Lack of regular physical exercise is associated with reduced cardiovascular parasympathetic function and increased sympathetic activity in sedentary individuals (6,26). Prolonged bed rest also has pronounced effects on autonomic nervous system balance, with documented increased sympathetic activity and decreased parasympathetic activity (8). In contrast, exercise training induces decreased cardiac sympathetic nervous system activity and increased cardiac parasympathetic measures (4,10,12,30,32). However, there have been contradictory findings showing that exercise training does not affect the autonomic nervous system activity of the participants (5,15). The duration and intensity of the training programs have varied widely between studies, which may partially explain these disparate results. Specifically, exercise training influences cardiac rhythm. It induces sinus bradycardia in resting conditions, and a slower increase in heart rate at any degree of submaximal oxygen uptake resulting from a shift towards parasympathetic dominance (29).
Evidence suggests that a sedentary lifestyle, as well as systematic exercise withdrawal, result in negative mood symptoms, especially feelings of fatigue and depressive symptoms (3,7,19,20). The role of autonomic nervous system activity in the development of these symptoms is not well understood. Some evidence suggests that an elevated sympathovagal balance (higher sympathetic nervous system activity relative to parasympathetic nervous system activity) may predispose individuals to depressive mood symptoms in response to reduced exercise levels (11). However, the changes that occur in autonomic nervous system activity during the course of exercise withdrawal have not been systematically reported.
Autonomic nervous system activity changes can be evaluated using heart rate variability (HRV) analysis through measurement of beat-to-beat variations in R-R interval length (16). The R-R interval length is the interval represented by the space between the R peaks of two consecutive heart beats in the electrocardiogram (ECG). In addition to the noninvasive methodology, an additional advantage of HRV analysis is the possibility of repeated measurements within the same participant. Using HRV analysis, this investigation examined the role of autonomic nervous system activity in the development of negative mood symptoms (including fatigue) that occurs with systematic exercise withdrawal. It has been previously documented that stimulation of the vagus nerve (which would increase parasympathetic activity) causes a reduction in depressive symptoms (21,28). Therefore, there might be a direct relationship between autonomic nervous system activity and mood. To investigate this relationship, it was hypothesized that (a) the exercise-withdrawn group would show greater negative mood symptoms at the conclusion of the intervention, (b) changes in negative mood symptoms in response to exercise withdrawal would be predicted by low levels of parasympathetic activity at baseline, and (c) changes in sympathovagal balance would occur across the 2-wk exercise-withdrawal protocol, resulting in a shift toward more sympathetic activity relative to parasympathetic activity.
Participants (N = 40, age 31.3 ± 7.5 yr, 55% women), recruited via newspaper and local advertisements, were eligible if they had engaged in aerobic exercise for at least 30 consecutive minutes a day, at least 3 d·wk−1, during the past 6 months (25). Exclusion criteria were 1) age < 18 or > 45; 2) use of prescription antiinflammatory or anticoagulant medications, other than over-the-counter medications (i.e., aspirin, acetaminophen, ibuprofen); 3) history of cardiovascular disease; 4) hypertension (blood pressure > 140/90 mm Hg); 5) obesity (body mass index (BMI) > 30 kg·m−2); and/or 6) currently under treatment for psychiatric or psychological disorder. Participants were randomized, independently of baseline autonomic nervous system measures, to either stop (N = 20) or to continue (N = 20) regular aerobic exercise activities, as described previously (3). The duration of the study was set at 2 wk to optimize evaluation of mood changes after exercise withdrawal (3,20). Previous experiments have demonstrated that 2 wk of exercise withdrawal is sufficient for the induction of negative mood symptoms.
The research was approved by the institutional review board of the Uniformed Services University of the Health Sciences, and all participants provided written informed consent. Assessments were obtained during baseline and follow-up 2 wk after baseline. Electrocardiograms were recorded continuously and were analyzed offline for HRV assessments. Questionnaires for negative mood and fatigue were completed after an initial 10-min rest period at baseline and at the 2-wk follow-up. The 10-min rest period was used when the participants first arrived to allow acclimation to the laboratory environment. To evaluate protocol adherence, all participants were equipped during the course of the study with an ambulatory activity monitor (Actiwatch, Mini-Mitter Co, Bend, OR).
