Twenty-four subjects participated in the study. Subjects were divided into four age and gender groups (20M: 6 males 19.3 ± 0.5 yr; 20F: 6 females 20.2 ± 1.0 yr; 40M: 6 males 42.3 ± 2.2 yr; and 40F: 6 females 41.7 ± 1.9 yr). Each subject was medically examined before beginning the experiment to ensure their capability of participating safely in the rigorous training involved. Each subject read a complete description of the testing and training protocol, and the associated risks involved, and they then signed a consent form approved by the Office of Human Research (Simon Fraser University).
The experimental design required a subject’s physiological monitoring before and after a standardized/quantified, 12-wk training program. Each subject avoided strenuous physical activity in the 48 h before a physiological monitoring session. Each subject had adequate sleep and avoided caffeinated beverages before the monitoring session. During physiological testing, heart rate was measured continuously, with a three-lead electrocardiograph (Lifepak 8 Cardiac Monitor, Physio-Control, Redmond, WA). Heart rate was recorded in a quiet laboratory during 10 min of supine rest and 7 min of submaximal cycling at the beginning and end of the 12-wk endurance training program. During supine rest, each subject breathed at a rate of 15 breaths per minute in order to reduce the effect of respiration on HRV. The submaximal cycling work rate was set at 180 W for males and 150 W for females in order to produce a similar heart rate response in both genders.
Each subject was interviewed to ensure that they were capable of completing the training protocol. Subjects were all physically active and had been running regularly in the month before the start of the training program. The training program (Fig. 1) started with 4 wk of training and was followed by a 2-wk taper. Training sessions took place 4× wk−1 and consisted of runs that were 45–60 min in duration, at an intensity of 70–90% of a subject’s maximum heart rate. After 4 wk of training, there was a 2-wk taper period. The first week of taper was at 75% of the 4th training week, and the second week of taper was at 50% of the 4th training week. The second 4-wk block of training was 20% more than the first block. The 20% increase consisted of longer training sessions at a similar intensity. This second block of training was again followed by a second 2-wk taper. The first week of the second taper was 75% of the 10th training week, whereas the final week of taper was 50% of the 10th training week.
Quantification of training.
The training impulse (TRIMP) method was used to measure the total effort (dose) of each subject’s training session accurately (3). The training impulse is assessed from both the duration of effort, and its relative intensity, to determine the total effort of a training session. In the present study, each subject’s TRIMPS were calculated on a weekly basis to ensure that the appropriate amount of training was being completed. Average training heart rate for each training session was recorded via telemetry (Polar Vantage XL, Kempele, Finland) at 1-min intervals, in serial daily files for weekly downloading into a computer.
Once a week during the study, each subject performed a 2-mile criterion performance (CP) run. These timed, maximal intensity runs provided a measure of the relative relationship between accumulating fitness and fatigue during the training and taper periods. Each subject’s improvement in running performance after the 12-wk training program was determined from the difference in time between the criterion performance at the beginning and end of the training program (Table 1).
The analog output ECG data was processed through an A-D board (DATAPAC, Run Technologies, Rolling Hills, CA) at 1000 Hz and analyzed off-line. The ECG data was visually reviewed for anomalies and movement artifact. The beat-by-beat variability of heart rate was evaluated by coarse-graining spectral analysis (CGSA) to extract the fractal components from the harmonic components of the power spectrum (29).
In the present study, data length was set at 256 beats in order to analyze fractal and harmonic power at a constant data length and allow for a reasonable amount of time spent in each experimental condition. A data length of 256 beats lasted approximately 5 min in the resting condition and 3 min in the submaximal cycling condition. Each experimental condition was approximately twice as long as the 256-beat window in order to ensure a stable signal. This time interval allowed for comparison of results to the majority of other spectral analysis studies.
