Clinical and Experimental Pharmacology and Physiology (2006) 33, 868871

doi: 10.1111/j.1440-1681.2006.04456.x

Beilin Festschrift

Blackwell Publishing Asia

EXERCISE AND BLOOD PRESSURE: APPLYING FINDINGS FROM

THE LABORATORY TO THE COMMUNITY SETTING

Exercise
KL
Cox and blood pressure change

Kay L Cox
School of Medicine and Pharmacology University of Western Australia, Western Australian Institute for
Medical Research, Perth, Western Australia, Australia

SUMMARY
1. During the 1980s, there was growing epidemiological
evidence that aerobic training lowered blood pressure (BP). Early
intervention studies had not always supported this conclusion.
Such studies were limited by methodological shortcomings. Few
studies had used normotensive individuals or women alone.
2. Exercise training is an attractive lifestyle strategy in the
prevention of hypertension. In our studies in men, vigorous
intensity cycling did not lower resting BP.
3. Vigorous exercise reduces day time ambulatory BP and, in
combination with calorie restriction, had a synergist effect to
reduce night-time and 24 h BP.
4. Exercise is a positive and effective adjunct to other lifestyle
measures in the prevention of hypertension. Adherence to either
supervised or unsupervised moderate-intensity exercise is
sufficient to reduce BP in the short and long term.
5. Further studies need to be performed to evaluate the
clinical significance and mechanisms for the increase in resting
BP observed in older women.
6. Older individuals with hypertension should be monitored
when taking up a swimming programme.
7. Exercise induced changes in BP seen in the laboratory are
also observed in the community setting.
Key words: alcohol, blood pressure, exercise, intensity, lifestyle,
overweight, swimming, walking.

had numerous methodological limitations, such as lack of a control

group, control of other lifestyle factors, including diet and alcohol
intake, confounders such as age, gender and body composition and
standardization of BP measurement.4 Furthermore, studies used a
variety of modes of exercise training, had varying lengths of sessions,
frequency of training and exercise intensity, making comparisons
between studies difficult.
Although it was recognized that exercise training may be a potentially potent intervention for the initial treatment of high BP, few
studies had examined the role of exercise training in the prevention
of hypertension in those with normal BP. A study by Jennings et al.5
was pivotal in setting the standard of rigour in research in this area
and the direction that the research may take for us and others. This
is demonstrated by the following quote:
Epidemiological evidence suggesting that physical training lowers blood
pressure has recently been confirmed by well-controlled exercise
intervention studies. Jennings et al. (1986)5 showed that 3 training
sessions of 1 h per week at 6070% maximum exercise capacity lowered
systolic blood pressures 1012 mmHg over 4 weeks compared with
sedentary periods. Whether such vigorous training is needed to produce
this effect is not clear.6

The present paper will focus on the work performed by our group
on exercise training and BP and the transition and application of
exercise training from the laboratory to the community setting.

BACKGROUND
INTRODUCTION
The early to mid 1980s saw the emergence of growing epidemiological evidence that long-term aerobic exercise was associated with
lower blood pressure (BP).1 Early intervention studies using exercise
training to reduce BP in hypertensives had mixed results, with some
reporting reductions in BP,2 but not others.3 However, these studies
Correspondence: Dr Kay L Cox, University of Western Australia, School
of Medicine and Pharmacology (Royal Perth Hospital Unit), GPO Box
X2213, Perth, WA 6001, Australia. Email: kaycox@cyllene.uwa.edu.au
Presented at a symposium in honour of Lawrie Beilin held in conjunction
with the High Blood Pressure Research Council of Australia Annual Scientific
Meeting, Melbourne, 67 December 2005. The papers in these proceedings
have been peer reviewed.
Received 19 January 2006; accepted 6 March 2006
2006 Blackwell Publishing Asia Pty Ltd

The study by Jennings et al.5 was an important stimulus for work

in this area by providing five key prompts. First, the study provided
an example of a well-controlled study with clearly defined protocols,
particularly in relation to the exercise regimen. Second, reductions
in BP were seen within 4 weeks, whereas previous studies had
required much longer interventions. Long training studies are difficult
to conduct and expensive to resource; thus, a shorter effective
training regimen provided a feasible and practical option. Third, the
study used normotensive subjects, which highlighted the potential
for exercise training to be used in the prevention of hypertension.
Fourth, the study posed the question, could exercise be combined
with other lifestyle interventions to produce an additive or synergistic
effect? Fifth, we questioned whether sedentary individuals could
tolerate this exercise intensity and whether the exercise regimen
could be maintained in the longer term. Because our work had a

Exercise and blood pressure change

focus on the prevention of hypertension, the present paper highlights
four studies in individuals with normal BP that were undertaken,
two in men and two in women, to investigate the effects on BP of
combining lifestyle factors and modifying predominantly laboratory
exercise regimens for application to the community setting.

