Content uploaded by David John Bishop
Author content
All content in this area was uploaded by David John Bishop on Aug 01, 2014
Content may be subject to copyright.
ORIGINAL ARTICLE
Effect of wearing an ice cooling jacket on repeat sprint
performance in warm/humid conditions
R Duffield, B Dawson, D Bishop, M Fitzsimons, S Lawrence
.............................................................................................................................
Br J Sports Med
2003;37:164–169
Objective: To examine the effect of cooling the skin with an ice jacket before and between exercise
bouts (to simulate quarter and half time breaks) on prolonged repeat sprint exercise performance in
warm/humid conditions.
Methods: After an initial familiarisation session, seven trained male hockey players performed two
testing sessions (seven days apart), comprising an 80 minute intermittent, repeat sprint cycling exercise
protocol inside a climate chamber set at 30°C and 60% relative humidity. On one occasion a skin
cooling procedure was implemented (in random counterbalanced order), with subjects wearing an ice
cooling jacket both before (for five minutes) and in the recovery periods (2 × 5 min and 1 × 10 min)
during the test. Measures of performance (work done and power output on each sprint), heart rates,
blood lactate concentrations, core (rectal) and skin temperatures, sweat loss, perceived exertion, and
ratings of thirst, thermal discomfort, and fatigue were obtained in both trials.
Results: In the cooling condition, chest (torso) skin temperature, thermal discomfort, and rating of thirst
were all significantly lower (p<0.05), but no significant difference (p>0.05) was observed between condi-
tions for measures of work done, power output, heart rate, blood lactate concentration, core or mean skin
temperature, perceived exertion, sweat loss, or ratings of fatigue. However, high effect sizes indicated
trends to lowered lactate concentrations, sweat loss, and mean skin temperatures in the cooling condition.
Conclusions: The intermittent use of an ice cooling jacket, both before and during a repeat sprint
cycling protocol in warm/humid conditions, did not improve physical performance, although the per-
ception of thermal load was reduced. Longer periods of cooling both before and during exercise (to
lower mean skin temperature by a greater degree than observed here) may be necessary to produce
such a change.
I
t is well documented that heat strain, as a result of an
elevated core body temperature, is a major cause of reduced
exercise performance and heat related illnesses.
1–4
Under-
lying all heat related illnesses is the inability of the body to rid
itself of the excess heat produced.
5
When team sports, which
require repeated high intensity short sprints, are played in hot/
humid conditions for 70–120 min (plus warm up and extra
time), the possibility of pla yers succumbing to some form of
heat illness increases.
6
However, by precooling the body, it may
be possible to lessen the impact of hot and/or humid conditions
on both exercise performance and athlete wellbeing.
7
In
particular, cooling between exercise bouts at quarter/half time
breaks in games has not been evaluated, and it is at these times
that cooling procedures may be of most practical use.
Precooling should lower skin temperature, but it may also
result in small decreases in core temperature. Precooling can
therefore delay the rise in both core and skin temperatures
during exercise, increasing the time to the so called critical
temperatures where performance may be affected.
89
Along
with a widening of this “margin of safety” there may also be a
delay in the redirection of blood flow to the skin, which may
theoretically allow a higher muscle blood flow,
10 11
which,
although not experimentally documented, may potentially
provide a greater aerobic energy supply and/or a reduction in
muscle and blood lactate concentrations.
To gain these potential physiological improvements, differ-
ent methods of precooling have been attempted: immersion in
cold water,
12
exposure to cold ambient temperatures in a
climate chamber,
8
and direct skin cooling with the use of water
perfused suits and wet towels.
10 13
Relatively recently, cooling
jackets have come to prominence as a simple and efficient field
method by which athletes can reduce skin temperature and
attempt to counter the negative effects of heat strain.
Precooling methods have been shown to limit the effects of
hyperthermia and produce improvements in exercise perform-
ance in all out, six minute rowing ergometer tests,
10
in 2000 m
rowing ergometer time trials,
14
and in cycle and treadmill exer-
cise of longer durations (30 minutes plus).
15 16
However,
precooling failed to elicit any exercise improvements in a step-
wise cycle ergometer exercise test to exhaustion.
8
Although
research supports the ergogenic qualities of cooling body tem-
peratures before continuous aerobic exercise in hot conditions,
a recent report
17
failed to find any significant improvement in
exercise performance for an intermittent soccer specific
exercise protocol performed in cool conditions, after a 60
minute precooling intervention. Although both core and skin
temperature were significantly lowered before exercise, this did
not translate into performance improvement, nor any differ-
ences in oxygen consumption, perceived exertion, heart rate, or
substrate metabolism at the finish of the 90 minute intermit-
tent running protocol when compared with a non-cooling con-
dition. However, other authors
10
have found improvements in
mean power output for a maximal 70 second cycle ergometer
test after whole body precooling. This result shows the poten-
tial of precooling for improving exercise performance in short
duration, high intensity exercise, where the effects of
thermoregulation on performance are less influential.