Measurements and Instruments
Participants were asked to wear a three-lead ECG monitor (Marquette Medical Systems Series 8500; Milwaukee, WI) for 10 min while sitting quietly. Seven electrodes were positioned on the patient's chest, corresponding to a modified V5, V1, and VF position. Each ECG monitor was automatically calibrated at the beginning of each recording (1 mV = 10 mm).
Autonomic nervous system indices for sympathovagal balance were based on HRV analyses in accordance with guidelines for standardization (28). Frequency domain HRV indices were obtained using spectral analysis by fast Fourier transformation (33), categorized into high-frequency (hf; 0.15-0.40 Hz) and low-frequency (lf; 0.04-0.15 Hz) bands. The power of each frequency band was logarithmically transformed to avoid undue influences of extreme values in parametric statistical analyses and was expressed in ln(ms2). A semiautomatic software program (MARS PC 6.01, GE Medical Systems Information Technologies, 2003) was used for HRV analyses. Before HRV analyses, ECGs were examined by a trained technician to ensure that all abnormal intervals (including supraventricular beats) were removed to obtain an unbiased index of HRV without irregularities (33).
hf HRV is considered to be a relatively pure measure of parasympathetic cardiac input (33). Heuristically, lf HRV has been considered a measure of sympathetic drive to the heart, but evidence suggests that lf HRV may also reflect some degree of vagal cardiac control (33). The ratio of lf/hf was therefore used as the primary index of "sympathovagal balance" to reflect the concurrent and relative influence of both neural axes in cardiac control (23,33).
Fatigue assessments were based on the multidimensional fatigue inventory (MFI) (31). The MFI consists of 20 items and has good internal consistency (Cronbach α = 0.84). The MFI is composed of five scales: general fatigue, physical fatigue, mental fatigue, reduced activity, and reduced motivation, and each of the scales demonstrates good internal consistency (Cronbach α > 0.65) (31). For nonclinical samples, the MFI total score ranges from 35 to 45 (31). Negative mood was assessed with the profile of mood states (POMS) (18). The total mood disturbance score (POMS-TMD) was calculated by summation of the five negative affect scales (fatigue, depression, tension, anger, confusion) and subtraction of the vigor scale. Normative values for the POMS-TMD score ranges from 0 to 40 (22) Depressive symptoms were assessed using the Beck depression inventory II (BDI-II), which has good internal consistency (Cronbach α = 0.92-0.93) (2). Scores greater than 10 on the BDI-II are considered indicative of the possible presence of depression (2).
Initial activity level.
The Aerobics Center Longitudinal Study Physical Activity Questionnaire (13) was used to assess initial physical activity to establish whether physical activity levels of participants were equivalent at the time of study enrollment. The questionnaire evaluates leisure-time exercise participation (including intensity of activities) during the previous 6 months. Reported exercise participation per week was converted to energy expenditure (kcal) estimates by using established metabolic equivalent of task (MET) values for each activity (1).
To document protocol adherence, ambulatory physical activity levels were assessed using an actigraph accelerometer (Actiwatch, Mini-Mitter Co, Bend, OR) throughout the duration of the study (14 d). Participants in the exercise-withdrawal group were instructed not to engage in their usual aerobic exercise activities and were informed that the actigraph accelerometer enabled the investigators to verify protocol adherence. The actigraph is a wristwatch-sized (37 × 29 × 9 mm), lightweight (17 g) device that has been validated previously to assess whole-body movements during daily life activities (24,27). Actigraph signals are based on a piezoelectric sensor that generates a voltage when the device undergoes a change in acceleration (14). Care was taken for proper placement of the actigraph by using a standardized mounting and positioning protocol on the participant's nondominant wrist (14,24).
Activity counts were summed over 5-min epochs and recorded continuously (24 h·d−1). Peak activity levels were defined as the highest 5-min period during the entire 2-wk observation period. Average activity values were calculated as the mean of all 5-min epochs over the course of the study (14). On the basis of previous research, it was hypothesized that a between-groups difference of peak activity would occur during the exercise-withdrawal period, and that the effects on average activity levels would be less (14). The current study removed the participant's high-intensity exercise (peak activity on the actigraph), but it did not alter the daily physical activity of the participants (i.e., walking to work, shopping at a grocery store-which would be reflected as average activity levels on the actigraph) (14).
