Medicine & Science in Sports & Exercise:
BASIC SCIENCES: Original Investigations
Influence of Short-Term Endurance Exercise Training on Heart Rate Variability
LEE, C. MATTHEW1; WOOD, ROBERT H.2; WELSCH, MICHAEL A.2
1Department of Kinesiology, San Francisco State University, San Francisco, CA; and
2Department of Kinesiology, Louisiana State University, Baton Rouge, LA
Address for correspondence: Matthew Lee, Ph.D., Department of Kinesiology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132-4161; E-mail: email@example.com.
Submitted for publication June 2002.
Accepted for publication February 2003.
LEE, C. M., R. H. WOOD, and M. A. WELSCH. Influence of Short-Term Endurance Exercise Training on Heart Rate Variability. Med. Sci. Sports Exerc., Vol. 35, No. 6, pp. 961–969, 2003.
Purpose: To examine the influence of 2 wk (eight sessions) of endurance training on cardiac autonomic modulation, as measured by heart rate variability (HRV).
Methods: Twenty-four males (mean age: 23.1 yr) were randomized to an exercise (EX;N = 12) or control group (CT;N = 12). EX trained for eight sessions (4× wk−1, 40 min, 80–85% HRreserve) on a cycle ergometer. ECG tracings were collected during 5 min of paced breathing (12 breaths·min−1 (PB)), 5 min of spontaneous breathing (SB1), 5 min of 70° head-up tilt (TILT), and a second 5-min period of spontaneous breathing (SB2). Data were collected before (test 1), during (tests 2–4), and 48 h after (test 5) the 2-wk period. HRV was reported as the standard deviation of RR intervals, and as natural logarithm of the normalized units (NU) of high- and low-frequency power (lnHF and lnLF).
Results: EX exhibited a significant increase in peak oxygen consumption (8%). During PB and TILT conditions, ANOVA revealed a group × time interaction such that EX exhibited lower lnLFNU and lnLF/lnHF during test 5 compared with test 1.
Conclusion: These data suggest that eight endurance exercise-training sessions performed over 2 wk enhance the relative vagal modulation of the heart during PB and TILT, but not during SB.
Chronic endurance exercise training may alter cardiac autonomic balance as reflected by changes in heart rate variability (HRV). Studies involving young apparently healthy individuals suggest that endurance-training programs generally result in a shift toward a greater vagal modulation of the heart (2,6,15). For example, De Meersman (6) reported a significantly elevated respiratory sinus arrhythmia after 8 wk of running, and Levy et al. (15) reported significant elevations in the standard deviation of normal RR intervals after 6 months of endurance training. More recently, Al-Ani et al. (2) reported that 6 wk of cycling exercise at an intensity of 85% heart rate maximum elevated high-frequency (HF) modulations of the heart, believed to be associated with vagal activity (1,4). Although the exact mechanism of the shift in sympathovagal balance is not fully understood, evidence suggests the changes may be mediated through central, cardioreflex, and/or end-organ changes (7,8,10).
Interestingly, the time course of exercise-training induced changes in autonomic modulation of the heart has not been extensively studied. Most studies that have approached this issue have examined plasma catecholamines and/or heart rate at rest and during physical activity (17,18,26). From these findings, investigators have concluded reductions in heart rate and plasma norepinephrine (NE) and epinephrine (E) are evident within 6–10 training sessions (17,18,26). However, the inferences that can be drawn from these studies are somewhat limited, inasmuch as measures of heart rate data do not allow for inferences about the specific activity of the sympathetic and parasympathetic systems. Furthermore, plasma NE does not necessarily solely reflect the rate of release from the axon terminals. Rather, it is influenced by binding and rate of clearance. Thus, it is of interest to evaluate the time course of training-induced autonomic changes using HRV, which may be more sensitive to changes in sympathovagal balance (19) than HR or plasma catecholamines. Additionally, because poor HRV profiles are associated with an increased risk of cardiac arrhythmic events in patients with heart disease as well as among the general population (5,24), elucidation of a time course of endurance training-induced alterations in HRV may provide clinicians with useful information regarding exercise programming in healthy and at-risk populations.
Of concern when utilizing HRV to generate inferences about the autonomic modulation of the heart is the observation that HRV parameters appear to be influenced by acute laboratory conditions. For example, Hayano et al. (11) reported that paced breathing results in a shift in the HRV power spectrum toward the HF range, a phenomenon thought to be a result of the respiratory sinus arrhythmia. Moreover, autonomic reactivity as observed under stressors such as head-up tilt may provide insight beyond that which is gleaned from resting measures alone. Furlan et al. (9) reported that head-up tilt, which elicits a sympathetic outflow, results in a shift toward the LF range. At present, however, it is not clear as to how HRV measured under different laboratory conditions might influence the sensitivity of HRV to detect training induced alterations in autonomic modulation of the heart.
