Eur J Appl Physiol (1995) 70:75-80

Springer-Verlag 1995

Stephen H. Boutcher' Phyllis Stein

Association between heart rate variability and training response

in sedentary middle-aged men

Accepted: 27 July 1994

Abstract The effect of exercise training on heart rate

variability (HRV) and improvements in peak oxygen
consumption (gO2peak) was examined in sedentary
middle-aged men. The HRV and absolute and relative
gO2pea k of training (n = 19) and control (n = 15) subjects were assessed before and after a 24-session moderate intensity exercise training programme. Results indicated that with exercise training there was a significantly increased absolute and relative gO2peak
(P < 0.005) for the training group (12% and 11% respectively) with no increase for the control group. The
training group also displayed a significant reduction in
resting heart rate; however, HRV remained unchanged.
The trained subjects were further categorized into high
(n = 5) and low (n = 5) HRV groups and changes in
gO2pea k were compared. Improvements in both absolute and relative l?O2peak were significantly greater
(P > 0.005) in the high HRV group (17% and 20%
respectively) compared to the low HRV group (6% and
1% respectively). The groups did not differ in mean age,
pretraining oxygen consumption, or resting heart rate.
These results would seem to suggest that a short aerobic
training programme does not alter HRV in middle-aged
men. Individual differences in HRV, however, may be
associated with l/O2peak response to aerobic training.
Key words Heart rate variability" Aerobic training
Peak oxygen consumption
Introduction

One of the most pronounced cardiovascular adaptations to aerobic exercise is a lowered resting heart rate
S.H. Boutcher (I~)
P.O. Box 1144, Department of Biomedical Science,
University of Wollongong, Wollongong, NSW 2500, Australia
P. Stein
Washington University Medical Center, St. Louis, MQ, USA

(HR). For instance, it has been reported that trained

runners often possess resting HR of less than 50 beats" min-1 (Costill 1986). Similarly, lower resting HR
have been shown to occur in sedentary individuals after
they have been exposed to aerobic training (Maciel
et al. 1985). One mechanism which has been proposed
to explain the relationship between training and resting
bradycardia has been an increase in parasympathetic
or vagal tone (Kenney 1985; Smith et al. 1987). However, support for the influence of increased vagal tone
on the heart after training is equivocal.
For instance, results of some studies using pharmacological blockade (e.g. atropine) have indicated that
trained compared to untrained subjects possess higher
vagal tone (Ekblom et al. 1973; Smith et al. 1987).
However, other blockade studies have not detected
differences in vagal tone between these groups (Lewis
et al. 1980; Maciel et al. 1985). Furthermore, studies
using heart rate variability (HRV) as an indirect assessment of vagal tone have also produced equivocal results. For example, Maciel et al. (1985), Reiling and
Seals (1988), and Furlan et al. (1993) have not found
that trained individuals possessed greater influence of
vagal tone on the heart, whereas De Meersman (1992)
and Dixon et al. (1992) have.
Factors that may contribute to the inconsistency of
these results are the variety and validity of methods
used to assess vagal sensitivity, the predominant use of
cross-sectional studies, and the use of young individuals as subjects who already possess average or
above average levels of influence of vagal tone on the
heart. For example, it is possible that training may not
substantially influence parasympathetic outflow in individuals who already possess high vagal tone (e.g.
young, healthy men). Thus, the existence of a maximum
may offset the potential increase in vagal tone generated through aerobic training. Consequently, individuals possessing lower levels of vagal tone may show
a greater resting vagal response to training. As vagal
tone has been shown to decrease with age (Hellman

76

and Stacey 1976) the training of older individuals may

demonstrate greater changes in HRV.
A related issue to the exercise parasympathetic influence relationship is whether pretraining vagal tone
influences fitness improvement in response to aerobic
training. For instance, some individuals show substantial increases in maximal oxygen consumption
(l?O2max) after training, whereas others show little improvement (Bouchard 1984). Past research examining
this issue has focused on phenotype factors such as
muscle fibre distribution (Costill et al. 1976), substrate
availability (Dalrymple et al. 1973), and heart size (Personen et al. 1975). The association between pretraining
HRV and l?O2max response to training, however, has
not been examined.
Therefore, the purpose of the present study was to
assess the influence of aerobic training on HRV of
middle-aged men undertaking an aerobic training programme. A second purpose was to examine the association between pretraining HRV and peak oxygen consumption (VO2,peak) response to aerobic training.