Thus far, no research has investigated the influence of cool-
ing both before and between exercise bouts on intermittent
repeat sprint exercise (as occurs in team sport games where
there are established quarter/half time breaks). As team sports
have only a small time frame between the cessation of the
warm up and the start of game play, the opportunity for an
initial precooling exposure before the start of play may only be
of a short duration. Although some athletes have used the
cooling jacket during warm up activities (possibly attenuating
See end of article for
authors’ affiliations
.......................
Correspondence to:
Dr Duffield, Exercise
Physiology Laboratory,
Department of Human
Movement and Exercise
Science, University of
Western Australia,
Crawley, WA 6009,
Australia;
robd@cyllene.uwa.edu.au
Accepted 16 May 2002
.......................
164
www.bjsportmed.com
the rise in skin temperature), in order to isolate the effects of
a precooling procedure, a short non-exercise period was used
here as a typical representation of the time between warm up
cessation and the start of the game. Anecdotal reports have
indicated that the ice cooling jackets have also been used in
team sports—for example, field hockey, Australian football—
during the designated game breaks. Therefore, whether such a
short precooling application and/or additional cooling during
game breaks is effective in improving exercise performance is
yet to be determined. Accordingly, the objective of this study
was to investigate whether cooling before and during high
intensity exercise of an intermittent nature (consisting of
multiple repeated short sprints with designated quarter and
half time breaks) could improve exercise performance.
METHODS AND PROCEDURES
Sample
The seven male subjects used in this study were all first grade
field hockey players (and therefore familiar with intermittent
exercise activity) and were tested near the end of their
competitive season. All gave their informed consent to partici-
pate in the study after approval for the procedures was granted
by the University of Western Australia human ethics
committee. Mean (SD) values for age, mass, height, sum of six
skinfolds,
18
and body surf ace area (Dubois nomogram) were as
follows: 20.2 (2.2) years, 69.3 (6.7) kg, 179.1 (3.7) cm, 47.4
(15.7) mm, and 1.88 (0.09) m
2
.
Procedure
Research design
Subjects were required to attend three testing sessions, each of
which was separated by seven days. Initially, subjects engaged
in a familiarisation session, followed by either the control
condition (no ice cooling jacket) or the experimental
condition (with the ice cooling jacket). The order of these last
two sessions was randomised so as to avoid any order effect. In
each session subjects performed an 80 minute intermittent
sprint cycling exercise protocol wearing running shoes, t shirt,
and shorts (standard hockey clothing). They were asked to
refrain from any vigorous exercise for 24 hours before under-
taking any of the testing sessions, and to avoid any food, drink,
cigarettes, or caffeinated products during the two hours before
the start of a testing session.
Exercise protocol
The exercise pattern was an intermittent repeated sprint
cycling protocol for 80 minutes, consisting of four identical 15
minute quarters (mimicking the duration of a team sport
game). Five minute rest periods were allowed between the
first and second quarters and between the third and fourth
quarters, with a 10 minute half time break. The protocol con-
sisted of a five second sprint every minute, followed by 55 sec-
onds of recovery at varying intensities (thus, each sprint
occur red on the minute). The rationale behind the exercise
protocol was to attempt to simulate the type of intermittent,
repeat sprint activity that occurs in many team sports. The
continually changing nature of the recovery intensity level
used in this protocol was an attempt to simulate the continual
change in exercise intensity observed during intermittent
activity—that is, walking to jogging to sprinting , back to jog-
ging again, and so on. However, in an attempt to induce a
greater degree of metabolic strain, three extra sprints at 2.5,
7.5, and 12.5 minutes were added. The work done and power
produced on these extra sprints was not recorded. Figure 1
outlines the exercise and measurement protocol.
Experimental procedures
The control testing session began with weighing the subject to
obtain their nude mass. Standardised prehydration then
occur red with the ingestion of 500 ml water, during which the
subject was made ready with all measuring devices. Subject
preparation took place in an air conditioned environment of
20–25°C dry bulb temperature. A rectal probe was positioned
10 cm beyond the anal sphincter, and skin thermistors were
placed on the sternum, chest (3 cm to the right of the left nip-
ple), right forearm, right anterior quadriceps, and right poste-
rior calf. A Polar heart rate transmitter and accompanying LSD
receiver were attached. The subject then entered the climate
chamber (already stable at the preset 30°C ambient tempera-
ture and 60% relative humidity, chosen to represent typical
outdoor environmental conditions encountered in the Atlanta
summer Olympics) and sat on the cycle ergometer. Seat height
was measured and kept constant for each testing session. The
subject was strapped to the bike with a waist strap to prevent
standing up, and toe clips with heel straps were used to stabi-
lise the feet in the pedals. Once prepared, the subject remained
seated in the warm/humid climate for 15 minutes while
instructions were given about the exercise protocol, measures
to be taken, and the need for maximal effort on each sprint.
