The Effect of Swimming Patients with Ischemic Heart Disease

The Effect of Swimming on Patients
with Ischemic Heart Disease

SHELDON MAGDER, M.D., DAG LINNARSSON, M.D., AND LENNART GULLSTRAND, PHYS. ED.

Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016 SUMMARY Swimming is frequently recommended for cardiac rehabilitation, but little is known of its
physiologic consequences in ischemic heart disease. Eight males who had had a myocardial infarction 8-17
months before the study were exercised to exhaustion or angina with 10 W/min-1 ramp on a cycle ergometer in

sitting and supine positions. Oxygen uptake (V02) was continuously measured to monitor the physiologic

power requirement. All eight patients were taking fl blockers and four were taking digoxin. During sitting

cycling, angina occurred in four and ST depression in five; during supine cycling, angina occurred in five and
ST depression in six. VO2 was then measured while they swam at their own comfortable speed (mean 0.43
m/sec-') in a swimming flume at water temperatures of 25.5°C and 18°C. In six, the water speed was
gradually increased until they were limited by symptoms. Comfortable swimming at 25.5°C was 87% (1.28
1/min-m) and at 18°C 89% (1.30 1/min-m) of sitting peak V02, while heart rates were 92% and 91% respec-
tively. The mean peak VO2 and heart rate did not differ significantly between bicycle and swim tests (peak VO2
sitting 1.49 ± 0.23, supine 1.42 ± 0.24, 25.5°C 1.60 ± 0.17, 18°C 1.52 ± 0.19 1/min-'). Only two patients
reported angina while swimming in warm water and one in cold water, although ST depression occurred in six
in both swims. The subjective comfort and large muscle groups involved make swimming a good exercise, but
the high relative energy cost and failure to identify ischemic symptoms indicate caution in cardiac patients,
especially if their swimming skills are poor.

THE PHYSIOLOGY of swimming in fit normal sub- Preference was given to patients with a history of
jects and trained swimmers, both young and old, has
been described,1-3 but the pathophysiology of swim- angina or ST depression on a work test. All eight were
ming in patients with ischemic heart disease has been taking f, blockers and four were taking digoxin. These
studied little.4' 5 Our understanding of this condition were continued throughout the study. Further patient
raises theoretical concerns. Immersion in water6 and a
horizontal body position both increase the central data are given in table 1.
blood volume,7 which could stress the limited reserves
of patients with ischemic heart disease.8 The large con- A 12-lead ECG was obtained before the sitting cy-
tribution of work from the arms, which results in a
higher peripheral vascular resistance than leg work at cling test and the six chest leads were recorded on a
the same work load,9 the compressive effect of the
water on the extremities" 6 and the decreased skin multichannel recorder during both cycling tests. Heart
blood flow" 6 all increase the left ventricular afterload
and thus might decrease the subjects' peak work rate was determined by means of a cardiotachometer.
level.10 The swimming flume, a kind of "swimming
treadmill," offered a controlled and safe environment During swimming, the ECG signal from a CM6
for evaluating such patients."" Oxygen uptake (VO2)
was used as an objective measure of the amount of bipolar lead was transmitted telemetrically to an FM
work performed, and the ECG as an indicator of receiver and recorded by an Elema Mingograph 14.1
ischemic changes. The swimming results were related
to the subjects' performance in the sitting and supine Significant ST depression was assessed as greater than
positions on a bicycle ergometer. Further, swimming 1 mm ST depression 0.08 second after the J point.
tests were performed with water temperatures of 18°C
and 25°C to assess the effect of different water In the first two swimming tests, V02 was measured
temperatures. by the Douglas-bag method.' In all other cycling and
swimming tests, V02 and other gas-exchange
Patients and Methods variables were measured continuously. In the con-