Methods

Table 1 Biometric data

Age (years)
Height (cm)
Mass (kg):
pretraining
Mass (kg):
post-training

Trained (n = 19)

Untrained (n = 15)

Mean
46.2
178.6
87.1

Mean
45.0
179.6
83.9

85.6

SEM
1.4
1.19
1.4
3.3

84.9

SEM
1.4
1.9
2.4
2.3

assessed by a Rayfield gasometer, that was interfaced to an Apple II

computer. Expired gas was firstly channeled into a 7-1 mixing chamber
and then analysed using a calibrated Applied Electrochemistry S-3A
oxygen analyser and a Beckman LB-3 carbon dioxide analyser. Outputs from the analysers, which were calibrated before each test, were
continuously integrated to the computer using Rayfield REP200c software. Oxygen consumption was determined at 15-s intervals.

Experimental procedures
All procedures were approved by a Human Investigation Committee. The subjects were tested at the same time of day for both preand post-testing and were instructed to refrain from eating, smoking,
and ingesting caffeine or alcohol for at least 3 h before the test.

Subjects

Maximal exercise capacity treadmill protocol

Middle-aged, male sedentary subjects were randomly assigned to
either a training (n = 25) or control group (n = 15). Table 1 shows
subject characteristics. All the subjects were screened by a physician
before undertaking a maximal exercise treadmill test. A total of 19
subjects completed the exercise programme of the original 25 subjects in this group. 1 subject demonstrated supraventricular
tachycardia at maximal exercise, whereas the remaining 5 dropped
out of the study because of scheduling conflicts and health problems.
The subjects that completed the training programme were then
categorized into high (n = 5) and low (n = 5) HRV groups (Table 3)
based on their pretraining resting HRV.

Apparatus

Electrocardiogram and respiration

The electrocardiogram (ECG) and respiratory pattern were recorded using a Grass physiograph (model 7D) that was linked to an
A-D converter (R.C. Electronics, model ISC-67) and a 386 PC
computer collecting at 1,000 samples, s- 1. For ECG, two electrodes
were positioned on the fifth rib either side of the chest and a third to
the upper back to the right of the third vertebra of the thorax.
Respiratory pattern and rate were assessed by recording the temperature of respired gas. A nasal thermistor (Grass model TCT 1R) that
was attached to the main port of a plastic mask produced a signal in
millivolts in response to the change in temperature of the respired
gas.

The subjects reported to an exercise physiology laboratory and

performed a 1202pe~k treadmill test (Balke protocol). The subjects
walked on a treadmill at 3.3 mph and the gradient was increased by
1% every minute until the subject felt exhausted (Balke and Ware
1959).

HR V assessment
Prior to the maximal exercise test the subjects underwent resting HR
and HRV assessment. The subjects completed a health questionnaire and then ECG electrodes, the mask containing the thermistor,
and headphones were attached. These procedures took approximately 30 min. All explanations and instructions were tape-recorded,
and the same tape used for pre- and post-test. The subjects were
instructed to perform paced breathing (i.e. to breathe in and out
continuously to a count of 4 s in and 4 s out) and they practised at
the beginning of the session until they could produce a continuous
breathing pattern with a consistent tidal volume. The voltage associated with this tidal volume was recorded and then monitored
during the test to ensure that the subjects produced a voltage
commensurate with the tidal volume they had recorded during the
practice. During the test, the breathing pattern was also monitored
and abnormal respiratory patterns (e.g. coughing, breath-holding)
were excluded from the data analysis. All the subjects displayed
similar breathing patterns before and after the training programme.
The assessment period consisted of 15 min of quiet sitting during
which ECG was recorded for 64 s beginning at min 5, 10, and 12, and
for 90 s beginning at min 13.