Therefore, the objectives of this investigation were to 1) examine adaptations in cardiac autonomic modulation as measured by HRV under different conditions, including paced breathing and head-up tilt, in response to 2 wk of exercise training; and 2) evaluate a possible time course of any training-induced alterations in autonomic of the heart. It is hypothesized that participants undergoing 2 wk of high-intensity exercise training will exhibit a greater relative vagal modulation of the heart, as identified by an increase in HF power, and a concurrent decrease in LF power, expressed as normalized units. Furthermore, in accordance with previous studies evaluating endurance training on markers of autonomic activity, it is hypothesized that these training-induced alterations will be evident by the sixth training session.
Twenty-four healthy college-aged males (mean age: 23.1 yr) were recruited to participate in the study. This number of participants was selected based on an a priori power analysis to give 80% power. All participants provided written informed consent, and all procedures were approved by the institutional review board of the host site. All participants met the following inclusion criteria: 1) male gender, 2) 18–30 yr of age, 3) an energy expenditure of less than 2000 kcal·wk−1 for the previous 3 months, and 4) a peak oxygen consumption (V̇O2peak) between 30 and 45 mL·kg−1·min−1. Exclusion criteria included: 1) presence of any documented cardiovascular, metabolic, or neurological disease; 2) regular consumption of any medication that influences the cardiovascular system; and 3) participation in regular physical activity (an energy expenditure of greater than 2000 kcal·wk−1).
Sequence of tests.
On the initial visit to the laboratory, each participant gave written informed consent, completed the Health Status Questionnaire (14) and Aerobics Center Longitudinal study Physical Activity Questionnaire (20), was assessed for body composition, and completed a symptom-limited graded exercise test (SL-GXT) for assessment of V̇O2peak. Participants were then assigned to either an exercise-training group (EX) or to a control group (CT) via stratified random selection in an attempt to minimize any initial group difference in V̇O2peak.
After the initial assessment, the EX group underwent 2 wk of endurance exercise training, whereas the CT group was asked to maintain their previous level of physical activity. Training was performed 4 d·wk−1 (Monday, Tuesday, Thursday, and Friday). To evaluate the time course of any changes in HRV that occurred during the course of the training protocol, electrocardiograms (ECG) were collected, and HRV was evaluated in both groups 2× wk−1 (Monday and Thursday) and at the same time of day as their previous ECG collection, with the EX group being evaluated before their exercise session. On the Monday immediately after the training period, all participants underwent a posttesting session, during which HRV, V̇O2peak, and body composition were reassessed, and the physical activity questionnaire was completed. The 5 d in which HRV was assessed were identified as tests 1 through 5, respectively.
Assessment of peak oxygen consumption.
A SensorMedics Vmax 29c pulmonary gas exchange system (Yorba Linda, CA) was used to evaluate the participant’s V̇O2peak during a SL-GXT on an 818E Monark cycle-ergometer (Stockholm, Sweden). A Polar Electro heart rate monitor (Oulu, Finland) and mercury sphygmomanometer and stethoscope were used to assess heart rate and arterial blood pressure. The SensorMedics system was calibrated before each test and was used to sample expired air via a mouthpiece. Before the SL-GXT, participants were asked to sit upright on the cycle ergometer while resting heart rate, blood pressure, and cardiorespiratory measures were recorded. After a brief resting period on the cycle-ergometer, the participant was asked to begin cycling at 70 rpm at a workload of 0.5 kp. The workload was increased by 0.5 kp every 2 min. Heart rate and ratings of perceived exertion (RPE) were recorded every minute, and arterial blood pressure was recorded every stage. The participant was encouraged to continue cycling until he could no longer maintain a pace of 70 rpm, at which time the test was terminated. Cardiorespiratory variables measured during the SL-GXT included oxygen consumption (V̇O2), minute ventilation (V̇E), respiratory frequency (RF), tidal volume (V̇T), volume of expired carbon dioxide (V̇CO2), and the respiratory quotient (RQ). Three of the following criteria had to be met to ensure that the participant achieved a maximal effort: 1) failure of heart rate to increase with further increases in intensity, 2) plateau of oxygen uptake with increased workload, 3) RQ of > 1.15, and 4) an RPE > 17.
Assessment of body composition.
