Content uploaded by Craig Heller
Author content
All content in this area was uploaded by Craig Heller on Nov 11, 2014
Content may be subject to copyright.
doi: 10.1152/japplphysiol.00093.2005
99:972-978, 2005. First published 5 May 2005;J Appl Physiol
Dennis A. Grahn, Vinh H. Cao and H. Craig Heller
aerobic exercise endurance in a hot environment
Heat extraction through the palm of one hand improves
You might find this additional info useful...
23 articles, 14 of which you can access for free at: This article cites
http://jap.physiology.org/content/99/3/972.full#ref-list-1
2 other HighWire-hosted articles: This article has been cited by
http://jap.physiology.org/content/99/3/972#cited-by
including high resolution figures, can be found at: Updated information and services
http://jap.physiology.org/content/99/3/972.full
can be found at: Journal of Applied Physiology about Additional material and information
http://www.the-aps.org/publications/jappl
This information is current as of June 6, 2013.
http://www.the-aps.org/.
© 2005 the American Physiological Society. ISSN: 8750-7587, ESSN: 1522-1601. Visit our website at
year (monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright
physiology, especially those papers emphasizing adaptive and integrative mechanisms. It is published 12 times a
publishes original papers that deal with diverse area of research in appliedJournal of Applied Physiology
by guest on June 6, 2013http://jap.physiology.org/Downloaded from
Heat extraction through the palm of one hand improves aerobic exercise
endurance in a hot environment
Dennis A. Grahn, Vinh H. Cao, and H. Craig Heller
Department of Biological Sciences, Stanford University, Stanford, California
Submitted 27 January 2005; accepted in final form 29 April 2005
Grahn, Dennis A., Vinh H. Cao, and H. Craig Heller. Heat
extraction through the palm of one hand improves aerobic exercise
endurance in a hot environment. J Appl Physiol 99: 972–978, 2005. First
published May 5, 2005; doi:10.1152/japplphysiol.00093.2005.—In situ-
ations where the accumulation of internal heat limits physical perfor-
mance, enhanced heat extraction from the body should improve
performance capacity. The combined application of local subatmo-
spheric pressure (35–45 mmHg) to an entire hand (to increase blood
volume) and a heat sink (18 –22°C) to the palmar surface were used to
draw heat out of the circulating blood. Subjects walked uphill (5.63
km/h) on a treadmill in a 40°C environment. Slopes of the treadmill
were held constant during paired experimental trials (with and without
the device). Heat extraction attenuated the rate of esophageal temper-
ature rise during exercise (2.1 ⫾ 0.4° and 2.9 ⫾ 0.5°C/h, mean ⫾ SE,
with and without the device, respectively; n ⫽ 8) and increased
exercise duration (46.1 ⫾ 3.4 and 32.3 ⫾ 1.7 min with and without the
device, respectively; n ⫽ 18). Hand cooling alone had little effect on
exercise duration (34.1 ⫾ 3.0, 38.0 ⫾ 3.5, and 57.0 ⫾ 6.4 min, for
control, cooling only, and cooling, and subatmospheric pressure,
respectively; n ⫽ 6). In a longer term study, nine subjects participated
in two or four trials per week for 8 wk. The individual workloads
(treadmill slope) were varied weekly. Use of the device had a
beneficial effect on exercise endurance at all workloads, but the
benefit proportionally decreased at higher workloads. It is concluded
that heat can be efficiently removed from the body by using the
described technology and that such treatment can provide a substantial
performance benefit in thermally stressful conditions.
arteriovenous anastomoses; venous plexus; aerobic capacity; cardiac
drift; heat stress
A RISING BODY TEMPERATURE during exercise can be a primary
factor limiting performance, especially endurance, in a hot
environment (1, 7, 12, 15, 17, 19, 21, 23). Studies have shown
that precooling can increase subjects’ endurance for exercise or
work in a hot environment (see Ref. 16 for review). The effect
of precooling is most likely due to creation of a greater heat
sink in peripheral tissues for metabolically produced heat (7,
13, 24). Therefore, the effectiveness of precooling is most
prominent in endurance events lasting 30 – 40 min, and 4 –16%
increases in endurance have been reported (5, 13, 20). Precool-
ing maneuvers that have been used, however, require special-
ized equipment, such as cold rooms or water baths, and the
treatments typically last from 30 min to ⬎1 h. Thus application
of such maneuvers is not practical under many circumstances.
We have been exploring performance benefits that could be
obtained from using a portable device to continuously extract
heat from the body core during endurance exercise in a hot
environment.
The heat extraction technology takes advantage of adapta-
tions for heat transfer that are features of certain nonhairy skin
surfaces. The arteriovenous anastomoses (AVAs) and venous
plexuses in the palms of the hands and the soles of the feet are
effective mechanisms for heat dissipation when core body
temperature rises (3, 4, 9, 10, 17). A device previously de-
scribed (11) is used to apply a 35- to 45-mmHg subatmospheric
pressure to an entire hand to draw blood into the hand and
increase the filling of the venous plexus underlying the palmar
surface. A heat sink applied to that palm extracts heat and cools
the venous blood. In the present study, we used the device in
an attempt to slow the rate of core temperature rise of individ-
uals engaged in aerobic exercise in a hot environment.