We selected eight males, ages 49-66 years, from a tinuous system, expired air passed through a wide-
cardiac rehabilitation program. They all had had a
myocardial infarction 8-17 months previously. bore, low-resistance valve to a 6-1 mixing box and was
analyzed by a Centronic 200 MGA mass spec-
trometer. Argon was added just distal to the ex-
piratory valve and also analyzed by the mass spec-
trometer so that ventilation (VE) could be assessed by
an argon dilution method.12 A respiratory pump was
used to calibrate the VE monitoring system. The data
from the mass spectrometer and the heart rate were

recorded by a four-channel tape recorder. V02, CO2

output, VE, heart rate and oxygen pulse could then be
calculated off-line on a Hewlett Packard 9830 A
calculator and graphed by an HP 9866 plotter. The
following formulas were used12:

From the Department of Clinical Physiology, Thoracic Clinics, YE k k X VAR(O.9907 FAR) 1/min BTPS
Karolinska Hospital, and the Departments of Medical Engineering FAR FAR
V02 = kSTPD X VE(IO02 - FEO 2) 1/min STPD
and Physiology III, Karolinska Institutet, Stockholm, Sweden.
Supported by the Canadian Heart Foundation.
Address for correspondence: Sheldon A. Magder, M.D., Car-

diology Division, Royal Victoria Hospital, 687 Pine Avenue West, 02 pulse = VJO,/HR (ml/beat)
Montreal, Quebec, H3A lAl, Canada.

Received March 17, 1980; revision accepted September 3, 1980. where FAR = the difference of argon concentration

Circulation 63, No. 5, 1981. 979

980 CI RCULATION VOL 63, No 5, MAY 1981

TABTLE 1. Patient Data

Age Weight Height Months Mledication

Pt (years) (kg) (cm) Angina ASAT* from
(SIU) AMI

1 49 76 183 No 3.9.5) 8 Metoprolol 109 mg b.i.d.

2 60 79 169 NYHA class I 3.08 17 Pindolol 5) mg b.i.d.,

digoxin 1).25 mg

3 59 75) 176 NYHA class I 5.82 9 Propraniolol 20 mg b.i.d.,

digoxin 0.25 mg

4 66 77 165 NYHA class II 1.80 9 Metoprolol 100 mg b.i.d.

5 54 79 173 No 2.20 13 Metoprolol 100 mg b.i.d.

6 62 79 176 No 6.3.5 15 Aprenolol 10) mg b.i.d.,

7 .58 71 169 No 2.30 digoxin 0.25) mg

8 6,5 78 178 NYHA class II 3.00 16 MIetoprolol 100 mg b.i.d.,

*Highest level recorded dturinig acuite mvocardial infarction. digoxin 0.25 mg

Abbreviations: ASAT = aspartate-arginine amino transferase; SIU 12 Aprenolol 100 mg b.i.d.
acute myocardial infarction.
staiudard international u-nits; AMI

Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016 from air, FIO = inspired oxygen, and FEO = mixed easy communication with the subject, and requires a
relatively low swimming speed to elicit a maximal
02~~~~~~~~~~~~~~~~ physiologic response. The subjects were not spe-
cifically trained in swimming nor was swimming a
expired oxygen concentration, and HR = heart rate. part of their rehabilitation program; swimming skills
were not assessed before the test.
The peak VO, was assessed by manually integrating
A defibrillator and dry plastic matting were kept at
the last 30 seconds of the continuous tracing. the side of the pool. One of the investigators was
On two different days within 3 weeks of the swim- designated as the "dry" person. In the event of a car-
diac arrest he was to remain dry so he could safely
ming test, the patients performed both sitting and defibrillate the patient.

supine work tests on an electrically braked cycle The data were first analyzed by Friedman's non-
parametric test for analysis of variance. Differences
ergometer. The patients exercised until they could go between exercises were then analyzed by the Wilcox-
no further or developed grade 6 symptoms of dyspnea on's rank sum test for paired data. Values are given as
mean ± SD.
or angina as judged by a Borg scale of 1-10.13 In the
Results
sitting test they sat for 3 minutes at rest, pedaled for 3
minutes at a load of 20 W and then against a work The swimming flume and mouthpiece were well
tolerated by all subjects and there were no medical
load that increased by 10 W each minute. During the complications. The only arrhythmias that occurred
were occasional ectopic beats on entering the water, a
supine test, the patients first rested for 2 minutes, had run of bigeminy in one subject in both swims and a
their feet on the pedals for another 2 minutes, and then short run of bigeminy after exercise in another subject.

