Physical Fitness and Performance

Resting Heart Rate Changes after Endurance

Training in Older Adults: A Meta-Analysis
GUOYUAN HUANG1, XIANGRONG SHI2, JANE A. DAVIS-BREZETTE1, and WAYNE H. OSNESS3
1
Department of Physical Education, University of Southern Indiana, Evansville, IN; 2Department of Integrated
Physiology, University of North Texas Health Science Center, Fort Worth, TX; and 3Department of Health, Sport, and
Exercise Science, University of Kansas, Lawrence, KS

ABSTRACT
HUANG, G., X. SHI, J. A. DAVIS-BREZETTE, and H. WAYNE OSNESS. Resting Heart Rate Changes after Endurance Training
in Older Adults: A Meta-Analysis. Med. Sci. Sports Exerc., Vol. 37, No. 8, pp. 1381–1386, 2005. Purpose: Question remains regarding
endurance training and changes in resting heart rate (HR) among older individuals. The objective of this meta-analysis was to determine
the effects of controlled aerobic training on resting HR among sedentary older adults. Methods: Studies were identified by a systematic
computer database search, hand article search, and cross-reference. The inclusion criteria were (i) controlled clinical trials, (ii)
endurance exercise as the only intervention, (iii) a nonexercise control group, (iv) within-group mean ages of subjects ⱖ60 yr, (v) a
measure of changes in resting HR, (vi) studies published in English journals. Results: Outcome was derived from 13 studies with a
total of 651 individuals in 14 control (N ⫽ 241) and 16 exercise groups (N ⫽ 410). The pooled standardized effect size by a fixed-effect
model showed an upper moderate effect of ⫺0.58 ⫾ 0.08 (mean ⫾ SEM, 95% CI ⫽ ⫺0.74 to ⫺0.42). This homogeneity effect was
statistically significant (P ⫽ 0.001). The magnitude of net change averaged ⫺6 bpm (⫺2 to ⫺12 bpm), representing an 8.4% reduction.
Greater and statistically significant decrease of resting HR among the elderly was found in the studies with training length more than
30 wk. Conclusions: This meta-analytic investigation supports the efficacy of endurance exercise training in decreasing HR at rest in
older adults. This training induced adaptation may have protective benefits for cardiovascular aging. A longer exercise training length,
probably more than 30 wk, may be needed for older individuals to be more effective in terms of resting HR reduction. Key Words:
ELDERLY, AEROBIC, CARDIOVASCULAR, CHANGE, SYSTEMATIC

B
oth longitudinal and cross-sectional studies (26,28) exercise on resting HR, as well as the magnitude of such
have reported that endurance exercise training pro- changes, among older adults.
duces a dominant vagal control of the heart and a In addition, for those reported exercise effect on resting
resting bradycardia in younger adults. It has been considered HR, this training adaptation may be found only in those with
one of the hallmarks of aerobic exercise training. The ex- continued conditioning and longer training time. Resting
isting studies in older population, however, have presented HR in old master athletes who were still competing was
discrepant results with regard to changes of resting heart rate much lower than those who had ceased competing (22).
(HR) as a result of endurance training. Some studies have Subjects younger than 60 yr old may be able to attain a
shown little or no changes in resting HR after endurance significant decrease in resting HR in a short training time
exercise training in older individuals (2,32) despite a sig- (23,25). However, other research indicated that relatively
nificant resting bradycardia established in master athletes brief periods of endurance training exert little lowering
(1,22). The others indicate a greater variation of the mag- effect on resting HR (32). These discrepancies imply that
nitude in resting HR of sedentary old individuals after endurance exercise training may take longer period of time
aerobic training (11,23). The disagreements have raised and/or a greater volume to be effective in elderly people.
questions regarding the efficacy and direction of endurance The purpose of this study was to use meta-analytic
method to determine the effects of controlled endurance
exercise training on resting HR among sedentary older
Address for correspondence: Guoyuan Huang, Ph.D., University of South-
ern Indiana, 8600 University Boulevard, Evansville, IN 47712; E-mail:
adults (age ⱖ 60 yr), the magnitude of the resting HR
[email protected]. changes, and the specific role of endurance exercise com-
Submitted for publication November 2004. ponents, such as intensity, length, and volume, on such
Accepted for publication March 2005. changes. Meta-analysis has been frequently applied in clin-
0195-9131/05/3708-1381/0 ical medicine and health related fields to improve traditional
MEDICINE & SCIENCE IN SPORTS & EXERCISE® methods of narrative review by systematically aggregating
Copyright © 2005 by the American College of Sports Medicine information and quantifying its impact. Furthermore, meta-
DOI: 10.1249/01.mss.0000174899.35392.0c analysis may increase statistical power for primary end
1381
points and for subgroups, resolve uncertainty when reports ing HR, exercise training characteristics, and primary out-
disagree, improve estimates of effect size, and answer ques- comes. Reliability checks were performed throughout the
tions not posed at the start of individual trials. The functions abstraction process. The studies for the reliability testing
of meta-analysis are particularly applicable and useful to were randomly selected from the list. Discrepancies were
randomized controlled trials and investigations in the el- resolved after reviewing, checking, and discussing the re-
derly population, because such studies are often too small to lated information in the article. Because of their research
detect clinically important differences and too easy to be design, some investigations enabled the calculation of mul-
confounded to increase the generalizability when they are tiple study treatment effects. If a study contained more than
presented individually. one experimental group, then each group mean was included
and treated as a single data point.
