Heat loss from the human head during exercise
W. RASCH, P. SAMSON, J. COTE, AND M. CABANAC
Department of Physiology, Lava1 University, Quebec GlK 7P4, Canada
RASCH,W.,P. SAMSON,J. COTE,ANGM. CABANAC.H~~~~OSSis capable of removing this metabolic heat from the brain
from the human head during exercise. J. Appl. Physiol. 71(2):
590-595, 1991.-Evaporative and convective heat loss from
head skin and expired air were measured in four male subjects
at rest and during incremental exercise at 5, 15, and 25°C am-
bient temperature (T,) to verify whether the head can function
as a heat sink for selective brain cooling. The heat losses were
measured with an open-circuit method. At rest the heat loss
from head skin and expired air decreased with increasing T,
from 69 t 5 and 37 t 18 (SE) W (5°C) to 44 & 25 and 26 t 7 W
(25°C). At a work load of 150 W the heat loss tended to increase
with increasing T,: 119 k 21 (head skin) and 82 t 5 W (respira-
tory tract) at 5°C T, to 132 t 27 and 103 * 12 W at 25°C T,.
Heat loss was always higher from the head surface than from
the respiratory tract. The heat losses, separately and together
(total), were highly correlated to the increasing esophageal tem-
perature at 15 and 25°C T,. At 5°C T, no correlation occurred.
The results showed that the heat loss from the head was larger
than the heat brought to the brain by the arterial blood during
hyperthermia, estimated to be 45 W per 1°C increase above
normal temperature, plus the heat produced by the brain, esti-
mated to be up to 20 W. The total heat to be lost is therefore
~65 W during a mild hyperthermia (+1 “C) if brain tempera- loses the excessive heat. Intuition
ture is to remain constant. It is concluded that the human head,
from a strictly physical point of view, can function as a heat
sink with a heat loss capacity higher than the heat produced by
the brain and received from the arterial blood during hyper-
thermia.
human selective brain cooling; latent heat loss; sensible heat
loss; tympanic temperature; esophageal temperature; incre-
mental exercise
THE EXISTENCE of a mechanism in animals (2) that cools
the brain selectively during hyperthermia is well ac-
cepted by the community of thermal physiologists. How-
ever, as soon as it is hypothesized that this mechanism
might also exist in humans (7), doubt is cast on the basis
of several arguments (24, 25, 30, 36). The most fre-
quently quoted argument is that the human species does
not possess a rete mirabile on the internal carotid artery
in its course through sinus cavernosus. Therefore it is
believed that no heat exchange exists between venous
return and arterial blood. According to this point of view,
cooling exists, but it is exclusively provided by thermal
convection from the carotid blood (3, 24, 36).
There are also arguments leading to the opposite con-
clusion. As in any system with constant mass and specific
heat, the local temperature of the brain depends on the
balance among heat production, heat inflow, and heat
outflow. The brain is more thermogenetic than most
other tissues (l), and its heat production is estimated to
be from 13 (1) to 20 W (3). At rest, the carotid blood flow
590 0161-7567191 $1.50 Copyright
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (134.058.253.030) on December 1, 2018.
Copyright © 1991 the American Physiological Society. All rights reserved.
to prevent hyperthermia (27). To achieve this, estimated
from the blood flow and the specific heat value of blood,
it is sufficient for the arterial blood to be 0.3-0.5°C cooler
than the venous blood. Yet the proponents of human
selective brain cooling have never said that there was a
need for selective brain cooling during resting conditions
and normothermia. In that regard, the human species is
no different from the animal species in which the cooling
of the rete, by returning cool venous blood, takes place
only during hyperthermia (18, 19, 21, 23). The main
threat to brain homeothermia is the heat introduced into
the brain by the arterial blood during passive and exer-
cise-induced hyperthermia. In these cases, not only is the
arterial blood unable to cool the brain, but in addition, it
constitutes a thermal threat to the very survival of the
most heat-susceptible nervous tissue (5, 31). Each 1°C
increase in the arterial blood temperature brings -45 W
of heat to the brain. In that case, when the arterial blood
temperature is elevated, one may wonder how the brain
and the hypothesis of
selective brain cooling imply that this heat jeopardizing
the brain must be eliminated by the head itself.
The aim of this experiment was, therefore, to verify
this very point by measuring the amount of heat lost by
the head during rest and exercise under moderate envi-
ronmental conditions sufficient to create a mild hyper-
thermia [ +l”C in esophageal temperature (T,,)]. As far
as we know there is no information on global heat loss
from the heads of human hyperthermic subjects.
METHODS
Subjects. Four adult men in good health served as sub-
jects. Details of their physical characteristics are given in
Table 1. During the experimental sessions they were all
dressed in cotton jogging trousers, T-shirts, and training
shoes. All sessions took place in a climatic chamber be-
tween 9:00 and 11:00 A.M. The subject’s head was covered
with a plastic hood connected to a hose and a high-ve-
locity aspiration pump (Fig. 1). Another hose with a
pump was connected to a mouthpiece. The subject’s nose
was closed by a noseclip. The upper part of the body was
covered by a plastic cape to keep convective heat lost by
the trunk from flowing into the hood covering the head.
Protocol. Moderate hyperthermia was obtained by ex-
ercising the subjects on a cycle ergometer (Bosch, EGR
551). Three sessions were conducted, each at different
ambient temperatures (T,): 5,15, and 25°C. After 10 min
of rest on the ergometer, the subject started to pedal at 50
W work load. The work load was increased by steps of 15
0 1991 the American Physiological Society
 HEAT LOSS FROM THE HUMAN HEAD 591