Peak oxygen consumption

Treatment
The subjects wore a Hans-Rudolph face mask while walking on
a Quinton model Q65 treadmill. Oxygen consumption was recorded
through standard open-circuit spirometry. Inspired ventilation was

The intervention consisted of 24 supervised exercise sessions. The

subjects exercised three times each week at an intensity of 60% of

77
HR range (determined through baseline and maximal exercise HR).
During exercise the subjects wore pulse monitors (Uniq Heartwatch,
model 8799) to ensure they exercised at the appropriate intensity.
The exercise session consisted of a 0.25-mile walking warm-up,
a series of stretches, an aerobic exercise period, a 0.25-mile cooldown, and a repeat of the stretching. The aerobic period was 20 min
for the first 3 sessions, 25 min for the next 3 sessions, and 30 rain
from the 9th through to the 25th sessions. Most subjects walked and
jogged for the aerobic period but 4 used an exercise cycle all or part
of the time.

Determination of measurements

with a 51-point cubic polynomial. Next a band-pass filter was used

to remove sources of variance below the two major oscillatory HR
spectrum components. One of these components is termed high
frequency and is synchronized with respiration and typically occurs
at frequencies at 0.12 Hz and above. The other component, called
Mayer waves, is termed low frequency and is centred at around
0.10 Hz and below. The slow frequency characteristics of the Mayer
waves have been shown to correspond to the slow oscillations
present in arterial pressure variability (Furlan et al. 1993). Finally,
the natural logarithm of the band-passed variance (in milliseconds
squared) was calculated and used as high and low frequency
measures of HRVts. These estimates of HRVu appear as a linear
scale ranging from 0 (minimal HRVu) to 10 (maximal HRV,~).
A more detailed explanation of this technique can be found in
Porges (1985).

Interbeat interval and HR

Interbeat interval (IBI) was determined beat-by-beat as the difference in the time of peak voltage between two successive R-wave. The
HR was calculated from IBI. Data was stored on a hard disk and IBI
was calculated by the Datapac analysis program (Run Technology,
Calif.)

Calculation of HR V by peak-to-trough method

The HRV was calculated by the peak-to-trough method (HRVpt)
and was determined by subtracting the smallest from the largest IB1
within each 8-s respiratory cycle. This number in milliseconds was
then averaged across respiratory cycles (Grossman 1983).

Peak oxygen consumption

The gO2pea k was determined as the highest 1-min average oxygen

consumption in litres per minute. The end point was the inability to
continue because of severe fatigue. Maximal HR for both groups
exceeded age-estimated maximal HR and did not differ between
groups before or after training. Similarly, maximal HR within
groups was similar before and after training.

Results

The aerobic training results and the relationship between high and low group HRV a n d ~/~O2peak change
are reported separately for HRV, HR a n d gO2pea k.

Calculation of HRV by time series method

The HRV was also calculated using a time series method (HRVts)
through the M X E D I T software package (Delta-Biometrics, Inc.,
Bethesda, Md.). The IBI were converted into time-based data by
sampling successive 200-ms intervals. The IBI were then plotted on
a personal computer, and edited to remove and interpolate artefact
and outlying values. This technique has been developed to control
for the statistical limitations when extracting HRV amplitude that is
super-imposed on a fluctuating baseline trend (see Porges 1985).
Leakages of non-stationarities and slow periodicities into the frequency component were avoided by detrending the time-based data

Aerobic training and HRV

Treadmill gO2peak

Analysis of variance revealed a significant increase,

P < 0.05, in both absolute and relative treadmill
gO2pea k for the training group compared to the control
group (see Table 2). The training group showed a 12%