Data analysis. The standardized effect size for an
METHODS
individual study was calculated according to procedures
Study selection. This research followed the general developed by Hedges and Olkin (16). The standardized
outlines, steps, and recommendations for conducting meta- effect size for each measure was estimated from reported
analysis (7,12). Specific training on computer searches of means and standard deviations in the included studies. It is
data sources was performed before literature search. A ref- the standardized mean difference between interventional
erence librarian/bibliographer who has expertise in health, and control groups expressed in standard deviation units.
biomedical, and exercise sciences was consulted. System- Because all cited studies in this meta-analysis were the
atic computer database searches were performed in elec- controlled clinical trials with comparative parallel control
tronic databases including PubMed–Medline, Sport Discuss, group, net changes in these outcomes were calculated as the
HealthSTAR, Current Contents/Chemical Medicine, and difference (exercise minus control) of the changes (prein-
Dissertation Online. A broad rather than a more specific tervention minus postintervention) in these mean values.
search was chosen to locate relevant articles. The initial Study quality was assessed using a three-item question-
search strategy was to use key words either alone or in naire developed by Jadad et al. (19). Publication bias was
various combinations. Extensive hand searching was con- examined using a funnel plot, by plotting the trials’ effect
ducted including professional books, professional journals estimates against their sample size, to detect the likely
specific in exercise and aging, and position statements of presence of bias in meta-analysis (8). In the absence of bias,
national or international organizations. Cross-referencing the plot should thus resemble a symmetrical inverted funnel.
was performed from the bibliographies of published reviews If the plot shows an asymmetrical and skewed shape, bias
and original codeable articles. Completeness of the refer- may be present (8). The Q-statistic was used to examine
ence list and potential missing articles were sought from homogeneity of results for the standardized effect sizes (16).
experts. Pooled estimates of effect sizes were calculated by using
Only studies that meet the following criteria were se- both the fixed-effect and random-effect models (16). Nine-
lected into the present study: randomized and nonrandom- ty-five percent confidence intervals were calculated for both
ized controlled trials that included a nonintervention control individual standardized effect sizes and pooled results for all
group; aerobic exercise and/or conditioning as the only outcomes from the approach.
exercise intervention or treatment; older human subjects Subgroup analyses were performed on categorical vari-
apparently sedentary before study initiation with mean age ables using independent t-test. Combined effect size
of subjects in the study equal to and/or greater than 60 yr; changes and overall net changes in resting HR were exam-
English-language studies published and/or indexed as jour- ined when data were partitioned according to study design,
nal articles from 1980 to 2002; reported measure of changes subject physical characteristics, and exercise training char-
in resting HR; and training lasting a minimum of 2 wk and, acteristics. Specifically, these analyses included: study year,
frequency, intensity, and duration of the training regimen randomization, country, within-study mean age of subjects,
reported in quantifiable terms. Studies that included a group subject age span or range within the sample, sample size of
that both performed endurance exercise and were adminis- study, gender, initial blood pressure status, exercise length,
tered diet or weight loss were excluded. Relevant abstracts frequency, duration, and intensity. Computations and statis-
from publications, conference proceedings, or dissertations tics related to effect sizes were performed in the Compre-
were checked but excluded. Evaluation of the completeness hensive Meta Analysis (Biostat, Inc.). The Statistical Pack-
and independence of each study’s data was done by reex- age for the Social Sciences (SPSS) application software for
amining all included studies and the extracted information. Windows was used for all other statistical analysis. Unless
Data extraction. A data abstraction form, coding otherwise noted, all data were reported as mean ⫾ standard
sheet, is the primary research tool in a meta-analysis. Thus, deviation. Statistical significance was set at P ⱕ 0.05.
a codebook with operational definitions was developed and
applied during the coding process for data extraction. The
RESULTS
codebook was refined and revised several times to consol-
idate the diverse outcomes and characteristics of studies. Search results. After checking the titles and abstracts
The coded variables included study characteristics, subject and/or reading the entire text, we identified 19 trials that
physical characteristics, assessment characteristics of rest- potentially met the all of the criteria for inclusion. Of these
1382 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org
studies, six were subsequently excluded, because of multi- TABLE 1. Initial physical characteristics of subjects.