Data are presented as means ± standard deviations or percentages as appropriate. To compare the exercise-withdrawal group with the control group, 2 × 2 mixed model analyses of variance (ANOVA) were conducted, where the exercise-withdrawal group was compared with the control group (two-level between-subject factors), and assessments over the two study visits (baseline vs follow-up) were included as two-level within-subject factors. Significant main and interaction effects were further examined by using independent and paired t-tests. Data analyzed by ANOVA did not use change scores; instead, they used repeated-measures data for the dependent variables.
To evaluate whether HRV parameters were predictive of mood symptoms at follow-up, hierarchical regression analysis examined mood symptoms at follow-up as the dependent variable and included baseline lf/hf HRV as the primary predictor variable; potentially confounding factors (gender, age, weight, baseline fitness level, and baseline mood symptoms) were adjusted for.
To examine whether changes in HRV were associated with the magnitude of mood symptoms, changes between lf/hf HRV values and scores on questionnaires (MFI, POMS, and BDI-II) at follow-up and baseline were calculated and compared. These change scores were then correlated with changes in HRV and changes in mood symptoms. Multivariate regression analyses were performed when significant bivariate associations were found, to adjust for the aforementioned variables. Data analyzed by correlations and regressions used change scores for the dependent variables. When analyses were statistically significant, the baseline levels were entered to control for the potential association between change scores with the baseline starting point.
Participant characteristics and manipulation check.
Characteristics of the exercise-withdrawal and control groups are shown in Table 1. The groups were comparable with respect to age, gender, BMI, activity level, HRV indices, fatigue, negative mood, and depressive symptoms at baseline (Table 1).
To establish whether the experimental group adhered to the exercise-withdrawal protocol, ambulatory actigraphy data were examined for peak and average activity levels. The exercise-withdrawal group engaged in significantly less peak activity levels (12,794 ± 7778 activity counts per 5 min) than the control group (18,300 ± 6214 activity counts per 5 min; P = 0.02), confirming adherence to the exercise-withdrawal procedure. Consistent with prior studies (14), no statistically significant differences were observed between the experimental and control groups on average daily activity levels, although a trend in the expected direction was observed (976 ± 249 vs 1138 ± 285 activity counts per day, respectively; P = 0.07). Thus, the experimental group engaged in lower high-intensity activity levels than the control group.
Exercise withdrawal and increases in fatigue and negative mood.
Exercise withdrawal resulted in significantly higher self-reported levels of fatigue (MFI) compared with the control condition (Fig. 1; P = 0.03). This finding was supported by a significant interaction between group status and time (F interaction(2,72) = 8.25; P = 0.001). The MFI subscales, accounting for the effects of 2-wk exercise withdrawal, were: general fatigue (10.7 ± 3.7 vs 7.7 ± 3.4; P = 0.01), physical fatigue (8.6 ± 3.4 vs 6.6 ± 2.16; P = 0.03), and activity-related fatigue (8.5 ± 3.5 vs 6.3 ± 2.4; P = 0.02; exercise withdrawal vs control, respectively).
Negative mood levels increased during exercise withdrawal, such that the experimental group displayed higher scores than controls at follow-up (16.0 ± 28.5 vs −1.8 ± 18.8; P = 0.03). In addition, a significant interaction between group status and time on POMS-TMD (F interaction(2,68) = 6.07; P = 0.004) was observed.
Depressive symptoms as documented by the BDI-II revealed similar results. The withdrawal group displayed higher BDI-II scores than did controls at follow-up (4.7 ± 3.3 vs 1.4 ± 2.6; P = 0.01); this was supported by a significant group-by-time interaction for BDI-II scores (F interaction(2,68) = 7.48; P = 0.001).
Baseline HRV and development of fatigue and negative mood.
The relationship between baseline HRV measurements and subsequent changes in negative mood during the 2 wk of the study were examined. Because increases in negative mood occurred only in the experimental group (see above), the remaining analyses focus on the exercise-withdrawal group only.