After this period, resting measures of heart rate, core and skin
temperatures, lactate concentration, and rating scales for per-
ceived exertion, thermal comfort, thirst, and fatigue were
taken. The subject then performed a five minute warm up,
consisting of three minutes of cycling at 75 W, before increas-
ing to 100 W for the final two minutes. Within the last two
minutes there were two practice sprint starts of 3–4 seconds
duration (at 3.5 and 4.5 minutes). At the end of the warm up,
core and skin temperatures, heart rate, and rating scale meas-
ures were again obtained. After the warm up, a five minute
recovery period, in which the subject stretched (remaining on
the bike, but with feet freed) occurred. In the final minute of the
stretching period, the subject was strapped back into the pedals,
and body temperatures, heart rate, and ra ting scale measures
were again taken. The subject then began the exercise protocol,
as shown in fig 1, with all measures recorded at the appropriate
Figure 1 Intermittent exercise and testing protocol. Six second sprints are indicated by thin black lines, and recovery intensities by black
shading. T, Temperature; H, heart rates; L, blood sample for lactate measurement; W, water; J, ice cooling jacket; rating scales.
Effect of precooling on performance 165
www.bjsportmed.com
times, including work done (kJ) and power output (W)
recorded on each sprint performed (except those occurring on
the half minute—that is, 2.5, 7.5, and 12.5 minutes). Strong
verbal encouragement was given to elicit maximal effort on
each sprint. En vironmental temperatures (dry, wet, and black
bulb) were also measured and recorded every 20 minutes.
At the end of each quarter and just before the start of the
next, skin and core temperatures, heart rate (before and after
the sprint—that is, both the last and first sprints of the respec-
tive quarters), rating scales, and a blood sample from a finger
tip for lactate measurements were taken. The subject was given
a five minute break at the end of the first and third quarter
during which they remained seated on the bike and were
allowed to ingest 250 ml water. During a 10 minute break after
the second quarter of the test, the subject was again allowed to
ingest 250 ml water and to dismount the bike in order to move
around (but remained within the climate chamber). After the
fourth quarter the exercise test concluded, and, once final core
and skin temperatures, heart ra te, rating scales, and a blood
sample had been obtained, the subject left the climate chamber
to be reweighed. All leads were removed and all unevaporated
sweat was removed with a towel before weighing. The subjects
was then free to leave the laboratory.
The experimental and control sessions were identical, apart
from the application of the ice cooling jacket during the rest
periods. The jacket was used during the break for stretching
after the warm up (five minutes) and during each break in the
exercise protocol (2 × 5 and 1 × 10 minute breaks). The jacket
was donned once temperatures, heart rates, and rating scale
measures had been obtained—that is, 30–40 seconds into the
rest period. It was then removed about 15–20 seconds before
the exercise was started again. Apart from the use of the
jacket, all measurements, procedures, and items of equipment
used were kept standard across all testing sessions. The pock-
ets on the ice cooling jacket were filled with cubed ice before
the start of the test, and then the jackets were placed in an
insulated cooler in the climate chamber. Additional ice was
stored in a different cooler within the chamber so that extra
ice could be added to the jacket at the half time break.
Measures and apparatus
Cooling was achieved with an ice cooling jacket (AIS,
Canberra, Australia), constructed from material similar to that
of a wetsuit and shaped in the form of a vest, with a zipper on
the front, allowing easy application and removal. Multiple
pockets allowed the placement and removal of cold substances
(ice/ice packs) to cool the torso region. The cooling jacket was
worn over the exercise attire. Exercise was performed on an
Evolution Track Cycle (Adelaide, South Australia, Australia),
with performance (power output and work done on each
sprint) measured by CEDAA computer software developed at
the Western Australian Institute of Sport. Heart rates were
measured with a Polar Sports Tester (PE4000) (Kempele,
Finland). Blood lactate concentrations were measured in mic-
rosamples of blood taken from a finger. They were determined
with an Analox LM3 Multi Stat Analyser (Sheffield, Yorkshire,
UK). Technical error of measurements (TEMs) ranged from
0.12 to 0.83 mmol/l for blood lactate concentrations in the
range 1–10 mmol/l. Skin and rectal temperatures were
recorded by either a Grant 1200 Series Squirrel Meter/Logger
(Cambridge, UK) or a Yellow Springs Instruments model 46
TUC Telethermometer (Yellow Springs, Ohio, USA). Thermistor
leads used were Grant EU-U2 Surface Probes (skin thermis-
tors) and a Grant Rectal probe for the Squirrel (Cambridge,
UK), with YSI model 408 skin thermistors and a YSI model 401
rectal probe for the Telethermometer. Mean skin temperatures
were calculated using the method of Ramanathan
19
:
Mean skin temperature = (0.3 × sternum temperature) +
0.3 × forearm temperature) + (0.2 × quadriceps
temperature) + (0.2 × calf temperature).