Spectral analysis decomposes a heart rate signal to its constituent frequency components and quantifies the relative power (squared amplitude) of the components. Spectral power can be calculated from a fast Fourier transform and used to provide information on the changes that occur in neural control of heart rate (29). From the harmonic component, total spectral power (PTot) and integrated power in both the low-frequency region (0.04–0.15 Hz, PLo) and high-frequency region (0.15–0.5 Hz, PHI) may be calculated. Given the frequency distribution of nervous system activity at the sinus node, indicators of parasympathetic and sympathetic activity, or sympathovagal balance, control of heart rate can be estimated (15). Conclusions drawn about sympathetic nervous system activity from low-frequency power are controversial due to disagreement about the low-frequency cut-off point and the influence of parasympathetic activity in the low-frequency region. Even with its limitations, judicious use of spectral analysis can provide useful insights into autonomic cardiovascular control (15).
Statistical analysis of variables was performed using the JMP-IN statistical package (SAS Institute Inc.). Total group (N = 24) mean and subgroup (N = 6) mean results from the resting and submaximal exercise experimental conditions in both pre- and posttraining conditions were determined. All data are quoted as group means ± standard deviation. A three-way repeated measures ANOVA was used to determine the main effects of training, age, and gender. Main effects and interactions (P < 0.05) are quoted as mean ± standard error of the mean.
All subjects improved their 2-mile criterion run time performance after training (Fig. 2). There was a group mean (N = 24) improvement in criterion run performance from the start of training to the end of the program (Table 1). A three-way ANOVA determined that the 20-yr-old group (1.07 ± 0.34 min) had a larger increase in criterion run performance compared with the 40-yr-old group (0.49 ± 0.33 min).
During supine rest, there was a total group mean (N = 24) decrease (P < 0.0001) in heart rate between pre- (1019.6 ± 181.0 ms, 58.8 ± 9.5 beats·min−1) and posttraining (1069.4 ± 171.6 ms, 56.1 ± 9.0 beats·min−1). Training, age, and gender interactions indicated that there was a decrease (P < 0.02) in heart rate for the 20F, 20M, and 40 M groups but not in the 40F group after training (Table 2).
After training there was a significant (P < 0.05) total group mean increase in total spectral power during supine rest. An example of the difference in spectral power between pre- and posttraining is illustrated in Figure 3. In addition to the training effects (Table 2), there were some main age effects for the HRV spectral analysis results during fixed breathing at rest. There was an age effect (P < 0.01) in the standard deviation of R-R interval. The 40-yr-old group showed a smaller (49.49 ± 6.06 ms) standard deviation of R-R interval compared with the 20-yr-old group (75.37 ± 6.06 ms). There was an age effect (P < 0.01) for high-frequency power (Fig. 4). The 40-yr-old group had a smaller HF power (108.69 ± 115.76 ms2) compared with the 20-yr-old group (644.47 ± 115.76 ms2). There was an age effect (P < 0.05) for total power after training. The 40-yr-old group had a smaller total power (628.09 ± 274.45 ms2) compared with the 20-yr-old group (1586.32 ± 274.45 ms2). The 40-yr-old group showed a smaller PNS indicator (0.15 ± 0.03) compared with the 20-yr-old group (0.38 ± 0.03). The 40-yr-old group had a larger SNS indicator (2.71 ± 0.71) compared with the 20-yr-old group (0.11 ± 0.71).
There were two significant age and training interactions. The 20-yr-old group had a large increase (P < 0.01) in HF power between pre- (362.26 ± 82.61 ms2) and post- (926.68 ± 82.61 ms2) training. The 20-yr-old group posttraining HF power (926.68 ± 82.61 ms2) was larger (P < 0.01) than the 40-yr-old group posttraining HF power (168.19 ± 82.61 ms2). The 20-yr-old group pretraining HF power (326.26 ± 82.61 ms2) was larger (P < 0.01) than the 40-yr-old group pretraining HF power (49.18 ± 82.61). The 20-yr-old group had a large increase (P < 0.01) in total power between pre- (1292.18 ± 146.69 ms2) and post- (2191.72 ± 146.69 ms2) training. It should be noted that the large increase in total power in the 20-yr-old group was mainly caused by the change in the females. The 20-yr-old group pretraining total power (1161.82 ± 59.85 ms2) was larger (P < 0.01) than the 40-yr-old group pretraining total power (528.94 ± 59.85 ms2). The 20-yr-old group posttraining total power (2010.82 ± 59.85 ms2) was larger (P < 0.001) than the 40-yr-old group posttraining total power (727.24 ± 59.85 ms2). The 20-yr-old group had a large increase (P < 0.01) in total power between pre- (1161.82 ± 59.85 ms2) and post- (2010.82 ± 59.85 ms2) training. The 40-yr-old group also had an increase (P < 0.05) between pre- (528.94 ± 59.85 ms2) and post- (727.24 ± 59.85 ms2) training. There were no significant gender and training interactions.