started pedaling at an initial load of 20 W, which was The VO, during comfortable speed swimming was
increased by 10 W each minute. 1.28 ± 0.12 1/min in warm water and 1.30 ± 0.20
1/min in cold water (table 2) (p = NS). This was
The swimming studies were performed in the swim- 87 ± 19% in warm and 89 ± 18% in cold water of the
peak .10, achieved during cycling. When the mean
ming flume of the G.I.H. in Stockholm" on two VO2 during comfortable speed swimming was com-
pared with the highest peak 102 obtained in any of the
separate days at water temperatures of 25.5°C and four tests, comfortable speed swimming required
18°C, respectively. On the first day, the subjects were 81 ± 16% at 25.5°C and 83 i 17% at 18°C of the
peak O2. In three subjects in warm water and an ad-
timed as they swam two lengths of a 25-meter pool at ditional two in cold water, the comfortable speed swim
their own comfortable speed. From this we calculated required a peak .102 that was the same as or higher
than that during cycling (fig. IA).
their water speed and will refer to this as their comfor-
table speed. Arterial blood pressure was measured The peak VO, during swimming was higher than or
before the swimming tests while the subjects were sub-
merged to the level of the neck, but not during swim- equal to the peak .102 during cycling in the sitting
ming. After standing for 2 minutes in the water, the
patients swam at comfortable speed for 4 minutes. In position in seven of the eight subjects (fig. 1), but there
the first two subjects in whom Douglas bags were
used, the comfortable speed was continued for up to 7

minutes. Because in three of these four studies the two
subjects were limited by symptoms before 7 minutes,
these data were included in the analyses of the peak

response. In the other six, after the first 4 minutes of

comfortable speed, the water speed was increased by
0.05 m/sec until the subjects had to stop.

By their own choice, all subjects swam the breast

stroke with their head out of the water. The breast

stroke can load the oxygen transport system to ap-
proximately the same extent as front crawl,14 allows

SWIMMING AND ISCHEMIC HEART DISEASE/Magder et al. 981

TAB3LE 2. Oxygen Uptake, Heart Rate, Ventilation and Oxygen Poise During JIJaximal Cycling. Swimming and comfortable Speed

Swimming

Cycle ergometer Maximal swimming Comfortable speed swimming

Sitting Supine 25.50C 18.0°C 25.5`C 18.0°C
(n = 7) (n 8) (n=8) (n=8)
(n 8 (n

V02 1.49 - 0.23 1.42 - 0.24 1.60 - 0.17 1.52 - 0.20 1.28 - 0.12 1.30 - 0.20
(1/min STPD) (108- 12) (101 - 7)
(100 = 11) (87 - 19) (89 = 18)
Heart rate
100 - 13 113 - 16 112 - 15 106 - 13 102 - 10 99 - 14
(beats/min) (101 - 12) (102 - 8) (97 - 11)
(92 - 9) (91 12)
Ventilation
49 11 50 11 53 17 52 14 39 9 37 8
(l/min BTPS)
(105 - 22) (107 - 26) (110 - 23) (81 - 16) (79 19)
Oxygen pulse
(ml/beat) 14.5 - 2.6f 13.6 1.3 14.6 1.4 14.7 2.5 13.3 - 1.8 12.5 1.5
(95 i10) (100 13) (102 16) (93 15) (88 13)

Values are meani sD. MIean percentages of the peak sitting values are giveii in parenthesis.
Abbreviation: V02 = oxygen uptake.

Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016 was no statistically significant difference between the pulse did not differ between cycling and swimming
(fig. 1, table 2).
mean peak V02 of 1.60 ± 0.17 1/min during swim-
ming at 25.5°C and 1.52 ± 0.19 !/min at 18°C com- There was no significant difference in the peak and

pared with the sitting cycling peak VO2 of 1.49 + 0.23 comfortable speed V02 between the swims in warm
1/min (table 2). The mean peak V02 of 1.60 ± 0.17 and cold water (fig. 3). Most subjects swam for a

1/min during warm water swimming was significantly shorter time and reached a lower speed in cold water,
greater than the mean peak of 1.42 + 0.24 1/min
but this difference was not significant. The mean blood
(p < 0.01) during supine cycling (table 2, fig. 1). pressure while resting in the water was 1 13 ± 12 mm
The peak V02 during swimming increased linearly Hg before the cold water swim and 101 ± 6 mm Hg
before the warm water swim (p < 0.05). Blood
with the maximal water speed in the subjects who
pressures during swimming were not obtained. The
swam against progressively increasing water speeds maximal heart rate did not differ between warm and
(peak V02 = 0.79 ± 1.02 X water speed at 25.5°C, cold water swims (fig. 3).
r = 0.89, p < 0.01 ). There was no clear relationship of
the level of V02 to water speed during the comfortable The commonest reason subjects stopped swimming
speed swimming. The V02 could remain the same dur- was shortness of breath. When asked how they
ing the initial increase in water speed (fig. 2). This sub- perceived the work of swimming in cold water as com-
pared with warm water, five reported no difference,
ject increased his speed from 0.35 to 0.75 m/sec before three felt the cold water swim was easier, and none felt
he had any increase in VO2. This cannot be accounted the warm water swim was easier. Four subjects had
for by a lag in the measuring system, for even with the angina and ST depression during sitting cycling,
large increase in demand at the onset of exercise, he whereas five had angina and six had ST depression
reached his steady-state value in less than 2 minutes. during supine exercise (table 3). The same six subjects

The maximal heart rate, ventilation and oxygen

TAIILE 3. Heart Rate at Onset of ST Depression

Cycle ergometer Swimming
Cold
Sitting Supine Warm
Angina* STj Heart
Heart Heart rate Heart

rate rate (beats/ rate
min) Angina* STI
(beats/ (beats/ (beats/
min)
Pt Angina* STI min) Angina* STI min)

1 0 1 120 2 1 115 0 1 120 0 1 96
(bigeminy)
2 6 1 100 6 s80 0 1 98
0 1 88
3 6 1 100 6 82 0 1 88
0 1 60
1 75
4 5 1 78 6 1 115 5 J 88 0 J 83

5 0 No 5 No 0 1 100 4 1 105

6 0 No 0 No 0 No 0 No -

7 0 No - 0 (ectopics) 0 No 0 No
1 93 (bigeminy)
(ectopiCs) (bigeminy)
0 1 86
8 5 1 85 5 5 1 79

*Borg scale (1-10).13

Abbreviation: STI = ST-segment depression.

982 CIRCULATION VOL 63, No 5, MAY 1981

OXYGEN UPTAKE HEART RATE
beats. min-1
2.0 liter.min-1 STPD 1 A fn

l5U .

Swimming d Swimming 0

1.8- 130
o 180C
* 25C °0 / 0 18°C
120 * 25C
1.6 - 00 8
0
1.4 0 0
0
110
1.2
8/ 0

100

1.0 90 0

Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016 0.8 Cycle ergometer 80 Csiytctlieng ergometer
sitting
`08 1. ,01 9 aa a a a
. 1.4 8. 2.0 __
80
90 100 110 120 130 140

OXYGEN PULSE OXYGEN UPTAKE
ml -beat-1 2.0 liter. min-1 STPD

Swimming Swimming 0

o 180C 1.8 0
* 250C
16 0 o 180C
14 * 25C

12 0 1.6 0
0
0
Q 1.4 0

0 0

t2

ergometer 1.0
Cycle ergometer

had ST depression during both swims, but only two studied directly, we used V02 as a measure of the
had angina in warm water and one (a different patient) energy cost and related this to the maximal Vo2 on a
in cold water. The heart rate at which ST depression cycle ergometer. This approach has been used to
occurred was similar during swimming and cycling
(sitting 97 ± 16 beats/min, supine 93 ± 17 beats/min; assess the power requirement of other activities.15 By
swim at 25.50C, 96 ± 14 beats/min, swim at 18°C, this method, although comfortable speed swimming
86 ± 15 beats/min, but only five had ST depression only required a VO2 of 0.85-1.47 1/min, it averaged
during sitting cycling, compared with six subjects in 87% in warm water and 89% in cold water of the peak
the other three tests. V02 during sitting cycling. Similarly, in cold water,