Sweat loss was estimated from the change in nude body mass
during exercise accounting for any fluid ingested and urine
voided. Relevant psychological states were assessed by rating
scales consisting of a rating of perceived exertion (RPE)
scale,
20
a 16 point bipolar scale for rating of thermal comfort
(with anchors of 0.0 (unbearably cold) through to 8.0
(unbearably hot)), a 10 point bipolar scale for rating of thirst
(with anchors of 0 (not thirsty) through to 10 (extremely
thirsty)) (both of which were constructed similarly to the Borg
scale), and a fatigue/vigour scale, which were subcomponents
of the profile of mood states (POMS) checklist.
21
All
psychological variables were recorded in the two minutes
before the start of exercise and again immediately after the
completion of each quarter. A climate chamber located within
the Western Australian Institute of Sport laboratory was used
for the testing, with a radiant heat source (3000 W) being
simulated by a heat lamp located 2.5 m to one side of the sub-
ject. Measures of the environmental conditions were regularly
made by recording wet, dry, and black bulb temperatures
using a Reuter Stokes RSS-212 Wibget heat stress monitor
(Cambridge, Ontario, Canada).
Statistical analysis
Analysis of the means of the data over the four quarters for per-
formance, core and skin temperatures, heart rates, and lactate
concentrations were conducted using a two way (condition v
time) analysis of variance with repeated measures. Where
appropriate, Schef fe post hoc comparisons were used. Overall
performance measures and swea t losses were analysed using a
dependent samples t test. Analysis of the rating scale measures
(RPE, thermal comfort, thirst, and fatigue/vigour) was also car-
ried out using a two way repeated measures analysis of
variance. Statistical significance was set at p<0.05. Effect sizes
(Cohen’s d = X
Exp
−(X
Con
/SD
Con
))
22
were also calculated for all
measured variables. Effect size results were interpreted as
described by Christensen and Christensen,
23
with effect sizes of
<0.2 classified as small, 0.4–0.6 as medium, and >0.8 as large.
RESULTS
Exercise performance
Table 1 shows the mean total work done and power produced
in each quarter and across the whole test. In addition, the
mean work done per sprint in each quarter is also presented.
Although five of the seven subjects performed more work and
produced higher power levels in the cooling condition, there
were no significant differences between the two conditions for
work done or power output, either quarter by quarter or over
the whole test (p>0.05). In addition, there was no significant
decline in work completed over time (no significant difference
recorded between the first and last quarter work scores)
(p>0.05), nor for the mean work done per sprint in each
quarter (p>0.05). Overall, the effect sizes were small for all
work and power variables (d<0.3).
Temperatures
Table 2 gives the mean core, skin, and chest temperatures
recorded throughout the test. No significant differences, as well
as small effect sizes, w ere found between the mean core
temperatures recorded throughout the exercise test in the two
conditions (p>0.05 and d<0.3). The change in core tempera-
ture across the test (from the start of the test, including the
warm up) was not significant between conditions (1.4°C no
cooling v 1.2°C cooling; p>0.05), as was the change in core
temperature between the end of the warm up and start of the
test (−0.06°C no cooling v −0.08°C cooling). However, there was
a tendency for a slight reduction in the rise in core temperature
in the cooling condition, with a lowered mean change in core
temperature as the experiment continued (p = 0.072, d = 0.7).
Although a trend for lower mean skin temperatures in the
cooling condition was noticeable (at the start of exercise and
each quarter), no significant differences were found between
166 Duffield, Dawson, Bishop, et al
www.bjsportmed.com
the two conditions for either mean skin temperatures or the
change in mean skin temperatures across time, including
mean skin temperature at the start of the test—that is, after
the warm up (34.2 (0.7) cooling, 34.5 (0.6) no cooling)
(p>0.05). However, very large effect sizes were found for mean
skin temperature after half time (d = 2.3) and for the change
in each period after quarter time (0.7<d<1.4), with reduced
temperatures being noted in the cooling condition (table 2). In
addition, a significant difference (p<0.02) was found for the
mean chest temperature—that is, under the ice cooling jacket.
Post hoc comparisons showed that at the start of each quarter
there was a significant difference (p<0.05) between the con-
ditions for mean chest temperature, with temperatures in the
cooling condition not reaching control levels until the very end
of the quarter (d>4.0 at the start of each quarter).
Heart rate
There were no differences observed between the cooling and
non-cooling conditions for heart rates before or after the
sprint (p>0.05), which ranged from an average of 145 to 157
beats/min for each of the quarters.
Lactate concentration
Although no significant difference in the means of the two
conditions was observed (p>0.05), slightly lower blood lactate
concentrations were recorded in the cooling condition at the
end of each quarter and at the start of the next quarter (table
3). Starting concentrations were similar (1.8 mmol/l for the
non-cooling condition compared with 1.6 mmol/l for the cool-
ing condition), but at the end of the exercise protocol lactate
concentrations were 5.1 mmol/l (non-cooling) and 4.3 mmol/l
(cooling) respectively. Large effect sizes at the start and end of
each quarter were found, with lactate concentrations being
lower in the cooling condition (0.7<d<2.6) (apart from after
the 10 minute half time break).