The hypothesis to be tested in these studies was that manip-
ulation of heat balance by enhancing heat loss from the hand
can increase the endurance capacity of individuals exercising at
a fixed workload in a hot environment. To test the hypothesis,
it was first necessary to establish that use of the heat extraction
method during exercise in a hot environment affected core
temperature. This was accomplished by measuring esophageal
temperature (T
es
) in a subset of the subject population (only a
limited number of subjects would tolerate the esophageal
thermocouple probe placement). Besides an increase in core
temperature, another concomitant of steady-state aerobic exer-
cise in a hot environment is cardiac drift (see Ref. 6 for
review). At the onset of exercise, heart rate rises to an appro-
priate level for the particular workload. However, as core
temperature rises, heart rate also rises, even though the work-
load is held constant. In the present study, we compared the
rates of rise of heart rate with and without heat extraction, and
we used a specific heart rate target as the exercise end point for
comparisons of endurance. The protocols were designed to
address three specific questions: 1) Does the continuous use of
the heat extraction device attenuate core temperature rise
during fixed-load exercise? 2) Does use of the heat extraction
device improve fixed-load exercise endurance? 3) Is the effect
of heat extraction on exercise duration workload dependent?
MATERIALS AND METHODS
Subjects
A total of 26 subjects participated in the studies. The physical
characteristics [gender, age, height, weight, and maximal O
2
con
-
sumption (V
˙
O
2 max
), when available] of each subject are tabulated in
Table 1. Eight of the subjects (6 men and 2 women) tolerated the
placement of an esophageal thermocouple probe. Ten male and eight
female subjects participated in a short-term study (an acclimation/
assessment session followed by sets of paired trials conducted over a
Address for reprint requests and other correspondence: D. A. Grahn, Dept.
of Biological Sciences, Stanford Univ., Stanford, CA 94305-5020 (E-mail:
DAGrahn@stanford.edu).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
J Appl Physiol 99: 972–978, 2005.
First published May 5, 2005; doi:10.1152/japplphysiol.00093.2005.
8750-7587/05 $8.00 Copyright
©
2005 the American Physiological Society http://www. jap.org972
by guest on June 6, 2013http://jap.physiology.org/Downloaded from
6-day period). Seven male and two female subjects participated in a
long-term study (a 2-wk acclimation and baseline assessment period
followed by 8 wk of experimental trials). Informed consent was
obtained from each subject using an instrument approved by the
Stanford University Institutional Review Board. Each subject was
assigned an alphanumeric identifier, which was used thereafter in
accordance with Health Insurance Portability and Accountability Act
guidelines. Subjects wore their own light exercise clothing and foot-
wear.
Facilities
Trials were conducted on stationary treadmills (model 60, Quinton;
model 9800, Nordic Track). V
˙
O
2 max
tests were administered in a 23°C
room. The acclimation and heat stress trials were conducted in a
2.44 ⫻ 3.35 ⫻ 2.44-m (width ⫻ length ⫻ height) temperature-
controlled environmental chamber. The ambient conditions inside the
environmental chamber were 40.0 ⫾ 0.5°C. Relative humidity inside
the chamber was 20 –25% at the start of all exercise trials. However,
when evaporative water loss from the subjects exceeded the capacity
of the room heating-ventilation-air conditioning system, relative hu-
midity rose as high as 45% by the end of a trial.
Monitoring Equipment
For the V
˙
O
2 max
tests, respiratory gases were measured using a
respiratory gases/metabolic analysis system (Parvomedics, Salt Lake
City, UT). Heart rate monitors/data loggers (Polar, Kempele, Finland)
were used to record and collect heart rate data at 5-s intervals. T
es
was
measured with a Mon-a-therm general-purpose temperature probe
(model 503-0028, Mallinckrodt Medical, St. Louis, MO). The probes
were self-inserted through the nose or mouth to a depth of 38–39 cm.
The probes were connected to a thermocouple transducer/data logger
(model OM-4000, Omega Engineering, Stamford, CT), which re-
corded temperature data at 1-s intervals. Water loss was determined
by subtracting postexercise nude weight from preexercise nude
weight. Nude weight was determined using a commercially available
cargo scale (model c-12, OHAUS) located in a private changing room.
At the end of each trial, heart rate and temperature data were
downloaded from the data loggers to a central desktop computer and
transferred to a spreadsheet (Microsoft Excel) for subsequent offline
analysis. Hand-noted data logs were also tabulated for each trial.
Subject identifier, date, treatment, pre- and postexercise nude weights,
exercise duration, and miscellaneous comments were recorded on the
data sheets, along with heart rate measurements, at 3-min intervals.
Heat Extraction Device
The heat extraction device (AVAcore Technologies, Ann Arbor,
MI) consisted of a rigid chamber into which a hand could be inserted
through an elastic structure that formed a flexible airtight seal around
the wrist. The rigid chamber was connected to a pressure sensor, a
pressure relief valve (cracking pressure ⫺45 mmHg), and a vacuum
source [the building in-house system or a commercially available
vacuum pump (1/10th horsepower; model SR-0015-VP, Thomas
Industries, Louisville, KY)]. A water trap consisting of a 1,000-ml
filter flask (VWR) was plumbed into the vacuum line upstream of the
vacuum pump. Activation of the vacuum pump created a slight
subatmospheric chamber pressure (⫺40 mmHg). Inside the chamber,
the palm rested on a curved metal surface that was maintained at 22°C
or 18°C (⫾0.5°C) by perfusion of the temperature-controlled water
beneath it. The hand interface was tethered via Tygon tubing (8-mm
bore, 3-mm wall) to a temperature-controlled heated/refrigerated cir-
culating water bath (model RM 6, Lauda, Konigshofen, Germany) that
regulated the temperature of the circulating water. The hand interface
device was suspended from the ceiling by an elastic cord so that the
subject could maintain normal arm movements while walking.