Table 2 Heart rate (HR) absolute and relative peak oxygen consumption (gO2peak) heart rate variability peak-to-trough (HRV;0 and heart
rate variability time series (HRVts) of training and control subjects before and after training
Pretraining

Post-training

HR

VO2peak

gO2pea k

(beats"
min - *)

( l ' m i n - i)

(ml" kg- 1.
min - 1)

HRVpt
(ms)

High
HRVt~
(ms 2)

Low
HRV,~
(ms 2)

HR

WO2pea
k

[/O2peak

(beats.
rain - 1)

( l m i n - 1)

(ml.kg- 1.
min ~ 1)

Training
(n = 19)
Mean 71.3
SEM 2.3

2.9
0.01

34.3
2.24

169.5
20.1

5.92
0.32

4.82
0.25

67.9 a
2.3

3.3 b
0.16

Controls
(n = 15)
Mean 73.1
SEM 2.1

2.9
0.01

34.2
2.42

155.7
16.5

5.49
0.28

4.32
0.31

77.3
2.5

2.9
0.10

HRVpl
(ms)

High
HRVt~
(ms 2)

Low
HRVts
(ms 2)

38.6 b
2.6

181.9
21.6

5.87
0.31

4.72
0.27

34.2
2.48

128.9
20.0

5.31
0.35

3.84
0.38

High HRVt~, high frequency component of the spectrum power; Low HRVts, low frequency component of the spectrum power
ap < 0.05 significantly lower than pretraining
bp < 0.05 significantly greater than pretraining

78
Table 3 Heart rate and absolute and relative ltO2peak of high medium and low HRV groups after training
Pretraining

High HRV

Post-training

HR
(beats. m i n - 1)

1202peak
(1. rain- 1)

l)'O2peak
(ml' kg- 1. rain- 1)

HR
(beats- m i n - 1)

gO2pea k
(1. m i n - 1)

~O2peak
(ml" kg- 1. rain- 1)

Mean
69.4

SEM
2.5

Mean
2.9

SEM
0.18

Mean
32.6

SEM
2.6

Mean
64.7

SEM
2.8

Mean
3.5*

SEM
0.3

Mean
41.0"

SEM
3.9

70.0

2.0

3.1

0.2

36.1

2.2

65.6

1.7

3.37

0.3

40.0

2.6

74.1

2.8

2.8

0.2

34.1

1.9

73.1

1.2

2.97

0.2

34.4

1.3

(n = 5)
Medium HRV
(n = 9)
Low HRV

(n = 5)
*P < 0.05 Significantly greater than pretraining
For definition see Table 2

gain in absolute and a 11% gain in relative treadmill

gO2pea k, These results indicated a clear improvement
in aerobic fitness for the training group (Table 2). No
significant change in body mass had occurred after the
training period for either of the two groups.

icantly greater than that of the low HRV group [mean

4.56 (SEM 0.34) ms2: mean 3.6 (SEM 0.14) ms 2 respectively] and medium HRV group subjects [mean 5.7
(SEM 0.18) ms2: mean 4.8 (SEM 0.2) ms 2 respectively]
(P < 0.05).

Heart rate

Heart rate

Post-test HR were significantly lower for the training

group, P < 0.05 (Table 2). Mean baseline pretest HR
for the training group and the control group were not
statistically significant. On post-test, however, baseline
HR was significantly lower for the training group, and
not significantly different from the pretest for the controls (Table 2).

Although there was a trend for the high HRV group to

exhibit lower post-training HR no significant difference
existed between groups (Table 3).

H R variability

Means and standard errors for pre and post HRVpt and
for the high and low frequencies of HRVts for the
training and control groups are shown in Table 2. Analysis of variance, P > 0.05, did not reveal any significant differences before and after training for either
group. The H R V p t and the high frequency of the H R V t s
were significantly correlated (r > 0.90; P < 0.05) both
before and after training.