Change in body mass
Mean decrease in body mass across the test for the
non-cooling condition was 2.10 (0.51) kg whereas in the cool-
ing condition it was 1.70 (0.30) kg. Although not significant
(p>0.05), a large effect size (d = 0.8) was noted.
RPE
Mean (SD) RPE in the non-cooling condition was 11.9 (3.6)
and 11.0 (3.2) in the cooling condition (p>0.05). Likewise, no
Table 1 Work done and power produced in each quarter and over the whole test for the cooling and non-cooling
conditions, also including mean work done per sprint in each quarter
Total mean work (kJ) Mean work per sprint (kJ) Mean power (W)
Cooling Non-cooling Cooling Non-cooling Cooling Non-cooling
Quarter 1 59.4 (8.9) 59.5 (10.2) 3.7 (0.5) 3.7 (0.6) 937 (135) 933 (140)
Quarter 2 60.0 (8.1) 57.5 (11.3) 3.7 (0.5) 3.6 (0.7) 939 (126) 903 (173)
Quarter 3 58.4 (7.7) 56.8 (10.5) 3.6 (0.5) 3.6 (0.7) 912 (118) 888 (152)
Quarter 4 58.7 (7.7) 56.3 (10.6) 3.6 (0.5) 3.5 (0.6) 899 (164) 899 (164)
Test total 239.8 (33.9) 230.2 (41.9) – – 928 (123) 906 (155)
Values are mean (SD).
Note: no significant differences were recorded between conditions.
Table 2 Core, skin, and chest temperature at the start and finish of each quarter (Q) (°C) for the cooling (Cool) and
non-cooling (Non cool) conditions
Q1 Q2 Q3 Q4
Start Finish Start Finish Start Finish Start Finish
Core
Cool 37.4 (0.3) 38.0 (0.3) 38.2 (0.3) 38.5 (0.3) 38.4 (0.3) 38.5 (0.3) 38.6 (0.3) 38.7 (0.3)
Non cool 37.5 (0.3) 38.1 (0.3) 38.2 (0.3) 38.6 (0.3) 38.5 (0.3) 38.6 (0.4) 38.6 (0.3) 38.8 (0.3)
Mean skin
Cool 34.2 (0.7) 34.3 (0.4) 34.9 (0.6) 34.1 (0.6) 34.2 (0.7) 33.6 (0.8) 34.4 (0.5) 33.6 (0.6)
Non cool 34.5 (0.6) 34.5 (0.7) 35.2 (0.8) 34.2 (0.6) 34.9 (0.3)† 33.9 (0.7) 35.0 (0.9)† 34.0 (0.7)
Chest
Cool 30.1 (1.2) 33.2 (1.5) 28.9 (3.3) 33.5 (1.4) 26.0 (4.3) 32.6 (1.9) 25.5 (4.6) 32.8 (1.8)
Non cool 34.5 (0.6)* 34.2 (1.2) 35.3 (1.2)* 34.1 (1.0) 35.4 (0.7)* 33.9 (1.1) 35.4 (1.1)* 33.7 (1.2)
Values are mean (SD).
*Significant difference between conditions (p<0.05).
†Large effect size between conditions (d>0.8).
Table 3 Blood lactate concentrations recorded at the start and end of each quarter for cooling (Cool) and non-cooling
(Non-cool) conditions
Q1 Q2 Q3 Q4
Before After Before After Before After Before After
Cool 1.6 (0.5) 5.3 (1.6)* 4.3 (1.0) 5.2 (1.5)* 3.6 (0.9) 4.2 (0.8)* 3.5 (0.4)* 4.4 (0.7)*
Non-cool 1.8 (0.4) 6.2 (0.7) 4.9 (1.0) 5.8 (0.8) 3.7 (0.7) 5.0 (0.5) 4.3 (0.3) 5.1 (0.9)
Values are mean (SD).
Note: There were no significant differences between conditions (p>0.05).
*Large effect size between conditions (d>0.8).
Effect of precooling on performance 167
www.bjsportmed.com
differences for the final rating (17.1 (2.0) for non-cooling and
15.0 (3.0) for cooling) at the end of exercise was noted
(p>0.05; d = 1.05). However, after the half time break, mod-
erate to strong effect sizes (d = 0.7–1.2) were found for low-
ered RPEs in the cooling condition.
Thermal comfort
Mean (SD) ther mal comfort over the whole test for the non-
cooling condition was 5.9 (0.7) compared with 5.0 (1.3) for
the cooling condition (p>0.05). However, significant differ-
ences (p<0.05) between the cooling condition and the
non-cooling condition were found at 0, 20, 45, and 60 minutes:
2.9 (1.2) v 5.2 (0.6), 3.8 (0.1) v 5.4 (0.7), 3.8 (1.1) v 5.6 (0.5)
and 4.0 (1.2) v 5.9 (0.6).