There were several training effects during submaximal exercise (Table 3). After training there was a significant (P < 0.05) total group mean increase in total spectral power during submaximal exercise. An example of the difference in spectral power between pre- and posttraining during submaximal exercise is illustrated in Figure 5. There were no age or gender main effects during submaximal exercise. There was a significant age and training interaction for heart rate during submaximal exercise. The 20-yr-old group had a decrease (P < 0.01) in submaximal heart rate between pre- (147.0 ± 0.6 beats·min−1) and post- (137.4 ± 0.6 beats·min−1) training. The 40-yr-old group also had a decrease (P < 0.01) in submaximal heart rate between pre- (144.3 ± 0.6 beats·min−1) and post- (138.2 ± 0.6 beats·min−1) training. The 20-yr-old group pretraining heart rate (147.0 ± 0.6 beats·min−1) was higher (P < 0.05) than the 40-yr-old group pretraining heart rate (144.3 ± 0.6 beats·min−1). However, there was no difference in posttraining heart rate between the two age groups.
In the present study, a standardized 12-wk training program produced positive physiological adaptations in the different age and gender groups. There was a significant total group mean decrease in heart rate both at rest (5%) and during submaximal exercise (6%) after training. After training total spectral power and high-frequency power increased in both age groups at rest. The endurance training thus improved cardiovascular function both at rest and during a submaximal exercise work rate.
The beneficial results of endurance training evidenced in this study support research of other studies that have indicated that an optimal training stimulus is a step increase above the level of training currently being practiced (2). Quantitative calculation of TRIMPS for each subject ensured that all subjects undertook the optimal amount of training and the most effective taper. A taper after heavy training has been found to be an effective way to ensure recovery and is generally accepted as an integral part of optimal preparation for competition (6). If a subject is monitored during a heavy training period, resting and exercise heart rate may be elevated, and heart rate variability power may be reduced. Thus, it is important to monitor the physiological benefits of an endurance-training program during a taper period.
The unique experimental design of this study attempted to match the fitness levels of older and younger subjects and then expose them to the same training program. The fastest time a subject could run 2 miles was used to establish initial fitness levels. The pretraining run times for the 20-yr-old males was very similar to the 40-yr-old males (Table 1). The pretraining run times for the 20-yr-old females was 1.5 min slower than the 40-yr-old females. Thus, the 40-yr-old females were marginally better runners at the start of the training program. Because the 20-yr-old females were initially less aerobically fit compared with the 40-yr-old females, it could be concluded that the 20-yr-old females should demonstrate greater cardiovascular improvements during the 12-wk training period. All four of the age and gender groups (N = 6) had an improvement in criterion run time after the training program, with the 20-yr-old females having the largest increase in performance (Fig. 2). A q2-way ANOVA was used to determine whether there were any age or gender effects in criterion run performance improvement. The 20-yr-old group had a larger increase (P < 0.001) in criterion run performance compared with the 40-yr-old group.
There was a total group mean reduction in resting heart rate between pre- (58.8 ± 9.5 beats·min−1) and posttraining (56.1 ± 9.0 beats·min−1) measurements (Table 2). These results are similar to other studies that report a small, but significant, reduction in resting heart rate in both older and younger, male and female, adults after endurance training (5,28). Thus, the reduction in resting heart rate with endurance training does not differ quantitatively or qualitatively with gender or age.
Research studies indicate that endurance training leads to an anticipated decrease in resting heart rate, with an increase in stroke volume that results in an unchanged resting cardiac output (26). The reduction in resting heart rate after endurance training may also be partly due to a decrease in intrinsic heart rate (1,20,23).