Lange skinfold calipers (Cambridge, MD) were used to collect skinfold measures at the following sites on each participant: 1) chest, 2) tricep, 3) abdominal, 4) suprailiac, 5) subscapular, 6) midaxillary, and 7) thigh. The sum of the seven skinfold sites was used to estimate percent body fat using previously established equations (3).
Endurance exercise training consisted of 40 min of cycling on an 818E Monark cycle ergometer (Stockholm, Sweden). Each exercise bout consisted of a 5-min “warm-up” period (cycling with no added resistance), followed by 30 min of cycling at a resistance that elicited a heart rate of 80–85% HRreserve, and ended with a 5-min “cool-down” period (cycling with no added resistance). Exercise intensity was closely monitored using a Polar Electro heart rate monitor.
All procedures were performed in a controlled laboratory setting (23–24°C; ∼760 torr). Upon reporting to the laboratory, the participant was asked to lie supine on a tilt table. Ag/AgCl electrodes were then arranged on the participant’s anterior side in a standard three-lead configuration, and the electrodes were connected to the Biopac MP100 data acquisition system (Santa Barbara, CA). The AcqKnowledge ACK100 software program (Santa Barbara, CA) was used to collect a continuous ECG signal at 500 Hz. Additionally, a Cosmed K4b2 pulmonary gas exchange system (Rome, Italy) was used to continuously monitor RF and VT during each testing session. The ECG and respiratory data were collected during a 5-min period of paced breathing at a frequency of 12 cycles per minute (PB), during a 5-min period in which the participant breathed spontaneously (SB1), during a 5-min period of 70° head-up tilt (TILT) and during a subsequent 5-min period of supine rest (SB2). Arterial pressure was assessed at the midpoint of each 5-min period.
HRV was evaluated in accordance with established guidelines (23). ECG were visually inspected for nonsinus beats and none were found in any of the participants. The ECG were subsequently plotted as a tachogram of heart period which was evaluated for the mean and standard deviation of all normal RR intervals (SDNN). Spectral analysis of HRV was derived via a 1024-point linear fast Fourier transformation using a Hamming window. The resultant power density spectrum was then analyzed for total power (0.00–0.40 Hz), LF (0.04–0.15 Hz), and HF (0.15–0.40 Hz). LF and HF were further normalized (LFNU and HFNU) to quantify sympathovagal balance. Systolic (SBP) and diastolic (DBP) arterial blood pressures were initially recorded and were used to calculate mean arterial pressure (MAP) using the equation: MAP = DBP + 1/3 (SBP − DBP).
All statistical analyses were performed with the SAS statistical package (Cary, NC). The Shapiro-Wilk test of normality was performed on all variables and those that violated assumptions of normality were transformed using the natural log (ln) algorithm. All tests were considered significant at the 0.05 level.
Descriptive and peak cardiorespiratory measures.
Independent t-tests were used to examine any initial group differences in age and height. A mixed-model, two-way ANOVA with repeated measures was used to examine main effects of group, time, or group × time interactions on body weight, body composition, physical activity, and cardiorespiratory measures at peak exercise during the SL-GXT. When a significant group × time interaction was found, a one-way repeated measures ANOVA was used to test for simple effects and identify any prepost differences within each group.
HRV, arterial pressure, and respiratory measures.
HRV indices, arterial pressures, and respiratory measures collected each testing session (test 1 through test 5) were analyzed with a mixed model, two-way repeated measures ANOVA to examine differences between groups and conditions. Similarly, two-way repeated measures ANOVA was used to analyze these variables collected during each condition (PB, SB1, TILT, and SB2) in an effort to examine differences between groups and testing sessions across time. When a significant group × condition or group × time interaction was found, a one-way repeated measures ANOVA was used to test for simple effects and identify differences between conditions, or across time, within a group. When the test for simple effects was significant, post hoc contrasts were used to identify which conditions differed from each other, or in the case of the longitudinal analysis, which testing sessions differed from test 1. Similarly, when a significant main effect of condition or time was found in the absence of significant interactions, contrasts were used to identify differences between conditions, or identify which testing sessions significantly differed from test 1. Lastly, bivariate correlations were performed within the EX group by calculating Pearson correlation coefficients to examine the relationship between initial values and changes in HRV, peak cardiorespiratory variables, and body composition over the course of the investigation.
