583
Effect of Regular Exercise on 24-Hour
Arterial Pressure in Older
Hypertensive Humans
Douglas R. Seals and Mary Jo Reiling
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
The experimental goals were to determine if regular low-intensity aerobic exercise reduces
24-hour arterial blood pressure in middle-aged and older (aged 50 years or older) humans with
mild diastolic (90-105 mm Hg) essential hypertension and, if so, whether this is accurately
reflected by changes in casual recordings made at rest Fourteen subjects walked 3 - 4 days/wk
for 6 months, with 10 exercising an additional 6 months; 12 other subjects served as
nonexercising controls. In the exercising subjects, maximal oxygen consumption increased
7-14% (p<0.05) with little or no change in body weight or fat Conventional casual readings of
systolic, mean, and diastolic arterial pressure at rest were lower (5-10 mm Hg,p<0.05) in all
body positions after 6 months of exercise and changed little thereafter. Casual recordings made
during additional circulatory measurements showed 6-month decreases of only half this
magnitude and were specific to a particular blood pressure phase and body position; however,
all changes were significant after 12 months of exercise. The reductions in arterial pressure at
rest were associated with decreases in heart rate (p<0.05) and cardiac output (/?<0.05).
Ambulatory-determined 24-hour arterial pressure was unchanged after 6 months of exercise,
but mean levels were slightly lower (4 mm Hg, p<0.05) after 12 months due to reductions in
daytime (7 mm Hg,/?<0.05) and nighttime (4 mm Hg, NS) systolic pressure; diastolic pressure
was unchanged throughout the year of training. In the controls, conventionally recorded casual
blood pressure levels were lower after 6 months (p<0.05), but no other changes were observed
in any other variable over the 12 months of study. We conclude 1) regular low-intensity aerobic
exercise at best produces only small reductions in 24-hour levels of arterial pressure in
middle-aged and older humans with mild (diastolic) essential hypertension and 2) trainingassociated changes in casually determined blood pressure at rest are dependent on the
measurement conditions and, most importantly, do not necessarily reflect the magnitude or
even the direction of changes in arterial pressure throughout an entire day. (Hypertension
1991;18:583-592)
A rterial blood pressure rises with advancing age
/ \
in industrialized societies.12 This contributes
A. \ - to the increased risk of cardiovascular disorders such as stroke and ischemic heart disease in
older people.3-4 Lowering chronic levels of arterial
pressure via drug therapy reduces cardiovascular risk
in older patients with essential hypertension,5-7 but
the associated side effects are generally greater than
From the Department of Exercise and Sport Sciences and the
Department of Physiology, Arizona Health Sciences Center, University of Arizona, Tucson, Ariz.
Supported by grant AG 06537 from the National Institute on
Aging, National Institutes of Health. During the period of this
study D.R.S. was supported by Research Career Development
Award AG 00423 from the National Institute on Aging.
Address for correspondence: Douglas R. Seals, PhD, Department of Exercise and Sport Sciences, 228B McKale Center,
University of Arizona, Tucson, AZ 85721.
Received March 13, 1991; accepted in revised form July 1,1991.
in younger patients.8 Thus, nonpharmacological interventions have been recommended as an initial
approach for older patients with mild-to-moderate
blood pressure elevations.6
Regular physical exercise lowers arterial pressure
at rest by 7-9% (8-10 mm Hg) on average in young
and middle-aged hypertensive humans,9"12 but its
influence in older people is less well documented.
Recently Hagberg et al13 reported that 3 months of
low-intensity aerobic exercise training reduced arterial pressure at rest by 8-12 mm Hg in middle-aged
and older men and women with mild essential hypertension; an additional 6 months of training produced
further (4-8 mm Hg) reductions. These decreases in
arterial pressure were as great or greater than those
observed in subjects who performed higher intensity
exercise. Interestingly, training had a much smaller
effect on blood pressure determined at rest during
584
Hypertension
Vol 18, No 5 November 1991
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
experimental sessions in which additional circulatory
measurements (i.e., cardiac output) were made.
In both young and older patients with essential
hypertension, levels of arterial blood pressure recorded acutely in the clinic or laboratory under
resting conditions (i.e., casual readings) often overestimate levels measured throughout the entire day
under freely behaving (ambulatory) conditions14-15;
the latter correlate better with end organ damage
and therefore with patient prognosis.1617 In longitudinal studies, casually determined pressures can decrease over time without corresponding changes in
ambulatory-measured levels.18 Thus, it is possible
that intervention-induced changes in casually recorded arterial pressure at rest may not accurately
reflect the magnitude or even the direction of
changes in levels throughout the entire day. This
could be the case in studies examining the influence
of physical training on blood pressure. For example,
two recent investigations reported little or no decrease in daytime or nighttime ambulatory-determined arterial blood pressure after training in young
and middle-aged normotensive or borderline hypertensive subjects1920; in one of these studies,19 casually
measured levels decreased to a much greater extent
than the ambulatory pressures.
The primary aims of the present investigation were
to determine if regularly performed low-intensity aerobic exercise reduces arterial blood pressure throughout the entire day in middle-aged and older men and
women with mild essential (diastolic) hypertension and,
if so, whether this is accurately reflected by changes in
casually determined blood pressure at rest measured
under at least two different experimental conditions.
Our secondary aims were to determine, in this population, whether extending the length of exercise training
would produce further reductions in arterial pressure
and to gain insight into the systemic hemodynamic
mechanisms underlying training-associated decreases
in arterial pressure.