ple utilization of the same groups or subjects included in Exercise Control
these publications, or because of other treatment regimens, Variable N Mean SD N Mean SD
which were mixed with aerobic training as part of the
Age (yr) 16 66.6 5.0 14 67.0 5.4
intervention, or missing data of the interested outcome val- Height (cm) 11 165.6 8.0 10 164.9 8.8
ues. Thus, 13 studies (2,3,5,13–15,17,20,24,27,30,31,34) Weight (kg) 14 72.6 7.1 12 69.7 6.9
BMI (kg䡠m⫺2) 9 25.9 0.5 9 25.2 0.6
were included in the final analysis for initial and final Fat (%) 4 27.8 7.3 3 24.2 2.5
resting HR assessment. A total of 651 older subjects were FFM (kg) N/A N/A N/A N/A N/A N/A
RHR (bpm) 16 74.6 6.5 14 71.5 5.6
involved in 16 exercise and 14 control groups, composed of SBP (mm Hg) 13 137.8 13.5 11 135.6 10.8
410 and 241 subjects, respectively. DBP (mm Hg) 13 82.5 8.6 11 81.6 6.3
Characteristics of included studies. The studies were V̇O2max (L䡠min⫺1) 5 1.856 0.338 5 1.850 0.371
V̇O2max (mL䡠kg⫺1䡠min⫺1) 9 24.3 3.0 8 24.6 3.9
published from nine different peer-reviewed journals. The pub- FVC (L) 2 2.436 0.626 2 2.465 0.495
lication dates ranged from 1985 to 2001. Ten studies were FEV1 (L䡠s⫺1) 2 1.712 0.622 2 1.687 0.487
conducted in the United States, two in Australia, and one in N, number of groups reporting data; SD, standard deviation; BMI, body mass index;
FFM, fat free mass; RHR, resting heart rate; SBP, systolic blood pressure; DBP, diastolic
Korea. The majority of the investigations were conducted in blood pressure; V̇O2max, maximum oxygen consumption; FVC, forced vital capacity;
universities or colleges. Approximately 85% of the primary FEV1, forced expiratory volume at 1 s.
authors worked in academia and 15% in clinical settings. Most
studies reported written informed consent from subjects. Seven
(54%) were randomized controlled trials whereas the other six sumption (V̇O2max), percentage of oxygen uptake reserve
(46%) were nonrandomized trials. Only three studies reported (V̇O2R) or HR reserve (HRR), maximal HR per minute, etc.,
using matching procedures. Six studies (46%) reported no with the use of %HRmax and %HRR most common. The mean
treatment for control groups. Three (23%) applied the waiting intensity of training using %HRmax ranged from approximately
list. Four other studies (31%) did not report the status of control 65 to 85% (74 ⫾ 7%) in the eight groups, %V̇O2max ranged
groups. All studies appeared to use an analysis-by-protocol from approximately 53 to 73% (64 ⫾ 8%) in the three groups,
approach in the analysis of their outcome data. The dropout %HRR ranged from approximately 45 to 78% (62 ⫾ 12%) in
rate of the studies ranged from 0 to 23% (mean ⫾ SD, 10 ⫾ the six groups, and HRmax ranged from approximately 100 to
8%), from 0 to 17% (mean ⫾ SD, 5 ⫾ 6%) in the control 115 bpm (108 ⫾ 7 bpm) in two groups. Among those six
groups and 0 to 30% (mean ⫾ SD, 13 ⫾ 11%) in the exercise studies for which data were available, compliance, defined as
groups. The number of subjects who completed each study the percentage of endurance training sessions that the exercise
ranged from 10 to 110 in exercisers (mean of 32 ⫾ 30) and groups attended, ranged from 75 to 97% (88 ⫾ 7%). Of the 13
from 6 to 36 in the controls (mean of 19 ⫾ 11). studies, eight (62%) reported the training sessions supervised,
There was only one study (8%) reporting data based on whereas the remaining five studies (39%) did not report such
gender. Four studies included only female subjects, and all information. The primary methods for intensity monitoring
others used mixed gender. Only one study provided race/ included one or more of the following: subjects self palpation
ethnicity of the subjects (8%). Thus, it was unable to further of radial or carotid pulses, monitored by HR device, rating of
assess changes in resting HR according to ethnicity and perceived exertion, ECG, and checked by investigators and/or
gender. Eleven (85%) reported that all subjects were study staff. The majority of studies used HR measurement to
healthy, whereas the other two reported subjects having monitor intensity levels.