The lf/hf HRV was positively correlated with the increase in MFI score (r = 0.64; P = 0.004) (Fig. 2). Specifically, increases in the subscales of physical fatigue (r = 0.51; P = 0.03), activity-related fatigue (r = 0.56; P = 0.02), and mental fatigue (0.68; P = 0.002) were related to baseline lf/hf. The results revealed the same pattern when statistically adjusting for fatigue levels at baseline (partial r= 0.64; P = 0.005). Further analyses revealed that the predictive value of the lf/hf ratio was primarily a result of hf HRV as a predictor of ΔMFI (r hf HRV, ΔMFI = −0.60; P < 0.01; r lf HRV, ΔMFI = −0.24; P = 0.33).
With respect to negative mood, the correlation between baseline lf/hf HRV and POMS-TMD increases demonstrated a trend in the same direction (r = 0.47; P = 0.06). Changes in depressive symptoms (ΔBDI-II) were also predicted by baseline lf/hf HRV (r = 0.53; P = 0.03). The relationship between baseline lf/hf HRV and the increase in depressive symptoms remained significant after adjusting for baseline BDI-II levels (partial r = 0.62; P = 0.01). The predictive value of the lf/hf ratio was primarily a result of hf HRV as a predictor of ΔBDI-II (r hf HRV, ΔBDI-II = −0.44; P = 0.08; r lf HRV, ΔBDI-II = −0.18; P = 0.50).
To further examine the independent predictive power of baseline HRV on the future development of symptoms, hierarchical regression analyses were performed. Baseline lf/hf HRV predicted changes in MFI scores at follow-up (R 2 change = 0.41; β = 0.74; P = 0.008) over and above a model that already included gender, age, weight, baseline fitness level, and baseline MFI score. Results in the same direction were found for the POMS-TMD (R 2 change = 0.14; P = 0.14) and BDI-II (R 2 change = 0.23; P = 0.04).
Exercise withdrawal and changes in HRV.
The exercise-withdrawal and control groups displayed no significant change in hf HRV, lf HRV, or the lf/hf HRV during the 2 wk (Table 2). Specifically, there were no statistically significant interactions between group status and time for hf HRV (F interaction (1,25) = 0.33; P = 0.57), lf HRV (F interaction (1,25) = 1.31; P = 0.26), and lf/hf HRV (F interaction (1,26) = 0.26; P = 0.61). Also, the cross-sectional correlations of HRV and mood symptoms at baseline and follow-up were not significant (baseline: r < 0.05; P > 0.85; follow-up: r < 0.25; P > 0.40).
The present investigation documents that baseline HRV measures predict the development of subsequent negative mood and fatigue in response to exercise withdrawal. The exercise-withdrawal paradigm did not result in significant changes in HRV-derived indices of autonomic measures over time. These data suggest that individual differences may exist in the vulnerability to develop negative emotional consequences after a reduction in routine exercise level.
The observation that baseline HRV parameters predict the development of fatigue and depressive symptoms in response to exercise withdrawal is consistent with a report by Glass et al. (11). These investigators documented that development of negative mood symptoms was related to lf/hf ratio, which may represent an increase in parasympathetic activity and/or a decrease in sympathetic activity. Further inspection of the data revealed that the predictive value of the lf/hf ratio primarily reflected the association of lower hf HRV with larger increases in fatigue; this finding supports the importance of reduced parasympathetic activity. In previous research, it has been demonstrated that stimulation of the vagus nerve (increase in parasympathetic activity) caused a decrease in depressive symptoms, indicating a direct connection between autonomic activity and future development of negative mood. The current study documents a relationship between baseline HRV and the development of negative mood symptoms, but causal inferences cannot be made, because of the correlational nature of these associations. Additional investigations with larger samples and more extreme exercise-withdrawal paradigms may further clarify the role of sympathetic versus parasympathetic parameters in the development of negative mood symptoms.
Previous exercise-withdrawal studies have not reported on change in HRV during exercise withdrawal. This investigation showed that 2 wk of voluntary withdrawal of peak exercise levels did not result in significant reductions in HRV parameters (i.e., lf HRV, hf HRV, or lf/hf ratio). However, additional analyses demonstrated that changes in lf/hf HRV from baseline to follow-up were predictive of increases in MFI scores at follow-up (r = −0.61, P = 0.02). These results suggest that the HRV measures consistent with a relative vagal withdrawal over time might have resulted in paradoxically less negative mood symptoms. These findings are not readily interpretable, because of the overall lack of change in HRV; future research with longer exercise-withdrawal periods may reveal significant HRV changes, which would enable investigation of the relationship between changes in HRV parameters and development of negative mood symptoms.
