Heat Loss and Blood Flow of the Feet Under
Hot and Cold Couditionsl
LOIS H. LOVE,
From the Department
of Physiology,
of Pennsylvania,
Philadelphia,
Penmylvania
University
A
Received for publication
February
17, Ig48.
l Presented
to the Graduate
School of Arts and Sciences of the University
of Pennsylvania,
in partial fulfillment
of the requirements
for the degree of Doctor
of Philosophy.
Part of the
expense of this investigation
was defrayed
by a grant from the Life Insurance
Medical
Research
Fund to Dr. H. C. Bazett.
20
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
LTHOUGH THE IMPORTANCE of the extremities
in thermo-regulation
is well established little has been done to follow progressive changes in these
areas during relatively long exposures to heat and cold. There are available
the results of Burton (I) showing a gradual increase in finger blood flow
during several days of exposure to heat and a gradual decrease during exposure to cold. In this same paper it is shown that the superficial veins of the
forearm become fully dilated in the heat and fully constricted in the cold only
after some time. There are also observations on the amount of convective
and radiant heat loss from the hand which show that the proportion of heat
lost by the different pathways can change during the course of exposure (2).
The experiments to be reported here were designed to extend these
relatively long-range observations.
Measurements were limited to the foot
and included the evaporative and non-evaporative heat loss, blood flow and
skin temperature of subjects living in a controlled temperature room for
periods up to two weeks. Except for several short experiments done to confirm part of the results the same two subjects were used throughout.
Only
two room conditions (approximately
33OC. and ZIOC.) were used and the
subjects were exposed to both of these temperatures during the summer and
again during the winter.
Blood flow measurements were made by the venous occlusion technique.
The heat exchange of the foot was obtained by the use of a calorimeter similar to that described by Forster, Ferris and Day (3) for use on the hand.
The two were not combined into one instrument as was done by Forster et al.
but the calorimetric measurements were made on one foot, while the plethysmograph was used on the other. Such an arrangment has the disadvantage
that the two feet were not sub jetted to the same conditions due to the sup-
JdY
BLOOD
I948
AND
HEAT
EXCHANGE
IN
THE
FEET
21
pression of vaporization and the absence of air movement in the plethysmograph. In spite of this, the arrangement used was preferred because otherwise blood flow measurements could not have been interspersed at will with
calorimetric determinations.
Stopping the air movement in the calorimeter
to measure blood flow would cause a rise of temperature of such extent that
further calorimetric determinations
could be made only after the air circulation had been restored for some time. In view of the large variation in
blood flow during any one day many observations were necessary. It was
thought best to obtain these simultaneously with the calorimetric data even
though it necessitated exposing the two feet to different conditions.
OF
CONSTRUCTION
AND
USE
OF
THE
APPARATUS
The Calorimeter.
This was a double-walled
copper vessel (capacity, 7.5 liters)
shaped like a boot. Air entered through two openings at the level of the toes and left
by a single opening at the back of the ankle. The volume rate of air flow was 50 liters/
The effective air
minute.
The estimated linear air velocity was 2 to 3 meters/minute.
movement was probably greater due to turbulence.
The foot was sealed into the calorimeter and cork stops kept it from touching the metal wall.
The calorimeter was suspended in air on a framework.
The air movement in the
Standardization
was
room was constant and was estimated at about 35 meters/minute.
accomplished with an electrically heated coil fitted into a boot. The heat loss to the air
was measured with a thermocouple of high sensitivity with one junctio’n in the inlet air
stream and one in the outlet.
A similar thermocouple was used to measure the thermal
gradient across the walls. The constant relating this gradient to the heat loss across
the walls was determined empirically.
A series of measurements over a period of a year
with different room conditions gave estimates of the heat recovered which ranged from
gg to 104 per cent of the heat input.
Evaporative
heat loss was determined gravimetrically.
For most purposes the
evaporative heat loss has been calculated from the mass of water lost using the constant
for the heat of vaporization at the average surface temperature of the skin. For more
exact purposes allowance must be made for the expansion of the water vapor to the volume
corresponding to the relative humidity attained in the calorimeter, as has been emphasized by Murlin and Burton (4) and Hardy and DuBois (5). Calculations of this type
have been made on the data only in comparisons of the insulating value of air at different
temperatures to make certain that apparent differences were not due to such factors.
In making these corrections the cooling of water vapor from the temperature of
the skin to that of the calorimeter has been considered negligible.