Endurance exercise is a physiological perturbation that significantly affects autonomic nervous activity. Research indicates that long-term endurance training increases heart rate variability, increases parasympathetic activity, and decreases sympathetic activity in the human heart at rest (5,8,10,11,20,23). The contribution of the sympathetic nervous system to the lower resting HR is probably small, as resting sympathetic activity is already low. These training-induced autonomic changes, coupled with a possible reduction in intrinsic heart rate, will decrease resting heart rate (1,5,10,20). Results from the present study confirm that endurance training will significantly increase heart rate variability, increase parasympathetic activity, and decrease sympathetic activity in the human heart at rest.
During supine rest in the present study, there was a total group mean increase in total power between pre- and posttraining (Table 2). There was also an increase in the standard deviation of the R-R interval between pre- and posttraining. Therefore, the 12-wk endurance-training program of the present study significantly increased the total group HRV in both the frequency (total power) and time domain (SD).
Heart rate variability is considered to be highly reproducible and is a good method for determining cardiovascular dynamics (15). Studies that have investigated the effect of exercise on HRV at rest, however, have provided some inconclusive results. Cross-sectional studies have determined that endurance-trained individuals have a higher HRV at rest in both the time and frequency domain (8,11). Longitudinal studies have determined that endurance training increases HRV at rest in both the time and frequency domain (19,25). Other longitudinal studies, however, have reported that endurance training does not significantly increase HRV at rest (14).
The parasympathetic division of the autonomic nervous system decreases heart rate mainly through its influence on the SA node. Parasympathetic stimulation influences the SA node by decreasing its rate of depolarization to threshold, causing the SA node to fire less frequently. Research has indicated that long-term endurance training increases parasympathetic activity in the human heart at rest (5,8). Athletes have a lower resting heart rate, and a more rapid recovery of heart rate after exercise, due to enhanced parasympathetic activity produced by long-term endurance training (21).
Most spectral analysis studies, including the present report, support the theory that endurance training enhances parasympathetic activity and thus contributes to resting bradycardia (10,20). In the present study, total group mean high-frequency power (0.15–0.50 Hz) increased after training (Table 2). Total group mean PNS indicator (HF power/total power) also increased between pre- and posttraining. Most cross-sectional studies have concluded that endurance-trained individuals have increased parasympathetic control of the heart at rest (8,11,23).
Endurance training seems to reduce the efferent sympathetic neural outflow to the SA node (4,8,23). Human studies that have used cardiac autonomic blockade to investigate the effect of training on autonomic balance report a decreased sympathetic control of heart rate after endurance training (23). The present study supports the theory that endurance training reduces sympathetic activity at rest. After training, total group mean low-frequency power was lower than before training (Table 2). Total group mean SNS indicator (LF power/HF power) also decreased between pre- and posttraining. The results from the present study indicate that a decrease in sympathetic activity after training is smaller than the increase in parasympathetic activity.
When the total group was divided into age and gender subgroups, there were some significant group differences. Individual variability, and the small age and gender group sample size, limited the number of significant group differences revealed after training. The 20F, 20M, and 40M groups demonstrated more spectral analysis changes after training compared with the 40F group (Table 2). The 20F group showed the greatest change in heart rate and total power with training. The 40F group produced the smallest increase in total power after training. Age and gender results from this study are similar to the spectral analysis results from a cross-sectional study by Gregoire et al. (11). In the Gregoire et al. (11) study, CGSA was used to determine that endurance-trained young (18–30 yr) females (N = 10) had a higher total power (5396 ± 940 ms2) than young (18–30 yr) males (3060 ± 791 ms2), older (40–55 yr) females (3193 ± 1310 ms2), and older (40–55 yr) males (1770 ± 411 ms2), respectively.