Methods
Subjects
A total of 34 subjects (24 men and 10 women) aged
50-74 years were enrolled in this investigation after
providing their written, informed consent. All procedures and protocols used in this study had been
approved by the Human Subjects Committee at the
University of Arizona. Eight of the subjects dropped
out due to illness or lack of compliance with the
requirements of the study. Twenty-six subjects completed at least two rounds of testing separated by 6
months of either regular aerobic exercise (n = 14; nine
men and five women) or unchanged physical activity
levels (nonexercising controls) (n = 12; 10 men and
two women). Eighteen of these subjects (10 exercise
and eight controls) participated in an additional 6
months of study and completed a third round of
testing. At enrollment every subject had an average
level of diastolic arterial blood pressure at rest be-
tween 90 and 105 mm Hg based on repeated casual
recordings (see session 1). The subjects were free of
overt disease based on a medical history, physical
examination, and electrocardiogram at rest and during maximal treadmill exercise. They were within
20% of ideal body weight based on their body mass
index and had no orthopedic complications that
would prohibit their participation in testing or physical training. Only one (training group) subject
smoked; her responses were not different from those
of the other exercising subjects. Before enrollment,
four subjects in the exercise group (diuretics only)
and seven controls (diuretics, n=3; /3-blockers, n=4)
were taking antihypertensive medications. Each of
these subjects stopped their medications at least 2
weeks before the first session in a particular round
of testing and remained off drugs throughout the
testing period. Most of the subjects resumed drug
therapy immediately after the final session of each
round of testing to ensure proper blood pressure
control; however, three subjects (one exercise, two
controls) did not continue therapy after their initial
round of testing because their arterial pressure
remained at a satisfactorily low level. No subject
had performed any regular exercise for at least 2
years before enrollment. Subjects were instructed
not to alter their diet during the period of study. To
document this, a comprehensive food frequency
questionnaire was administered and analyzed during each round of testing.
Experimental Procedures
Except where indicated, all subjects underwent
each of the following procedures during each round
of testing. For subjects in the physical training group,
these procedures were performed on days other than
those on which they exercised.
Maximal O2 Consumption
Maximal O2 consumption was measured by open
circuit spirometry during graded treadmill walking as
described in detail previously2122 and was used to
document training-induced adaptations resulting in
an increase in aerobic exercise capacity. Subjects
were required to exercise to the point of subjective
exhaustion. All subjects attained 1) a plateau in O2
consumption despite further increases in work load,
2) a respiratory exchange ratio (CO2 production/O2
consumption-hyperventilatory index of exercise effort) of more than 1.15, and 3) their age-predicted
maximal heart rate. Concerning the latter two variables, the levels achieved over the three rounds of
testing were similar within and between the two
groups (1.20±0.02-1.22±0.03 and 168±4-171±5
beats/min and 1.19±0.02-1.21±0.02 and 170±4173±4 beats/min for the exercising subjects and
controls, respectively).
Estimation of Body Fat
Body fat was estimated from skinfold thickness at
six sites using a Harpenden caliper. Body density and
Seals and Reiling Exercise and 24-Hour Arterial Pressure
percent body fat were estimated using the equations
of Jackson and Pollock23 and Siri,24 respectively.
Determination of Arterial Blood Pressure: Casual
Recordings at Rest
Session 1. Arterial pressure only. In the initial round
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
of testing, these sessions were used to determine
whether the subject's level of diastolic pressure met
the criterion for enrollment and, if so, to establish the
baseline level of arterial pressure from which values
obtained in subsequent testing periods could be
compared. In this initial round (i.e., before enrollment), subjects were studied once or twice per week
for 3-6 weeks to ensure that a stable level of arterial
pressure was obtained. Consistent recordings over a
period of 3 weeks were required before a subject was
allowed to continue with testing. The readings from
these 3 weeks were averaged and used as the baseline
arterial pressure. During subsequent rounds of testing, similarly stable, multiple recordings were obtained over a 2-3 week period and averaged. Measurements were made by the auscultatory method
over the brachial artery using a Hawksley randomzero sphygmomanometer (Hawksley and Sons, West
Sussex, England). During the first session of the
initial testing period, arterial pressure was determined in both arms, and the arm with the highest
reading was used for all subsequent recordings. The
first and fifth phases of the Korotkoff sounds were
used to determine the systolic and diastolic arterial
pressures, respectively. All measurements conformed
strictly to American Heart Association guidelines.25
To document that changes in arterial pressure observed over time were not specific for a particular
body position, measurements were made with the
subject in the supine, sitting, and standing positions.
Briefly, in each session subjects rested in the supine
position on a padded table in a quiet laboratory
(~24°C) for 15 minutes, after which time three or
more recordings were made over a 10-15-minute
period. The back of the table was then raised, and
subjects sat upright for 5 minutes followed by at least
three readings over 10-15 minutes. Subjects then
stood for 5 minutes, and three or more measurements were obtained in this position.
Session 2. Arterial pressure during systemic hemody-
namic measurements. The aims of this session were 1)
to gain insight into the systemic hemodynamic mechanisms responsible for training-induced reductions in
arterial pressure at rest and 2) to obtain an additional
set of arterial pressure recordings made at rest but
under somewhat different experimental conditions
than those obtained in the sessions described above;
this would indicate whether any training-induced
reductions observed at rest were measurement condition-specific. Subjects rested in the supine position
for 20 minutes after which cardiac output (Colliers
CO2 rebreathing technique26), heart rate, and arterial
pressure were measured three to five times over a
30-45-minute period. The arterial pressure measurements were made before the rebreathing maneu-
585
vers when the subject was resting quietly. The table
then was adjusted, and the subjects sat upright for
—10 minutes, at which point these measurements
were repeated. At least three consistent (i.e., ± ~ 1
1/min) values of cardiac output were obtained in each
position for each subject and averaged. Stable measurements could not be obtained in two of the
controls; thus, their hemodynamic data were deleted
from subsequent analysis. Cardiac output was divided by the corresponding heart rate to calculate
stroke volume and into the corresponding mean
arterial pressure to calculate systemic vascular resistance. Because the CO2 rebreathing technique appears to underestimate absolute (i.e., 1/min) levels of
cardiac output at rest in hypertensive older subjects,13 the data on cardiac output, stroke volume,
and systemic vascular resistance are presented as
percent changes from levels obtained during the
initial round of testing.