Discussion the heart rate was 92% in warm water and 91 % of the
maximal heart rate during sitting cycling. This high
Knowledge of the physiologic power requirement is percentage of the maximum response was partly
essential in analyzing the effects of swimming in because our subjects' maximal exercise capacity was
patients with ischemic heart disease, for this gives reduced by their previous myocardial infarctions, but
some idea of the relative stress of the exercise. Because was also related to the subjects' skill. For example,
the power requirement of swimming cannot easily be
the best swimmer, patient 1 (fig. 2), used only 68%

(V02 = 1.07 1 /min) the first day and 54% (V02 = 0.85

SWIMMING AND ISCHEMIC HEART DISEASE/Magder et al. 983

VENTILATION OXYGEN UPTAKE

liter min-1 BTPS liter min-1

Swimming

70 0 1.5

o 18°C ( 0
60 * 25C 0
1.0
50 0
0.5
40 0
0 Swimming speed
30
m sec-1 Time, min
20 1 1 1. 1. A 1a .1..La

0 2 4 6 8 10 12 14 16 18 20

Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016 Cycle ergometer FIGURE 2. Original computer-generated recording of oxy-
sitting
gen uptake (kO2) shown together with swimming speed in

one subject. There was no increase in V02 as the water speed

was increased from 0.35 to 0.70 m/sec, and then V02 in-

creased rapidly.

FIGURE 1. Peak values for oxygen uptake, ventilation, these studies concluded that only those subjects who
heart rate and oxygen pulse during the final 30 seconds of were specifically trained in swimming would obtain
maximal effort. Individual values from swimming tests are equal peak VO2 during swimming and cycling. This
shown as a function ofsitting or supine ergometric exercise. cannot explain our results, for our subjects only per-
The line of identity is also shown.
formed light arm exercise during their rehabilitation
1/min) the second day of his sitting cycling maximum program and had not swum for at least 1 year. Our
VO2, while the worst swimmer, patient 3, used 108%
(V02 = 1.47 1/min) the first day and 103% (V02 = subjects probably achieved equal mean peak V02 dur-
1.40 1 /min) the second day of his sitting cycling ing swimming and cycling because they were less
maximum. This is a reflection of the wide range of limited by symptoms during swimming. This is sup-
mechanical efficiency during swimming which, in con- ported by the fact that only those with evidence of
trast to cycling, is highly dependent on the subject's ischemia had their highest peak VO2 during swim-
skill."4' 16 The mechanical efficiency also varied con-
siderably in individual subjects. An example is given in ming.
figure 3, where the subject doubled his swimming
speed before he had any increase in his Vo2 and then Except in the one subject who achieved a higher
his V02 increased rapidly with the increasing speed. peak heart rate during swimming, the increased peak
V02 was probably due to peripheral rather than cen-
A major factor in swimming is the body build for tral factors, because the peak stroke volume has been
this determines the sinking force" 16 and the drag shown to be similar during swimming and other forms
force," 16 both of which must be overcome during swim- of exercise.4' 16 This increased peripheral use of oxygen
ming. We could not reliably weigh our subjects in the is probably not due to an increased intrinsic capacity
water because they could not perform a full expiration for 02 extraction, because the arms perform a major
while completely submerged, but they all floated easily portion of the work of swimming'" and there is no
and so the sinking force was probably not a major fac- need for postural muscles,' making the total muscle
tor. Drag forces become increasingly important at involvement less than in cycling.
higher speeds than those our subjects attained.'
Our results differ from those of Heigenhauser et
An unexpected finding was that the maximum Vo2 al.,4 who found, in a group of patients who had had a
obtained during swimming was at least as great as that
during sitting cycling in seven of eight subjects. In myocardial infarction, that the peak VO2 during
studies in younger, untrained (for swimming) normal swimming was 21% lower than during cycling. This
subjects,' the maximum VO2 during swimming was was due to a lower stroke volume during swimming
about 10% less than cycling. Only elite swimmers at- than cycling. A possible explanation for the
tained their cycling maximum. In a study of healthy differences between their study and ours is that their
old men,3 a higher maximum Vo2 was obtained during subjects may have failed to reach a maximal effort
swimming than during stationary cycling, but the sub- during swimming. The ramp format protocol we used
jects all participated in a swimming program. Both of during swimming brings out a maximal effort in short
time. Also, none of their patients had angina or ECG
evidence of ischemia, which were present in six of our
subjects. Fletcher et al.5 also found that swimming
required a high Vo2 compared with other exercises,
but they did not compare peak values.