ALCOHOL RESTRICTION AND EXERCISE

Previously, it was demonstrated that 6 weeks of alcohol restriction
reduced systolic and diastolic BP by 4 and 1.5 mmHg, respectively,
in normotensive men.7 We hypothesised that, with the combination
of alcohol restriction and exercise, there would be an independent
fall in BP with both alcohol restriction and exercise and, further, that
there would be an additive effect of alcohol restriction and exercise
in combination.8 Healthy, but inactive, 2045-year-old men (n = 75)
with high normal BP were randomized to a 2 2 factorial design
study. The men were randomized to one study arm in which they
were asked to maintain their normal alcohol intake or to a lowalcohol arm, where they were restricted to light-alcohol beer (0.9%
alcohol, v/v). Within each of the arms, they were further allocated
to a light exercise control group or a vigorous exercise group. The
vigorous exercise group was prescribed the same protocol described
by Jennings et al.5 This was three sessions per week of stationary
cycling for 30 min at 6070% of the individuals previously determined maximum workload. A post hoc analysis of the exercise
intensity from heart rate recordings showed the subjects were working
at 80% of heart rate reserve, which is considered vigorous exercise.
The light exercise group completed the same sessions, but they
pedaled against no load. All sessions were completed as supervised
group sessions and the 72 who completed the study had 100%
adherence to the exercise.
For all the studies described in this paper, resting BP was measured with a Dinamap 1846SX recorder (Critikon, Tampa, FL, USA).
Mean BP for each visit was calculated from 10 supine readings taken
over 20 min and five standing readings over 5 min. In this study, BP
was evaluated at baseline, weekly and post intervention.
There was a significant main effect relative to control of alcohol
restriction on BP, with reductions of 4.7 and 1.9 mmHg in supine
systolic and diastolic BP, respectively. This effect was consistent
with our previous findings.7 Despite a 10% increase in fitness with
vigorous exercise, there was no significant effect of vigorous exercise
on systolic or diastolic BP. The reason why we were not able to
replicate the reductions in BP using the same exercise protocol
reported by Jennings et al.5 is not known. It is possible that differences between the studies, such as design, a higher exercise intensity
and the use of a light exercise control, could have contributed to the
difference in results between the two studies.

869

Again a two-way factorial study design was used and those allocated to the vigorous exercise group used the same exercise regimen
as in the previous study. The light exercise protocol used the same
stationary cycling programme as the previous study for 2 days a
week, whereas the third session comprised a slow walk or a stretching programme on alternate weeks. In addition to the resting BP
measurements, we also measured 24 h ambulatory BP at baseline
and the end of intervention using the Accutracker II monitor
(Suntech Medical Instruments, Raleigh, NC, USA).
Calorie restriction reduced bodyweight by 9.5 kg, but there was
no significant change with vigorous exercise. Vigorous exercise
increased fitness by 24% with no change in fitness seen in the light
exercise group. Falls in resting systolic and diastolic BP were seen
with calorie restriction, but not with vigorous exercise. The calorie
restriction main effect on supine systolic and diastolic BP was 5.6
and 2.4 mmHg, respectively. However, when adjusted for the change
in alcohol consumption, only the reduction in systolic BP was
significant. Changes in resting systolic and diastolic BP with vigorous,
relative to light, exercise were 1 and 0.5 mmHg, respectively.
Using a conventional analysis of ambulatory BP, there was a
significant reduction in 24 h and day time systolic and diastolic BP
with calorie restriction, but not vigorous exercise. However, when
we used a time series analysis (TSA) of the ambulatory BP data, a
comparison of models examining the effect of calorie restriction in
the light exercise groups showed that the fall in systolic BP from
baseline was significantly greater in the low calorie group and that
this effect was greater during the day time. In the vigorous exercise
groups, calorie restriction was also associated with a greater fall in
BP and this was sustained throughout the 24 h period. Comparison
of models of the effects of exercise intensity with subjects having
normal caloric intake showed that there was a significantly greater
fall associated with vigorous exercise for systolic BP during the day
time. For subjects with calorie restriction, there was a greater fall
in the vigorous exercise group that was significantly greater during
the night-time. There was no significant effect of calorie restriction or
vigorous exercise when a similar comparison of models for diastolic
BP was used.
This study was of particular interest in that although we were
unable to demonstrate an effect of vigorous exercise on resting BP or
24 h ambulatory BP analysed by conventional methods, when a more
powerful TSA approach was used we demonstrated a decrease in
day time ambulatory BP with either vigorous exercise or calorie
restriction. Furthermore, although the two approaches did not have
an additive effect on day time BP, when combined they produced a
synergistic effect to also lower night-time BP, which resulted in BP
reduction over the 24 h period for those assigned to both calorie
restriction and vigorous exercise. This synergistic effect had not been
reported previously.