Ambulatory Measurements
Levels of arterial blood pressure throughout an
entire day were obtained using a noninvasive ambulatory monitor (model 5200, Spacelabs Inc., Redlands, Wash.). The monitor operates by the auscultatory method when Korotkoff sounds are audible
(phase five for diastolic) and by oscillometry when
not; extraneous noise (body movement) is differentiated from pressure signals by an internal microprocessor. Ambulatory readings from this system correlate well with ambulatory measurements made
during direct (intra-arterial) recordings.27'28 On a
weekday in which no other measurements (or training) were performed (typically in the morning) subjects were oriented to the monitor and fitted with a
standard, properly sized blood pressure cuff; in their
second (or third) rounds of testing, subjects reported
to the laboratory at the same time and day of the
week as during this initial trial. The ambulatory
system was calibrated against conventional sphygmomanometer-determined readings and programmed to
inflate automatically every 15 minutes between 6:00
AM and midnight and every 60 minutes between
midnight and 6:00 AM; thus, 78 recordings were
possible over a 24-hour period. To reduce noise, each
subject was instructed to stop and relax the arm
throughout the inflation/deflation cycle and to record
the time, location, body position, and state of activity
at each reading. Recordings were stored on magnetic
tape and subsequently analyzed and printed using a
desktop computer (IBM PC). Each reading was
edited both by computer and manually, and outliers
(systolic pressure less than 70 or more than 260
mm Hg, diastolic less than 40 or more than 150, and
heart rate less than 40 or more than 150 beats/min)
were deleted. Averages were then calculated for each
hour and also for the entire daytime (6:00 AM to 6:00
PM), evening/nighttime (6:00 PM to 6:00 AM), and all
day (24-hour) time periods. Over the three rounds of
testing, meaningful readings were obtained on 8590% of the total possible recordings. To ensure that
586
TABLE 1.
Hypertension
Vol 18, No 5 November 1991
Selected Characteristics of Subjects in the Training and NonexercUing Control Groups
Age
Group (n)
Exercise (14)
Control (12)
Exercise (10)
Control (8)
(y)
61±2
61±2
63±2
61±3
Height
(cm)
172.4±2.2
175.7±1.7
171.9+1.9
175.5±1.9
Weight (kg)
Initial
76.5+2.7
76.9 ±3.5
76.4 ±3.2
73.6±4.0
6 mo
12 mo
Estimated fat |{%)
12 mo
6 mo
Initial
76.1±2.8
25.7+1.8
76.1 ±3.6
22.4±1.6
76.1±3.1 75.0±3.0* 27.4±2.2
72.7+43 73.2±43 21.5±1.2
Vo^max (ml/kg/min)
Initial
27J±1.8
25.5 + 1.7
31J±2.1
22.4 ±1.6
27.0±2.2 26.4±23 25.2±1.8
21.9±1.3 22.1±1.2 32.2±2.8
6 mo
12 mo
293±1.8*
31.9±2.1
27.5±1.9* 28.6±2.2*
32.7±2.6 33.2±3.1
All data are mean±SEM. Vc^max, maximum oxygen consumption.
><0.05 vs. initial value.
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
differences in daily activity levels did not confound
the interpretation of serial measurements, on the
second and third rounds of testing subjects performed the same activities as during their initial
24-hour time periods; the activity logs were checked
to ensure that this procedure was followed. It was
determined that one training group subject performed a substantially greater amount of physical
activity during the day of his second round of testing.
All of his ambulatory data were deleted from subsequent analysis. In addition, in one control group
subject the magnetic tape was accidently erased while
he was working, and a repeat trial was not possible.
Physical Activity Levels
After completing the initial round of testing, the 14
subjects in the exercise group performed lowintensity aerobic training for approximately 6
months. Subjects were instructed to walk at least 3
days/wk for at least 30 min/day at an intensity associated with an exercise heart rate of 40-50% of their
individually determined heart rate reserve (calculated as heart at rest+0.40-0.50 [maximal heart
rate-heart rate at rest]).22 Subjects actually exercised for an average of 27.6 ±1.8 (mean±SEM)
weeks, 3.5 ±0.1 days/wk, for 41 ±4 min/day, at 47 ±2%
of heart rate reserve. On completion of the second
round of testing, 10 of these subjects agreed to
continue training for an additional 6 months; in
exchange, these subjects were allowed to gradually
increase their exercise intensity and duration if they
so chose. This subgroup walked an average of
3.5 ±0.1 days/wk for 43 ±5 min/day at an intensity of
46 ±2% of heart rate reserve during their initial 6
months of training and 3.6±0.2 days/wk for 50±2
min/day at an intensity of 57 ±3% of their heart rate
reserve during the final period (30.3 ±2.4 weeks) of
training. The 12 nonexercising controls maintained
their normal physical activity patterns throughout the
6-month period between the first and second rounds
of testing as determined by analysis of their individual activity logs; subsequently, eight of these subjects
continued to maintain their normal activity levels for
an additional 6 months. Due to the length of the
study period, to ensure the highest possible compliance with their respective assignments subjects were
allowed to choose between training and maintained
physical activity. This strategy was effective in eliminating the "crossover" effect associated with training
studies described previously by Holloszy et al29 (i.e.,
failure of some randomly assigned training group
subjects to exercise regularly and the covert initiation
of training by some control subjects).
Statistics
Changes in a variable from the initial round of
testing over time within a group were evaluated by
analysis of variance. Differences between two specific
time points were determined by Scheffe post hoc
analysis. Differences of /?<0.05 were considered significant. All data are presented as mean±SEM.
Results
Subject Characteristics
Body weight and estimated body fat were unchanged
after 6 months of regular aerobic exercise, but maximal
O2 consumption was increased 7% (/?<0.05) (Table 1).
In the subset of subjects who completed 1 year of
training, body weight decreased slightly (~1 kg;
/><0.05), but estimated body fat was unchanged during
the final 6 months. In this subgroup, maximal O2
consumption was increased by 9% and 14% after 6 and
12 months of training, respectively (both p<0.05).