Spectral analysis results in the present study indicate that the younger subjects (20-yr age group) had a larger autonomic adjustment to training, whereas the older subjects had smaller autonomic changes to the same exercise stimulus (Table 2). Individual age and gender group differences indicate that the 40-yr-old females incurred the smallest autonomic adaptation to training. These results indicate that age influences plasticity of cardiovascular adaptations to endurance exercise. Gender may effect the autonomic response to training, but the effect is smaller than that of a 20-yr-age gap. The small subgroup sample size used in the study limits specific quantification of age and gender training differences. The 20-yr-old females had the lowest initial fitness level and demonstrated the largest improvement in criterion run performance (Fig. 2) and high-frequency power (Fig. 4). The 20-yr-old males demonstrated a larger increase in criterion run performance (Fig. 2) and high-frequency power (Fig. 4) compared with the 40-yr-old males, even though they had similar initial fitness levels. Therefore, the results seem to indicate that younger individuals have a greater autonomic adaptation to endurance training.
Early training studies report that for the same training stimulus, a younger adult has a greater cardiovascular adaptation than an older individual (13). Thus, it was thought that an individual past 60 yr of age would show only a minor adaptation to endurance training. However, subsequent published data have demonstrated that the older individual does in fact maintain an ability to adapt to endurance exercise training, evidenced by an increase in oxygen transport power and capacity (4,24). Research also shows that females have the same qualitative cardiovascular pattern of response as males to training but generally do not reach the same absolute levels of maximum oxygen uptake (16).
In the present study, the submaximal work rate was set at 180 W for males and 150 W for females in order to create a similar exercise heart rate in both genders. If the gender groups exercised at the same work rate, the females would have a significantly greater heart rate. During submaximal exercise, there was a decrease in total group mean heart rate between pre- (145.7 ± 11.9 beats·min−1) and post- (137.6 ± 11.6 beats·min−1) training (Table 3). To account for the effect of a change in resting heart rate on exercise heart rate, the difference between resting and exercise heart rate was calculated before and after training. There was a total group mean decrease in heart rate from rest, during submaximal exercise after training. Before training, the total group mean heart rate difference between rest and submaximal exercise was 88.1 ± 10.7 beats·min−1. After training the total group mean heart rate difference between rest and submaximal exercise was 82.8 ± 10.8 beats·min−1.
During exercise, most studies report a progressive withdrawal of parasympathetic activity with increasing exercise intensity. However, quantification of autonomic control during exercise has not been very reliable, with inconsistent findings in several studies (30). The inconsistent results obtained in spectral analysis studies may be due to a number of factors. A potential limitation of spectral analysis for the evaluation of physiologic responses to submaximal exercise is that nonstationary conditions may increase noise in the data. Spectral analysis of HRV has not yet been standardized, and different analyses yield varying results. A rise in core temperature during exercise will affect the low-frequency oscillations that reflect thermoregulatory control. The incremental nature of exercise produces complex changes on cardiovascular variables and respiration. Reduced HRV total power during exercise is probably the main reason for inconsistent spectral analysis results during exercise. As exercise intensity increases, there is a reduction HRV total power. This reduction in total power, coupled with individual variability at a given exercise intensity, makes measurement of autonomic control during exercise less accurate and more variable than HRV at rest, and during orthostatic stress.
In the present study, during submaximal exercise, there was a total group mean increase in total power and standard deviation of the R-R interval after the training program (Table 3). When the difference between rest and submaximal exercise was calculated, there was a total group mean pre- to posttraining increase from rest in: high-frequency power, total harmonic power, fractal power, total power, and PNS indicator. These results indicate that endurance training will increase HRV total power during submaximal exercise at a constant work rate. The increase in submaximal exercise HRV total power after endurance training is mainly due to an increase in high-frequency power. The increase in high-frequency power during submaximal exercise is due to less withdrawal of parasympathetic activity after an endurance-training program. Improved cardiovascular function during submaximal exercise, after endurance training, is demonstrated by a decreased heart rate and increased heart rate variability.