TABLE 1. Characteristics of subjects used in study from sensors to computer was calibrated for sensible and
latent heat flows. An electrical resistor with known heat
Subject Weight, kg Height, cm Age, yr production was used as a heat source for sensible heat
JC 59.5 170 21
transfer. For latent heat transfer calibration, a moist
JCt 65 168 50 cloth was weighed before and after exposure under the
HH 81 186 29 ventilated hood. Latent heat loss measured by the sys-
GB 75 180 26 tem was 8% smaller than the amount computed from the
Mean + SD 70.1 + 9.7 176 ? 8.5 31.5 + 12.8 weight loss of the humid cloth. This precision was consid-
ered reasonably small and contrary to the hypothesis to
be checked. The sensible heat loss given by the sensors
min to 100 and then to 150 W. Each work load was main- and the system was 12% less than the known value from
tained for 15 min, a duration sufficient to reach a steady the heat source, which suggests that some heat was lost
state of the respiratory and circulatory functions and a by radiation and was not measured by the system. To
quasi-steady state of temperature regulation. A 5min verify whether heat was transferred to the environment
recovery (at 50 W) was performed by each subject after through the plastic hood and thereby lost from our re-
the heaviest work load. Altogether the muscular work cordings, we taped heat-flux disk sensors on the plastic
lasted 50 min. hood. The heat-flux read from the sensors was then ex-
Recordings. Latent (evaporative) and sensible heat loss trapolated to the whole surface of the hood. The results
from the surface of the head and from the expired air was showed that heat lost by radiation and convection to the
measured with an open-circuit method. Air was aspirated whole plastic hood and then to the environment varied
through the two hoses at a flow of 100 1. min-’ through from 2 to 5 W, dependent on T, but independent of work
the hose connected to the mouthpiece and at 250 1 min-’ l
load. The amount thus estimated explains why our read-
through the hose connected to the hood. The second aspi- ings were less than the actual heat lost by the head.
ration created a mild draft of -0.8 me s-’ over the head Again, this is contrary to the hypothesis to be checked.
and face surface. Rotameters (KL Engineering) mea- Therefore our methods provided conservative measure-
sured these airflows and fed the data into an Apple II ments of both respiratory and skin heat loss.
computer. Temperature and relative humidity were re- T,, was obtained from a thermocouple inserted 0.42 m
corded from thermocouples and psychrometers (Omega) beyond the naris (9,17,32). Tympanic temperature (T,)
placed in the air line before and behind the mouthpiece. was recorded from a copper-constantan thermocouple.
The temperature and humidity in ambient air and in air Accurate placing of the thermocouple was evidenced by
circulated over the head were also recorded. Ambient hu- acute pain, transmission of noise when touching the wire,
midity was measured with an evaporimeter (EPl, Ser- and, above all, checking that T,, at rest was equal to or
voMed, Sweden). From these parameters and barometric higher than T,, (4). The auditory canal was then filled
pressure, the latent and sensible heat loss from the ex- with cotton wool and taped. Forehead temperature (T,,)
pired air and from the head skin could be calculated. was measured with the use of a thermocouple taped to
The accuracy of the whole system of measurements the skin.