CALORIE RESTRICTION AND EXERCISE

To further investigate the combination of lifestyle factors to reduce
BP, we investigated the independent and combined effects of 16 weeks
of calorie restriction and vigorous exercise in 60 overweight
sedentary men aged 2050 years.9 We used a dietary intervention that
reduced total energy intake by 10001500 kcals/day, comprised of
15% protein, 30% fat and 55% carbohydrate. When used previously,
we demonstrated a fall in resting BP.10 We hypothesized that calorie
restriction and exercise would independently reduce BP and that,
when combined, there would be an additive effect.

EXERCISE INTENSITY AND LONG-TERM

BP CHANGE IN WOMEN
Most of the previous work evaluating the effects of exercise on BP
by us and others had used predominantly men. Hagberg,11 in a review
paper of hypertensives, highlighted that of the 25 studies reviewed
only three had used female subjects only. This posed the question
of whether women had the same BP response to exercise as men.
Furthermore, in earlier studies the exercise programme was usually
highly supervised and conducted in a centre-based laboratory setting.

2006 Blackwell Publishing Asia Pty Ltd

870

KL Cox

In this setting, in the previous two studies we achieved a 100%

adherence to the number of exercise sessions and compliance to the
exercise intensity prescription. What was not known was, would
participants who had less supervision adhere to the prescribed number
of sessions in an exercise programme, would they comply with the
exercise intensity and would they be able to continue with this in
the long term. These were important questions to be resolved before
exercise could be considered an effective lifestyle strategy in the
prevention of hypertension.
In a study investigating the optimal strategies for initiating
and maintaining a regular exercise programme in 126 previously
sedentary older women (4065 years), we took the opportunity to
investigate the effects of 6 and 18 months of exercise on resting
BP.12,13 Women were randomly assigned to either a closely supervised centre-based or a minimally supervised home-based exercise
programme for an initial 6 months. All subjects participated in a
minimally supervised home-based exercise programme for a further
12 months. The women were further randomized to exercise at either
moderate intensity (4055% maximum heart rate reserve (HRres))
or vigorous intensity (6580% HRres). They completed three 30 min
sessions/week with four walks, one aerobics session and one circuit
session prescribed for each fortnight. Resting BP was measured
using the previously described protocol.
After 6 months, adherence to exercise was higher in the centrebased group (84 vs 63%); however, at 18 months, adherence was
similar in both the centre- and home-based groups (73 vs 79%,
respectively). The prescribed intensity of exercise did not influence
adherence rates (76% for the moderate and 71% for the vigorous
intensity group at 6 months and 79 vs 72%, respectively, at 18 months;
an approximate mean of 2.3 sessions/week). After 6 months, there
was a significant fall in systolic BP of 2.7 mmHg with moderate but
not vigorous exercise. At 18 months, the fall in systolic BP with
moderate exercise (2.8 mmHg) was maintained and there was a
significant fall of 2.70 mmHg in diastolic BP with moderate intensity
exercise, but not with vigorous intensity exercise (0.21 mmHg for
systolic BP and 1.8 mmHg for diastolic BP).
This study demonstrated that even in a largely normotensive population, regular moderate intensity exercise maintained in the long
term achieved sustained falls in resting systolic and diastolic BP. This
study highlighted that, in older women, moderate intensity exercise
is well accepted, sustainable long term and has the health benefit of
reduced BP. Such a finding has important public health implications
for the promotion of exercise as a lifestyle and lifelong strategy for
the prevention of hypertension.