hypertension. All 13 studies provided information on co- Effect of aerobic training on resting HR. The ini-
morbidities. Of the seven studies (54%) that reported such tial physical characteristics of the subjects for the exercise
information, three reported that none of the subjects were and control groups were described in Table 1. Figure 1
taking medication during the study, and four reported that presented the results of the standardized effect size in resting
some of subjects (range from 19 to 46%) were taking HR for individual study and the combined effect size pool-
medication. All studies performed premedical screening for ing all studies. Across all designs and categories, the pooled
subjects. Eleven studies (85%) reported that subjects were standardized effect size by a fixed-effect model showed an
sedentary before taking part in the study. average effect of ⫺0.58 ⫾ 0.08 (mean ⫾ SEM, 95% CI ⫽
The length of training for the 13 studies varied from approx- ⫺0.74 to ⫺0.42). This effect size for the net change was
imately 8 to 45 wk (25.1 ⫾ 11.6 wk). Approximately 90% of high moderate and statistically significant (P ⫽ 0.001).
the studies used walking as the primary training modality. Study quality of the included studies ranged from 1 to 3
Many exercise programs included one or more of the follow- (2 ⫾ 1). Potential publication bias was not identified. No
ing: walking on graded treadmill or outdoor track or in large statistically significant heterogeneity was observed for the
mall, jogging, indoor track running, and cycling /Monark er- effect (Q ⫽ 10.30, P ⫽ 0.801).
gometer. Most studies trained subjects 3 d䡠wk⫺1, and the mean Table 2 provided a description of each included interven-
frequency of exercise ranged from 1.8 to 4.0 times per week tion in terms of major characteristics of study, subjects, and
(3.0 ⫾ 0.6). Duration of training averaged 40.7 ⫾ 7.3 min per exercise parameters. Pre- and postintervention values and
session with mean range from 20 to 51 min. The use of exercise the average changes in resting HR for exercise and control
intensity varied in the 13 studies, expressed as percentage of groups in each study were showed in Table 3. After inter-
maximal HR (HRmax), percentage of maximal oxygen con- vention, control groups had no changes in resting HR with
AGING, AEROBIC EXERCISE, AND HEART RATE Medicine & Science in Sports & Exercise姞 1383
FIGURE 1—Net changes in resting heart rate (HR) response to intervention. The black circle and horizontal line correspond to standardized effect
size of the change outcome and 95% confidence intervals for each group. The rectangle with the straight line represents the pooled effect size of the
overall mean change for resting HR by fixed or random models and 95% confidence intervals.

a difference of 0.66 ⫾ 2.22 bpm (mean ⫾ SEM, 95% CI ⫽ ing HR among older adults. Our study demonstrates that
⫺3.89 to 5.22, P ⫽ 0.767), but a statistically significant aerobic exercise training has a statistically significant effect
change was found in exercise groups with a mean difference on resting HR in sedentary older adults whose within-study
of ⫺5.50 ⫾ 2.10 bpm (mean ⫾ SEM, 95% CI ⫽ ⫺9.79 to mean age was 60 yr and older. Endurance training resulted
⫺1.22, P ⫽ 0.014). Overall net change in resting HR in in a moderate combined effect size across all 13 studies
these mean values was ⫺6.16 ⫾ 0.97 bpm (mean ⫾ SEM, assessed in a total of 651 elderly subjects. The actual mag-
95% CI ⫽ ⫺8.15 to ⫺4.18) and statistically significant nitude of an average reduction was approximately 6 bpm,
(P ⫽ 0.001). Subgroup analysis showed that trials with the ranging from ⫺2 to ⫺12 bpm, or representing a decrease of
length of exercise training more than 30 wk had a greater approximately 8.4% in resting HR as a result of aerobic
and statistically significant decrease in net change of resting endurance training. This finding suggests that endurance
HR than trials with short training less than 30 wk (mean ⫾ training is able to enhance vagal control of the heart and to
SD, ⫺8.37 ⫾ 3.22 bpm vs ⫺4.86 ⫾ 1.69 bpm, P ⫽ 0.011). reduce the resting HR in older adults.
No statistically significant differences were observed in Although it is confirmed that resting HR can be substantially
other subgroup analyses. reduced consequent to aerobic exercise training, the magnitude
of its change appears to show greater variations. Some studies
found a decrease of 10 –11 bpm in resting HR in middle-aged
DISCUSSION
men after aerobic training over a period of 12 wk (4). In older
The main finding of this meta-analytic investigation con- women who had trained for 9 –12 months, Spina et al. (29)
firms the efficacy of controlled endurance training on rest- reported a decline of 10 bpm in resting HR. Several other

TABLE 2. Characteristics of included exercise interventions.