Experimental Protocols
Pretrial assessments of physical condition. For the V
˙
O
2 max
test,
subjects were equipped with the heart rate-monitoring equipment,
snorkel mouthpiece, and nose plug from the respiratory gas analysis
system. Once equipped, the subjects stood on the idle treadmill for 5
min. After 5 min of baseline data collection, the speed of the treadmill
was increased by 3.2 km/h at 3-min intervals until O
2
consumption
stabilized for 30 s or until subjective exhaustion. V
˙
O
2 max
and maxi
-
mum heart rate were noted for each subject.
Baseline assessments of individual physical performance capacities
were conducted in the hot room and required that the subjects start
walking on a level treadmill at 5.63 km/h for 3 min; then the slope of
the treadmill was increased by 2% at 3-min intervals. Elevations of the
slope continued until the subject attained a heart rate that was 90% of
the estimated age-specific maximum (221 ⫺ age) or, if available, the
heart rate attained in a prior V
˙
O
2 max
test. The slope of the treadmill in
the subsequent experimental trials was initially set at 60 –65% of the
slope at which the subject reached 90% maximum heart rate in these
baseline trials.
Standard daily experimental routines. The subjects arrived at the
laboratory 30 min before the start of a trial. Preexercise nude weight
was measured, and a heart rate monitor was attached to the subject.
Table 1. Gender, age, weight, and V
˙
O
2max
of individual subjects
Subject
No. Gender
Age,
yr
Height,
cm
Weight,
kg
V
˙
O
2max
,ml
O
2
䡠min
⫺1
䡠kg
⫺1
Single-paired-trial subjects
3a M 23 185 91
3b F 23 170 55
3c M 58 180 80
3d M 53 185 109
3e M 36 170 89
3f M 36 185 76
3g M 49 180 69
3h F 29 163 63
3i F 37 165 71
3k M 47 173 69
3l M 50 183 84
3m M 52 180 97
3n F 48 168 60
3o M 20 175 73
3p F 55 173 55
3q F 54 160 64
3r F 60 170 64
Core temperature effects
3a M 23 185 91
3b F 23 170 55
3c M 58 180 80
3d M 53 185 109
3e M 36 170 89
4c M 20 178 128
4f M 20 170 84
4j F 20 170 62
Multiple-paired-trial subjects
4a M 21 179 78 48
4b M 21 183 85 56
4c M 20 178 128 37
4d M 23 183 76 46
4f M 20 170 84 48
4g M 21 175 83 53
4h F 23 155 47 44
4i M 19 180 69 62
4j F 20 170 62 44
V
˙
O
2max
, maximal O
2
consumption. Subject 4i participated in only 2 trials/
wk.
973HEAT EXTRACTION FROM THE HAND IMPROVES ENDURANCE
J Appl Physiol • VOL 99 • SEPTEMBER 2005 • www.jap.org
by guest on June 6, 2013http://jap.physiology.org/Downloaded from
The subjects then rested in a 23°C room for 30 min or until heart rate
had stabilized at ⬍70 beats/min. The subjects then moved into the hot
room and performed the designated exercise task. Stop criteria for
exercise were 90% of maximum heart rate, 120 min of exercise, or
subjective exhaustion. On completion of the exercise, the subjects
returned to the 23°C room where they sat quietly for 30 min. After the
30-min recovery period, nude body weight was again measured.
Before leaving the facility, the subjects consumed a volume of water
or a sports drink equivalent in mass to the amount of body weight lost
during the exercise trial. All trials for an individual subject were
conducted at the same time of day.
Does continuous use of the heat extraction device attenuate core
temperature rise during fixed-load exercise? Eight subjects partici-
pated in these trials after completing the baseline performance capac-
ity assessment. These trials were part of the larger study on endurance
(see Does use of the heat extraction device improve fixed-load exer-
cise endurance?), but they included only the subjects who tolerated
placement of an esophageal thermocouple probe. The subjects were
equipped with a heart rate monitor and esophageal thermocouple
probe. The slope of the treadmill for the experimental trials was set at
60 – 65% of the slope at which the subject reached 90% of his/her
estimated maximum heart rate. The trials consisted of the subjects
walking on the treadmill at 5.63 km/h at their predetermined slope.
Each subject participated in two experimental trials, one with and one
without the heat extraction device. The order of the treatments was
randomized. The experimental trials were initiated ⱖ2 days after the
acclimation/assessment trial and were separated by ⱖ2 days.
Does use of the heat extraction device improve fixed-load exer-
cise endurance? Each of the 18 subjects participated in a minimum of
three activities: one baseline assessment and two experimental trials.