Peak oxygen consumption

Pre to post increases in both absolute and relative

rO2pea k of the high HRV group were significantly
greater (P < 0.05) than that of the low HRV group. The
high HRV group increased absolute and relative
1202peak by 17% and 20% respectively, whereas the
low HRV group increased l/O2p~,k by 6% and 1%
respectively (Table 3). Thus, higher HRV was associated with a greater response to aerobic training. For
all trained subjects pre-training resting HRV was significantly correlated with improvements in VO2pe~k
after training (r = 0.66; P < 0.01) but was not correlated with pre-training absolute or relative 1202p~,k
(r = < 0.20; P > 0.05).

High versus low HRV gO2pea k response

Heart rate variability

Discussion

HRVpt of the high HRV group [mean 266.9 (SEM 28.2)

ms] was significantly greater than that of the low HRV
group [mean 86.7 (SEM 7.5) ms] and medium HRV
subjects [mean 139 (SEM 3.9) ms] (P < 0.05). Similarly,
both the high and low frequency components of the
H R V t s of the high HRV group [mean 7.70 (SEM 0.24)
ms2: mean 6.13 (SEM 0.3) ms 2 respectively] were signif-

The present study attempted to assess the influence of

aerobic training on HRV and the relationship between
1202veak increase and pretraining HRV. Results agree
with previous reports that have suggested that aerobic
training does not increase HRV (Maciel et al. 1985;
Reiling and Seals 1988). The data also indicated that
those individuals possessing higher HRV responded to

79
aerobic training with greater increases in ]~O2peak compared to those possessing low HRV.
In the present study those individuals completing the
aerobic training programme recorded a significant increase in both absolute and relative gO2pea k and
a significant decrease in resting HR. However, HRV
was not significantly increased after training. Thus,
these data support the results of prior research that
has demonstrated no HRV change with exercise in
young men (Reiling and Seals 1988). It is noteworthy
that in the present study the pre-exercise HRV level
was substantially lower than that of subjects in the
aforementioned studies. Thus, the notion that a maximal vagal tone may exist to limit the increases in vagal
tone to training, was not supported. As the training
regimen in the present study was relatively short,
it may be that changes in vagal tone only occur after
longer training periods. For instance, Seals and
Chase (1989) have conducted a 30-week walk/jog
programme with middle-aged men and have found that
when HRV was expressed as the difference between
the minimal and maximal IBI there was no difference
between the groups after the training programme.
However, when expressed as the standard deviation
of the IBI interval, HRV was increased slightly but
significantly.
The equivocal results of the Seals and Chase (1989)
study may have been a function of the sensitivity of the
HRVptmethod (Byrne and Porges 1993; Lawrence et
al. 1992) or that the training programme needed to be
longer. For instance, De Meersman (1993), has examined HRV of active and sedentary elderly subjects and
has found that the HRV of runners aged between 55
and 65 years was significantly higher than that of a sedentary control group. However, l/O2m,x of the active
group was not significantly higher than that of sedentary subjects and appeared to be extremely low for
regularly exercising individuals of this age. In contrast,
cross-sectional studies comparing resting HRV of
highly trained younger athletes to untrained individuals have not found differences in resting HRV
(Furlan et al. 1993; Realing and Seals 1988).
These data would seem to suggest that other mechanisms besides an increase in vagal outflow may be
responsible for the decrease in resting HR occurring
through aerobic training. For instance, several studies
have indicated that the resting bradycardia of trained
individuals may be mediated by adaptations that result
in a decrease in intrinsic HR (Lewis et al. 1980; Sutton
et al. 1967). Also it has been suggested the decreased
sympathetic activity and responsiveness of the heart
may also influence the bradycardia response (Hammond et al. 1987). Similarly, the potential role of atrial
stretch receptors cannot be excluded. As stroke volume
is significantly increased through aerobic training, it is
feasible that the increased filling of the heart may bring
about what Frick et al. (1967) have termed a "reverse
Bainbridge effect."