The expansion of the
water vapor is not negligible, since the relative humidity of the calorimeter was estimated
sometimes to be as low as one per cent, but the amount of heat absorbed by the gas in
The maximal amount of heat that
this expansion cannot be calculated with certainty.
could be absorbed is that involved if the process took place very slowly against an external
pressure which at all times was only inf?nitesimally less than the gas pressure (i.e., in a
thermodynamically
reversible isothermal expansion).
In this case the heat absorbed
would be equal to p/v or RT In pl/pz. The minimal heat absorbed is that involved on
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
DETAILS
22
LOIS H. LOVE
Voltme I
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
the assumption that the expansion of the gas takes place against a pressure at all times
equal to the final pressure. In this case the heat absorbed is equal to paV. At very
low humidities there is considerable difference between the heats calculated in these
two ways. For example, if the relative humidity of the calorimeter was one per cent and
the temperature
27OC., the upper limit is 0.153 Cal/gram and the lower limit 0.033
Cal/gram.
No information is available to determine what value should be used between
these limits, so that both corrections have been calculated, whenever the correction has
been applied.
The Plethysmograph.
The plethysmograph was a copper vessel of 3.5 liters capacity
shaped roughly like a boot. The plethysmograph and calorimeter were the same height.
It was air filled and air transmission was used throughout.
Optical records from a
Frank capsule were obtained on paper moving at 1.7 cm/second.
Calibrations
were
made after each experiment by introducing measured amounts of water. This was done
with the foot in place and the circulation occluded.
The excursion was linearly related
to the amount of water introduced
&about
5 per cent) and there was no overshoot
unless the water was introduced at a rate in excess of the largest blood flows.
No water bath was used since it was found that changes in the base line were so slow
as to be insignificant in the few seconds required for a measurement.
The occlusion pressure was IOO mm. Hg. A similar pressure was found to be necessary for the dependent foot by Abramson et al. (6). The apparent venous pressure was
regularly found to be 30 to 40 mm. Hg. The artefact produced by inflating the cuff was
recorded.
Its maximum duration was determined by the method of Wright and Phelps
(7) to be one second. The blood flow was measured from the first pulse after the end of
the artefact time, for, as has been found by Wright and Phelps (7) and Christensen and
Nielsen (8), the pressure rise after venous occlusion is often not linear.
Skirt Temperattire Measurements and their Probable Accuracy.
Skin temperatures
were obtained with 36-gauge copper-constantan
thermocouples of low sensitivity which
could be read to =t o.IOC. They were attached to a single Kipp and Zonen galvanometer
through a selector switch. Recording was photographic.
They were held to the skin
with adhesive or Scotch tape. Four were used on each foot. The locations were as
follows: I) dorsal surface of the middle phalanx of the middle toe; 2) the center of the
sole; 3) the Tendo-Achilles
at the back of the ankle; and 4) the outside of the ankle just
below the external malleolus.
These positions were selected as a result of experiments in
which the foot was divided into eight transverse segments with two thermocouples
in
each segment. The arithmetical mean of the four points designated was found to agree
closely with the average skin temperature determined with the 16 thermocouples under
conditions in which the temperatures covered a range between 17.6”C. for the toe and
26.5OC. for the back of the ankle (9).
In some later experiments the skin temperature was measured with a I3c-cm.
length of 3%gauge Hytemco wire which was used as a resistance thermometer.
It was
wrapped in two loops parallel to the long axis of the foot and cemented to the skin with
a band of cement which was not more than 5 mm. wide. This arrangement should give
a better average surface temperature since evaporation was suppressed in only a narrow
zone over which a large thermal gradient was less likely than with the thermocouple
coverings.
A series of simultaneous measurements was made by the two methods.
In the cold the resistance thermometer gave values which averaged o.8”C. below the
thermocouple values. In the heat this difference increased to 1.3Oc.
July 1938
Routine
below :
BLOOD
AND
of the Experiments.
HEAT
EXCHANGE
The characteristics
IN
THE
FEET
23
of the two subjects are given
Subject R
Subject
Y
21 years
Age.......................................
23
Height. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . 178
160 cm.
Weight....................................
51.1
kgm.
59-o
Surface area (DuBois formula) . . . . . . . . . . . . . . .
1.51
m2
I-74
1200
cc.
Average foot volume in heat. . . . . . . . . . . . . . . . . 1400
1100 cc.
Average foot volume in cold.. . . . . . . . . . . . . . . . . 1300
Average foot area (IO). . . . . . . . . . . . . . . . . . . . . . .
0.082
0.075 m2
Throughout
the experiments the subjects wore shorts, undershirt and sandals.