In the present study, there was a significant age and training interaction for heart rate during submaximal exercise. The 20-yr-old group had a decrease in submaximal heart rate between pre- (147.0 ± 0.6 beats·min−1) and post- (137.4 ± 0.6 beats·min−1) training. The 40-yr-old group also had a decrease in submaximal heart rate between pre- (144.3 ± 0.6 beats·min−1) and post- (138.2 ± 0.6 beats·min−1) training. However, the 20-yr-old group pretraining heart rate (147.0 ± 0.6 beats·min−1) was higher than the 40-yr-old group pretraining heart rate (144.3 ± 0.6 beats·min−1). There were no significant gender effects during submaximal cycling after training. These results indicate that during submaximal exercise a younger adult may have a greater autonomic adaptation to endurance training than an older individual.
The rate and magnitude of cardiovascular adaptation in response to endurance exercise training depends on the type, intensity, frequency, and duration of the exercise stimulus, as well as genetic factors. The beneficial adaptation in heart function attributed to endurance exercise training includes an increase in: maximal cardiac output, stroke volume, diastolic filling, and left ventricular volume overload hypertrophy (4,9). These central adaptations to endurance training are coupled with peripheral adaptations in skeletal muscle to increase maximal oxygen uptake. Long-term adaptation to endurance exercise is thus a very complex process involving structural, metabolic, humoral, and neural changes.
Limitations of the present study.
This study was constrained to the use of spectral analysis of heart rate variability. Although CGSA allows for simultaneous evaluation of the harmonic components of the parasympathetic and sympathetic nervous system, it is possible that the amount of information contained in the HRV signal cannot be completely extracted by a single approach. Important physiological adaptations that occur with endurance training were not measured in this study. Additional physiological information such as the change in V̇O2max, ventricular size, baroreceptor sensitivity, catecholamine levels, and muscle sympathetic nerve activity during rest and exercise may enable future studies to achieve greater insight into the cardiovascular adaptations induced by endurance training. Although the present data provides important information about spectral changes induced by endurance training in young and older healthy subjects, analysis is limited by the small sample size.
This research was supported was supported by the Heart and Stroke Foundation of B.C. & Yukon.
1. Areskog, N. Effects and adverse effects of autonomic blockade in physical exercise. Am. J. Cardiol. 55: 132D–134D, 1985.
2. Banister, E. W., R. H. Morton, and J. R. Fitz-Clarke. Clinical dose response effects of exercise. In: Physiology and Pathophysiology of Exercise Tolerance, J. M. Steinacker and S. A. Ward (Eds.). London: Plenum Press, 1996,
3. Banister, E. W., J. B. Carter, and P. C. Zarkadas. Training theory and taper: validation in triathlon athletes. Eur. J. Appl. Physiol. 79: 182–191, 1999.
4. Blomqvist, C. G., and B. Saltin. Cardiovascular adaptations to physical training. Ann. Rev. Physiol. 45: 169–189, 1983.
5. Chen, C., and S. E. Dicarlo. Endurance exercise training-induces resting bradycardia: a brief review. Sports Med. Training Rehabil. 8: 37–77, 1997.
6. Coyle, E. F., W. H. Martin, D. R. Sinacore, M. J. Joyner, J. M. Hagberg, and J. O. Holloszy. Time course of loss of adaptations after stopping prolonged intense endurance training. J. Appl. Physiol. 57: 1857–1864, 1984.
7. Demeersman, R. E. Heart rate variability and aerobic fitness. Am. Heart. J. 125: 726–731, 1993.
8. Dixon, E. M., M. V. Kamath, N. Mccartney, and E. L. Fallen. Neural regulation of heart rate variability in endurance athletes and sedentary controls. Cardiovasc. Res. 26: 713–719, 1992.
9. Douglas, P. S., and M. O’Toole. Aging and physical activity determine cardiac structure and function in the older athlete. J. Appl. Physiol. 72: 1969–1973, 1992.
10. Goldsmith, R. L., D. M. Bloomfeld, and E. T. Rosenwinkel. Exercise and autonomic function. Coron. Artery Dis. 11: 129–135, 2000.
11. Gregoire, J., S. Tuck, Y. Yamamoto, and R. L. Hughson. Heart rate variability at rest and exercise: influence of age, gender and physical training. Can. J. Appl. Physiol. 21: 455–470, 1996.