The change in mood that was found in the present investigation was not accompanied by an overall change in HRV. Therefore, other factors may play a role in the development of mood changes in relation to exercise deprivation. For example, the exercise-withdrawn group was denied a pleasurable activity (exercise). The increase in negative mood could be related to the removal of a pleasurable activity rather than a physiological cause. In addition to the lack of pleasurable activities, a lack of exercise has been associated with an increase in levels of circulating inflammatory factors (17), and increases in circulating inflammatory factors may have resulted in depressive symptoms (9).
Exercise withdrawal was partial and was limited to discontinuation of high-level aerobic activities rather than all physical activity. This relatively moderate reduction in overall activity may have attenuated the results' related changes in HRV parameters. A related limitation is that 2 wk of high-activity exercise withdrawal may not be adequate to induce the same biological and behavioral concomitants as would long-term reduced exercise levels among sedentary individuals. Thus, although experimental control of exercise levels and ambulatory assessments of protocol compliance are strengths of the present investigation, the results may not be generalizable to other populations, such as sedentary individuals or individuals exposed to complete withdrawal of physical activity (e.g., those confinement to bed rest).
Another potential limitation is that sympathetic nervous system activity measures were not assessed directly (e.g., by using peroneal nerve activity assessments). Such measures may be useful in detecting subtle autonomic nervous system changes that may emerge in response to exercise-withdrawal paradigms as used in this study.
This study suggests that autonomic nervous system measures could potentially be used for risk stratifying populations exposed to exercise withdrawal. In the current investigation, initial HRV (decreased parasympathetic activity and/or increased sympathetic activity) was predictive of developing negative mood symptoms in response to exercise withdrawal. It is important to note that the mood changes that occurred remained within the normal range of scores on the questionnaires (2,18,31). Vulnerability factors for negative mood after exercise withdrawal may also include fitness level, emergence of initial somatic symptomatology, low-grade inflammatory measures, and lower cortisol levels (3,11). These factors may, therefore, be important to assess in cases where reduction of habitual exercise levels can be anticipated and planned (e.g., postsurgical bed rest, space travel). Future research may identify novel interventions to combat the fatigue and depressive symptoms characteristic of exercise withdrawal. Such interventions could use both pharmacological and/or behavioral strategies to specifically target the vulnerability factors mediated by autonomic nervous system-related processes.
This work was supported in part by a grant from the Charles E. Dana Foundation and from the NIH (HL58638 and T32 HL69751). We thank Micah Stretch and Rachel Hoult for their help in conducting the experiments and data collection.
The opinions and assertions expressed herein are those of the authors and are not to be construed as reflecting the views of the USUHS or the U.S. Department of Defense.
1. Ainsworth, B. E., W. L. Haskell, M. C. Whitt, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med. Sci. Sports Exerc.
32(9 Suppl):S498-S504, 2000.
2. Beck, A. T., R. A. Steer, and C. Brown. In: The Beck Depression Inventory Manual
. 2nd Ed. San Antonio, TX: The Psychological Corporation, pp. 1-38, 1996.
3. Berlin, A. A., W. J. Kop, and P. A. Deuster. Depressive mood symptoms and fatigue following exercise withdrawal: the potential role of decreased fitness. Psychosom. Med.
4. Blomqvist, C. G., and B. Saltin. Cardiovascular adaptations to physical training. Annu. Rev. Physiol.
5. Boutcher, S. H., and P. Stein. Association between heart rate variability and training response in sedentary middle-aged men. Eur. J. Appl. Physiol. Occup. Physiol.
6. Buchheit, M., C. Simon, A. Charloux, S. Doutreleau, F. Piquard, and G. Brandenberger. Heart rate variability and intensity of habitual physical activity in middle-aged persons. Med. Sci. Sports Exerc.
7. Conboy, J. K. The effects of exercise withdrawal on mood states in runners. J. Sport Behav.
8. Crandall, C. G., K. A. Engelke, J. A. Pawelczyk, P. B. Raven, and V. A. Convertino. Power spectral and time based analysis of heart rate variability following 15 days simulated microgravity exposure in humans. Aviat. Space Environ. Med.