984 CI RCULATION VOL 63, No 5, MAY 1981

OXYGEN UPTAKE HEART RATE

2.0 liter min-1 STPD 140 beats min-

Swimming U Swimming U
warm
warm

1R.i o comfortable ._-, o comfortable
speed
U speed

1.6 * max. speed U 120 * max.speed ti
1.4
O, U
I 0

0

1.2 00

1O

0.8 Swimming Swimming
cold cold
Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016
90 100 110

VENTILATION OXYGEN PULSE
liter min-1 BTPS
Swimming n zu m,elve. beat1

warm U Swimming
warm
o comfortable U U 18 a
speed U
o comfortable
* max. speed speed

16 U max.speed 0]

0 14 ma
0
12
0

,b

10 Swimming
cold
Swimming _ Ia ...... a
aa
cold

iInU 41 0z 1d A4 4I0D 4100 zU0 a'

FIGURE 3. Individual values for oxygen uptake, ventilation, heart rate and oxygen pulse during thefinal 30 seconds of swim-
ming at comfortable and maximal speed. Determinations at 25.5 °C water temperature (warm) are shown as afunction of those
obtained at 18°C (cold).

An important aspect of swimming in subjects with acted as a counterirritant and blocked the subjects'
ischemic heart disease is the occurrence of signs and
symptoms of exercise-induced ischemia. The same six perception of their angina. It is unlikely that the ex-
subjects who developed ST depression during supine perimental situation itself distracted the subjects, for
cycling had ST depression during both swims; and there was a defibrillator at the poolside and they were
constantly asked how they felt. Although seven of the
even though five had angina during the cycling test, eight subjects expressed a preference for swimming
only two had angina in warm water and one other sub- over cycling, the failure to recognize symptoms must
ject had angina in cold water (table 3). Perhaps the be kept in mind when patients with ischemic heart dis-
sensation of dyspnea, which was the commonest com- ease are being considered for a swimming program.
plaint, overrode any perception of their more typical
angina. Alternatively, either the water passing over To understand the pathophysiology of angina in
the chest or the rhythmic arm movement may have these patients, we must consider the three major fac-
tors that determine myocardial oxygen consumption:

SWIMMING AND ISCHEMIC HEART DISEASE/MWagder et al. 985

Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016 heart rate, contractility and wall tension.19 The heart have shown that subjects with normal ventricles reach

rate accelerated rapidly during swimming and reached a lower peak VQ0 while taking A blockers, while sub-