Plastic

FIG. 1. General diagram of experimental setup showing

subject connected to systems used to measure heat loss
from head skin and respiratory tract.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (134.058.253.030) on December 1, 2018.

Copyright © 1991 the American Physiological Society. All rights reserved.
592 HEAT LOSS FROM THE HUMAN HEAD

Rest with the use of a gradient calorimeter. The sensible heat

I loss was 35 W at 5OC T,, 24 W at 15°C T,, and 14 W at
25OC T,. At rest we obtained, at identical T,s, values of
29,35, and 12 W. Both series of measurements are there-
FOREHEAD
fore in the same order of magnitude. This confirms that
our methods were reliable: the airflow washed out most
33.5 - r of the sensible heat lost by the head, and the plastic hood
worked as a greenhouse to trap convective and radiative
33.0 - heat losses. However, at 15”C, there was a difference be-
ESOPHAGEAL tween Froese and Burton’s (14) value of 24 and our value
37.6 -
of 34 W. The explanation could be that one of our sub-
37.4 - J jects had, at this T,, an elevated sensible heat loss com-
pared with the other three subjects during this session.
37.2 - TYMPANIC Without this value the mean sensible heat loss at 15°C
.
f was 27 W, but we have not excluded this value because of
37.0 - its small influence on the total heat loss.
36.8 In this study, latent heat loss from the head skin was
always higher than the sensible heat loss, and this was
36.6 t true at every work load at all three T,s. The total heat
loss at 150 W reached from -200 to 250 W. This is a
200 r considerable proportion of the heat produced in the body
HEAD SKIN at this work load. If the heat production is equal to basal
150 metabolic rate (BMR) plus the heat produced during
/
work, this heat load can be estimated to be 70 W (BMR)
100 / plus three times the external work load because of the
RESPIRATORY efficiency of the biological engine. This gives a heat load
50 of -70 + 450 = 520 W. Our results indicate that, under
Ta=25”C the given circumstances (forced convection), two-fifths
(Subject: H.H.)
0 I I I I to one-half of this heat was lost from the head skin and
0 5 10~25303540455055 the respiratory tract. This result confirms that a power-
Time (min) ful heat sink exists on the head (29).
FIG. 2. Example of session at 25°C ambient temperature (T,) on Varene et al. (35) measured total respiratory heat loss
subject HH to illustrate protocol of sessions and recorded variables. in men at rest and found it to be on the order of 8 W. This
Temperatures were measured with thermocouples. Heat loss was com- is only one-third of our results. A small contribution to
puted on-line.
this difference could be found in the different conditions
RESULTS of the subjects. The subjects of Varene et al. (35) were
reclining, and it can be assumed that their ventilation
Figure 2 provides one example of a session and shows rate was lower than in our subjects, who were sitting on
the protocol of the sessions and the typical evolution of an ergometer. However, the main explanation might be
the variables measured during the experiment. The re- found in the different methods used in the two experi-
cordings are from one subject at 25OC T,. In Fig. 3 the ments. Varene et al. (35) were interested primarily in
total cephalic heat loss is given as a function of increas- comparing heat loss from the mouth vs. heat loss from
ing work load at the three different T,s. The total ce- the nose, and therefore measured air temperatures
phalic heat loss equals heat loss from the head plus heat within the natural cavities of mouth and nose. This may
loss from the respiratory tract. This total cephalic heat
loss was negatively correlated to the increasing T, at rest 300
and at 50 W, but the influence of T, during the two heavi-
est work loads tended to be reversed. The total cephalic
z F •zl 5°C TT
L 250
heat loss at 5, 15, and 25OC T, is plotted as a function of t!
T,, in Fig. 4. The results of cephalic heat loss at rest and 2 200
at 150 W are given in Fig. 5. Here the total cephalic heat I-
loss is divided into latent and sensible heat loss from the 4 150
head and latent and sensible heat loss from the respira- I
tory tract. The evolution of T,, and T,, during the ses- d 100
sions at 25°C T,, at which the highest body temperatures
were achieved in the present experiment, is shown in Fig. F 50
6. The temperatures are plotted as a function of time of
0
exercise performed.
0 50 100 150
WORK LOAD (W)
DISCUSSION
FIG. 3. Total cephalic heat loss (heat loss from head plus respira-
Heat lossfrom the human head. Froese and Burton (14) tory tract) at different ambient temperatures (5, 15, and 25°C) and
studied sensible heat loss from the human head at rest increasing work load. Values are means + SE; n = 4.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (134.058.253.030) on December 1, 2018.