On day 1, baseline assessments of individual physical performance
capacities were conducted. The slope of the treadmill for the subse-
quent experimental trials was set at 65% of the slope at which the
subject reached 90% of his/her age-adjusted maximum heart rate in
the baseline assessment. This workload was selected because it
resulted in exercise durations of 20 –45 min before the subjects
reached the stop criterion (90% of maximum heart rate).
The experimental trials consisted of the subjects walking on the
treadmill at 5.63 km/h at their predetermined slope until heart rate
reached 90% of the age-adjusted maximum heart rate. The experi-
mental trials were initiated ⱖ2 days after the acclimation/assessment
trial and were separated by ⱖ2 days. All subjects performed a
minimum of two experimental trials: one without the heat extraction
device and one with the heat extraction device worn and activated.
The order of the treatments was randomized. Six subjects participated
in an additional trial to assess the effect of the subatmospheric
pressure in the operation of the heat extraction device. For these trials,
the heat extraction device was worn and cool water circulated through
the device, but the subatmospheric pressure was not applied.
Is the effect of heat extraction on exercise duration workload de-
pendent? To minimize the confound of acclimation in this series of
experiments, each subject spent 2 wk acclimating to the experimental
environment before participating in 8 wk of experimental trials.
Acclimation to the experimental conditions entailed six 1-h exer-
cise bouts in the hot environment over a 10-day period, during which
the subjects walked on a treadmill at 5.63 km/h at a self-selected
treadmill slope. Initial V
˙
O
2 max
tests and baseline performance assess
-
ments were conducted on each subject 2– 4 days after the acclimation
period. The V
˙
O
2 max
test and baseline assessments were separated by
24 h. For the baseline physical performance capacity assessment, the
stop point was when the subject’s heart rate reached 90% of the
maximum heart rate that had been attained in the V
˙
O
2 max
trials.
Eight subjects participated in four exercise trials per week: two
with and two without cooling treatment. Trial days were Mondays,
Tuesdays, Thursdays, and Fridays. The Monday and Tuesday trials
and Thursday and Friday trials were paired for treatments, with the
treatments ordered randomly. Because of scheduling conflicts, one
subject participated in only two trials per week (Thursdays and
Fridays). On each day, the subjects walked on a level treadmill at 5.63
km/h for 3 min before the slope of the treadmill was increased to the
individual’s predetermined slope. The treadmill slope was set for each
individual so that his/her 90% maximum heart rate would be reached
in a specified exercise period. The individual subjects’ slopes were
adjusted weekly, but the individual subjects’ treadmill slopes re-
mained constant throughout a given week. The targeted exercise
duration (time to reach 90% of age-adjusted maximum heart rate) for
the control trials was 45– 60 min during weeks 1 and 2, 35–45 min
during week 3, 25–35 min during week 4, 15–25 min during week 5,
10 –20 min during week 6, 20–30 min during week 7, and 35– 45 min
during week 8. The slopes selected for each individual corresponded
to ⬃50 –85% of the slopes attained at 90% maximum heart rate in the
baseline assessments.
Data Analysis
Endurance times for all trials were tabulated, and raw heart rate and
T
es
data were plotted for each trial using Microsoft Excel software.
The raw 5-s interval heart rate data were plotted and screened for
artifact and then sorted by 30-s intervals. T
es
data were treated in a
manner similar to the heart rate data. The raw 1-s interval T
es
data
were plotted and screened for artifact and then sorted by 30-s intervals
for regression analysis. For display purposes in Fig. 1, these data are
plotted in 3-min intervals. Most of these curves were characterized by
an initial rapid rise, a break point corresponding to the onset of
vasodilation, and a final linear increase (Fig. 1). T
es
data from the final
linear section of each curve were subjected to a regression analysis
(available as a graph tool in Microsoft Excel) to determine the best-fit
slope of the T
es
change over time. The rates of T
es
change data were
tabulated and sorted according to individual and treatment and sub-
jected to descriptive statistical analysis and a post hoc paired t-test
(available as a statistical analysis tool in Microsoft Excel).
For the exercise endurance trials, exercise duration records were
verified using the heart rate data. Results from the sets of paired trials
were included in subsequent data analysis only if the subject reached
the 90% maximum heart rate stop criterion in both of the trials.
Exercise durations were sorted by treatments, trial day, and subject,
and means ⫾ SE were calculated for each treatment group. All main
effects of factors “treatment,” “order,” “slope,” and “subject” were
analyzed by one- or two-way ANOVA (Proc GLM in SAS/STAT
version 8.02, SAS Institute, Cary, NC), with repeated measures where
appropriate. Post hoc paired t-tests were used for statistical analysis of
the exercise duration results.
RESULTS
Does Continuous Use of the Heat Extraction Device
Attenuate Core Temperature Rise During
Fixed-Load Exercise?
Cooling treatment attenuated the rate at which T
es
increased
during exercise (Fig. 1). The rise in T
es
during exercise was
characterized by an abrupt linear rise in core temperature early
in the exercise bout. At 10 –20 min into the exercise bout, a
deflection point in the T
es
vs. time trace could be discerned;
then the linear rise in T
es
continued, but at a lower rate. Cooling
had little effect on the rise in T
es
early in the exercise bouts but
substantially attenuated the rise in T
es
in the later bouts: 2.1 ⫾
0.4 and 2.8 ⫾ 0.5 (SE)°C/h for cooling and control, respec-
tively (n ⫽ 8, P ⬍ 0.005). The attenuation of the late phase of
exercise T
es
rise was consistent in all subjects (P ⬍ 0.005,
paired t-test; Fig. 1).