The data comparing the gO2pea k improvement

of the high and low HRV subjects suggested that the
level of pretraining vagal tone, as measured by HRV,
may be associated with the magnitude of the [/O2pea k
response. The high HRV subjects recorded significantly
greater increases in both absolute and relative
1202veak compared to that of the low HRV subjects.
These data provide preliminary evidence to suggest
that pretraining levels of HRV may be an indicator
of cardiovascular response to aerobic training.
Althoug.h in this study resting HRV was not correlated
with VO2veak, Kenny (1988) has found that
resting HRV was highly correlated to l/O2max in
college-age men. Thus, the ease of HRV assessment
together with the present absence of genetic markers
suitable to type an individual's sensitivity to training
would seem to warrant further research into the
suitability of HRV to predict the aerobic fitness
response.
The mechanisms underlying the relationship between HRV and the training response are at present
unclear. Indeed, the mechanisms underlying HRV itself
are poorly identified. For instance, cardio-respiratory
generators, chemoreceptor and baroreceptor reflexes,
pulmonary stretch reflexes, central vagal drive, and
metabolic factors have all been believed to influence the
genesis ofHRV (Berntson et al. 1993; Daly 1985; Grossman 1983; Richter and Spyer 1990). From a cardiovascular adaptation perspective it is possible that the
association between higher HRV and improvement in
1202p~ak may reflect the ability of the cardiovascular
system to adapt to aerobic training. As an increase in
stroke volume is generally necessary for an increase in
1202max, elasticity of the left ventricle would seem to be
an important prerequisite for an enhanced training
response. Thus, normaI or high HRV may reflect the
distensibility of the left ventricle, aorta, and other blood
vessels. As the pericardium has also been found to play
a restricting role in the ability of the left ventricle to
distend (Mitchell et al. 1968) it is also possible that
greater HRV may be associated with greater pericardial compliance.
In summary, these data extend recent findings
concerning the relationship between aerobic training
and HRV. The failure to find an increase in HRV after
training in middle-aged men suggests that even with
subjects that have lower pre-exercise HRV, exercise
does not result in an HRV increase. Thus, these results
suggest that a short aerobic training programme does
not alter HRV. In contrast, improvement in gO2pea k
was significantly greater for the high HRV compared
to the low HRV group. Thus, individual differences
in HRV may be associated with the lzO2v~ak response
to aerobic training. The mechanisms underlying
training-induced resting bradycardia and the relationship between HRV and VO2peak response appear
to be important factors for consideration in future
research.

80

References
Balke B, Ware RW (1959) An experimental study of fitness of air
force personnel. U.S. Armed Forces Med J 10:675-682
Berntson GC, Cacioppo JT, Quigley KS (1993) Respiratory sinus
arrhythmia: autonomic origins, physiological mechanisms, and
psychophysiological implications. Psychophysiology 30:183 196
Bouchard C (1984) Genetics of aerobic power and capacity. In
Mallina R, Bouchard C (eds) Sport and human genetics. Human
Kinetics, Champaign, Ill., pp 59 88
Byrne EA, Porges SW (1993) Data-dependent filter characteristics of
peak-valley respiratory sinus arrhythmia estimation: a cautionary note. Psychophysiology 30:397 404
Costill D (1986) Inside running: basics of sports physiology. Benchmark Press, Indianapolis, pp 15
Costill D, Fink WJ, Pollock ML (1976) Muscle fiber composition
and enzyme activities of elite distance runners. Med Sci Sports
8:96-100
Dalrymple RH, Kastenschmidt LL, Cassens RG (1973) Glycogen
and phosphorylase in developing red and white muscle. Growth
37:19-34
Daly M de B (1985) Interactions between respiration and circulation. In: Cherniack NS, Widdicombe JG (eds). Handbook of
physiology: the respiratory system 11 American Physiological
Society, Bethesda, Md., pp 529-594
De Meersman RE (1992) Respiratory sinus arrhythmia alteration
following training in endurance athletes. Eur J Appl Physioi
64:434-436
De Meersman RE (1993) Heart rate variability and aerobic fitness.
Am Heart J 125:726-731
Dixon EM, Kamath MV, McCartney N, Fallen EL (1992) Neural
regulation of heart rate variability in endurance athletes and
sedentary controls. Cardiovasc Res 26:713 719
Ekblom B, Kilbom A, Soltysiak J (1973) Physical training, bradycardia, and autonomic nervous system. Scand J Clin Lab Invest
32:251-256
Frick MH, Elovainio RO, Somer T (1967) The mechanism of bradycardia evoked by physical training. Cardiologia 51:46-54
Furlan R, Piazza S, Dell'Orto S, Gentile E, Cerutti S, Pagani M,
Malliani A (1993) Early and late effects of exercise and athletic
training on neural mechanisms controlling heart rate. Cardiovasc Res 27:482-488
Grossman P (1983) Respiration, stress, and cardiovascular function.
Psychophysiol 20:284-300