Copyright © 1991 the American Physiological Society. All rights reserved.
 HEAT LOSS FROM THE HUMAN HEAD 593

300 r q Respiratory, sensible

0 Respiratory, latent
l Ta=S”C ggl Head, sensible
250 - n Head, latent

200 - 120 REST

100
z
3; 80

4 60
a 2
I- 40
O !i!
I- 20

0’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ m ’ ’ ’

300 WORK LOAD 150 W

300 r c
n 250
3 2
- 250 ; 200

E
0 200
g: 150

5 100
I
2 150 50
W
I 0
J 100 15 25
5
la (“C)
FIG. 5. Sensible and latent heat loss from head skin and respiratory
tract at rest and at work load equal to 150 W at 5, 15, and 25°C T,.
Values are means; n = 4. Note that 2 ordinate scales are different.

300 r those used for basal metabolic measurements. In addi-

tion, cerebral blood flow is shown to increase 31-58%
250 during moderate exercise (50% maximal 0, consump-
tion) (34). Such an increase would raise the amount of
heat entering the head from 45 to 70 W by each 1°C
increase in arterial blood temperature, provided that ve-
nous temperature does not change. To this amount the
heat production of the brain should be added. Therefore
a if the head was to lose heat at the rate of 90 W, brain
I- temperature could remain constant even if arterial blood
O 50 -
I- . temperature increased by 1°C and cephalic blood flow
w 1 ’ 1 ’ 1 ’ ’ ’ ’ ’ ’ ’ ’ 4 increased by 58%. Yet, the technique used for measure-
$6.8 37.0 37.2 37.4 37.6 37.8 38.0 38.2 ments of cerebral blood flow by Thomas et al. (34) gives a
Tes (“C) measure of cortical blood flow in a restricted region of
middle cerebral artery territory and must be used with
FIG. 4. Total cephalic heat loss (heat loss from head plus respira-
tory tract) plotted against esophageal temperature (T,,), taken as index
caution as evidence for an increase in total cerebral blood
of body’s heat load, at 5, 15, and 25°C T,. Lines show linear regression. flow during exercise.
r, Coefficient of correlation. At rest and at the lowest work load (50 W) when hyper-
thermia was not yet obtained, the total heat loss was
have minimized their results, because the temperature of
inhaled air was already high in these loci, thus reducing
the difference between inhaled and exhaled flows and, in
turn, the calculated heat loss. We were interested in the
total heat loss from the head skin and from the respira-
tory tract and therefore recorded inhaled air tempera-
ture outside the mouth. It remains that if the relative
difference between our results and those of Varene et al.
(35) remains large in relative values, in absolute terms
the difference is not excessive. 36.8 -
Heat balance of the human head. A constant body tem- 36.6 " " " " ' "
perature requires that the heat loss equals the heat gain 0 5 10152025303540455055
(heat balance). The total heat loss from the head was TIME (min)
71 + 30 (SE) W (25°C T,) at rest. This value may seem
FIG. 6. Mean (GE) esophageal and tympanic temperatures during
high compared with the BMR (-70 W), but the subjects whole experimental session at 25°C T,. Subjects rested on cycle ergom-
were not fasted and were not reclining, and the environ- eter from time 0 to 10 min. Work load was 50 W between 10 and 25 min,
mental conditions during these experiments were not 100 W between 25 and 40 min, and 150 W between 40 and 55 min; n = 4.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (134.058.253.030) on December 1, 2018.