Ratings of thirst
Significant differences (p<0.05) were found between the con-
ditions over time, with the mean (SD) rating of thirst over the
whole test in the non-cooling condition being 3.1 (2.1) and in
the cooling condition 2.9 (1.3).
Fatigue and vigour scale
No significant differences (p>0.05) were found between the
non-cooling and cooling conditions for either the fatigue scale
(fatigued, exhausted, weary, worn out, bushed) or the vigour
scale (vigorous, energetic, lively, active, full of pep), with no
trends evident. However, in both conditions subjects reported
increased feelings of fatigue (on all fatigue scales) over time
(p<0.05), and lowered ratings on the energetic, active, and full
of pep scales over the duration of the exercise protocol
(p<0.05).
DISCUSSION
The aim of this study was to investigate whether the use of an
ice cooling jacket for brief periods before and during intermit-
tent, repeated sprint exercise of a team sport type duration
could improve exercise performance. The use of cooling proce-
dures in designated “breaks” during prolonged exercise tests
has not been previously examined. The results indicated that
no significant differences existed between the cooling and
non-cooling conditions in measures of cycling sprint perform-
ance, core and mean skin temperature, blood lactate
concentration, heart rate, change in body mass, RPE, or
fatigue/vigour scale ratings. Because of small subject numbers,
effect size analysis was used to indicate whether trends noted
in the data were small, moderate, or large. Effect sizes were
moderate to large for lower lactate concentrations, sweat loss,
and mean skin temperatures in the cooling trial. Significantly
lower values (p<0.05) were found for the cooling condition
for measures of chest (torso) temperature, rating of thermal
comfort, and rating of thirst.
Significant improvements in continuous, submaximal, and
maximal aerobic exercise performance (6–35 minutes)
14 15
after 10–15 minutes of precooling with ice jackets have been
reported. In contrast, Drust et al
17
did not find any
improvement in distance covered over a 90 minute intermit-
tent non-motorised treadmill running protocol (after a 60
minute cold shower, with a 15 minute break before the start).
However, in their study, subjects ran at set speeds (in normo-
ther mic conditions) for most of the test protocol, thus not
per mitting any large degree of differentiation in distance cov-
ered (as a performance measure) between cooled and
non-cooled conditions. Although our study did provide for a
greater amount of maximal sprinting throughout the exercise
protocol, the results again did not show any significant differ-
ences between the conditions for work done or power
produced either after the initial cooling or with the additional
cooling at the designated breaks (although five of the seven
subjects did improve their exercise performance).
The absence of a notable reduction in core temperature after
warm up may explain why the use of ice cooling jackets did not
alter performance in this study. The three following studies all
reported improved performance after precooling and all showed
a significant reduction in core temperature. Yates et al
14
used an
ice cooling jacket for 10 minutes, w hich produced a small
reduction (−0.5°C) in rectal temperature after the warm up,
before a 2000 m rowing ergometer test. Similarly, Myler et al
11
used an eight minute precooling period (ice wa ter towels)
before the start of the same test, which was sufficient to produce
a significant lowering of core (tympanic) temperature by 0.9°C.
More recently, Marsh and Sleivert
10
reported that rectal
temperature was lowered by 0.3°C by cold water immersion (for
30 minutes) before a 70 second maximal cycle ergometer sprint.
Therefore, a significant decrease in core temperature may be
necessary for precooling to improve exercise performance.
As core temperature is strictly regulated,
24
it is likely that the
cooling intervals used here of five minutes (and to a lesser
extent, the 10 minute half time break) were not of sufficient
duration to elicit any significant change in this variable. How-
ever, a short cooling duration before exercise (five minutes)
was chosen for use here to simulate a more realistic game spe-
cific situation, where little time exists between the end of the
warm up and the start of the game. Laboratory precooling
methods of the sort and duration used by Schmidt and
Bruck
8
or Marsh and Sleivert,
10
although successful at
reducing core temperature, are not practical for a field compe-
tition setting. Moderate to large effect sizes (d = 0.5–1.0)
recorded here did indicate a trend for a reduction in the rise in
core temperature when the ice jacket was worn before exercise
(and also during the half time break). Therefore, it is likely
that a greater length of initial cooling may have allowed a sig-
nificant reduction in core temperature before starting the test,
and therefore a greater likelihood of performance enhance-
ment. Although it is noted that the environmental and
exercise load used here did not increase mean core tempera-
ture to greater than 39°C, the impact of precooling may have
been increased with a greater hyperthermic load. However, an
exercise protocol involving alternating intermittent work and
recovery periods (as used here to simulate team sport activity)
will limit the rise in core temperature, which might otherwise
be achieved with a continuous exercise protocol. Also, cooling
of the head and neck regions may assist in improving the
impact of precooling, as the head is an area of great heat loss
because of a lack of vasoconstriction in the cranial area.