Citation Na Ageb (yr) Gender RCTc Length (wk) Description of Exercise Programd
Blumenthal et al. (2) 31 67.7 Mixed Yes 16 Cycle ergometer, brisk walking, jogging; 70% HRR; 45 min; 3 ⫻ wk⫺1
Braith et al. (3) [1] 19 66.0 Mixed Yes 26 Walking, jogging/uphill treadmill; 70% HRR; 40–45 min; 3 ⫻ wk⫺1
[2] 14 65.0 Mixed Yes 26 Walking, jogging/uphill treadmill; 80–85% HRR; 35–40 min; 3 ⫻ wk⫺1
Cononie et al. (5) 17 72.0 Mixed Yes 26 Brisk walking, uphill treadmill walking, jogging; 50–85% V̇O2max; 25–40 min; 3 ⫻ wk⫺1
Hagberg et al. (13) [1] 11 64.0 Mixed Yes 37 Walking/home; 53% V̇O2max; 51 min; 3.1 ⫻ wk⫺1
[2] 10 64.0 Mixed Yes 37 Walking/treadmill walking, jogging, cycle ergometer; 73% V̇O2max; 51 min; 2.5 ⫻ wk⫺1
Hamdorf et al. (14) 18 82.4 Female Yes 26 Walking/outdoor; 56.5% HRR; 20 min; 1.8 ⫻ wk⫺1
Hamdorf et al. (15) 30 64.1 Female Yes 26 Walking; 74% HRmax; 35 min; 1.9 ⫻ wk⫺1
Hill et al. (17) 87 64.0 Mixed No 45 Walking/uphill treadmill walking, indoor track running; 60–80% HRmax; 30–50 min; 3–5 ⫻ wk⫺1
Kohrt et al. (20) [1] 53 63.7 Male No 45 Walking/treadmill, indoor track running, cycle/rowing ergometers; 80% HRmax; 46 min; 4 ⫻ wk⫺1
[2] 57 64.0 Female No 43 Walking/treadmill, indoor track running, cycle/rowing ergometers; 79% HRmax; 45 min; 3.9 ⫻ wk⫺1
Seals et al. (24) 10 62.0 Mixed No 26 Cycling, graded treadmill walking or jogging; 80–90% HRmax; 30–45 min; 3.6 ⫾ 0.2 ⫻ wk⫺1
Shin (27) 14 67.5 Female No 8 Walking/outdoor track; 40–60% HRR; 30–40 min; 3 ⫻ wk⫺1
Tanaka et al. (30) 11 62.0 Mixed No 26 Walking; 45% HRR; 42 min; 3.4 ⫻ wk⫺1
Whitehurst/Menendez (31) 18 68.0 Female No 8 Walking/outdoor or large mall; 70–80% HRmax; 40 min; 3.3 ⫻ wk⫺1
Wood et al. (34) 10 69.1 Mixed Yes 12 Cycling/ergometer, treadmills; 60–70% HRmax; 45 min; 3 ⫻ wk⫺1
a
number of exercise subjects; b mean values presented; c randomized control trials; d mode, intensity, duration, and frequency.

1384 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org

TABLE 3. Resting heart rate (bpm) results.
Exercise Control

Citations N Pre-Mean ⴞ SD Post-Mean ⴞ SD Db N Pre-Mean ⴞ SD Post-Mean ⴞ SD Db

Blumenthal et al. (2) 31 74.0 ⫾ 11.3 70.0 ⫾ 13.4 ⫺4.0 31 70.0 ⫾ 13.7 68.0 ⫾ 14.2 ⫺2.0
Braith et al. (3) [1] 19 71.0 ⫾ 8.0 66.0 ⫾ 7.0 ⫺5.0 11 65.0 ⫾ 8.0 66.0 ⫾ 8.0 1.0
[2] 14 69.0 ⫾ 9.0 63.0 ⫾ 7.0 ⫺6.0 11 65.0 ⫾ 8.0 66.0 ⫾ 8.0 1.0
Cononie et al. (5) 17 66.0 ⫾ 13.5 61.5 ⫾ 13.5 ⫺4.5 12 67.0 ⫾ 14.0 66.0 ⫾ 11.5 ⫺1.0
Hagberg et al. (13) [1] 11 87.0 ⫾ 9.0 74.0 ⫾ 10.0 ⫺13.0 9 75.0 ⫾ 6.0 76.0 ⫾ 5.0 1.0
[2] 10 73.0 ⫾ 6.0 65.0 ⫾ 9.0 ⫺8.0 9 75.0 ⫾ 6.0 76.0 ⫾ 5.0 1.0
Hamdorf et al. (14) 18 74.4 ⫾ 8.9 71.8 ⫾ 9.3 ⫺2.6 20 72.7 ⫾ 7.6 75.4 ⫾ 7.6 2.7
Hamdorf et al. (15) 30 75.5 ⫾ 11.0 69.9 ⫾ 8.5 ⫺5.6 36 75.2 ⫾ 8.5 74.7 ⫾ 8.0 ⫺0.5
Hill et al.17 (17) 87 85.0 ⫾ 19.0 76.9 ⫾ 12.6 ⫺8.1 34 79.2 ⫾ 12.4 78.9 ⫾ 12.8 ⫺0.3
Kohrt et al. (20) [1] 53 81.0 ⫾ 12.0 73.0 ⫾ 11.0 ⫺8.0 19 79.0 ⫾ 14.0 75.0 ⫾ 12.0 ⫺4.0
[2] 57 86.0 ⫾ 13.0 79.0 ⫾ 13.0 ⫺7.0 16 79.0 ⫾ 10.0 83.0 ⫾ 13.0 4.0
Seals et al. (24) 10 68.0 ⫾ 9.0 67.0 ⫾ 8.0 ⫺1.0 10 62.0 ⫾ 5.0 66.0 ⫾ 10.0 4.0
Shin (27) 14 72.1 ⫾ 7.3 69.1 ⫾ 7.3 ⫺3.1 13 73.6 ⫾ 8.1 76.9 ⫾ 8.1 3.2
Tanaka et al. (30) 11 70.0 ⫾ 13.3 68.0 ⫾ 9.9 ⫺2.0 11 66.0 ⫾ 6.6 66.0 ⫾ 9.9 0.0
Whitehurst/Menendez (31) 18 72.9 ⫾ 8.4 68.3 ⫾ 8.6 ⫺4.6 13 72.9 ⫾ 5.5 73.3 ⫾ 4.9 0.4
Wood et al. (34) 10 67.8 ⫾ 9.3 62.2 ⫾ 9.4 ⫺5.6 6 64.1 ⫾ 5.0 64.8 ⫾ 8.9 0.7
N, number of subjects; Pre-, preintervention; Post-, postintervention; Db, difference between pre- and postintervention.