Copyright © 1991 the American Physiological Society. All rights reserved.
594 HEAT LOSS
negatively correlated to T, (Fig. 3). A coefficient of corre-
lation of 0.99 (P < 0.01) between heat loss and T, was
obtained both at rest and at 50 W. The situation changed
during the two heaviest work loads (Fig. 3), at which the
heat loss tended to increase with increasing T,. The ori-
gin of this trend to reverse the influence of T, at heavy
work load can be found in the core temperatures, as illus-
trated in Fig. 4. When the total heat loss was plotted as a
function of T,,, an indication of the body’s heat load, a
high correlation between heat loss and T,, was obtained
at 15°C T, (r = 0.97) and 25OC T, (r = 0.99). The situa-
tion at 5OC T, was different. At that temperature no sig-
nificant correlation between total heat loss and T,, (r =
0.70) was obtained, mainly for two reasons. First, the
heat loss was already elevated at rest because of the low
T,, and second, the low T, prevented the T,, from in-
creasing to levels of real hyperthermia. The same corre-
lation occurred when heat loss from the head skin only
was plotted as a function of T,,. The r value was 0.76 at
5°C 0.96 at 15°C and 0.99 at 25°C. The corresponding r
values for respiratory heat loss as a function of T,, were
0.61,0.97, and 0.97. This result shows that the amount of
heat lost by the head and the respiratory tract was largely
a thermoregulatory response dependent on heat load. At
each T,, and particularly at the lowest T,s at rest, the
heat lost from the head skin was responsible for the larg-
est part of the total heat loss (Fig. 5). This is probably
because there was little or no vasoconstriction in the for-
ehead in response to the cold (15, 33).
We have little information on how or whether the heat
loss from the respiratory tract participates in cooling of
the brain (16). Caputa (9) has recorded T,,. at four verti-
cal levels in the esophagi of human subjects during exer-
cise and during passive heating at 0 and 30°C T,. He
found that hyperventilation decreased only the higher
T,,, and those by only 2°C at O°C T,. This indicates that
most of the respiratory heat exchange takes place in the
upper airways. This heat loss could be transferred to the
intracranial space because the venous anastomosis of the
pterygoid plexus with sinus cavernous is large and quite
functional (12). Yet, even without this contribution, the
heat lost by the head alone (120-130 W) could compen-
sate for an increase of arterial blood temperature of -l-
2°C depending on cephalic blood flow, plus the heat pro-
duced by the brain (15-20 W). In addition, the fact that
the heat loss capacity of the head is much larger than the
need suggests that a large amount of cool venous blood is
derived directly from the trunk via the external jugular
veins.
Selective brain cooling. The existence of a powerful
heat sink on the head itself is a necessary but not a suffi-
cient condition for selective brain cooling in humans.
Some opponents of selective brain cooling in humans
propose that there is no need for the head to function as a
heat sink because the arterial blood sufficiently cools the
brain (3). This argument might be true at rest but is not
valid during hyperthermia, especially if cerebral blood
flow is increased during exercise (34). If selective brain
cooling does not exist, the brain temperature has to re-
main higher than arterial temperature to permit heat ex-
change. However, the brain is heat sensitive and does not
tolerate high temperatures (5), and core temperatures as