individuals16 highlighted a need for the efficacy of swimming to

reduce BP to be evaluated. One non-randomized controlled trial in
middle-aged hypertensive men and women reported a fall in BP after
10 weeks of swimming training.17
In further study using women, we compared the longer-term
effects of walking and swimming training on resting BP in previously sedentary, older individuals with normal or controlled BP.18
Another aim of that study was to compare a behavioural intervention
versus a usual care approach to promote adherence to the physical
activity mode. Sedentary women (n = 116) aged 5070 years were
randomly assigned to a walk or swim programme and then further
randomized to a behavioural intervention or a usual care programme.
Again, we used a supervised centre-based programme for 6 months
and then the women completed the same programme home based
and without supervision for a further 6 months. The women completed
three sessions per week, with each session comprised of 10 min
warm-up/cool down and 5 min stretching, before and after 30 min
swimming or walking at a moderate intensity (6070% HRRes).
Adherence to the number of sessions completed in the walking
and swimming grousp were similar at 6 and 12 months: 75%
(2.3 sessions/week) and 64% (1.9 sessions/week), respectively.
There was no significant difference in intensity between the walking
(59.7% HRRes) and the swimming group (60.9% HRRes). When a post
hoc analysis of the training intensity for the swim group was
calculated using the adjusted maximum heart rate for swimming,
the training intensity was significantly different from the walk group
(66.5% HRRes); however, both training intensities were still in the
moderate intensity classification.
There was a significant main effect of swimming relative to
walking to increase supine systolic BP by 4.4 mmHg after 6 months.
Diastolic BP also increased by 1.4 mmHg, but this was not statistically
significant (P = 0.07). The increase in systolic and diastolic BP with
swimming was maintained at 12 months; however, the falls in BP
seen with walking were not maintained and the difference between
the two groups of 3.1 mmHg systolic BP and 1.1 mmHg diastolic
BP was no longer significant (Fig. 1). It is possible that the reduction
in the number of sessions completed in the second 6 months was

COMPARISON OF THE EFFECTS OF WALKING

AND SWIMMING ON BP
Walking is the most popular activity performed by Australians for
exercise, followed by swimming. Our previous study demonstrated
that moderate walking reduces BP. Swimming is frequently recommended as a form of exercise to improve health. However, this
recommendation is based on the assumption that all aerobic exercise
produces the same health benefits and that the effects of swimming
are the same as walking, running and cycling. A lack of information
of the effects of swimming on BP makes this assertion tenuous.14
Earlier findings of a higher mean arterial BP for the same heart rate
with swimming compared with running15 and observed increases
in BP with immersion in water and with acute swimming in older

Fig. 1 Changes in resting systolic blood pressure (SBP) in the walking and
swimming groups after 6 (
) and 12 months ( ). There was a significant
increase in SBP with swimming relative to walking after 6 months. Data are
the mean SEM. **P < 0.01.

2006 Blackwell Publishing Asia Pty Ltd

Exercise and blood pressure change

responsible for the failure of the walking to maintain the falls in BP
and it could be speculated that with maintenance of the higher
number of sessions the difference between the two groups would
have still been apparent.
The increase in BP with swimming is a novel finding. The reasons
for this are unclear; however, the swimming environment is different
from that of land-based exercise and swimming training evokes
complex responses that may be a result of the different posture, the
effects of hydrostatic pressure, facial immersion, breath holding,
water temperature and the increased thermal conductivity of water.
The observed increase in BP may be a necessary adaptation to the
aquatic environment. The clinical relevance of this increase in BP
is also not known and requires further investigation, especially in
those with hypertension. However, the finding suggests that any
recommendation for older, sedentary novice swimmers to take up
swimming should be made with caution and that they should have
regular BP monitoring.

CONCLUSIONS
In a series of four randomized controlled trials, we have demonstrated
that the BP-lowering effects of exercise seen in highly supervised
research settings can be achieved in unsupervised community settings
in individuals with normal BP. We were not able to replicate earlier
work demonstrating that 4 weeks of cycling reduced resting BP, nor
were we able to demonstrate that a combination of exercise and other
lifestyle factors had an additive effect in lowering BP. However, we
did demonstrate an independent effect of vigorous exercise to lower
day time ambulatory BP and, when combined with calorie restriction, there was a synergistic effect to lower night-time BP and 24 h
BP. Exercise is a useful adjunct to other lifestyle measures in the
prevention of hypertension. Moderate intensity exercise is well
tolerated by older individuals and, whether supervised or unsupervised,
completed on a regular basis is sufficient to reduce BP in the short
and long term. The mode of exercise needs to be taken into account
when recommending exercise programmes and older individuals
with hypertension should be monitored when taking up a swimming
programme. Further studies need to be performed to evaluate the
clinical significance and mechanisms for the increase in resting
BP observed in older women and to asses whether this effect is
generalized to other populations.

