39). Therefore, changes in the respiratory components of HR variability, reflecting cardiac outflow, may also be due to altered respiratory patterns caused by physical training. Consequently, the HF power of the R-R interval may not be an ideal index to estimate vagal activity if the cardiac mechanical properties and breathing pattern are changed. The effects of altered cardiac mechanics and respiration on fractal scaling properties of HR fluctuation are not well known. Therefore, more studies of the physiological background, particularly of the effects of respiratory patterns, on the fractal HR dynamics are warranted. Analysis of cardiac autonomic regulation from 24-h recordings by frequency domain as well as by other measures may also have potential biases. Most importantly, many indexes may be influenced by physical activity during recordings (3). However, several studies have shown good short-term and long-term reproducibility of 24-h recordings when measured from surface ECGs (20, 24, 35). Furthermore, 24-h recordings have appeared to be a reliable and practical method that is not dependent on a subject’s cooperation and does not involve any placebo effect on the results (45, J Appl Physiol • VOL

48). This may result from a rather equal level of physical activity in individual subjects between the different 24-h periods. Therefore, 24-h HR variability measurement can be considered as an ideal method for assessing temporal changes in the autonomic regulation in individual subjects. In fact, reproducibility of 24-h HR variability measures may be even better than assessment of HR dynamics from short-term recordings during controlled situations (6, 50). In conclusion, the reduction of HR after aerobic training is associated with increased indexes of cardiac vagal outflow in healthy sedentary males. An 8-wk aerobic training intervention, including six 30-min sessions a week at an intensity of 70–80% of maximum HR, is a sufficient intervention to induce these effects on cardiac autonomic regulation. The authors appreciate the technical and financial support received from Polar Electro (Kempele, Finland) and the generous help from Heart Signal (Kempele, Finland). This research was funded by grants from the Ministry of Education (Helsinki, Finland), the Heart and Stroke Foundation of Ontario (no. T-3081, Toronto, Canada), and the Medical Council of the Academy of Finland (Helsinki, Finland). REFERENCES 1. Al-Ani M, Munir SM, White M, Townend J, and Coote JH. Changes in R-R variability before and after endurance training measured by power spectral analysis and by the effect of isometric muscle contraction. Eur J Appl Physiol Occup Physiol 74: 397–403, 1996. 2. Bernardi L, Passino C, Robergs R, and Appenzeller O. Acute and persistent effects of a 46-kilometer wilderness trail run at altitude: cardiovascular autonomic modulation and baroreflexes. Cardiovasc Res 34: 273–280, 1997. 3. Bernardi L, Valle F, Coco M, Calciati A, and Sleight P. Physical activity influences heart rate variability and very-lowfrequency components in Holter electrocardiograms. Cardiovasc Res 32: 234–237, 1996. 4. Bigger JT Jr, Steinman RC, Rolnitzky LM, Fleiss JL, Albrecht P, and Cohen RJ. Power law behavior of RR interval variability in healthy middle-aged persons, patients with recent acute myocardial infarction, and patients with heart transplants. Circulation 93: 2142–2151, 1996. 5. Boutcher SH and Stein P. Association between heart rate variability and training response in sedentary middle-aged men. Eur J Appl Physiol 70: 75–80, 1995. 6. Breuer HW, Skyschally A, Wehr M, Schulz R, and Heusch G. Poor reproducibility of parameters of heart rate variations. Z Kardiol 81: 475–481, 1992. 7. Byrne EA, Fleg JL, Vaitkevicius PV, Wright J, and Porges SW. Role of aerobic capacity and body mass index in the ageassociated decline in heart rate variability. J Appl Physiol 81: 743–750, 1996. 8. Davy KP, Miniclier NL, Taylor JA, Stevenson ET, and Seals DR. Elevated heart rate variability in physically active postmenopausal women: a cardioprotective effect? Am J Physiol Heart Circ Physiol 271: H455–H460, 1996. 9. Davy KP, Willis WL, and Seals DR. Influence of exercise training on heart rate variability in post-menopausal women with elevated arterial blood pressure. Clin Physiol 17: 31–40, 1997. 10. De Meersman RE. Heart rate variability and aerobic fitness. Am Heart J 125: 726–731, 1993. 11. Durnin J and Womersley J. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br J Nutr 32: 77–97, 1974. 12. Furlan R, Piazza S, Dell’Orto S, Gentile E, Cerutti S, Pagani M, and Malliani A. Early and late effects of exercise

95 • JULY 2003 •

www.jap.org

AEROBIC TRAINING AND R-R INTERVAL DYNAMICS

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25. 26. 27.

28.

29.

30.