There were no changes in any of these variables over
corresponding periods of time in the nonexercising
control subjects. There also were no changes across
time in any component of diet in either group.
Arterial Blood Pressure
Casual Blood Pressure at Rest
Session 1. Arterial pressure only. Six months of
regular exercise resulted in decreases in systolic
(range, 7-10 mm Hg), diastolic (5-7 mm Hg), and
mean (7-8 mm Hg) arterial pressure at rest in all
three body positions studied (p<0.05) (Table 2,
Figures 1 and 2). In the subset of subjects who
trained for 1 year, the decreases in arterial pressure
during the initial 6 months of training were slightly
greater than those for the whole group, but there
were no further reductions over the final 6 months of
exercise. Arterial blood pressure decreased to a
similar (6 month group) or slightly lesser (12 month
subgroup) extent in the nonexercising controls over
the same periods of time (p<0.05).
Session 2. Systemic hemodynamics. During the
initial round of testing, there were no consistent
Seals and Reiling
-10h
Exercise and 24-Hour Arterial Pressure
6 mo
12 mo
6 mo
12 mo
6 mo
12 mo
6 mo
12 mo
Diastolic
6 mo 12 mo
143±3
11011
110±2
94±1
I
14313
13914
11112
10912
9511
144±4
Mean
B|"B
Systolic
Mean
Systolic
6 mo 12 mo
Diastolic
93±1
9212
147±4
151±6
11112
11213
93±2
9411
flBflBBB
Casual
13714
10311
10311
88±1
Casual
9012
92l2
Casual
13713
587
Casual
88H
13513
13112
10312
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
FlGURE 1. Bar graphs show initial levels (top of each bar,
mean±SEM) and average changes in arterial blood pressure
in response to 6 (n=14) and 12 (n=10) months of regular
aerobic exercise recorded under three different experimental
conditions. Top panel: Casual recordings at rest during no
other measurements (average of three body positions). Middle
panel: Casual recordings during measurements of cardiac
output (average of supine and sitting positions). Bottom
panel: 24-hour ambulatory recordings. Note that the data
from the three conditions differed with respect to whether a
significant change occurred, the phase of arterial pressure that
was altered, and the magnitudes and timecourses of the
reductions. *p<0.05 vs. initial level. tp<0.05 only for supine
position.
differences in the absolute levels of arterial pressure determined in these hemodynamic measurement sessions compared with those obtained in the
sessions in which only blood pressure was measured
(Table 3, Figures 1 and 2). After 6 months of
aerobic training, there was a trend for small (3-4
mm Hg) reductions in systolic, diastolic, and mean
arterial pressure at rest in both the supine and
upright seated body positions; however, the changes
were statistically significant only for supine diastolic
and mean arterial pressure. In the subjects completing 1 year of exercise, trends similar to those
described for the entire group were observed during
the initial 6 months of training. In this subgroup, an
additional 6 months of regular exercise resulted in
further small (2-4 mm Hg) decreases in arterial
pressure such that the final levels of systolic, diastolic, and mean pressure were all significantly
(p<0.05) lower than the initial values in both body
positions. In contrast to the sessions in which only
blood pressure was measured, in these hemodynamic sessions there were no changes in arterial
pressure over time in the nonexercising controls.
Ambulatory Recordings
Depending on the time of day used for comparison,
the absolute levels of ambulatory-determined pressure
9112
8911
Ambul.
Ambul.
-10L
10011
-io L
FiGURE 2. Bar graphs show initial levels and average
changes in arterial blood pressure over 6 (n=12) and 12
(n=8) months of unchanged physical activity in the controls
recorded under three different experimental conditions (see
Figure 1 for details). Note that significant reductions were
observed over time during sessions in which only blood
pressure was measured at rest (top panel). In contrast, levels
of arterial pressure recorded at rest during other circulatory
measurements (middle panel) and ambulatory recordings
during 24 hours of normal activities (bottom panel) remained
constant over time. *p<0.05 vs. initial levels.
were slightly to markedly lower than the casually determined levels (Table 4, Figures 1 and 2). Average casual
levels at rest were most closely associated with daytime
ambulatory pressures and least associated with nighttime ambulatory levels; the latter were 9-14 mm Hg
lower than the corresponding levels during the day. Six
months of training did not significantly alter daytime,
nighttime, or 24-hour levels of arterial blood pressure;
however, there was a trend for a reduction in daytime
and 24-hour systolic and mean pressure. In the 1-year
training subgroup, an additional 6 months of regular
exercise resulted in a further, albeit slight (1-2 mm Hg)
trend for a lowering of daytime and 24-hour systolic
and mean pressure such that theirfinal(i.e., 12-month)
levels were significantly lower than then- initial values
(6-7 mm Hg for systolic, 4-5 mm Hg for mean; ah1
p<0.05). There was also a trend (3-4 mm Hg, NS) for
reductions in nighttime systolic and mean pressure
after 1 year of training compared with initial levels. In
contrast, the daytime, nighttime, and 24-hour levels of
diastolic pressure were unchanged over the 12 months
of regular exercise in this subgroup. Finally, the ambulatory-determined levels of arterial blood pressure were
quite constant over time in the nonexercising controls.