REFERENCES
1. Bjorntorp P. Effects of physical training on blood pressure in hypertension. Eur. Heart J. 1987; 8: 719.

871

2. Harris WE, Bowerman W, McFadden RB, Kerns TA. Jogging: An adult

exercise program. JAMA 1967; 201: 75961.
3. Kasch FW, Kulberg J. Physiological variables during 15 years of
endurance exercise. Scand. J. Sports Sci. 1981; 3: 5962.
4. Hagberg JM, Seals DR. Exercise training and hypertension. Acta Med.
Scand. Suppl. 1986; 711: 1316.
5. Jennings G, Nelson L, Nestel P et al. The effects of changes in physical
activity on major cardiovascular risk factors, hemodynamics,
sympathetic function, and glucose utilization in man: A controlled
study of four levels of activity. Circulation. 1986; 73: 3040.
6. Beilin LJ. Value of non-drug treatment and drug treatment in
hypertension. Drugs 1988; 36: 19.
7. Puddey IB, Beilin LJ, Vandongen R, Rouse II, Rogers P. Evidence for
a direct effect of alcohol on blood pressure in normotensive men: A
randomized controlled trial. Hypertension 1985; 7: 70713.
8. Cox KL, Puddey IB, Morton AR, Vandongen R, Beilin LJ, Masarei
JRL. The combined effects of aerobic exercise and alcohol restriction
on blood pressure and serum lipid-lipoproteins: A two way factorial
study in sedentary men. J. Hypertens. 1993; 11: 191201.
9. Cox KL, Puddey IB, Morton AR, Burke V, Beilin LJ, McAleer M.
Exercise and weight control in sedentary overweight men: Effects
on clinic and ambulatory blood pressure. J. Hypertens. 1996; 14: 779
90.
10. Puddey IB, Parker M, Beilin LJ, Vandongen R, Maserei JRI. Effects
of alcohol and calorie restriction on blood pressure and serum lipids
in overweight men. Hypertension 1992; 20: 53341.
11. Hagberg JM. Exercise, fitness and hypertension. In: Bouchard C,
Shephard RJ, Stephens T, Sutton JR, McPherson BD (eds). Exercise,
Fitness and Health. Human Kinetics Books, Champaign. 1990; 455
66.
12. Cox KL, Puddey IB, Burke V, Beilin LJ, Morton AR. Determinants
of change in blood pressure during S.W.E.A.T.: The Sedentary Women
Exercise Adherence Trial. Clin. Exp. Pharmacol. Physiol. 1996; 23:
5679.
13. Cox KL, Burke V, Morton AR, Gillum HF, Beilin LJ, Puddey IB.
Long-term effects of exercise on blood pressure and lipids in healthy
women 4065 years: The Sedentary Women Exercise Adherence Trial
(S.W.E.A.T.). J. Hypertens. 2001; 19: 173343.
14. Jennings GLR. Exercise and blood pressure: Walk, run or swim?
J. Hypertens. 1997; 15: 5679.
15. Holmer I. Physiology of swimming man. Acta Physiol. Scand. Suppl.
1974; 407: 155.
16. Itoh M, Araki H, Hotokebuchi N, Takeshita T, Gotoh K, Nishi K.
Increased heart rate and blood pressure response, and occurrence of
arrhythmias in elderly swimmers. J. Sports Med. Phys. Fitness 1994;
34: 16978.
17. Tanaka H, Bassett Jr DR, Howley ET, Thompson DL, Ashraf M,
Rawson FL. Swimming training lowers the resting blood pressure in
individuals with hypertension. J. Hypertens. 1997; 15: 6517.
18. Cox KL, Burke V, Beilin LJ, Grove JR, Blanksby BA, Puddey IB. Blood
pressure rise with swimming versus walking in older women: The
Sedentary Women Exercise Adherence Trial 2 (SWEAT 2). J. Hypertens. 2006; 24: 30714.

2006 Blackwell Publishing Asia Pty Ltd