studies have found more modest decreases in resting HR, a tion of 13 bpm in 60- to 69-yr-old persons with essential
mean reduction of 6.6 bpm from a review article of 18 pub- hypertension after exercising over 9 months at a moderate
lished studies (21) or a 5-bpm decrease in 11 middle-aged and intensity. Except for the length of training, however, the results
older men who had undergone endurance training for a period of our other subgroup analyses did not show significant dif-
of 30 wk (23). Still other research (33) indicated that 20 wk of ferences according to the different training components (fre-
intense endurance training resulted in only a small reduction in quency, duration, and intensity). Thus, the question still re-
resting HR, from ⫺2.4 to ⫺4.6 bpm, in 92 sedentary male and mains if a greater dose or volume of endurance training would
female individuals with ages from 50 to 65 yr. It is obviously accomplish the increased vagal-cardiac function among elderly
difficult to explain the discrepancy in results seen across these individuals, compared to that achieved in subjects less than 60
studies because of so many factors involved, such as study yr of age. It appears that more sound controlled clinical trials
design, subject age, sample size, outcome measurement meth- are needed to confirm the effect of longer training time and to
ods, and exercise training protocols. The present meta-analytic examine the effect of a greater training dose and/or volume on
study included 651 subjects. They were older individuals with resting HR among older adults aged over 60 yr.
average within-study age ranging from 61 to 83 yr. All 13 This study has limitations. One is that unpublished studies
studies were controlled clinical trials either randomized or were not included in the analyses. The possibility does exist
nonrandomized with parallel control groups. These older indi- that there are other significant sources of unpublished studies.
viduals exercised about an average of 40 min, three times per However, such studies generally cannot be located by a sys-
week, with intensity at around 75% of HRmax, or 64% of tematic search and, as such, violates one of the characteristics
V̇O2max or 57% of HRR. The training length averaged about 24 of the scientific methodology, that of replicability. Although
wk. All subgroup analyses, except for exercise training length, potential sources provided limited amount of information, we
did not find statistical differences between these variables and did search and review dissertations and reports to minimize
changes in resting HR. Given the fact that few longitudinal possible publication bias. Further, funnel plot is useful to detect
studies have included measurement of change in resting HR the likely presence of bias in meta-analysis (8). Our result of
among the elderly population and that the magnitude of such analysis appears to show the absence of bias. Another
changes was still not consistent and clear as a result of aerobic limitation involves not contacting authors of studies in order to
training, the finding of this meta-analysis may add new infor- identify further studies and gain missing data. However, our
mation for older adults regarding the direction and magnitude experience with contacting authors for other reviews has
of resting HR changes in response to endurance exercise. shown considerable time investment with a low response rate
Another important finding of this study is the greater and and few, if any, additional studies.