RESULTS
The blood volume data from these experiments have been given by
Increased values were found under warm conditions.
Spealman et al. (I I).
There were three periods of heat (17 days) which allow valid comparisons
of the averages of the different periods. Any differences in these averages
can be interpreted as evidence of acclimatization
to heat which would be
expected to be greatest during the first exposure in the summer and least
during the winter. Two periods (8 days) were spent in the cold but since
the second exposure lasted only z days no valid comparisons between the
periods can be made. The values obtained in the cold will be used only to
contrast the levels found in the heat.
Daily average values are given in table I and in figure I. Average
values for entire periods are compared in table 2.
Fluctuations in blood flow during any one day were large in the heat
(commonly amounting
to changes of IOO per cent) but were small in the
cold. Variations in the other measurements were small and no trends were
found to indicate that insufficient time was allowed for steady conditions to
be reached.
Levels of Heat Exchange in the Heat and Cold. The measured basal
under both conditions and 45
metabolism
was 36 to 42 Cal/m2/hr.
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
Each day began at about 7 a.m. At approximately
8 : 30 a.m., after a basal metabolism
test, the first subject moved to a nearby chair in which he sat while measurements were
made on his feet. Preparation
of the subject took about 45 minutes.
An additional
30 minutes elapsed before records were made. Twenty or more records of blood flow and
about IO observations of temperatures were made during the next period of about IOO
The morning experiment ended about 10:30 and was followed by breakfast.
minutes.
A light lunch was served at I :oo p.m. and shortly after this the experiment was repeated
on the second subject.
Subjects were alternated so that a fasting determination
was
made on each subject every other day.
Variations
in the room temperature
during the hours of a single experiment
amounted to less than 1°C. except on the afternoon of July 21 when a breakdown of the
air-conditioning
unit occurred.
LOIS
24
H. LOVE
Vohne I
Cal/m2/hr.
has been assumed as the probable metabolic level when the
subject sat up in the fasting morning experiments.
The area within the
calorimeter was about 5 per cent of the total body surface and lost 56
Cal/ms/hr. in the warmth and 29 in the cold. This implies that this area
lost 6.5 per cent of the total heat in the hot condition and 3.5 per cent in the
cold. Of the heat lost from the foot 51 per cent was evaporative loss in the
too
75
-
SUBJCCT
SUBJECT
Yd
R
COLD
SUMMER
HOT
COLD
HOT
WINTER
HOT
COLD
SUMMER
HOT
COLD
HOT
WINTER
of the
Fig. I. DAILY AVERAGE
VALUES
for subjects Y and R. Time of year and condition
room are indicated
at the bottom.
Uppermost
curve represents blood flow in cc/min.,
the middle
two curves the mean surface temperatures
of the feet (the warmer
being that in the plethysmograph) in “C., and the lower total and non-evaporative
heat loss in Cal/m2/hr.
heat and 27 per cent in the cold. Since the subjects were weighed at regular
intervals and a complete balance sheet of ingesta and excreta was kept,
comparisons of water loss may be made with that from the body as a whole
(see table 3). In both heat and cold the water loss from the foot often
heat losses can
equalled in intensity that from the body. Non-evaporative
be compared less accurately on the basis of the assumed heat production.
In the heat non-evaporative
loss from the body could have been at most 15
Cal/m2/hr., while that of the foot was 27. In the cold non-evaporative loss
from the body must have been at least 35 Cal/m2/hr. while that of the foot
was 22.
Blood flow in the heat was 54 cc/mm; it fell to 7 in the cold. Each cc.
of blood lost I to 2 calories in the heat and 4 to 7 in the cold. Since the
warmer foot was in the plethysmograph,
the flow of the other foot may have
been somewhat less, and the heat loss per unit volume of blood may be somewhat underestimated.
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
0
hOT
.7uly 1948
BLOOD
AND
HEAT
EXCHANGE
IN
THE
FEET
25
TABLE
I.
AVERAGEVALUESFOREACHDAY~
ROOM
D.B.
Subject
“C.
32.8
3w
33-I
33-3
33-I
33-3
OC.
27.8
28.0
28.6
28.6
28.6
28.0
OC.
cc/T?&.
35.8
34.1
36.1
37*7
36.7
40
40
42
35-6
36-S
356
8
36.9
36.8
72
20.3
8
21.1
14.9
14.5
25.1
26.9
5
g2
IS.7
16.3
15.8
25.8
24.8
112
20.8
20.8
20.3
12
20.3
153
7
4
5
8
I32
I4
32.2
27.5
26.4
28.3
28.0
IO
IS2
16
33.6
32.2
32.8
OC.