over 90% of the maximal value even during comfor- jects with reduced left ventricular function from
ischemia can sometimes obtain a higher peak VO2.21
table speed swimming. Thus, the ST depression There is also a decrease in the peak pressure-rate
usually occurred in the first 2 minutes of swimming product.2' Although the limitation to peak V02 and
and although the rapid acceleration made the exact cardiac output apply to any exercise, fi blockers could
determination of the onset of ST depression difficult, reduce some of the greater blood pressure rise per
work load that occurs during swimming and allow a
it appeared to be similar during swimming and cy-
greater peak VO2 than without. The other important
cling (table 3). Because the maximal heart rates were drug was digoxin, which does not affect the peak V02,22
but does make the ECG interpretation of ischemia
the same during swimming and cycling, there was a less dependable. Of the four subjects taking digoxin,
only two had ST depression with exercise, and this was
similar effect on myocardial oxygen demand. Con- associated in both with chest pain during sitting and
supine cycling tests. We therefore considered it valid
tractility could not be assessed in our study, but would to interpret their ST depression as ischemic.
also have been expected to be similar during swim-
In conclusion, we found that in patients with a
ming and cycling. Left ventricular wall tension would reduced exercise capacity due to ischemic heart dis-
ease, swimming, even at a comfortable speed, can re-
be expected to be higher during swimming than cy- quire near-maximal effort. In contrast to normal sub-
cling for at least two reasons. Swimming work results jects, good as well as poor swimmers achieved the
in a higher blood pressure, a major determinant of left
ventricular wall tension,19 at the same V02 as run- same peak V02 during swimming as they did during
cycling. There appeared to be a failure to identify
ning.17 Cycling, in turn, produces changes in blood
pressure similar to those with running.'8 This results ischemic symptoms. Swimming at 18°C and 25°C
from decreased skin blood flow,' 6 the compressive were equally well tolerated. Several recommendations
effect of water," 6 and the smaller muscle mass in- can be made. Because it has a high aerobic cost, swim-
volved in swimming compared with running and ming can be an effective method of training the oxygen
transport system, but because poor swimmers require
cycling.'7 Second, the increase in central blood volume a sudden large output from the oxygen transport
that occurs during swimming" would increase myocar- system, poor swimmers with ischemic heart disease
dial oxygen demand.'9 The increased volume would should avoid swimming as an exercise unless properly
also elevate left ventricular end-diastolic pressure in supervised. Because the heart rate at which ST depres-
hearts that are noncompliant from previous myocar- sion occurred was similar during cycling and swim-
dial infarction.8 This would in turn increase the ming, this heart rate can be obtained from a treadmill
or cycle test and used as a guide when swimming is
pulmonary venous pressure and might explain why used in a cardiac rehabilitation program. With these
our subjects complained more from dyspnea than guidelines, each patient must be treated individually
and the role of swimming in a rehabilitation program
angina. must be judged according to the degree and type of
There was no difference between the warm and cold disease and the patient's skill and interest.

water swims except for a shorter duration of the cold References

water swim and a moderately higher resting blood 1. Holmer I: The physiology of swimming in man. Acta Physiol
Scand [Suppl] 407, 1974
pressure in cold than in warm water. The potential im-
portance of blood pressure is illustrated by one subject 2. McArdle WD, Glaser RM, Magel SR: Metabolic and cardio-
who had a resting blood pressure of 135/70 mm Hg respiratory response during free swimming and treadmill walk-
ing. J Appl Physiol 30: 733, 1971
and maximal heart rate during exercise of 115
3. Kasch FW: Maximal oxygen uptake during free swimming and
beats/min in water at 25°C, but a resting blood stationary cycling. In Swimming Medicine IV, edited by
pressure of 200/100 mm Hg and maximal heart rate Ericksson B, Furberg B. Baltimore, University Park Press,

of 105 beats/min in water at 18°C. He was also the 1978, pp 143-146
only subject who had lower peak VO2 during swim- 4. Heigenhauser GF, Boulet D, Miller B, Faulkner JA: Cardiac
ming than during cycling; his peak VO2 was 1.83 dur-
ing sitting cycling and 1.62 in warm and 1.60 1/min in outputs of post-myocardial infarction patients during swim-
cold water swimming. The elevated resting blood ming and cycling. Med Sci Sports 9: 143, 1977
pressure in cold water was not associated with a lower 5. Fletcher GF, Cantwell JD, Watt EW: Oxygen consumption and
hemodynamic response of exercise used in training of patients
peak VO2, but, as a result, he reached his peak in a with recent myocardial infarction. Circulation 60: 140, 1979
6. Arborelius M Jr, Balldin UI, Lilja B, Lundgren LEG:
shorter time and at a lower heart rate. Hemodynamic changes in man during immersion with head
Although we did not study this, water temperature above water. Aerospace Med 43: 592, 1972
7. Bevegare SA, Holmgren A, Jonsson B: The effect of body posi-
close to body temperature could result in a lower tion on the circulation at rest and during exercise with special
diastolic pressure by causing peripheral vasodilation, reference to the influence on stroke volume. Acta Physiol Scand
and thus lead to decreased diastolic coronary artery 49: 279, 1960
perfusion and earlier ischemia during swimming, as
has been found to occur in sauna baths.20 Thus, water
that is too warm may be as detrimental, if not more
so, than water that is too cold.