statistically significant reductions in resting HR that were ob- The endurance training-induced reduction in resting HR in
served among older subjects who aerobically trained more than older individuals may have significant clinical implication. It is
30 wk. This implies that longer training time of aerobic exer- known that a lower resting HR as a result of exercise training
cise may have greater impact on lowering HR at rest among decreases myocardial oxygen and enhances cardiac mechanical
older adults. Some studies (23,25) have indicated that subjects efficiency (18). Given that the resting cardiac output averages
aged less than 60 yr old may be able to attain a significant 5 L䡠min⫺1 at rest and resting HR 70 bpm in most adults, a HR
decrease in resting HR in a short time. In the interventional decrease of approximately 6 bpm from our study would result
studies included in this meta-analysis, however, the trend was in an increase in stroke volume by approximately 7 mL䡠beat⫺1
observed that the longer the exercise training, the greater the or 448 mL䡠min⫺1 for those sedentary older individuals after
reductions in resting HR in older adults. Hagberg et al. (13) aerobic training. Moreover, cardiac reserve is apparently in-
found that aerobic exercise training elicited an average reduc- creased in response to various stresses that challenge the car-
AGING, AEROBIC EXERCISE, AND HEART RATE Medicine & Science in Sports & Exercise姞 1385
diovascular function (9,23). An augmented vagal cardiac mod- dividuals aged over 60 yr. The magnitude of the overall net
ulation associated with the training induced bradycardia change is modest in effect size, with a decrease of approxi-
improves the balance of the autonomic nervous system (9,28) mately 6 bpm, an 8.4% reduction. The analysis suggests that a
and helps maintain the cardiovascular homeostasis (35). None- longer length of exercise training, probably more than 30 wk,
theless, resting HR decreases little with advancing age but may be needed for the older adults to be more effective in
almost all these functions appear to decline with aging (6,10). reducing HR at rest. These findings may help assess the effects
Therefore, a lower resting HR after exercise training provides of exercise training in older people and design aerobic exercise
a nonpharmacological measure to benefit the cardiovascular programs for such populations. More research is needed in the
health and fitness in older individuals. future in confirming the effectiveness of endurance training in
In conclusion, this meta-analytic investigation supports the older adults and exploring quantitative relationship regarding
previous studies indicating that aerobic training can signifi- exercise components (intensity, duration, frequency, etc.) and
cantly influence resting HR in previously sedentary older in- changes in resting HR.

REFERENCES
1. ARBAB-ZADEH, A., E. DIJK, A. PRASAD, et al. Effect of aging and 19. JADAD, A. R., R. A. MOORE, D. CARROLL, et al. Assessing the
physical activity on left ventricular compliance. Circulation 110: quality of reports of randomized clinical trials: is blinding neces-
1799–1805, 2004. sary? Control Clin. Trials 17:1–12, 1996.
2. BLUMENTHAL, J. A., C. F. EMERY, D. J. MADDEN, et al. Effects of 20. KOHRT, W. M., M. T. MALLEY, A. R. COGGAN, et al. Effects of
exercise training on cardiorespiratory function in men and women gender, age, and fitness levels on response of V̇O2max to training
older than 60 years of age. Am. J. Cardiol. 67:633–639, 1991. in 60-71 yr olds. J. Appl. Physiol. 71:2004–2011, 1991.
3. BRAITH, R. W., M. L., POLLOCK, D. T. LOWENTHAL, J. E. GRAVES, 21. POLLOCK, M. L. The quantification of endurance training pro-
and M. C. LIMACHER. Moderate- and high-intensity exercise lowers grams. Exerc. Sports Sci. Rev. 1:155–188, 1973.
blood pressure in normotensive subjects 60 to 79 years of age. 22. POLLOCK, M. L., C. FOSTER, D. KNAPP, J. I. ROD, and D. H. SCHMIDT.
Am. J. Cardiol. 73:1124–1128, 1994. Effect of age and training on aerobic capacity and body compo-
4. BYRD, R. J., D. P. SMITH, and C. B. SHACKLEFORD. Jogging in sition of master athletes. J. Appl. Physiol. 62:725–731, 1987.
middle-aged men: effect on cardiovascular dynamics. Arch. Phys. 23. SEALS, D. R., and P. B. CHASE. Influence of physical training on
Med. Rehabil. 55:301–305, 1974. heart rate variability and baroreflex circulatory control. J. Appl.
5. CONONIE, C. C., J. E. GRAVES, M. L. POLLOCK, M. I. PHILLIPS, C. Physiol. 66:1886–1895, 1989.
SUMNERS, and J. M. HAGBERG. Effect of exercise training on blood 24. SEALS, D. R., B. F. HURLEY, J. M. HAGBERG, et al. Effects of
pressure in 70- to 79-yr-old men and women. Med. Sci. Sports training on systolic time intervals at rest and during isometic
Exerc. 23:505–511, 1991. exercise in men and women 61 to 64 years old. Am. J. Cardiol.