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (134.058.253.030) on December 1, 2018.
Copyright © 1991 the American Physiological Society. All rights reserved.
FROM THE HtJMAN HEAD
high as 41.9”C in marathon runners (22) and 44°C during
therapeutic whole-body hyperthermia in patients (20)
have been measured.
Selective brain cooling has been demonstrated in
many species of mammals (dog, cat, sheep, antelope, and
reindeer), in which panting provides a powerful heat sink
at a short distance from the brain (2, 18, 19, 21, 23).
These species have a vascular heat exchange rete mira-
bile lying in lakes of venous blood, which use cooled ve-
nous blood from the nasal mucosa to reduce the tempera-
ture of the carotid arterial blood before it enters the
brain. However, neither the carotid rete nor panting is a
prerequisite for selective brain cooling. Rabbits have no
rete (10) and monkeys do not pant (2), yet they both have
selective brain cooling.
Humans neither have a carotid rete nor do they pant to
lose heat. During hyperthermia, when vasodilation and
sweating on the head skin occur, evaporation cools the
blood in underlying skin (27). The cooling effect is ele-
vated by air convection, as in the present experiment by
the aspiration of air over the head. In Fig. 2, a typical
evolution of T,, is shown; it can be seen that T,, reached a
plateau during the last 15 min of work, which indicates
that the cold resulting from the evaporation was not
“trapped” in the skin due to active vasodilation in the
forehead (13). If trapping were the case, a further reduc-
tion in T,, would have been seen. It may be assumed that
this cold was transferred elsewhere. One destination may
be to the intracranial space via the angularis oculi veins.
By the use of Doppler techniques it has been shown (11)
that facial venous blood entered the intracranium via
these veins during hyperthermia, whereas no flow or a
weak outflow could be recorded during hypothermia. So
far there is no information on the magnitude of venous
drainage of the skin that can enter the cranial cavity. An
indirect approach was brought by Nagasaka et al. (26),
who compressed the angulari oculi veins. The inflow of
cool venous blood via angular veins seems to be impor-
tant because T,, increased faster than T,, during com-
pression. However, selective brain cooling persisted dur-
ing compression (26).
A thermal countercurrent exchange between the cold
venous blood and the hot arterial blood entering the head
can take place partly within the cavernous sinus, which is
traversed by the internal carotid, and along the interface
separating internal jugular and carotid vessels. In addi-
tion, the same reversible venous flow was also recorded
in the mastoid and parietal emissary veins (6). Selective
brain cooling may take place through the whole external
calvaria, and there is no reason that some conductive
cooling should not also take place from the cool skin to
the brain. It has been shown that the temperature of the
human frontal dura was 1°C lower than T,, and was in-
fluenced by face fanning (5). The area where subcalvaria
temperature was recorded was not very rich in emissary
veins. The hypothesis that the cooling was due to direct
conductive heat transfer through the calvaria cannot be
discarded.
Provided that T,, is a good index of intracranial tem-
perature (4,8), a T,, lower than T,, during hyperthermia
is an indication of selective brain cooling. Our experi-
ment provides another example of T,, being lower than
 HEAT LOSS FROM THE HUMAN HEAD 595