12. Ingram, D. K. Age-related decline in physical activity: generalization to nonhumans. Med. Sci. Sports Exerc. 32: 1623–1629, 2000.
13. Kasch, F. W., W. H. Phillips, J. E. Carter, and J. L. Boyer. Cardiovascular changes in middle-aged men during two years of training. J. Appl. Physiol. 34: 53–57, 1973.
14. Loimaala, L., H. V. Huikuri, P. Oja, M. Pasanen, and I. Vuori. Controlled 5-mo aerobic training improves heart rate but not heart rate variability or baroreflex sensitivity. J. Appl. Physiol. 89: 1825–1829, 2000.
15. Pagani, M., and A. Malliani. Interpreting oscillations of muscle sympathetic nerve activity and heart rate variability. J. Hypertens. 18: 1709–1719, 2000.
16. Raven, P. B., B. L. Drinkwater, and S. M. Horvath. Cardiovascular responses of young female track athletes during exercise. Med. Sci. Sports 4: 205–209, 1972.
17. Ryan, S. M., A. L. Goldberger, S. M. Pincus, J. Mietus, and L. A. Lipsitz. Gender-, and age-related differences in heart rate dynamics: are women more complex than men? J. Am. Coll. Cardiol. 24: 1700–1707, 1994.
18. Scheuer J., and C. M. Tipton. Cardiovascular adaptations to physical training. Ann. Rev. Physiol. 39: 221–251, 1977.
19. Schuit, A. J., L. G. Van Amelsvoort, T. C. Verheij, et al. Exercise training and heart rate variability in older people. Med. Sci. Sports Exerc. 31: 816–821, 1999.
20. Shi, X., G. Stevens, B. Foresman, S. Stern and P. Raven. Autonomic nervous system control of the heart: endurance exercise training. Med. Sci. Sports Exerc. 27: 1406–1413, 1995.
21. Shin, K., H. Minamitani, S. Onishi, H. Yamazaki, and M. Lee. The power spectral analysis of heart rate variability in athletes during dynamic exercise: part II. Clin. Cardiol. 18: 664–668, 1995.
22. Sinnreich, R., J. D. Kark, Y. Friedlander, D. Sapoznikov, and M. H. Luria. Five minute recordings of heart rate variability for population studies: repeatability and age-sex characteristics. Heart 80: 156–162, 1998.
23. Smith, M. L., D. Hudson, H. Graitzer, and P. Raven. Exercise training bradycardia: the role of autonomic balance. Med. Sci. Sports Exerc. 21: 40–44, 1989.
24. Spina, R. J. Cardiovascular adaptations to endurance exercise training in older men and women. Exerc. Sport Sci. Rev. 27: 317–332, 1999.
25. Stein, P. K., A. A. Ehsani, P. P. Domitrovich, R. E. Kleiger, and J. N. Rottman. Effect of exercise training on heart rate variability in healthy older adults. Am. Heart. J. 138: 567–576, 1999.
26. Stratton, J. R., W. C. Levy, M. D. Cerqueira, R. S. Schwartz, and I. B. Abrass. Cardiovascular responses to exercise: effects of aging and exercise training in healthy men. Circulation 89: 1648–1655, 1994.
27. Tulppo, M. P., T. H. Makikallio, T. Seppanen, R. T. Laukkanen, and H. V. Huikuri. Vagal modulation of heart rate during exercise: effects of age and physical fitness. Am. J. Physiol. 274: H424–H429, 1998.
28. Wilmore, J. H., P. Stanforth, J. Gagnon, et al. Heart rate and blood pressure changes with endurance training: the Heritage Family Study. Med. Sci. Sports Exerc. 33: 107–116, 2001.
29. Yamamoto, Y., and R. L. Hughson. Coarse-graining spectral analysis: new method for studying heart rate variability. J. Appl. Physiol. 71: 1143–1150, 1991.
30. Yamamoto, Y., R. L. Hughson, and J. C. Peterson. Autonomic control of heart rate during exercise studied by heart rate variability spectral analysis. J. Appl. Physiol. 71: 1136–1142, 1991.