6. EBERT, T. J., B. J. MORGAN, J. A. BARNEY, T. DENAHAN, and J. J. 55:797–800, 1985.
SMITH. Effects of aging on baroreflex regulation of sympathetic 25. SHELDAHL, L. M., M. T. EBERT, B. COX, and F. E. TRISTANI. Effect
activity in humans. Am. J. Physiol. 263:H798–H803, 1992. of aerobic training on baroreflex regulation of cardiac and sym-
7. EGGER, M., G. D. SMITH, and A. N. PHILLIPS. Meta-analysis: prin- pathetic function. J. Appl. Physiol. 76:158–165, 1994.
ciples and procedures. Br. Med. J. 315:1533–1537, 1997. 26. SHI, X., G. H. STEVENS, B. H. FORESMAN, S. A. STERN, and P. B.
8. EGGER, M., G. D. SMITH, M. SCHNEIDER, and C. MINDER. Bias in RAVEN. Autonomic nervous system control of the heart: endurance
meta-analysis detected by a simple, graphical test. Br. Med. J. exercise training. Med. Sci. Sports Exerc. 27:1406–1413, 1995.
315:629–634, 1997.
27. SHIN, Y-H. The effects of a walking exercise program on physical
9. EKBLOM, B., A. KILBOM, and J. SOLTYSIAK. Physical training, bra-
function and emotional state of elderly Korean women. Pub.
dycardia, and autonomic nervous system. Scand. J. Clin. Lab.
Health Nurs. 16:146–154, 1999.
Invest. 32:251–256, 1973.
28. SMITH, M. L., D. L. HUDSON, H. M. GRAITZER, and P. B. RAVEN.
10. FAGARD, R., L. THIJS, and A. AMERY. Age and the hemodynamic
Exercise training bradycardia: the role of autonomic balance. Med.
response to posture and to exercise. Am. J. Geriatr. Cardiol. 2
(2):23–30, 1993. Sci. Sports Exerc. 21:40–44, 1989.
11. FORTNEY, S., C. TANKERSLEY, J. T. LIGHTFOOT, et al. Cardiovascular 29. SPINA, R. J., T. OGAWA, T. R. MILLER, W. M. KOHRT, and A. A.
responses to lower body negative pressure in trained and untrained EHSANI. Effect of exercise training on left ventricular performance
older men. J. Appl. Physiol. 73:2693–2700, 1992. in older women free of cardiopulmonary disease. Am. J. Cardiol.
12. GLASS, G. V. Primary, secondary and meta-analysis of research. 71:99–104, 1993.
Educ. Res. 5:3–8, 1976. 30. TANAKA, H., M. J. REILING, and D. R. SEALS. Regular walking
13. HAGBERG, J. M., S. J. MONTAIN, W. H. MARTIN III, and A. A. increases peak limb vasodilatory capacity of older hypertensive
EHSANI. Effect of exercise training in 60- to 69-year-old persons humans: implications for arterial structure. J. Hypertens. 16:423–
with essential hypertension. Am. J. Cardiol. 64:348–353, 1989. 428, 1998.
14. HAMDORF, P. A., and R. K. PENHALL. Walking with its training 31. WHITEHURST, M., and E. MENENDEZ. Endurance training in older
effects on the fitness and activity patterns of 79-91 year old women: lipid and lipoprotein responses. Physician Sportsmed. 19:
females. Aust. N. Z. J. Med. 29:22–28, 1999. 95–98, 1991.
15. HAMDORF, P. A., R. T. WITHERS, R. K. PENHALL, and M. V. HASLAM. 32. WILMORE, J. H., P. R. STANFORTH, J. GAGNON, et al. Endurance
Physical training effects on the fitness and habitual activity patterns of exercise training has a minimal effect on resting heart rate: the
elderly women. Arch. Phys. Med. Rehabil. 73:603–608, 1992. HERITAGE study. Med. Sci. Sports Exerc. 28:829–835, 1996.
16. HEDGES, L. V., and I. OLKIN. Statistical Methods for Meta-Analy- 33. WILMORE, J. H., P. R. STANFORTH, J. GAGNON, et al. Heart rate and
sis. San Diego, CA: Academic Press, 1985, pp. 75–203. blood pressure changes with endurance training: the HERITAGE
17. HILL, R. D., M. STORANDT, and M. MALLEY. The impact of long- Family study. Med. Sci. Sports Exerc. 33:107–116, 2001.
term exercise training on psychological function in older adults. J. 34. WOOD, R. H., R. REYES, M. A. WELSCH, et al. Concurrent cardio-
Gerontol. 48:P12–17, 1993. vascular and resistance training in healthy older adults. Med. Sci.
18. ISHIBASHI, Y., T. SHIMADA, S. NOSAKA, et al. Effects of heart rate on Sports Exerc. 33:1751–1758, 2001.
coronary circulation and external mechanical efficiency in elderly 35. WRAY, D. W., K. J. FORMES, M. S. WEISS, et al. Vagal cardiac
hyertensive patients with left ventricular hypertrophy. Clin. Car- function and arterial blood pressure stability. Am. J. Physiol. Heart
diol. 19:620–630, 1996. Circ. Physiol. 281:H1870–1880, 2001.

1386 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org