All of our patients were taking d blockers and four
were taking digoxin. We continued these medications
for safety's sake and because this is the state under
which such patients will be exercising. Studies in nor-
mal subjects and patients with ischemic heart disease

986 CIRCULATION VOL 63, No 5, MAY 1981

8. Khaja F, Sanghvi V, Mark A, Parker SO: The effect of volume 16. DiPrampero PE, Pendergast DR, Wilson DW, Rennie DW:
expansion on the anginal threshold. Circulation 43: 824, 1971 Energetics of swimming in man. J Appl Physiol 37: 1, 1974

9. Wilken DEL, Charlier AA, Hoffman JIE, Guz A: Effects of 17. Holmer 1, Stein EM, Saltin B, Ekblom B, Astrand PO:
alterations in aortic impedance on the performance of the ven-
tricles. Circ Res 14: 283, 1964 Hemodynamic and respiratory responses compared in swim-
ming and running. J Appl Physiol 37: 49, 1974
10. Stenberg J, Astrand PO, Ekblom B, Royce J, Saltin B: 18. Hermansen L, Ekblom B, Saltin B: Cardiac output during sub-
Hemodynamic response to work with different muscle groups,
sitting and supine. J Appl Physiol 22: 61, 1967 maximal and maximal treadmill and bicycle exercise. J Appl
Physiol 29: 82, 1970
11. Astrand PO, Englesson S: A swimming flume. J Appl Physiol 19. Parmley WW, Tyberg SV: Determinants of myocardial oxygen
33: 514, 1972 demand. In Progress in Cardiology V, edited by Yu P, Goodwin
JF. Philadelphia, Lea and Febiger, 1976, pp 19-36
12. Davies NJH, Denison DM: The measurement of metabolic gas
exchange and minute volume by mass spectrometry alone. 20. Taggart P, Parkinson P, Carruthers M: Cardiac response to
Respir Physiol 36: 261, 1979 thermal, physical and emotional stress. Br Med J 3: 71, 1972

13. Borg G: The determination of subjective intensities in verbal 21. Bruce RA, Hossack MF, Kusumi F, Clark LS: Acute effects of
description of symptoms. Report from the Institute of Applied
Physiology, University of Stockholm 75: 22, 1976 oral propranolol on hemodynamic responses to upright exer-

14. Holmer I: Oxygen uptake during swimming in man. J Appl cise. Am J Cardiol 44: 132, 1979
Physiol 33: 502, 1972
22. Niederberger M, Bruce RA, Frederick R, Kusumi F, Marriott
15. Astrand PO, Rodahl K: Textbook of Work Physiology, 2nd ed. A: Reproduction of maximal exercise performance in patients
New York, McGraw-Hill, 1977, pp 449-480 with angina pectoris despite oubain treatment. Circulation 49:

309, 1974

Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016

Downloaded from http://circ.ahajournals.org/ by guest on August 28, 2016 The effect of swimming on patients with ischemic heart disease.
S Magder, D Linnarsson and L Gullstrand

Circulation. 1981;63:979-986
doi: 10.1161/01.CIR.63.5.979

Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1981 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539

The online version of this article, along with updated information and services, is located on
the World Wide Web at:

http://circ.ahajournals.org/content/63/5/979

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation 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 Circulation is online at:
http://circ.ahajournals.org//subscriptions/