Systemic Hemodynamics
Session 2
The reductions (or trends for reductions) in arterial blood pressure at rest after 6 months of training
588
Hypertension
Vol 18, No 5 November 1991
TABLE 2. Average Levels of Arterial Blood Pressure at Rest in the Training Group and Nonexercising Control Group in Casual Session 1
(Arterial Pressure Only)
Systolic (mm Hg)
6 mo
12 mo
Group (n)
Initial
Supine
Exercise (14)
Control (12)
Exercise (10)
Control (8)
Sitting
146±2
147 ±3
146±3
144±5
138±3*
14O±5*
138±3*
138±7
141±3
144±3
143±4
140±4
134±3*
134+4*
134±4*
132±6*
140±3
139±3
142±4
133±3
130±4*
131±4*
130±4*
128±6
Exercise (14)
Control (12)
Exercise (10)
Control (8)
Standing
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
Exercise (14)
Control (12)
Exercise (10)
Control (8)
Diastolic (mm Hg)
Initial
6 mo
137±3*
137±5*
94±1
95±1
93±1
94±1
87+1*
89±2*
86±1*
87±3*
133±3*
133±5*
92±1
95±1
92±2
93±1
87±2*
88±2*
86±2*
86+2*
130±4*
13O±5
94±1
96+1
94±1
95±1
87±2*
89±2*
87±2*
87±2*
Mean (mm Hg)
6 mo
12 mo
Initial
84±2*
87±2*
112±1
112±2
111±1
110±2
104±l*
106±3*
103±2*
104±4*
102±2*
104±3*
84±2*
86±1*
109±l
111±2
109±l
109±2
102+2*
104±2*
102±2*
101+3*
100±2*
102±2*
84±1*
89 ±2*
109±2
110±2
110±2
108±2
102±2*
103±2*
101+2*
101±3*
100±2*
103±2*
12 mo
All data are mean±SEM.
*p<0.05 vs. initial value.
were associated with small but statistically significant
decreases in heart rate and cardiac output and
unchanged levels of stroke volume and systemic
vascular resistance in both body positions (Figure 3,
only supine data shown). In the 1-year training
subgroup, the trend for slight, further reductions in
arterial pressure over the final 6 months of exercise
was associated with trends for further decreases in
heart rate and cardiac output. The hemodynamic
variables were quite constant over time in the nonexercising controls.
Ambulatory Recordings
There was a trend (NS) for a small (4 beats/min)
reduction in 24-hour heart rate after 6 months of
training in the entire exercise group; this was associated with a small (5 beats/min) but significant
(p<0.05) decrease in daytime heart rate and no
change in nighttime heart rate (Table 4). The same
trend was observed over the initial 6 months in the
1-year training subgroup; however, after the final 6
months of training, 24-hour heart rate had returned
toward the initial levels in these subjects. There were
no changes in ambulatory-determined levels of heart
rate over corresponding periods of time in the nonexercising controls.
Discussion
These findings provide the experimental basis for
two primary new conclusions regarding the influence
of physical training on blood pressure. First, regularly
performed low-intensity aerobic exercise, at best,
produces only small decreases in 24-hour levels of
arterial pressure in middle-aged and older patients
with mild diastolic essential hypertension. Second,
exercise training-induced reductions in casually de-
TABLE 3. Average Levels of Arterial Blood Pressure at Rest in the Training Group and Nonexercising Control Group in Casual Session 2
(Systemic Hemodynamic Measurements)
Group (n)
Supine
Exercise (14)
Control (10)
Exercise (10)
Control (7)
Silting
Exercise (14)
Control (10)
Exercise (10)
Control (7)
Systolic (mm Hg)
Initial
6 mo
12 mo
Initial
6 mo
150±5
144±4
154±6
140±5
146±4
147±5
147±5
143±6
145±4*
143±5
94±1
94±2
94±2
91±2
91±1*
93±2
91 ±2
90±2
144±4
142+5
148±6
138+.7
140±4
142 ±5
142±5
139 ±7
138±4*
138±5
92±2
91±1
89±2
89±1
88±2
91±2
87±2
89±2
All data are mean±SEM.
*p<0.05 vs. initial value.
Diastolic (mm Hg)
Mean (mm Hg)
6 mo
12 mo
12 mo
Initial
88±2*
89±2
113±2
111±2
114±3
108±3
109±2*
111±2
109±2*
108±3
107±2*
107±2
85±1*
88±2
109±2
108±2
109±3
106+3
106±2
108±2
105±3
105+3
103±2*
1O4±3
Seals and Reiling
Exercise and 24-Hour Arterial Pressure
589
TABLE 4. Average Levels of Arterial Blood Pressure and Heart Rate Determined From All-Day Ambulatory Recordings
24-Hour
Group (n)
6 mo
Initial
12 mo
Initial
Day
6 mo
131±3*
131±3
142±3
139±3
142±4
136±3
140±4
139±3
137±5
136±3
88±1
90±2
92+2
94±1
92±2
93±1
94±2
96±1
92±2
95±2
99±1*
100±l
107±l
106±l
108±2
104±l
106±2
107±l
104±2
105 ±1
82±3
79±3
84±2
79±3
86±3
79±3
79±2'
81±3
81±3
83±3
12 mo
Initial
Night
6 mo
12 mo
135±3*
135±4
129±3
129+4
128±4
123±2
127±3
128±4
126±4
122±3
124±4
125±4
91±1
94±2
83±2
85±2
82±2
82±2
85±2
86±2
83±2
83±1
82+2
84+3
103±l*
104+1
97±2
97±3
96±2
92±2
97±2
95±2
95±2
92±2
93±2
93±3
78±2
73±3
80±2
71 ±3
76±2
75±3
78±3
74±4
78±3
74±3
Arterial pressure (mm Hg)
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
Systolic
Exercise (13)
Control (11)
Exercise (9)
Control (8)
Diastolic
Exercise (13)
Control (11)
Exercise (9)
Control (8)
Mean
Exercise (13)
Control (11)
Exercise (9)
Control (8)
137±3
135±3
137±4
131±2
134±3
134±3
132±4
131±3
88±1
91±2
88±1
89±1
90±2
92±1
88±1
91±1
103 ±1
103±2
103 ±1
100±l
102±2
103±2
101±l
100±l
82±2
77±3
83±3
76±3
78±2
79±2
80±3
80±3
Heart rate (beats/min)
Exercise (13)
Control (11)
Exercise (9)
Control (8)
84±3
82+3
All data are mean±SEM.