Hammond HK, White FC, Brunton LL, Longhurst JC (1987) Association of decreased myocardial fi-receptors and chronotropic
response to isoproterenol and exercise in pigs following chronic
dynamic exercise. Circ Res 60:720 726
Hellman JB, Stacey RW (1976) Variation of respiratory sinus arrhythmia with age. J Appl Physiol 41:734-738
Kenney WL (1985) Parasympathetic control of resting heart rate:
relationship to aerobic power. Med Sci Sports Exer 17:
451 455
Kenney WL (1988) Endurance training increases vagal control of
heart rate. In: Dotson CO, Humphrey JH (eds). Exercise physiology: current selected research. AMS Press, New York, pp
59-65
Lawrence GP, Home PD, Murray A (1992) Repeatability of
measurements and sources of variability in tests of cardiovascular autonomic function. Br Heart J 68:205-211
Lewis SF, Nylander E, Gad P, Areskog NH (1980) Non-autonomic
component in bradycardia of endurance trained men at rest and
during exercise. Acta Physiol Scand 109:297-305
Maciel BC, Gallo L, Marin Neto JA, Lima Filho EC, Mancoy JC
(1985) Parasympathetic contribution to bradycardia induced by
endurance training in man. Cardiovasc Res 19:642-648
Mitchell JH, Mierzwiak DS, Wildenthal K, Willis WD, Smith AM
(1968) Effect on left ventricular performance of stimulation of an
afferent nerve from muscle. Circ Res 22:507 516
Personen E, Norio R, Sarna S (1975) Thickenings in the coronary
arteries in infancy as an indication of genetic factors in coronary
heart disease. Circulation 51:218-225
Porges SW (1985) Method and apparatus for evaluating rhythmic
oscillations in aperiodic physiological response systems. U.S.
Patent no. 4,510,994
Reiling MJ, Seals DR (1988) Respiratory sinus arrhythmia and
carotid baroreflex control of heart rate in endurance athletes and
untrained controls. Clin Physiol 8:511-519
Richter DW, Spyer KM (1990) Cardiorespiratory control. In: Loewy
AD, Spyer KM (eds) Central regulation of autonomic functions.
Oxford University Press, New York, pp 189-207
Seals DR, Chase PB (1989) Influence of physical training on heart
rate variability and baroreftex circulatory control. J Appl Physiol
66:1886-1895
Smith ML, Hudson DL, Graitzer HM, Raven PB (1987) Exercise
training bradycardia: the role of autonomic balance. Med Sci
Sports Exer 21:40-44
Sutton JR, Cole A, Gunning J, Hickie JB, Seldon WA (1967) Control
of heart rate in healthy young men. Lancet II: 1398-1400