974 HEAT EXTRACTION FROM THE HAND IMPROVES ENDURANCE
J Appl Physiol • VOL 99 • SEPTEMBER 2005 • www.jap.org
by guest on June 6, 2013http://jap.physiology.org/Downloaded from
Does Use of the Heat Extraction Device Improve
Fixed-Load Exercise Endurance?
Cardiovascular drift was observed in all trials. All subjects
(n ⫽ 18) reached the 90% maximum heart rate stop criterion.
The effects of treatment on the rate of cardiac drift were similar
to the effect on T
es
rise (Fig. 2).
According to reports from the
subjects, the 90% maximum heart rate provided a useful index
for impending subjective exhaustion. The combined applica-
tion of cooling and subatmospheric pressure increased exercise
duration by 43%: 46.1 ⫾ 3.4 and 32.3 ⫾ 1.7 (SE) min for
cooling and control, respectively (n ⫽ 18; Fig. 3A). A post hoc
t-test established that there was a significant effect of treatment
on exercise duration (P ⬍ 0.001).
Six subjects completed the trials under three experimental
conditions: control (no treatment), cooling only, and cooling
and subatmospheric pressure (Fig. 3B). In this subset of sub-
jects, cooling only (i.e., placing a hand in the heat extraction
device with cooling fluid circulating but without application of
subatmospheric pressure) provided little performance benefit
(34.1 ⫾ 3.0 and 38.0 ⫾ 3.5 min for control and cooling only,
respectively), but the combination of cooling and subatmo-
spheric pressure provided a substantial increase in endurance
(57.0 ⫾ 6.4 min, n ⫽ 6; Fig. 3B). ANOVA revealed a
significant effect of the factor treatment among groups (P ⬍
0.01, 1-way ANOVA). Post hoc t-tests established that treat-
ment with the fully activated device (i.e., cooling and subat-
mospheric pressure) resulted in a significant improvement in
Fig. 1. Effect of heat extraction on rate of esophageal temperature (T
es
) change during fixed-load exercise in a hot environment. 䊐, Control; ■, heat extraction
during exercise. T
es
vs. time plots are shown for each individual during a control and a cooling trial. In most individual runs, a break point in the curve indicates
vasodilation of the hand. Slopes of the curves (⌬T
es
/min) after these break points were determined by linear regression and are presented at bottom right. Symbols
for individual subjects are connected by lines. Symbols not connected by lines are treatment group average values (means ⫾ SE, n ⫽ 8, P ⬍ 0.005, paired t-test).
Fig. 2. Combined application of subatmospheric pressure and a cool thermal
load to a single hand attenuates rise in T
es
and cardiovascular drift during
fixed-load exercise in a hot environment in a single subject. Subject walked at
5.63 km/h up a 9% slope on a treadmill in a hot environment (40 ⫾ 1°C,
35– 45% relative humidity).
䊐, No treatment; F, 1 hand placed in the heat
extraction device in which water was circulating and a pressure differential
was maintained; ‚, 1 hand placed in the heat extraction device in which water
was circulating but no pressure differential was maintained. Stop criterion for
these trials was 90% of the subject’s age-adjusted maximum heart rate (i.e.,
163 beats/min). An abrupt drop in heart rate occurred at the end of exercise.
Effective heat extraction treatment extended exercise duration. BPM, beats/
min.
975HEAT EXTRACTION FROM THE HAND IMPROVES ENDURANCE
J Appl Physiol • VOL 99 • SEPTEMBER 2005 • www.jap.org
by guest on June 6, 2013http://jap.physiology.org/Downloaded from
performance over the other treatments (cooling only or control;
P ⬍ 0.005).
Is the Effect of Heat Extraction on Exercise Duration
Workload Dependent?
The combined application of cooling and subatmospheric
pressure to one hand increased exercise duration at all work-
loads (Table 2). ANOVA revealed significant effects of the
factors treatment and treadmill slope (P ⬍ 0.0001 for both
factors, 2-way ANOVA). The effect of treatment (cooling) was
affected by the slope at which the exercise was performed
[significant interaction between the main factors (P ⬍ 0.025)].
The exercise duration data from the entire set of paired trials
were grouped according to treatment and analyzed (n ⫽ 96).
Mean exercise duration with cooling was 48.0 ⫾ 2.9 min
compared with 30.7 ⫾ 1.5 min without treatment: a mean
increase of 56% (P ⬍ 2.0 ⫻ 10
⫺13
, 2-tailed paired t-test). The
magnitude of the cooling treatment effect was correlated with
endurance duration during the control condition (Fig. 4). This
relation appeared to be exponential. Seventy percent of the
variance in the improvement with the cooling treatment could
be accounted for by an exponential function fitted to the data:
y ⫽ 12.724e
0.0372x
, where y and x represent exercise duration
with and without treatment, respectively.