*p<0.05 vs. initial value.
termined blood pressure at rest are dependent, in
part, on the measurement conditions and, most importantly, do not always reflect the magnitude or
even the direction of changes in arterial pressure
throughout an entire day.
Exercise Group
6 mo
12 mo
Control Group
6 mo
12 mo
FIGURE 3. Bar graphs show changes from initial levels for
systemic hemodynamic variables (session 2) in the exercise
and control groups. Training-induced decreases in arterial
blood pressure (Table 3, Figures 1 and 2) were associated with
reductions in heart rate (HR) and cardiac output (CO), but no
significant changes in stroke volume (SV) and systemic vascular resistance (SVR). See text for further details. *p<0.05 vs.
initial levels.
Influence of Physical Training on 24 -Hour
Blood Pressure
In the present study, 6-12 months of regular
low-intensity aerobic exercise produced small increases in maximal oxygen consumption. These small
improvements in aerobic capacity are consistent with
previous observations21 and indicate that our exercise
stimulus was sufficient to produce physiologically
significant adaptations in circulatory or skeletal muscle function, or both. In spite of this training effect,
24-hour levels of arterial blood pressure were unchanged after 6 months of regular exercise. In a
subgroup that performed an additional 6 months of
slightly more intense exercise, 24-hour levels of mean
arterial pressure were slightly (4 mm Hg on average)
but significantly lower at the end of 1 year of training
compared with their initial levels. This reduction was
due primarily to a decrease in daytime systolic pressure, although there was also a trend for a reduction
in nighttime systolic pressure. In contrast, 24-hour
levels of diastolic pressure were unchanged throughout the entire year of training in these subjects. The
decreases in daytime and 24-hour systolic and mean
arterial pressure after 1 year of training were likely
the result of the exercise stimulus per se, because
24-hour levels of arterial pressure remained remarkably constant over the same period of time in the
nonexercising controls.
Little information is available concerning the effects of physical training on blood pressure through-
590
Hypertension
Vol 18, No 5 November 1991
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
out an entire day. In young and middle-aged normotensive men, Fortmann et al19 observed small (2-3
mm Hg) decreases in daytime and nighttime systolic
and diastolic pressure after a 1-year program of
exercise plus diet-induced weight loss. In a similar
population, Van Hoof et al30 reported that 4 months
of aerobic training lowered daytime diastolic pressure by 5 mm Hg but did not alter nighttime levels or
24-hour systolic pressure. Finally, Gilders et al20
recently observed no change in 24-hour levels of
arterial blood pressure after 16 weeks of regular
aerobic exercise in young and middle-aged normotensive or borderline hypertensive men and
women. In each of these studies, training produced
significant increases (13-14%) in aerobic capacity.
Taken together, the results of these investigations
and those from the present study indicate that lowto-moderate intensity aerobic exercise training has, if
any, only a small effect on 24-hour levels of arterial
blood pressure in normotensive or mildly hypertensive healthy humans. It is possible that high-intensity
training would evoke larger reductions in 24-hour
arterial pressure, but this postulate is not supported
by experimental findings on casual measurements,
which indicate that low-level exercise exerts as great
or greater hypotensive effect than more strenuous
training.11-13
Training-Induced Changes in Casual Versus
Ambulatory Measured Blood Pressure
In the present investigation, 6 months of regular
aerobic exercise was associated with significant reductions in conventionally measured systolic, mean,
and diastolic arterial blood pressure at rest, and the
changes were consistent among the three body positions studied. The data from the 1-year training
subgroup indicated that essentially no further change
occurred in this measure of arterial pressure with an
additional 6 months of exercise. The magnitudes of
the training-evoked reductions in these casual readings were similar to the average decreases reported
previously in young and middle-aged hypertensive
humans, especially when expressed as percent
change from initial levels.911-12 Compared with our
findings, Hagberg et al13 recently reported larger
absolute (i.e., mm Hg) reductions in arterial pressure
at rest after 9 months of low-intensity aerobic training in older subjects with mild diastolic hypertension.
However, the percent changes from initial levels to
the end of training were not markedly different in the
two studies, in part because of the higher initial levels
of systolic pressure in their subjects. Thus, based on
these casual measurements, the data from our exercising subjects per se and the findings of Hagberg et
al13 suggest that regular aerobic exercise has a blood
pressure-lowering effect in older hypertensive humans of similar magnitude to that observed in
younger hypertensive subjects.
Acute measurements of arterial pressure made
under slightly different experimental conditions,
however, provide a somewhat different picture. In
the present study, during the initial round of testing
the absolute levels of blood pressure determined at
rest during a protocol in which cardiac output was
also measured (i.e., session 2) were similar to the
other casually determined pressures. In spite of this,
reductions in the session 2 blood pressures over 6
months of training were 1) observed only for diastolic
and mean levels, not systolic; 2) significant only in the
supine position; and 3) 50% or less of the magnitude
observed in the other casual sessions. A similar
discrepancy in results between the two casual blood
pressure measurement conditions were observed by
Hagberg et al,13 but their intersession differences
were even greater than those observed here. Collectively, our results and those of Hagberg et al13
demonstrate that the magnitude of training-induced
reductions in casually determined arterial pressure at
rest is protocol-specific.
Do these casual measurements at rest accurately
reflect training-induced changes in arterial pressure
throughout an entire day under freely behaving conditions? In contrast to the reductions observed at rest
under laboratory conditions, ambulatory-measured
pressures were not altered after 6 months of training
in the present study. In a subset of the subjects,
24-hour levels of mean arterial pressure were lower
after 1 year of training. However, unlike the casually
recorded levels, these reductions in 24-hour pressure
were mediated solely by reductions in systolic pressure. This discrepancy was not likely due to differences in initial absolute levels (i.e., law of initial
baseline) because the ambulatory systolic pressures
were lower than, and the diastolic pressures (especially daytime) were similar to, the corresponding
casually determined levels.