Descriptive Statistics and Peak Cardiorespiratory Indices
Descriptive statistics for the participants can be seen in Table 1. There were no significant between group differences before the intervention. However, two-way ANOVA with repeated measures revealed a main effect of time on percent body fat (P < 0.01), indicating lower body fat for the entire participant pool after the 2-wk period. Furthermore, analysis of the physical activity questionnaire revealed a group × time interaction (P < 0.01) such that EX reported an increase in physical activity at the posttest. With respect to cardiorespiratory fitness (Table 2), ANOVA revealed group × time interactions on V̇O2peak (P < 0.01), V̇Epeak (P < 0.01), and V̇Tpeak (P < 0.05). These interactions were such that EX experienced increases in V̇O2peak (P < 0.01) and V̇Epeak (P < 0.05) across time, whereas these variables were unaltered in CT.
HRV, Arterial Pressure, and Respiratory Indices within the Testing Sessions
There were consistent main effects of condition on the mean RR interval. Figure 1 demonstrates the response of the mean RR interval during test 4, which is representative of the other tests. Contrasts revealed that these effects were such that the mean RR interval during TILT was lower than that during the other three conditions (P < 0.05), and also that the mean RR interval during SB2 was greater than that during the other conditions (P < 0.05). There was a main effect of condition on SDNN that was consistent across all testing sessions (Fig. 2). Post hoc contrasts revealed that SDNN during TILT was significantly lower than that during the other three conditions (P < 0.05).
Regarding the spectral indices, there were main effects of condition on lnLFNU (Fig. 3) and lnHFNU, and lnLF/lnHF during tests 1 through 4. Contrasts then revealed that lnLFNU and lnLF/lnHF were greater (and lnHFNU was lower) during TILT compared with the other three conditions (P < 0.05), and that lnLFNU and lnLF/lnHF were lower (and lnHFNU was greater) during PB compared with the other conditions (P < 0.05). Exclusively during test 5, there were group × condition interactions on lnLFNU (P < 0.01), lnHFNU (P < 0.01), and lnLF/lnHF (P < 0.01). Although CT exhibited similar responses in HRV indices as the previous tests, EX did not exhibit the significant elevation in lnLFNU and lnLF/lnHF and reduction in lnHFNU during TILT.
There were differences in arterial pressure according to test condition that were consistent across days and groups. The behavior of arterial pressure is illustrated in Figure 4, which shows pooled data from both groups and all testing sessions. The results indicate that on each day and for each group, TILT resulted in an elevation of DBP (P < 0.05) and MAP (P < 0.05), whereas SBP was not significantly altered. There were also differences in RF and V̇T according to test condition that were consistent across time and groups. Table 3 shows pooled data from both groups and all testing sessions. The results indicate that during each testing session and for each group, RF was lower (P < 0.05) and V̇T was greater (P < 0.05) during PB compared with the other conditions.
HRV, Arterial Pressure, and Respiratory Indices Across Time
ANOVA revealed a main effect of time (P < 0.01) on SDNN. Contrasts revealed that SDNN during test 5 was greater than during test 1 (P < 0.01). Regarding the spectral indices, there were group × time interactions on lnLFNU (P < 0.05) (Fig. 5) and lnLF/lnHF (P < 0.05) (Fig. 6). Tests for simple effects found that these variables were not altered across time in CT, whereas lnLFNU and lnLF/lnHF were lower during test 5 compared with test 1 (P < 0.01) in EX. Although not statistically significant, there was a trend for a group × time interaction for lnHFNU (P = 0.06), such that this variable was greater during test 5 compared with test 1 in EX. Furthermore, there were no main effects of group or time, nor were there any group × time interactions on arterial blood pressure or respiratory measures during PB.
There were no main effects of group or time, nor group × time interactions on the mean RR interval, HRV parameters, arterial blood pressure, or respiratory measures during SB1. During SB2, ANOVA revealed a main effect of time on lnLF/lnHF (P < 0.05). However, contrasts revealed that there were no differences in lnLF/lnHF between test 1 and any of the other tests. Additionally, there were no main effects of group or time, or group × time interactions on arterial blood pressure or respiratory measures during SB2.
ANOVA revealed group × time interactions on lnLFNU (P < 0.01) (Fig. 5), lnHFNU (P < 0.01), and lnLF/lnHF (P < 0.05) (Fig. 6). Tests for simple effects revealed no change in spectral parameters for CT but that EX displayed greater lnHFNU (P < 0.01), and lower lnLFNU (P < 0.01) and lnLF/lnHF (P < 0.01) during test 5 compared with test 1. Additionally, there were no main effects of group or time, or group × time interactions on arterial blood pressure or respiratory measures during TILT.