DISCUSSION
Common experience along with scientific studies in animals
and humans support the generalization that the ability to
sustain a high level of aerobic exercise is limited by core body
temperature and, therefore, indirectly, by high ambient tem-
perature (7, 18, 23). The corollary to that generalization is that
extraction of heat from the body core should increase the
capacity to sustain aerobic exercise. Studies using various
methods to achieve reductions in body heat content before
exercise onset (precooling) demonstrated that aerobic endur-
ance can be extended by these maneuvers, probably because of
Fig. 3. Effects of cooling with and without subatmospheric
pressure (SP) on exercise duration in a hot environment. Eigh-
teen subjects participated in cooling and control trials. A subset
of that group (n ⫽ 6) participated in an additional trial with
cooling only. A: exercise duration with treatment vs. exercise
duration in control condition shown as individual data points.
Three-trial subjects participated in control, cooling only, and
combined cooling and subatmospheric pressure tests. Heat
extraction had a significant effect on exercise duration (P ⬍
0.005, post hoc t-test). B: group data from the 6 subjects who
participated in 3 trials. Values are means ⫾ SE. Combined
cooling and subatmospheric pressure treatment resulted in a
significant improvement in performance over the other treat-
ments (P ⬍ 0.005, post hoc t-test). Results from cooling only
were marginally different from control (P ⬍ 0.05).
Table 2. Weekly summaries for multiple-paired-trial study
Week
No. n Slope
No. of
Paired
Trials
Exercise Duration, min
RatioCooling Control
1 9 9.6⫾1.0 13
a
76.6⫾8.0 48.7⫾3.7 1.57
2 9 9.6⫾1.0 6
b
77.5⫾11.0 50.4⫾4.8 1.54
3 9 12.3⫾1.4 14
c
71.8⫾7.8 41.8⫾3.5 1.72
4 9 13.8⫾1.3 15 41.3⫾7.4 28.3⫾4.7 1.46
5 9 14.9⫾1.4 17 26.6⫾3.9 18.9⫾1.4 1.41
6 6
d
16.6⫾1.9 11 23.8⫾1.4 17.2⫾2.0 1.39
7 6 15.5⫾2.4 11 32.0⫾5.8 24.7⫾5.0 1.30
8 5 12.3⫾2.4 8
e
65.0⫾6.6 37.4⫾2.2 1.74
Values are mean ⫾ SE; n, number of subjects. Paired trials represents pairs
of trials in which subjects reached 90% maximum heart rate stop criterion in
both trials; only data from paired trials in which individual subjects completed
experimental and control runs were included. Reasons for not completing trials
were exceeding 120 min of exercise, blisters, illness, and travel-related
absences. Ratio represents ratio of mean exercise duration with heat extraction
to mean exercise duration without heat extraction.
a
Four cooling trials were
terminated at 120 min.
b
Six cooling trials were terminated at 120 min; 5 trials
were cancelled because of subject illness.
c
One trial was terminated at 120 min,
3 were cancelled because of illness.
d
Three subjects quit the study because of
prior commitments.
e
One trial was terminated at 120 min.
Fig. 4. Effect of cooling on exercise duration: a comparison of 96 paired
treatment trials. Effect of cooling was affected by exercise duration during the
control condition: the longer the duration of exercise in the control condition,
the greater the cooling treatment effect. An exponential function (y ⫽
12.724e
0.0372x
, where y and x represent exercise duration with and without
treatment, respectively) accounted for 70% of the variance in the data.
976 HEAT EXTRACTION FROM THE HAND IMPROVES ENDURANCE
J Appl Physiol • VOL 99 • SEPTEMBER 2005 • www.jap.org
by guest on June 6, 2013http://jap.physiology.org/Downloaded from
creation of a greater sink for metabolic heat generated during
the exercise (see Ref. 16 for review). A recent study by Wilson
et al. (24) is an excellent example. Precooling was achieved by
immersion of the subjects up to the suprailiac crest in cool
(17.7°C) water for 30 min before exercise on a fixed bicycle at
60% V
˙
O
2 max
(24). The precooling created a negative heat load
on the body; as a result, the subjects experienced a doubling of
the exercise duration required to raise their body temperatures
0.5°C. Clearly, decreasing the rate of rise in body heat content
during exercise improves endurance.
Precooling techniques, however, have limitations. They can
require cumbersome equipment and significant time of appli-
cation just before exercise, e.g., 30 min in the study of Wilson
et al. (24). In addition, they have a finite benefit that is
determined by the amount of negative heat storage that is
possible and the rate of positive heat storage during exercise.
Thus they are optimally useful in events lasting ⬍1h.A
method that would enable repeated or continuous heat extrac-
tion during exercise would be of even greater benefit than
precooling in extending aerobic endurance and protecting
against heat stress.
Here we show that continuous extraction of heat from the
body through only one hand during exercise can increase
aerobic endurance by a large percentage. How is such a large
increase in physical performance possible from the cooling of
such a small area of the body surface? The answer resides in
the fact that the palms of the hands, the soles of the feet, and
some areas of the face in humans have circulatory adaptations
for the dissipation of metabolic heat. These adaptations consist
of AVAs that can shunt blood directly from arterioles to
venous plexuses, which act as radiators (3, 8, 9). The blood
cooled in such a venous plexus returns directly to the core of
the body. The heat extraction device we used in these studies
has the ability to enhance the heat exchange capacity of those
radiators by distending the venous plexus vessels through the
application of subatmospheric pressure. Furthermore, reflex
vasoconstriction of the AVAs is prevented by maintenance of
a heat-sink temperature above the threshold for local vasocon-
striction.