The results of the three previous studies on physical training and ambulatory blood pressure fail to
provide any consistent insight into this issue. The
data of Fortmann et al19 are similar to our findings in
that the magnitudes of the training-induced reductions in casually recorded pressures at rest were
much greater than the changes in ambulatory levels.
In contrast, Van Hoof et al30 reported 5 mm Hg
reductions in both casually determined and ambulatory-determined diastolic blood pressure and no
changes in systolic levels after training, suggesting a
good correlation between the two measures on average. Gilders et al20 observed no training-associated
changes in either casual or ambulatory pressures in
their study. Considered together, the present findings
and those from previous reports19-20-30 demonstrate
that casual readings at rest and 24-hour ambulatory
recordings can provide quite different information
regarding the influence of physical training on
chronic levels of arterial blood pressure, including 1)
whether a change occurred, 2) the particular phase of
the blood pressure cycle that was altered, and 3) the
magnitude of change.
The same could be said with respect to measurements over time in nonexercising control subjects. In
our investigation, the conventionally determined pres-
Seals and Reiling
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
sures at rest (Table 2) fell as much over 6 months in the
control subjects as in the training group. Most importantly, 24-hour ambulatory-recorded levels were unchanged over the same period of time in the control
subjects (Table 4). Similar discrepancies between the
two measurements have been reported previously for
control subjects in longitudinal studies involving both
physical training19 and other interventions.18 Furthermore, because the pressures in session 2 of the present
investigation did not change over time in the control
subjects, our data indicate that casual determinations
made under different experimental conditions also can
markedly influence conclusions drawn on this portion
of the study population.
As discussed in earlier reviews on this topic,911
several laboratories have reported reductions in casually recorded blood pressure levels in nonexercising controls of similar magnitude to that observed in
their training subjects.31"33 Insufficiently long screening periods before enrollment probably contributed
to some of these observations911 (i.e., blood pressure
decreased over time as subjects became more familiar with the measurement conditions). This clearly
was not the case in the present investigation because
we documented that blood pressure was stable in
every subject before they were allowed to continue in
the study. Thus, some common, unknown factor
specifically influenced this measure of arterial pressure in both of our groups. From an interpretive
standpoint, these data suggest that the decreases in
conventionally recorded arterial pressure observed in
our exercising subjects may not have been caused by
physical training but instead by a general effect
associated with study participation.
Systemic Hemodynamics
The decreases in casually determined blood pressure at rest after 6 months of training were associated
with small but significant reductions in heart rate and
cardiac output and no changes in stroke volume or
systemic vascular resistance (Figure 3). In the subgroup that completed 1 year of training, the further
slight reduction in arterial pressure over the latter 6
months of exercise was associated with similar trends
in heart rate and cardiac output. These observations
confirm the recent findings of Hagberg et al13 on
older men and women with mild diastolic hypertension and indicate that in this population, decreases in
arterial pressure at rest after low-intensity aerobic
training are mediated solely by bradycardia-evoked
reductions in cardiac output.
The magnitude of the decrease in heart rate at rest
was similar to the reduction we reported previously in
normotensive middle-aged and older men who underwent more strenuous aerobic training.22 In the present
study, this training-induced bradycardia at rest was
associated with a trend for a decrease in 24-hour heart
rate (Table 4). This trend was, in turn, solely the result
of a significantly lower daytime heart rate. The fact that
24-hour levels of heart rate were unchanged in the
Exercise and 24-Hour Arterial Pressure
591
control subjects indicates that this bradycardia was an
adaptation to the chronic exercise stimulus.
Clinical Significance
Cardiovascular disease increases dramatically with
advancing age and contributes greatly to morbity,
disability, and mortality in the elderly.34 Hypertension appears to be the key factor underlying this
increased risk because of its high prevalence in older
people.34 There is now convincing experimental support that lowering levels of arterial pressure at rest
reduces cardiovascular risk in elderly hypertensive
patients.5-7 However, because of the side-effects and
economic considerations associated with drug therapy, nonpharmacological interventions may be most
efficacious in older people with mild essential hypertension.6'8 The present findings (session 2 data) and
those of Hagberg et al13 demonstrate that regularly
performed low-intensity aerobic exercise lowers arterial blood pressure at rest in middle-aged and older
men and women with mild diastolic hypertension and
that the degree of reduction can be as much or more
than that observed in younger hypertensive subjects.
Such changes should result in a lowering of cardiovascular risk in these older subjects based on data
from clinical trials and epidemiological investigations
using casual recordings of blood pressure.2-7 Therefore, it is not clear how discrepancies in traininginduced blood pressure changes such as those observed in the present study affect the clinical
importance of this intervention. Our data and earlier
reports14-19 do indicate, however, that to obtain
accurate information on chronic levels of arterial
pressure over time, 24-hour ambulatory recordings
should be performed whenever possible along with
traditional casual readings.
Acknowledgments
We thank lisa Bare and Peter Chase for their
technical assistance during this investigation.
References
1. Harlan WR, Hull AL, Schmouder RL, Landis JR, Larkin FA,
Thompson FE: High blood pressure in older Americans: The
first national health and nutrition examination survey. Hypertension 1984;6:802-809
2. Schoenberger JA: Epidemiology of systolic and diastolic systemic blood pressure elevation in the elderly. Am J Cardioi
1986;57:45C-51C
3. Harris T, Cook EF, Kannel WB, Schatzkin A, Goldman L:
Blood pressure experience and risk of cardiovascular disease
in the elderly. Hypertension 1985;7:118-124
4. Vokonas PS, Kannel WB, Cupples LA: Epidemiology andriskof
hypertension in the elderly: The Framingham study. / Hypertens
1988;6(suppl 1):S3-S9
5. European Working Party on High Blood Pressure in the
Elderly: Mortality and morbidity results from the European
Working Party on High Blood Pressure in the Elderly trial.