T,, during hyperthermia only, as illustrated in Fig. 6. control in the human head, neck and upper chest. J. Physiol. Land.
During the last minute of exercise at 150 W the differ- 161: 298-312, 1962.
ence, A(T,, - T,,), was 0.40 t 0.14”C (SE). Our results 14* FROESE, G., AND A. C. BURTON. Heat losses from the human head.
J. Appl. Physiol. 10: 235-241, 1957.
are in agreement with those reported by Nielsen (28), 15 HERTZMAN, A. B., AND L. W. ROTH. The absence of vasoconstric-
who found T,, to be lower during outdoor bicycling than tor reflexes in the forehead circulation.
l
Effects of cold. Am. J. Phys-
during comparable indoor bicycling with no wind move- iol. 136: 692-697, 1942.
ment, whereas T,, attained the same level in the two situ- 16. HIRATA, K., T. NAGASAKA, AND Y. SUGANO. Effect of alternating
respiratory pathway on respiratory capacity, and tympanic and
ations.
forehead temperatures during exercise. J. Aerosp. Environ. Med.
In conclusion, it was found that in subjects exercising 15: 8-13, 1978.
at a mild hyperthermia the head provided enough heat 17. JAEGER, J. J., E. C. DEAL, JR., D. E. ROBERTS, R. H. INGRAM, JR.,
loss to dissipate the heat produced by the brain and re- AND E. R. MCFADDEN, JR. Cold air inhalation and esophageal tem-
ceived from the hyperthermic arterial blood. perature in exercising humans. Med. Sci. Sports Exercise 12: 365-
The intense heat loss capacity of the head should be 18 369, 1980.
JOHNSEN, H. K., A. S. BLIX, J. B. MERCER, AND K. D. BOLZ. Selec-
kept in mind when hyperthermic patients are treated. ’ tive cooling of the brain in reindeer. Am. J. Physiol. 253 (Regulatory
The head can be used as a powerful heat sink, provided Integrative Comp. Physiol. 22): R848-R853, 1987.
the patient is freely evaporating. Whether or not the ther- 19. JOHNSEN, H. K., AND L. P. FOLKOW. Vascular control of brain
apist believes in selective brain cooling, even a moderate cooling in reindeer. Am. J. Physiol. 254 (Regulatory Integrative
Comp. Physiol. 23): R730-R739, 1988.
air convection, as in the present experiment, can provide 2. KISCH, R., AND D. SCHMIDT. Erste experimentelle und klinische
a heat sink capable of returning a patient rapidly to nor- Erfahrungen
l
mit der Ganzkorper-Extrem-Hyperthermie. In: Ak-
mothermia. tuelle Probleme aus dem Gebiet der Cancerologie. Berlin: Springer-
Verlag, 1966, p. 53-70.
This work was supported by Defence and Civil Institute of Environ- 21* MAGILTON, tJ. H., AND C. S. SWIFT. Response of veins draining the
mental Medicine, Canada. nose to alar-fold temperature changes in the dog. J. Appl. Physiol.
Address reprint requests to M. Cabanac. 27: 18-20, 1969.
22. MARON, M. H., J. A. WAGNER, AND S. M. HORVATH. Thermoregula-
Received 18 April 1990; accepted in final form 28 March 1991. tory responses during competitive marathon running. J. Appl.
Physiol. 42: 909-914, 1977.
23. MERCER, J. B., H. K. JOHNSEN, A. S. BLIX, AND R. HOTVEDT.
REFERENCES Central control of expired air temperature and other thermoregula-
1. ASCHOFF, J., AND R. WEVER. Kern und Schale im Warmehaushalt tory effecters in reindeer. Am. J. Physiol. 248 (Regulatory Integra-
des Menchen. Naturwissenschaften 45: 477-485, 1958. tive Comp. Physiol. 17): R679-R685, 1985.
2. BAKER, M. A. A brain-cooling system in mammals. Sci. Am. 240: 24. MITCHELL, D., H. P. LABURN, M. J. M. NIJLAND, Y. ZUROVSKY,
130-139, 1979. AND G. MITCHELL. Selective brain cooling and survival. South Afr.
3. BRENGELMANN, G. I,. Brain cooling via emissary veins: fact or J. Sci. 83: 598-604, 1987.
fancy? (Commentary on Falk). Behau. Brain Sci. 13: 349-350,199O. 25. NADEL, E. R. Letters: comments on “Keeping a cool head.” News