Within the EX group, there were significant correlations between initial V̇O2peak and the change in lnLF/lnHF (r = 0.68, P < 0.05) and lnHFNU (r = −0.71, P < 0.05) during TILT. These were such that participants with lower initial V̇O2peaks exhibited a greater decrease in lnLF/lnHF and a greater increase in lnHFNU during TILT over the course of the investigation (test 1 vs test 5). However, initial V̇O2peak was not correlated with the change in any HRV parameter during PB nor was any initial HRV parameter correlated with the magnitude of change of any peak cardiorespiratory parameter over the training period. Furthermore, there were no significant correlations between the change in any HRV parameter during PB or TILT and the magnitude of change in any peak cardiorespiratory parameter or body composition.
This investigation examined the influence of 2 wk of exercise training, 4 d·wk−1, 40 minutes per session, at 80–85% HRreserve, on cardiorespiratory fitness and vagal modulation of the heart. Our findings suggest that it takes at least eight training sessions within 2 wk to alter indices of HRV. Additionally, we were only able to detect training-induced alterations in HRV during the PB and TILT conditions. The initial values obtained for the cardiorespiratory variables and HRV are consistent with other studies using participants of similar fitness levels (9,18). With respect to the influence of training, participants in the treatment group improved their cardiorespiratory fitness by 8% (V̇O2peak from 33.5 to 36.1 mL·kg−1·min−1), whereas the control group exhibited no such changes. The magnitude of the training effect is similar to findings from Mier et al. (18) and Hickson et al. (13) in response to 10 d of training.
Influence of breathing pattern and orthostatic challenge on HRV indices.
HRV was examined under a variety of conditions, both to examine its construct validity, as well as to examine the extent to which these conditions expose alterations in autonomic behavior following periods of exercise training. Of particular interest was the influence that breathing pattern might have on autonomic modulation of the heart. Therefore, paced and spontaneous breathing conditions were included. In addition, the autonomic adjustment to head-up tilt was of interest inasmuch as autonomic reactivity may provide information that is unique from resting autonomic modulation alone. To this end, there were some general observations regarding the influences of the different conditions on HRV.
With respect to the influence of breathing pattern, the data collected during all testing sessions revealed that lnLFNU and lnLF/lnHF were lower and lnHFNU greater during PB compared with the other conditions. These findings are consistent with previous reports that PB results in an elevation of power in the HF range of the power spectrum, speculated to be a result of an augmented respiratory sinus arrhythmia due to input from pulmonary stretch receptors and/or modulation of cardiomotor neurons in the central respiratory center (11). Our findings of lower RF and greater V̇T during PB are consistent with this hypothesis.
With respect to the orthostatic challenge, TILT resulted in a significant reduction of the mean RR interval (i.e., increase in heart rate). This was accompanied by increases in lnLFNU and lnLF/lnHF and decreases in SDNN and lnHFNU on most testing days. These findings are consistent with previous reports that TILT results in a greater sympathetic outflow and/or a vagal withdrawal (9). Furthermore, 5 min of TILT resulted in a consistent elevation of DBP and MAP. This response, which was observed on each test day, is further evidence of an increased sympathetic outflow during 5 min of TILT. Such a response appears to be mediated by the baroreflex mechanism (19). The behavior of HRV and hemodynamics described above are consistent with our understanding of cardiovascular regulation during such laboratory conditions, and therefore the results suggest some degree of construct validity.
Influence of short-term training on HRV indices.
Previous studies have demonstrated that endurance exercise training in young, healthy participants results in significant alterations in HRV that reflect an increased vagal modulation of the heart. For example, Al-Ani et al. (2) reported that 6 wk of cycling significantly elevates HF power. Furthermore, De Meersman (6) and Levy et al. (15) reported improvements in time domain indices of HRV after 8 wk and 6 months of training, respectively. However, there is limited information regarding the time course of these training-induced changes. Thus, it was our intention to examine if 2 wk of training could elicit such changes, and if so, to examine the time course of these alterations. The data from this investigation demonstrate that eight sessions of high-intensity exercise training performed over 2 wk is sufficient to alter HRV, but only as detected during the PB and head-up TILT conditions.
During PB we found significant group × time interactions on lnLFNU and lnLF/lnHF. Further analysis revealed that both of these variables were lower (16 and 9%, respectively) during test 5 compared with test 1 in EX, whereas there were no differences across time in CT. These findings suggest that our training protocol resulted in a shift in sympathovagal balance favoring a greater relative vagal modulation of the heart. Although the group × time interaction did not reach statistical significance for SDNN and lnHFNU during PB, there was a trend toward an interaction for each of these indices (P = 0.08 and P = 0.06, respectively). Therefore, measures of HRV during the paced breathing condition are generally in agreement with previous work indicating a shift in sympathovagal balance after exercise training. Additionally, a unique finding of this study is that it suggests that this effect may not be observed until eight training sessions into a 2-wk training period.