Lancet 1985;1:1349-1354
6. Sowers JR: Hypertension in the elderly. Am J Med 1987;
82(suppl lB):l-8
7. Coope J: Hypertension in the elderly. J Hypertens 1987;5(suppl
3):S69-S72
8. Ouslander JG: Drug therapy in the elderly. Ann Intern Med
1981;95:711-722
592
Hypertension
Vol 18, No 5 November 1991
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
9. Seals DR, Hagberg JM: The effect of exercise training on
human hypertension: A review. Med Sci Sports Exerc 1984; 16:
207-215
10. Tipton CM: Exercise, training, and hypertension, in Terjung R
(ed): Exercise and Sport Science Reviews. Lexington, Mass, DC
Heath, 1984, vol 12, pp 245-306
11. Hagberg JM, Seals DR: Exercise training and hypertension.
Ada Med Scand 1986;711:131-136
12. Hagberg JM: Exercise, fitness, and hypertension, in Bouchard
C, Shephard RJ, Stephens T, Sutton JR, McPherson BD (eds):
Exercise, Fitness, and Health. Champaign, 111, Human Kinetics
Books, 1990, pp 455-466
13. Hagberg JM, Montain SJ, Martin WH, Ehsani AA: Effect of
exercise training in 60- to 69-year-old persons with essential
hypertension. Am J Cardiol 1989;64:348-353
14. Floras JAS, Jones JV, Hassan MO, Osikowska B, Sever PS,
Sleight P: Cuff and ambulatory blood pressure in subjects with
essential hypertension. Lancet 1981;2:107-109
15. Torriani S, Waeber B, Petrillo A, Di Stefano R, Mooser V,
Scherrer U, Nussberger J, Hofstetter J-R, Brunner HR:
Ambulatory blood pressure monitoring in the elderly hypertensive patient. / Hypertens 1988;6{suppl 1):S25-S27
16. Pessina AC, Palatini P, Di Marco A, Mormino P, Fazio G,
Libardoni M, Mos L, Casiglia E, Dal Palu C: Continuous
ambulatory blood pressure monitoring versus casual blood
pressure in borderline hypertension. / Cardiovasc Pharmacol
1986;8(suppl 5):S93-S97
17. Pickering TG, Devereux RB: Ambulatory monitoring of blood
pressure as a predictor of cardiovascular risk. Am Heart J
1987;114:925-928
18. Marolf AP, Hany S, Battig B, Vetter W: Comparison of casual,
ambulatory and self-determined blood pressure measurement.
Nephron 1987;47(suppl 1):142-145
19. Fortmann SP, Haskell WL, Wood PD, Standford Weight
Control Project Team: Effects of weight loss on clinic and
ambulatory blood pressure in normotensive men. Am J Cardiol
1988;62:89-93
20. Gilders RM, Voner C, Dudley GA: Endurance training and
blood pressure in normotensive and hypertensive adults. Med
Sci Sports Exerc 1989;21:629-636
21. Seals DR, Hagberg JM, Hurley BF, Ehsani AA, Holloszy JO:
Endurance training in older men and women: I. Cardiovascular responses to exercise. J Appl Physiol 1984;57:1024-1029
22. Seals DR, Chase PB: Influence of physical training on heart
rate variability and baroreflex circulatory control. / Appl
Physiol 1989;66:1886-1895
23. Jackson AS, Pollock ML: Generalized equations for predicting
body density of man. BrJNutr 1978;40:497-504
24. Sin WE: Body composition from fluid spaces and density, in
Brozek J, Henschel A (eds): Techniques for Measuring Body
Composition. Washington, DC, National Academy of Science,
1961, pp 223-244
25. Recommendations for human blood pressure determination
by sphygmomanometers. Report of a special task force
appointed by a special Steering Committee, American Heart
Association: Frohlich ED, Grim C, Labarthe DR, Maxwell
MH, Perloff D, Weidman WH. Circulation 1988;77:
502A-514A
26. Collier CR: Determination of mixed venous CO2 tensions by
rebreathing. J Appl Physiol 1956;25-29
27. Casadei R, Parati G, Pomidossi G, GroppeUi A, Trazzi S, Di
Rienzo M, Mancia G: 24-Hour blood pressure monitoring:
Evaluation of Spacelabs 5300 monitor by comparison with intraarterial blood pressure recording in ambulant subjects. / Hypertens 1988;6:797-803
28. Graettinger WF, Lipson JL, Cheung DG, Weber MA: Validation of portable noninvasive blood pressure monitoring
devices: Comparisons with intra-arterial and sphygmomanometer measurements. Am Heart J 1988;116:1155-1160
29. Holloszy JO, Ehsani AA, Hagberg JM: Exercise training in
coronary heart disease, in Knuttgen HG, Poortmans J, Vogel
JA (eds): Biochemistry of Exercise. Champaign, 111, Human
Kinetics, 1982, vol 13, pp 87-96
30. Van Hoof R, Hespel P, Fagard R, lijnen P, Staessen J, Amery
A: Effect of endurance training on blood pressure at rest,
during exercise and during 24 hours in sedentary men. Am J
Cardiol 1989;63:945-949
31. Bonnano JA, Lies JE: Effects of physical training on coronary
risk factors. Am J Cardiol 1974^33:760-763
32. Kukkonen K, Rauramaa R, Voutilainen E, Lansimes E:
Physical training of middle-aged men with borderline hypertension. Ann Clin Res 1982;14(suppl):130-145
33. DePlaen JF, Detry JM: Hemodynamic effects of physical
training in established arterial hypertension. Acta Cardiol
198O;35:179-188
KEY WORDS • aging
•
hemodynamics
ambulatory blood pressure monitoring
Effect of regular exercise on 24-hour arterial pressure in older hypertensive humans.
D R Seals and M J Reiling
Hypertension. 1991;18:583-592
doi: 10.1161/01.HYP.18.5.583
Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1991 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://hyper.ahajournals.org/content/18/5/583
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Hypertension is online at:
http://hyper.ahajournals.org//subscriptions/