31.

and athletic training on neural mechanisms controlling heart rate. Cardiovasc Res 27: 482–488, 1993. Goldsmith RI, Bigger JT, Steinman RC, and Fleiss JL. Comparison of 24-hour parasympathetic activity in endurancetrained and untrained young men. J Am Coll Cardiol 20: 552– 558, 1992. Gregoire J, Tuck S, Yamamoto Y, and Hughson RL. Heart rate variability at rest and exercise: influence of age, gender, and physical training. Can J Appl Physiol 21: 455–470, 1996. Hautala A, Tulppo MP, Ma¨ kikallio TH, Laukkanen R, Nissila¨ S, and Huikuri HV. Changes in cardiac autonomic regulation after prolonged maximal exercise. Clin Physiol 21: 238– 245, 2001. Hayano J, Sakakibara Y, Yamada A, Yamada M, Mukai S, Fujinami T, Yokoyama K, Watanabe Y, and Takata K. Accuracy of assessment of cardiac vagal tone by heart rate variability in normal subjects. Am J Cardiol 67: 199–204, 1991. Hayano J, Sakakibara Y, Yamada M, Ohte N, Fujinami T, Yokoyama K, Watanabe Y, and Takata K. Decreased magnitude of heart rate spectral components in coronary artery disease. Its relation to angiographic severity. Circulation 81: 1217– 1224, 1990. Hayano J, Yamada A, Mukai S, Sakakibara Y, Yamada M, Ohte N, Hashimoto T, Fujinami T, and Takata K. Severity of coronary atherosclerosis correlates with the respiratory component of heart rate variability. Am Heart J 121: 1070–1079, 1991. Hughson RL, Sutton JR, Fitzgerald JD, and Jones NL. Reduction of intrinsic sinoatrial frequency and norepinephrine response of the exercised rat. Can J Physiol Pharmacol 55: 813–820, 1977. Huikuri HV, Kessler KM, Terracall E, Castellanos A, Linnaluoto MK, and Myerburg RJ. Reproducibility and circadian rhythm of heart rate variability in healthy subjects. Am J Cardiol 65: 391–393, 1990. Huikuri HV, Linnaluoto MK, Seppa¨ nen T, Airaksinen KE, Kessler KM, Takkunen JT, and Myerburg RJ. Circadian rhythm of heart rate variability in survivors of cardiac arrest. Am J Cardiol 70: 610–615, 1992. Iwasaki K, Zhang R, Zuckerman JH, and Levine BD. Association of cardiovascular variability with intensity of exercise training (Abstract). Med Sci Sports Exerc 32: S49, 2000. Jirka Z and Adamus M. Changes in ventilation equivalents in young people in the course of three years of training (Abstract). J Sports Med 5: 1, 1965. Kleiger RE, Bigger JT, Bosner MS, Chung MK, Cook JR, Rolnitzky LM, Steinman R, and Fleiss JL. Stability over time of variables measuring heart rate variability in normal subjects. Am J Cardiol 68: 626–630, 1991. Kuipers H. Training and overtraining: an introduction. Med Sci Sports Exerc 30: 1137–1139, 1998. Lazoglu AH, Glace B, Gleim GW, and Coplan NL. Exercise and heart rate variability. Am Heart J 131: 825–827, 1996. Levine BD, Lane LD, Buckey JC, Friedman DB, and Blomqvist CG. Left ventricular pressure-volume and FrankStarling relation in endurance athletes. Circulation 84: 1016– 1023, 1991. Levy WC, Cerqueira MD, Harp GD, Johannessen KA, Abrass IB, Schwartz RS, and Stratton JR. Effect of endurance exercise training on heart rate variability at rest in healthy young and older men. Am J Cardiol 82: 1236–1241, 1998. Loimaala A, Huikuri HV, Oja P, Pasanen M, and Vuori I. Controlled 5-mo aerobic training improves heart rate but not heart rate variability or baroreflex sensitivity. J Appl Physiol 89: 1825–1829, 2000. Maciel BC, Gallo L, Neto JAM, Filho ECL, Filho JT, and Manc¸ o JC. Parasympathetic contribution to bradycardia induced by endurance training in man. Cardiovasc Res 19: 642– 648, 1985. Ma¨ kikallio TH, Ristimae T, Airaksinen KEJ, Peng CK, Goldberger AL, and Huikuri HV. Heart rate dynamics in patients with stable angina pectoris and utility of fractal and complexity measures. Am J Cardiol 81: 27–31, 1998. J Appl Physiol • VOL