The influence of short-term training on the response to the orthostatic challenge revealed significant group × time interactions on lnLFNU, lnHFNU, and lnLF/lnHF. Specifically, the group undergoing exercise training displayed lower lnLFNU and lnLF/lnHF, and greater lnHFNU during test 5 compared with test 1, whereas CT exhibited no differences in these indices across time. Thus, an additional contribution of this investigation is that 2 wk of endurance exercise training may alter the autonomic response (elevated parasympathetic and/or decreased sympathetic response) to TILT. In further support of this hypothesis, during test 5 there were significant group × condition interactions on lnLFNU, lnHFNU, and lnLF/lnHF. These interactions were such that EX did not demonstrate the elevation of lnLFNU and lnLF/lnHF, or the reduction in lnHFNU during TILT that were present during the first 4 tests.
Interestingly, no statistically significant change in mean RR interval was noted during any of the conditions after the training protocol. However, it should be noted that the mean RR interval of the training group increased from 925 to 985 ms from test 1 to test 5 (P = 0.08), reflecting a reduction in heart rate from 65 to 61 bpm during PB. The magnitude of this change in mean resting heart rate is similar to those reported by Wilmore et al. (25).
Our ability to detect a training effect on HRV only during PB and TILT suggests that HRV may be a more sensitive measure of training-induced changes in autonomic modulation when collected under such conditions. In fact, a few studies that have reported training-induced changes in HRV used a paced breathing protocol and/or head-up tilt (6,12,27). The absence of any significant findings during SB further emphasizes the need for careful standardization procedures during collection of HRV data. Thus, the addition of either of these two conditions would be useful to any longitudinal investigation examining cardiac autonomic modulation.
The findings of the present investigation are somewhat consistent with Hedelin et al. (12), who recently reported a reduction in LF power during head-up tilt, and this reduction was significantly correlated with the change in V̇O2peak. Although the magnitude of change in V̇O2peak in our exercising participants was not correlated with changes in HRV, their initial V̇O2peak was significantly correlated with the change in lnLF/lnHF and lnHFNU during TILT. Thus, it would appear that alterations in cardiac autonomic modulation after training are more pronounced in those with a lower initial V̇O2peak. However, unlike Hedelin et al. the initial HRV profile did not constitute a favorable condition to improve V̇O2peak (i.e., we found no relationship between initial HRV indices and change in V̇O2peak). Nonetheless, the small differences between our findings and those of Hedelin may possibly be attributed to the apparent higher fitness level of their participants and our findings of training-induced alterations in HRV during TILT remain very similar. In contrast to the present study, however, Yamamoto et al. (27) recently evaluated HRV after 1 wk, 4 wk, and 6 wk of exercise training. They subsequently reported a significant increase in HF power after 1-wk of training, followed by no further increases over the rest of the training period. On the other hand, we did not report any changes in autonomic modulation until the second week of training. This finding is perplexing, in that the training protocol of Yamamoto et al. (27) was very similar to the present study (frequency: 4 d·wk−1; duration: 40 min; intensity: 80% V̇O2peak). However, Yamamoto et al. (27) gave no information regarding the participant’s habitual level of physical activity and their participants had an initial fitness level that surpassed those of our participants (V̇O2peak = 48.2 mL·kg−1·min−1). Nonetheless, together these results appear to indicate that sympathovagal balance under PB conditions is rapidly modified in the early phase of an exercise-training program. In contrast to our findings, Loimaala et al. (16) reported no significant changes in HRV after a 5-month jogging protocol (frequency: 4–6× wk−1; duration: 30 min; intensity: 75% of V̇O2max). Possible reasons for the discrepancy between these results and our findings include: 1) the lower exercise intensity used compared with our study, 2) the lack of a paced breathing protocol during collection of HRV, and 3) the lack of daily supervision by Loimaala et al. (16). This group only provided supervision during one exercise session a week, thus the researchers cannot be certain that the participants fully adhered to the training protocol.
Lastly, there are studies that have found changes suggesting heightened sympathetic activity after exercise training. Pichot et al. (21) reported that during a 3-wk high-intensity training protocol, there was a shift toward cardiac sympathetic dominance, marked by a loss of global HRV. At first glance, such findings might seem directly opposed to those of the present investigation. However, the participants in Pichot’s study engaged in 6–10 “exhaustive” exercise sessions a week, and thus there may have been an overtraining effect along with insufficient recovery between sessions, at least in terms of identifying a training adaptation on autonomic activity. This is somewhat supported by their data indicating that after a fourth “recovery” week, there was a shift back toward parasympathetic dominance, leading the authors to indicate that indeed there is an important aspect of a sufficient recovery period on autonomic modulation. Similarly, Furlan et al. (9) reported an increase in cardiac sympathetic outflow, as identified by elevated LF power, after an acute bout of exercise that persists up to 24 h after cessation of exercise. They further reported that this sympathetic elevation was absent 48 h after the exercise bout. Inasmuch as the present investigation allowed at least 48 h after the participant’s last exercise session before collection of HRV data, thereby optimizing the recovery period, the findings of the present study stand in agreement with Pichot et al. (21) and Furlan et al. (9).