25
Although mean skin temperature also showed no signifi-
cant difference between conditions, examination of the effect
size data again showed moderate effects of cooling after the
pretest and third quarter break and large effects after the half
time application of the cooling jacket. Large effect sizes
(d>0.7) were found for both a smaller change in mean skin
temperature across quarter s after the half time break and for
a lowered mean skin temperature at the start of each quarter
in the cooling condition. However, three of the four sites used
in the calculation of mean skin temperature were not in the
vicinity of the ice cooling jacket (which only covers the torso
region). The use of the ice cooling jacket significantly reduced
chest skin temperature (under the jacket while worn), with
temperatures not reaching those displayed in the non-cooling
condition until the end of the quarter (by which time the
jacket was reapplied). Smith et al
15
found skin temperatures
under the ice cooling jacket during the warm up to be 10°C,
compared with 33°C without the jacket. They hypothesised
that the increased time measured until the onset of fatigue
may have been a result of a decreased skin blood flow (thereby
allowing an increased muscle blood flow), because of the cold
temperatures produced on the torso skin. A greater amount of
heat could then potentially be lost per unit of blood flow to the
skin, helping to preserve central blood volume.
17
Achieving a
decrease in core temperature, in combination with a large
reduction in skin temperature by precooling, may provide an
168 Duffield, Dawson, Bishop, et al
www.bjsportmed.com
increased possibility of improving exercise performance after
such procedures. However, it is important that the working
muscles are not cooled by these procedures, as performance
will be reduced if this is the case.
26
It is difficult to compare the results of the change in mean
skin temperature found here with other studies, as cooling
strategies have varied widely. Studies such as that of
Olschewski and Bruck,
27
in which whole body ambient air
cooling was used (duration of 30 minutes), obtained greater
decreases in mean skin temperature (4°C difference between
conditions) than studies such as the present one, where only
the torso region was cooled (for only brief periods, while
within warm/humid conditions).
After precooling, reduced heart rates have been reported
during the early stages of prolonged exercise tests
816
and sig-
nificant reductions in sweat loss,
826
but no difference in blood
lactate concentrations.
7
We failed to find any significant
differences between conditions in heart rates, blood lactate
concentrations, or changes in body mass as a result of sweat
loss. However, although much discussion has focused on the
physiological effects of precooling, the role of psychological
influences cannot be discounted. Although difficult to quan-
tify, it is possible that lowered feelings of heat strain could
improve the comfort level of the athlete, which may in tur n
translate into reduced feelings of fatigue and a smaller
decline in performance. At worst, the ice cooling jacket used
here provided some relief from the heat, which, regardless of
the influence on performance, is still valuable in hot/humid
conditions, especially for people in an initial encounter or
without pr ior exposure/acclimation to these types of condi-
tions. Although cooling before and during exercise may
relieve the feelings of thermal discomfort and thirst and pro-
vide a general perception that the exercise load is easier,
because of the continual engagement in maximum effort
sprints, it is unlikely that cooling will reduce RPE in repeated
maximal sprint exercise. Correspondingly, other studies
incorporating maximal exercise effor ts have reported no
changes to RPE with the administration of precooling,
14 15 17
as, when exercise is maximal in intensity, RPE will reach a
limit that should be the same for both conditions. The results
here also showed no difference in the perception of fatigue or
increased vigour reported by the subjects after the use of the
ice cooling jacket. However, effect size data (d = 0.7–1.0)
suggested large differences between conditions on both sets
of scales after the application of the cooling jacket at half
time, indicating that the lengthened half time exposure to a
cold microclimate may have helped suppress feelings of
fatigue and exertion during the second half.
In conclusion, the use of an ice cooling jacket for a short
initial precooling exposure and for additional cooling during
breaks throughout the intermittent exercise protocol did not
improve exercise performance. Therefore, although it appears
likely that the initial precooling exposure (five minutes) was
not of sufficient duration to enhance exercise performance,
there was some evidence of improved subject comfort in the
cooling trial. Further trials of cooling both before and during
prolonged intermittent exercise protocols that simulate the
duration and intensity of typical team sports should be
conducted, with longer exposure periods than used here,
although applicability to the field competition setting must be
considered. The use of ice jackets with hoods or the application
of ice towels to the head should also be included to assist with
greater body cooling.
.....................
Authors’ affiliations
R Duffield, B Dawson, D Bishop, Exercise Physiology Laboratory,
Department of Human Movement and Exercise Science, University of
Western Australia, Crawley, WA 6009, Australia
M Fitzsimons, S Lawrence, Western Australian Institute of Sport,
Challenge Stadium, Mt Claremont, WA 6010, Australia
REFERENCES
1 Adams WC, Fox RH, Fry AJ,
et al
. Thermoregulation during marathon
running in cool, moderate, and hot environments.