371

32. Ma¨ kikallio TH, Seppa¨ nen T, Airaksinen KEJ, Koistinen J, Tulppo MP, Peng CK, Goldberger AL, and Huikuri HV. Dynamic analysis of heart rate may predict subsequent ventricular tachycardia after myocardial infarction. Am J Cardiol 80: 779–783, 1997. 33. Montano N, Cogliati C, Porta A, Pagani M, Malliani A, Narkiewicz K, Abboud FM, Birkett C, and Somers VK. Central vagotonic effects of atropine modulate spectral oscillations of sympathetic nerve activity. Circulation 98: 1394–1399, 1998. 34. Montano N, Ruscone TG, Porta A, Lombardi F, Pagani M, and Malliani A. Power spectrum analysis of heart rate variability to assess the changes in sympathovagal balance during graded orthostatic tilt. Circulation 90: 1826–1831, 1994. 35. Nolan J, Flapan AD, Goodfield NE, Prescott RJ, Bloomfield P, Neilson JM, and Ewing DJ. Measurement of parasympathetic activity from 24-hour ambulatory electrocardiograms and its reproducibility and sensitivity in normal subjects, patients with symptomatic myocardial ischemia, and patients with diabetes mellitus. Am J Cardiol 77: 154–158, 1996. 36. O’Leary DS and Woodbury DJ. Role of cardiac output in mediating arterial blood pressure oscillations. Am J Physiol Regul Integr Comp Physiol 271: R641–R646, 1996. 37. Pagani M, Montano N, Porta A, Malliani A, Abboud FM, Birkett C, and Somers VK. Relationship between spectral components of cardiovascular variabilities and direct measures of muscle sympathetic nerve activity in humans. Circulation 95: 1441–1448, 1997. 38. Peng CK, Havlin S, Stanley HE, and Goldberger AL. Quantification of scaling exponents and crossover phenomena in nonstationary heartbeat time series. Chaos 5: 82–87, 1995. 39. Penttila¨ J, Helminen A, Jartti T, Kuusela T, Huikuri HV, Tulppo MP, Coffeng R, and Scheinin H. Time domain, geometrical and frequency domain analysis of cardiac vagal outflow: effects of various respiratory patterns. Clin Physiol 21: 365–376, 2001. 40. Pikkuja¨msa¨ SM, Ma¨kikallio TH, Sourander LB, Ra¨iha¨ IJ, Puukka P, Skytta¨ J, Peng CK, Goldberger AL, and Huikuri HV. Cardiac interbeat interval dynamics from childhood to senescence: comparison of conventional and new measures based on fractals and chaos theory. Circulation 100: 393– 399, 1999. 41. Ruha A, Sallinen S, and Nissila¨ S. A real-time microprocessor QRS detector system with a 1 ms timing accuracy for the measurement of ambulatory HRV. IEEE Trans Biomed Eng 44: 159–167, 1997. 42. Schuit AJ, van Amelsvoort LG, Verheij TC, Rijneke RD, Maan AC, Swenne CA, and Schouten EG. Exercise training and heart rate variability in older people. Med Sci Sports Exerc 31: 816–821, 1999. 43. Seals DR and Chase PB. Influence of physical training on heart rate variability and baroreflex circulatory control. J Appl Physiol 66: 1886–1895, 1989. 44. Shi X, Stevens GH, Foresman BH, Stern SA, and Raven PB. Autonomic nervous system control of the heart: endurance exercise training. Med Sci Sports Exerc 27: 1406–1406, 1995. 45. Stein PK, Bosner MS, Kleiger RE, and Conger BM. Heart rate variability: a measure of cardiac autonomic tone. Am Heart J 127: 1376–1381, 1994. 46. Stein PK, Ehsani AA, Domitrovich PP, Kleiger RE, and Rottman JN. Effect of exercise training on heart rate variability in older adults. Am Heart J 138: 567–576, 1999. 47. Tapanainen JM, Seppa¨nen T, Laukkanen RT, Loimaala A, and Huikuri. HV. Significance of the accuracy of RR interval detection for the analysis of new dynamic measures of heart rate variability. Ann Noninvasive Electrocardiol 4: 10–18, 1999. 48. 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 93: 1043– 1065, 1996. 49. Tulppo MP, Hughson RL, Ma¨kikallio TH, Airaksinen KEJ, Seppa¨nen T, and Huikuri HV. Effects of exercise and passive

95 • JULY 2003 •

www.jap.org

372

AEROBIC TRAINING AND R-R INTERVAL DYNAMICS

head-up tilt on fractal and complexity properties of heart rate dynamics. Am J Physiol Heart Circ Physiol 280: H1081–H1087, 2001. 50. Tulppo MP, Ma¨ kikallio TH, Seppa¨ nen T, Airaksinen KEJ, and Huikuri HV. Heart rate dynamics during accentuated sympathovagal interaction. Am J Physiol Heart Circ Physiol 274: H810–H816, 1998. 51. Tulppo MP, Ma¨ kikallio TH, Seppa¨ nen T, Laukkanen RT, and Huikuri HV. Vagal modulation of heart rate during exer-

J Appl Physiol • VOL

cise: effects of age and physical fitness. Am J Physiol Heart Circ Physiol 274: H424–H429, 1998. 52. Tulppo MP, Ma¨ kikallio TH, Seppa¨ nen T, Shoemaker K, Tutungi E, Hughson RL, and Huikuri HV. Effects of pharmagological adrenergic and vagal modulation on fractal heart rate dynamics. Clin Physiol 21: 515–523, 2001. 53. Tzankoff SP, Robinson S, Pyke FS, and Brawn CA. Physiological adjustments to work in older men as affected by physical training. J Appl Physiol 33: 346–350, 1972.

95 • JULY 2003 •

www.jap.org