Thus, our findings of a training-induced reduction in sympathovagal balance under PB and a decreased sympathovagal response to TILT are in agreement with other research. Although the present investigation cannot offer an explanation as to where training adaptations may occur, the appearance of training effects under some conditions, but not all, seems to suggest that the training adaptations extend beyond changes at the end-organ, alone. Otherwise, it could be argued that pre- versus posttraining differences should have existed under any of the laboratory conditions (including spontaneous breathing). One possible mechanism for our findings is an increased blood volume, which has been documented as an early adaptation to exercise training (28). Spinelli et al. (22) recently reported that acute plasma volume expansion via saline infusion resulted in an increase in the HF range of the HRV power spectrum. They subsequently suggested that increases in plasma volume activate arterial and cardiopulmonary baroreceptors, which decrease sympathetic, and increase parasympathetic outflow from the cardiorespiratory center in the medulla. This could further explain how our participants were able to maintain arterial pressure during TILT with a lower sympathetic outflow after training. Other possible mechanisms include changes in receptor sensitivity at the cardiorespiratory center (8) or adaptations in sensory organelles (e.g., baroreceptors) and afferent nerve traffic (12), or both.
There are a few limitations to the present investigation that warrant discussion. The conditions during which HRV was assessed were performed in the same order each testing session. Although randomization of the conditions may have reduced the possibility of seeing an order effect of the conditions, we believed that the large hemodynamic changes elicited by TILT may have hindered the sensitivity to detect training-induced changes during PB. Thus in our protocol, the PB condition always preceded the TILT condition. Additionally, it is possible that the changes in cardiac autonomic modulation occurred earlier during the second week of training but were only detected during the final assessment as a result of the extra day of rest preceding this testing session. However, this hypothesis somewhat contradicts the findings of Furlan et al. (9) that 48 h allows for sufficient recovery of cardiac autonomic modulation.
In conclusion, our data suggest that eight endurance exercise-training sessions performed over 2 wk significantly alters cardiac vagal modulation. We reported changes in HRV during PB and TILT that are consistent with a shift toward greater parasympathetic and/or lesser sympathetic modulation of the heart after training. Furthermore, these data suggest that it takes at least eight exercise-training sessions within a 2-wk period to alter HRV. Our ability to only detect changes in autonomic modulation during PB and TILT suggests that either or both of these conditions should be included in any longitudinal investigation examining autonomic modulation of the heart. Lastly, in light of the fact that HRV is associated with risk for cardiac arrhythmic events, these findings may have important implications regarding exercise prescription in both healthy and at-risk populations.
1. Akselrod, S., D. Gordon, F. A. Ubel, D. C. Shannon, A. C. Barger, and R. J. Cohen. Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat-to-beat cardiovascular control. Science 213: 220–222, 1981.
2. Al-Ani, M., S. M. Munir, M. White, J. Townend, and J. H. Coote. 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. 74: 397–403, 1996.
3. American College of Sports Medicine. Guidelines for Exercise Testing and Prescription
. Philadelphia: Lippincott, Williams, & Wilkins, 2000, p. 66.
4. Berntson, G. G., J. T. Bigger, D. L. Eckberg, et al. Heart rate variability: origins, methods, and interpretive caveats. Psychophysiology 34: 623–648, 1997.
5. Dekker, J. M., E. G. Schouten, P. Klootwijk, J. Pool, C. A. Swenne, and D. Kromhout. Heart rate variability from short electrocardiographic recordings predicts mortality from all causes in middle-aged and elderly men: the Zutphen Study. Am. J. Epidemiol. 145: 899–908, 1997.
6. De Meersman, R. E. Respiratory sinus arrhythmia alteration following training in endurance athletes. Eur. J. Appl. Physiol. 64: 434–436, 1992.
7. De Schryver, C., and J. Mertens-Strythagen. Heart tissue acetylcholine in chronically exercised rats. Experientia 31: 316–318, 1975.