9. Dantzer, R. Cytokine, sickness behavior, and depression. Neurol. Clin.
10. 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.
11. Glass, J. M., A. K. P. F. Lyden, P. K. Stein, et al. The effect of brief exercise cessation on pain, fatigue, and mood symptom development in healthy, fit individuals. J. Psychosom. Res.
12. Goldsmith, R. L., J. T. Bigger, Jr., R. C. Steinman, and J. L. Fleiss. Comparison of 24-hour parasympathetic activity in endurance-trained and untrained young men. J. Am. Coll. Cardiol.
13. Kohl, H. W., S. N. Blair, R. S. Paffenbarger, Jr., C. A. Macera, and J. J. Kronenfeld. A mail survey of physical activity habits as related to measured physical fitness. Am. J. Epidemiol.
14. Kop, W. J., A. Lyden, A. A. Berlin, et al. Ambulatory monitoring of physical activity and symptoms in fibromyalgia and chronic fatigue syndrome. Arthritis Rheum.
15. 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.
16. Malliani, A., M. Pagani, F. Lombardi, and S. Cerutti. Cardiovascular neural regulation explored in the frequency domain. Circulation
17. Mattusch, F., B. Dufaux, O. Heine, I. Mertens, and R. Rost. Reduction of the plasma concentration of C-reactive protein following nine months of endurance training. Int. J. Sports Med.
18. McNair, D. M., M. Lorr, and L. F. Droppleman. In: Manual for the Profile of Mood States
. San Diego, CA: Educational and Industrial Testing Service, pp. 1-40, 1992.
19. Mondin, G. W., W. P. Morgan, P. N. Piering, A. J. Stegner, C.L. Stotesbery, and M. R. Trine. Psychological consequences of exercise deprivation in habitual exercisers. Med. Sci. Sports Exerc.
20. Morris, M., H. Steinberg, E. A. Sykes, and P. Salmon. Effects of temporary withdrawal from regular running. J. Psychosom. Med.
21. Nemeroff, C. B., H. S. Mayberg, S. E. Krahl, et al. VNS therapy in treatment-resistant depression: clinical evidence and putative neurobiological mechanisms. Neuropsychopharmacology
22. Nyenhuis, D. L., C. Yamamoto, T. Luchetta, A. Terrien, and A. Parmentier. Adult and geriatric normative data and validation of the profile of mood states. J. Clin. Psychol.
23. Pagani, M., and A. Malliani. Interpreting oscillations of muscle sympathetic nerve activity and heart rate variability. J. Hypertens.
24. Patterson, S. M., D. S. Krantz, L. C. Montgomery, P. A. Deuster, S. M. Hedges, and L. E. Nebel. Automated physical activity monitoring: validation and comparison with physiological and self-report measures. Psychophysiology
. 30:296-305, 1993.
25. Pollock, M. L., G. A. Gaesser, J. D. Butcher, et al. Position stand: the recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults. Med. Sci. Sports Exerc.
26. Pollock, M. L., A. Ward, and J. J. Ayres. Cardiorespiratory fitness: response to differing intensities and durations of training. Arch. Phys. Med. Rehabil.
27. Puyau, M. R., A. L. Adolph, F. A. Vohra, and N. F. Butte. Validation and calibration of physical activity monitors in children. Obes. Res.
28. Schachter, S. C. Vagus nerve stimulation: mood and cognitive effects. Epilepsy Behav.
5(Suppl 1):S56-S59, 2004.
29. Seals, D. R., and P. B. Chase. Influence of physical training on heart rate variability and baroreflex circulatory control. J. Appl. Physiol.
30. Shin, K., H. Minamitani, S. Onishi, H. Yamazaki, and M. Lee. Autonomic differences between athletes and nonathletes: spectral analysis approach. Med. Sci. Sports Exerc.
31. Smets, E. M., B. Garssen, B. Bonke, and J. C. De Haes. The Multidimensional Fatigue Inventory (MFI) psychometric qualities of an instrument to assess fatigue. J. Psychosom. Res.
32. Smith, M. L., D. L. Hudson, H. M. Graitzer, and P. B. Raven. Exercise training bradycardia: the role of autonomic balance. Med. Sci. Sports Exerc.
33. 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