J Appl Physiol
1971;38:1030–7.
2 MacDougall J, ReddanW, Layton C,
et al
. Effects of metabolic
hyperthermia on performance during heavy prolonged exercise.
J Appl
Physiol
1974;36:538–44.
3 Kozlowski S, Brzezinska Z, Kruk B,
et al
. Exercise hyperthermia as a
factor limiting physical performance: temperature effect on muscle
metabolism.
J Appl Physiol
1985;59:766–73.
4 Gonzalez-Alonso J, Teller C, Anderson SL,
et al
. Influence of body
temperature on the development of fatigue during prolonged exercise in
the heat.
J Appl Physiol
1999;86:1032–8.
5 Lloyd E. Temperature and performance. II. Heat.
BMJ
1994;309:587–9.
6 Nevill M. Team sports: problems and solutions.
Sports Science Update
2(1)
. Slough: Gatorade Sports Science Institute Europe, 1996:6–7.
7 MacDonald A, Booth J, Fogarty A,
et al
. Whole body pre-cooling:
thermal cardiovascular and metabolic consequences.
International
Conference on Physiological and Cognitive Performance in Extreme
Environments.
Canberra: Australian Institute of Sport. 2000:76–9.
8 Schmidt V, Bruck K. Effect of a precooling manoeuvre on body
temperature and exercise performance.
J Appl Physiol
1981;50:772–8.
9 Hessemer V, Langusch D, Bruck K. Effect of slightly lowered
temperatures on endurance performance in humans.
J Appl Physiol.
1984;57:1731–7.
10 Marsh D, Sleivert G. Effect of precooling on high intensity cycling
performance.
Br J Sports Med
1999;33:393–7.
11 Myler G, Hahn A, Tumulty D. The effect of preliminary skin cooling on
performance of rowers in hot conditions.
Excel
1989;6:17–21.
12 Beelen A, Sargent A. Effect of lowered muscle temperature on the
physiological response to exercise in men.
Eur J Appl Physiol
1991;63:387–92.
13 Sjodin A, Forslund A, Webb P,
et al
. Mild overcooling increases energy
expenditure during endurance exercise.
Scand J Med Sci Sports
1996;6:22–5.
14 Yates K, Ryan R, Martin D,
et al
. Pre-cooling rowers can improve
laboratory 2000m performance in hot/humid conditions.
Sports
Medicine Australia Conference Proceedings.
Canberra: Sports Medicine
Australia, 1996:370–1.
15 Smith J, Yates K, Lee H,
et al
. Pre-cooling improves cycling performance
in hot/humid conditions.
Med Sci Sports Exerc
1997;29:S1501–263.
16 Lee D, Haymes E. Exercise duration and thermoregulatory responses
after whole body precooling.
J Appl Physiol
1995;79:1971–6.
17 Drust B, Cable NT, Reilly T. Investigation of the effects of the pre-cooling
on the physiological responses to soccer-specific intermittent exercise.
Eur
J Appl Physiol
2000;81:11–17.
18 Gore CJ, Edwards DA.
Australian fitness norms.
Canberra: The Health
Development Foundation, 1992.
19 Ramanathan NLA. A new weighting system for mean surface
temperature of the human body.
J Appl Physiol
1964;19:531–3.
20 Borg GA. Psychological basis of physical exertion.
Med Sci Sports Exerc
1982;14:377.
21 Grove JR, Prapavessis H. Preliminary evidence for the reliability and
validity of an abbreviated Profile of Mood States.
International Journal of
Sports Psychology.
1992;23:93–109.
22 Petruzzello SJ, Landers DM, Salazar W. Exercise anxiety reduction:
examination of temperature as an explanation for affective change.
Journal of Sport and Exercise Psychology
1993;15:63–76.
23 Christensen JE, Christensen CE. Statistical power analysis of health,
physical education, and recreation research.
Res Q
1977;48:204–8.
24 Garland HO. Altered temperature. In:
Case RM, ed. Variations in
human physiology.
Manchester: Manchester University Press, 1985.
25 Hertzman AB, Roth LW. The absence of vasoconstrictor reflexes in the
forehead.
Am J Physiol
1942;136:692–7.
26 Sleivert GG, Cotter JD, Roberts WS,
et al
. The influence of whole-body
vs torso precooling on physiological strain and performance of high
intensity exercise in the heat.
Comp Biochem Physiol A Physiol
2001;128:657–66.
27 Olschewski H, Bruck K. Thermoregulatory, cardiovascular, and muscular
factors related to exercise after precooling.
J Appl Physiol
1988;64:803–11.
Take home message
Although the use of an ice cooling jacket reduced the per-
ception of thermal load, it did not improve exercise
performance in prolonged intermittent repeated sprint
cycling efforts in warm/humid conditions.
Effect of precooling on performance 169
www.bjsportmed.com