Under the conditions of these studies, application of subat-
mospheric pressure to the hand was critical for enhancing heat
transfer. Figure 2 is an example of the effect of heat extraction
using the device with and without application of subatmo-
spheric pressure. Three trials at the same workload and ambi-
ent conditions are shown. The heart rate curves determined the
end of exercise, and, in this example, ⬎60% more time was
required to reach that criterion when the heat extraction device
was used. When the device was used to cool without applica-
tion of subatmospheric pressure, there was, at best, a limited
beneficial effect. The results seem in direct conflict with those
of Selkirk et al. (22), who reported that 20 min of submersion
of both hands and arms up to the elbows in turbulently mixed
17°C water during intermissions between 1-h exercise bouts by
firefighters clad in full turnout gear provided a substantial
physical performance benefit (an ⬃32% improvement in en-
durance time). There were considerable methodological differ-
ences between that study and ours, the largest (aside from the
cooling techniques themselves) being the attire of the subjects,
the exercise regimens, and the timing of cooling. In our studies,
the subjects were clad in warm-weather exercise garb and were
continuously cooled throughout a single sustained exercise
bout. In the study of Selkirk et al., the subjects wore heavy
insulation layers during exercise-rest cycles (some of the in-
sulation layers were removed during the resting phases of the
cycles) and received cooling treatment only during some of the
resting phases. Despite the methodological differences, both
studies demonstrate substantial benefits from local cooling
under specific thermally stressful conditions. It is likely that
much of the cooling effect of forearm immersion reported by
Selkirk et al., Allsopp and Poole (2), and others is mediated
through the thermoregulatory vasculature in the hands.
In the studies reported here, we used heart rate as the
determining factor for marking maximum exercise duration.
Alternative metrics for comparing exercise duration could have
been T
es
or the point of perceived exhaustion. In preliminary
work not reported here, we routinely saw a good correlation
between perceived exhaustion and T
es
, consistent with the
study of Gonzalez-Alonso et al. (7), in which the initial core
temperatures of subjects were manipulated by precooling or
preheating just before exercise in a hot environment (40°C) at
60% V
˙
O
2 max
until volitional exhaustion. The main conclusion
of that study was that exhaustion occurred at the same high
core temperature regardless of the starting conditions. Another
result from that study was that, after 10 min of exercise,
changes in heart rate were closely related to changes in T
es
regardless of the starting conditions. Given these results and
the difficulty many subjects have with esophageal thermocou-
ples, we decided that heart rate was the best objective metric
for our comparisons of endurance.
One would expect that the ability of the heat extraction
device to increase endurance would depend on the intensity of
the exercise. If the AVAs are fully open and the parameters of
the heat extraction device are held constant, there should be a
maximum attainable level of heat extraction. Therefore, as
workload increases, that maximum level of heat extraction
should be a lesser and lesser proportion of the total heat
produced. As a result, core body temperature should rise faster
the higher the workload, even with the heat extraction device.
This expectation is fulfilled in the results shown in Table 2 and
Fig. 4. The intensity of the workload was altered by changing
the slope of the treadmill. At the initial slope of 9.3%, the
cooling trials were 50 –70% longer than the control trials, but
at the highest workloads the difference dropped to 30– 40%.
The effect of cooling on exercise endurance appears to be
exponentially related to workload (Fig. 4).
More work is required to fully characterize the effectiveness
of the heat extraction device in extracting heat from the body
core, but the data obtained so far indicate that this technology
has great potential for extending endurance of individuals
working in hot environments. The results of this study are of
interest not just to those who want to improve athletic perfor-
mance but also to those in professions that require high levels
of physical work in thermally stressful environments. This
would include military personnel, firefighters, construction
workers, and a variety of industrial workers. Extracting heat
from the body core is of value in protecting such individuals
from heat stress and may facilitate recovery from episodes of
hyperthermia.
ACKNOWLEDGMENTS
The authors are grateful to Paul Franken for assistance with statistical
analyses, Julie van Loben Sels for assistance with data collection, and Grace
977HEAT EXTRACTION FROM THE HAND IMPROVES ENDURANCE
J Appl Physiol • VOL 99 • SEPTEMBER 2005 • www.jap.org
by guest on June 6, 2013http://jap.physiology.org/Downloaded from
Hagiwara for critical input throughout the project, especially during prepara-
tion of the manuscript.
This document was cleared by US Defense Advanced Research Projects
Agency and approved for public release, distribution unlimited.
GRANTS
This project was supported in part by a grant from the US Defense
Advanced Research Projects Agency.
DISCLOSURES
Patents have been issued for the technology described in this manuscript [D.
Grahn and H. C. Heller (Inventors), Stanford University (Assignee)], and
Stanford University has entered into a licensing agreement with AVAcore
Technologies, Inc., for the commercialization of the technology. Included in
the license is a royalty agreement that grants Stanford University a percentage
of the net sales of the technology, which will be shared by the University and
the inventors. D. Grahn and H. C. Heller are founders of AVAcore Technol-
ogies but receive no ongoing compensation from the company.