8. De Souza, C. G. G. T., L. C. Michelini, and D. R. Fior-Chadi. Receptor changes in the nucleus tractus solitarii of the rat after exercise training. Med. Sci. Sports Exerc. 33: 1471–1476, 2001.
9. Furlan, R., S. Piazza, S. Dell’Orto, et al. Early and late effects of exercise and athletic training on neural mechanisms controlling heart rate. Cardiovasc. Res. 27: 482–488, 1993.
10. Hammond, H. K., F. C. White, L. L. Brunton, and J. C. Longhurst. Association of decreased myocardial beta-receptors and chronotropic response to isoproterenol and exercise in pigs following chronic dynamic exercise. Circ. Res. 60: 720–726, 1987.
11. Hayano, J., S. Mukai, M. Sakakibara, A. Okada, K. Takata, and T. Fujinami. Effects of respiratory interval on vagal modulation of heart rate. Am. J. Physiol. 267: H33–H40, 1994.
12. Hedelin, R., P. Bjerle, and K. Henriksson-Larson. Heart rate variability in athletes: relationship with central and peripheral performance. Med. Sci. Sports Exerc. 33: 1394–1398, 2001.
13. Hickson, R. C., J. M. Hagberg, A. A. Ehsani, and J. O. Holloszy. Time course of the adaptive responses of aerobic power and heart rate to training. Med. Sci. Sports Exerc. 13: 17–20, 1981.
14. Howley, E., and B. D. Franks. Health Fitness Instructor’s Handbook. Champaign, IL: Human Kinetics, 1997, pp. 34–36.
15. Levy, W. C., M. D. Cerqueira, G. D. Harp, et al. Effect of endurance exercise training on heart rate variability at rest in healthy young and older men. Am. J. Cardiol. 82: 1236–1241, 1998.
16. Loimaala, A., H. 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.
17. Meredith, I. T., G. L. Jennings, M. D. Esler, et al. Time-course of the antihypertensive and autonomic effects of regular endurance exercise in human subjects. J. Hypertens. 8: 859–866, 1990.
18. Mier, C. M., M. J. Turner, A. A. Ehsani, and R. J. Spina. Cardiovascular adaptations to 10 days of cycle exercise. J. Appl. Physiol. 83: 1900–1906, 1997.
19. Pagani, M., F. Lombardi, S. Guzzetti, et al. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ. Res. 59: 178–193, 1986.
20. Pereira, M. A., S. J. Fitzgerald, E. W. Gregg, et al. A collection of physical activity questionnaires for health-related research. Med. Sci. Sports Exerc. 29: S10–S14, 1997.
21. Pichot, V., F. Roche, J. M. Gaspoz, et al. Relation between heart rate variability and training load in middle-distance runners. Med. Sci. Sports Exerc. 32: 1729–1736, 2000.
22. Spinelli, L., M. Petretta, F. Marciano, et al. Cardiac autonomic responses to volume overload in normal subjects and in patients with dilated cardiomyopathy. Am. J. Physiol. 277: H1361–H1368, 1999.
23. 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.
24. Tsuji, H., M. G. Larson, F. J. Venditti, et al. Impact of reduced heart rate variability on risk for cardiac events: the Framingham Heart Study. Circulation 94: 2850–2855, 1994.
25. Wilmore, J. H., P. R. 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.
26. Winder, W. W., J. M. Hagberg, R. C. Hickson, A. A. Ehsani, and J. A. Mclane. Time course of sympathoadrenal adaptation to endurance exercise training in man. J. Appl. Physiol. 45: 370–374, 1978.
27. Yamamoto, K., M. Miyachi, T. Saitoh, A. Yoshioka, and S. Onodera. Effects of endurance training on resting and post-exercise cardiac autonomic control. Med. Sci. Sports Exerc. 33: 1496–1502, 2001.
28. Yang, R. C., G. W. Mack, R. R. Wolfe, and E. R. Nadel. Albumin synthesis after intense intermittent exercise in human subjects. J. Appl. Physiol. 84: 584–592, 1988.
AUTONOMIC NERVOUS SYSTEM; SYMPATHETIC; PARASYMPATHETIC; FITNESS; VAGAL MODULATION
©2003The American College of Sports Medicine
What does "Remember me" mean?
By checking this box, you'll stay logged in until you logout. You'll get easier access to your articles, collections,
media, and all your other content, even if you close your browser or shut down your
To protect your most sensitive data and activities (like changing your password),
we'll ask you to re-enter your password when you access these services.
What if I'm on a computer that I share with others?
If you're using a public computer or you share this computer with others, we recommend
that you uncheck the "Remember me" box.
Highlight selected keywords in the article text.
Data is temporarily unavailable. Please try again soon.