4. BRINNEL, H., AND M. CABANAC. Tympanic temperature is a core Physiol. Sci. 2: 33, 1987.
temperature in humans. J. Therm. Biol. 14: 47-53, 1989. 26. NAGASAKA, T., H. BRINNEL, K. HIRATA, Y. NODA, AND N. SUGI-
5. BRINNEL, H., M. CABANAC, AND J. R. S. HALES. Critical upper MOTO. Increase in venous flow through ophthalmic veins enhances
levels of body temperature, tissue thermosensitivity and selective selective brain cooling in hyperthermic humans. In: Thermal Physi-
brain cooling in hyperthermia. In: Heat Stress. Physical and Envi- ology 1969. Amsterdam: Excerpta Med., 1989, p. 205-210.
ronment. Amsterdam: Excerpta Med., 1987, p. 209-240. 27. NAREBSKI, J. Human brain homeothermy during sleep and wake-
6. CABANAC, M., AND H. BRINNEL. Blood flow in the emissary veins of fulness: an experimental and comparative approach. Acta Neuro-
the human head during hyperthermia. Eur. J. Appl. Physiol. Occup. biol. Exp. 45: 63-75, 1985.
Physiol. 54: 172-176, 1985. 28. NIELSEN, B. Natural cooling of the brain during outdoor bicycling?
Pfluegers Arch. 411: 456-461, 1988.
7. CABANAC, M., AND M. CAPUTA. Natural selective cooling of the
29. NUNNELY, S. A., D. C. READER, AND R. J. MALDONADO. Head-tem-
human brain: evidence of its occurrence and magnitude. J. Physiol.
perature effects on physiology, comfort, and performance during
Land. 286: 255-264, 1979.
hyperthermia. Aviat. Space Environ. Med. 53: 623-628, 1982.
8. CABANAC, M., M. GERMAIN, AND H. RRINNEL. Tympanic tempera-
30. SHIRAKI, K., S. SAGAWA, F. TAJIMA, A. YOKOTA, M. HASHIMOTO,
tures during hemiface cooling. Eur. J. Appl. Physiol. Occup. Physiol.
AND G. I,. BRENGELMANN. Independence of brain and tympanic
56: 534-539, 1987.
temperatures in an unanesthetized human. J. Appl. Physiol. 65:
9. CAPUTA, M. Thermal gradients in the esophagus of man during
482-486, 1988.
exercise and passive warming. J. Therm. Biol. 5: 249-251, 1980.
31. SMINIA, P., J. HAVEMAN, AND B. W. ONGERBOER DE VISSER. What
10. CAPUTA, M., W. KADZIELA, AND J. NAREBSKI. Significance of cra-
is a safe heat dose which can be applied to normal brain tissue? Int.
nial circulation for the brain homeothermia in rabbits. II. The role
J. Hyperthermia 5: 115-116, 1989.
of the cranial venous lakes in the defense against hyperthermia. 32. SNELLEN, J. W. Body temperature during exercise. Med. Sci.
Acta Neurobiol. Exp. 36: 625-638, 1976. Sports Exercise 1: 39-42, 1969.
11. CAPUTA, M., G. PERRIN, AND M. CABANAC. Ecoulement sanguin 33. STEEGMANN, A. T., JR. Human facial temperatures in natural and
reversible dans la veine ophtalmique: mecanisme de refroidisse- laboratory cold. Auiat. Space Enuiron. Med. 50: 227-232, 1979.
ment selectif du cerveau humain. C. R. Acad. Sci. Paris Ser. D 287: 34. THOMAS, S. N., T. SCHROEDER, N. H. SECHER, AND J. H. MIT-
1011-1014, 1978. CHELL. Cerebral blood flow during submaximal and maximal dy-
12. DANIELS, D. I,., L. F. CZERVIONKE, J. F. BONNEVILLE, F. CATTIN, namic exercise in humans. J. Appl. Physiol. 67: 744-748, 1989.
L. P. MARK, P. PECH, L. E. HENDRIX, D. F. SMITH, V. M. 35. VAR~NE, P., L. FERRUS, G. MANIER, AND J. GIRE. Heat and water
HAUGHTON, AND A. L. WILLIAMS. MR imaging of the cavernous respiratory exchanges: comparison between mouth and nose
sinus: value of spin echo and gradient recalled echo images. Am. J. breathing in humans. Clin. Physiol. Oxf. 6: 405-414, 1986.
Neuroradiol. 9: 947-952, 1988. 36. WENGER, C. B. More comments on “Keeping a cool head.” News
13. Fox, R. H., R. GOLDSMITH, AND D. J. KIDD. Cutaneous vasomotor Physiol. Sci. 2: 150, 1987.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (134.058.253.030) on December 1, 2018.

Copyright © 1991 the American Physiological Society. All rights reserved.