REFERENCES
1. Adams WC, Fox RH, Fry AJ, and MacDonald IC. Thermoregulation
during marathon running in cool, moderate, and hot environments. J Appl
Physiol 38: 1030 –1037, 1975.
2. Allsopp AJ and Poole KA. The effect of hand immersion on body
temperature when wearing impermeable clothing. J R Nav Med Serv 77:
41– 47, 1991.
3. Bergersen TK. A search for arteriovenous anastomoses in human skin
using ultrasound Doppler. Acta Physiol Scand 147: 195–201, 1993.
4. Bergersen TK, Eriksen M, and Walloe L. Local constriction of arterio-
venous anastomoses in the cooled finger. Am J Physiol Regul Integr Comp
Physiol 273: R880 –R886, 1997.
5. Booth J, Marino F, and Ward JJ. Improved running performance in hot
humid conditions following whole body precooling. Med Sci Sports Exerc
29: 943–949, 1997.
6. Coyle EF and Gonzalez-Alonso J. Cardiovascular drift during prolonged
exercise: new perspectives. Exerc Sport Sci Rev 29: 88 –92, 2001.
7. Gonzalez-Alonso J, Teller C, Andersen SL, Jensen FB, Hyldig T, and
Nielsen B. Influence of body temperature on the development of fatigue
during prolonged exercise in the heat. J Appl Physiol 86: 1032–1039,
1999.
8. Grahn DA, Brock-Utne JG, Watenpaugh DE, and Heller HC. Recov-
ery from mild hypothermia can be accelerated by mechanically distending
blood vessels in the hand. J Appl Physiol 85: 1643–1648, 1998.
9. Greenfield ADM. The circulation through the skin. In: Handbook of
Physiology. Circulation. Washington, DC: Am. Physiol. Soc., 1963, sect.
2, vol. II, chapt. 39, p. 1325–1351.
10. Hales JR, Jessen C, Fawcett AA, and King RB. Skin AVA and capillary
dilatation and constriction induced by local skin heating. Pflu¨gers Arch
404: 203–207, 1985.
11. Hsu AR, Hagobian TA, Jacobs KA, Attallah H, and Friedlander AL.
Effects of heat removal through the hand on metabolism and performance
during cycling exercise in the heat. Can J Appl Physiol 30: 87–104, 2005.
12. Kozlowski S, Brzezinska Z, Kruk B, Kaciuba-Uscilko H, Greenleaf
JE, and Nazar K. Exercise hyperthermia as a factor limiting physical
performance: temperature effect on muscle metabolism. J Appl Physiol 59:
766 –773, 1985.
13. Lee DT and Haymes EM. Exercise duration and thermoregulatory
responses after whole body precooling. J Appl Physiol 79: 1971–1976,
1995.
14. Leweke F, Bruck K, and Olschewski H. Temperature effects on venti-
latory rate, heart rate, and preferred pedal rate during cycle ergometry.
J Appl Physiol 79: 781–785, 1995.
15. MacDougal JD, Reddan WG, Layton CR, and Dempsey JA. Effects of
metabolic hyperthermia on performance during heavy prolonged exercise.
J Appl Physiol 36: 538 –544, 1974.
16. Marino FE. Methods, advantages, and limitations of body cooling for
exercise performance. Br J Sports Med 36: 89 –94, 2002.
17. Midttun M and Sejrsen P. Cutaneous blood flow rate with and without
arteriovenous anastomoses during exercise. Scand J Med Sci Sports 8:
84 –90, 1998.
18. Nybo L and Nielsen B. Hyperthermia and central fatigue during pro-
longed exercise in humans. J Appl Physiol 91: 1055–1060, 2001.
19. Nybo L, Jensen T, Nielsen B, and Gonzalez-Alonso J. Effects of marked
hyperthermia with and without dehydration on V
˙
O
2
kinetics during intense
exercise. J Appl Physiol 90: 1057–1064, 2001.
20. Olschewski H and Bruck K. Thermoregulatory, cardiovascular, and
muscular factors related to exercise after precooling. J Appl Physiol 64:
803– 811, 1988.
21. Sawka MN, Young AJ, Latzka WA, Neufer PD, Quicley MD, and
Pandolf KB. Human tolerance to heat strain during exercise: influence of
hydration. J Appl Physiol 73: 368 –375, 1992.
22. Selkirk GA, McLellan TE, and Wong J. Active versus passive cooling
during work in warm environments while wearing firefighting protective
clothing. J Occup Environ Hyg 1: 521–531, 2004.
23. Walters TJ, Ryan KL, Tate LM, and Mason PA. Exercise in the heat
is limited by a critical internal temperature. J Appl Physiol 89: 799 – 806,
2000.
24. Wilson TE, Johnson SC, Petajan JH, Davis SL, Gappmaier E,
Leutkemeier MJ, and White AT. Thermal regulatory responses to
submaximal cycling following lower-body cooling in humans. Eur J Appl
Physiol 88: 67–75, 2002.
978 HEAT EXTRACTION FROM THE HAND IMPROVES ENDURANCE
J Appl Physiol • VOL 99 • SEPTEMBER 2005 • www.jap.org
by guest on June 6, 2013http://jap.physiology.org/Downloaded from