38
40
44
35.3
35.6
24*9
26.1
23.5
25.9
25.1
23.3
24.1
24.8
Y
OC.
34*5
33.8
36-9
36.2
Cal/d/hr.
67
72
61
76
37-o
36.0
$2
58
21.9
24
37
28
23-4
22.7
21.3
yo of total
21
32
24
29
23
22
-.
I5
26
20
21
22.2
29
28
23.1
25
18
21
czo.
Cd/CC.
0.63
0.28
69
56
61
2.3
2.3
1.8
62
2.4
0.53
0.31
56
62
I-7
1.8
0.67
0.42
38
30
3*8
9-6
29
28
25
28
5.1
9.3
7-2
4.0
1.3
1.6
0.49
o-44
I*4
1.6
0.51
--
I.05
1.08
I
o-49
0.82
1.24
-36.4
37.0
36.5
36.4
47
50
50
52
3595
36.2
36.2
50
28
36.2
24
35-7
34-9
35 09
35.3
63
53
61
44
62
30
23
44
62
0.36
--
12
20.6
I$.0
2
24.4
24.1
7
7
23.1
23.2
23*3
21.5
23
26
19
20
17
23
0.60
3.7
5-I
0*57
--
32
4
s2
6
72
8
g2
33.3
28.3
36.2
36.8
36.6
36.0
36.6
36.4
35.4
57
39
35-7
36.0
29
37
48
48
43
36.2
35-s
36.0
36.3
36.1
3595
36.0
36.0
44
61
31
34
51
36.6
37*3
36.9
56
49
62
35
36
31
37
36-s
47
31
30
44
31
1.0
0.28
I*9
0.29
2.1
0.32
36
37
40
34
I-9
I*3
I.5
1.3
0.19
o-34
0.31
I 0.35
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
The temperature distribution
on the foot differed in the two conditions.
In the cold the toe had the lowest temperature.
In the heat the gradient
was reversed and the toe became the warmest point at a level about o.&.
above the average.
Non-evaporative
heat losses of all types may be grouped for rough purposes and the thermal insulation value of air be calculated from the ratio of
the temperature difference to heat transfer.
Such values, expressed in Clo
units (12), are shown in table 2 and indicate striking differences between the
hot and cold conditions.
The levels in the summer and winter differ somewhat (possibly because of slight differences in the room conditions since the
relationship of air and wall temperatures was certainly different) but the
26
LOIS
H. LOVE
Volume
I
Szlbject R
OC.
32.8
32.2
7
g2
9
20.5
OC.
34.9
35.0
36.6
35-S
36.9
--
:c/?ni?t.
33.3
34-3
36.4
359
35*5
36.2
25.0
8
23*9
22.1
7
6
6
8
25.0
22.6
23.2
24.4
22.8
22.0
20.5
II
122
20.3
20.3
13
142
15
16~
33.3
33.9
3393
32.3
27-S
26.4
28.3
28.0
36-3
36.8
36-7
36.2
I
20.6
15.0
24.9
25.4
--
6
36.8
36.8
87
60
36-9
36.4
36.9
36:s
36.3
99
74
92
22
3
42
5
62
7
8"
9
33*3
28.3
s
65
5
85
65
67
66
60
22.2
63
27
32
23
31
. ---
.
I5
30
23
26
8
17
=7
23
I4
22
21.6
23.4
21
12
39
30
35.3
35-4
35.2
35.3
36.0
36.2
35.5
35.2
61
18
54
II
55
58
22
23.4
23.2
22.0
35.8
35.9
36.4
35-9
36.1
36.0
35.8
35-7
35.9
37-4
36.8
37-4
36.9
36-s
8
53
69
CaJ/nG/hr
65
27.4
25.4
26.5
24.4
27.8
IO2
20.5
OC.
33*7
3492
353
353
14.9
14-S
157
16.3
15.8
15*3
20.5
“C.
32
49
53
69
36
54
25
yo of total
Cal/cc.
Clo.
0.19
55
2.0
0.20
65
I-7
I-3
o-34
0.26
57
88
73
2.5
1.6
0.70
37
28
38
29
43
23
o-95
0.91
0.91
0.81
0.99
0.76
7=
80
60
57
0.47
0.62
o-33
0.52
18
26
21.2
66
62
67
47
74
67
66
1 The first 16 days are those of the summer experiment;
experiment.
2 Values obtained
in the morning
with the subject fasting.
32
36
35
34
36
29
27
52
1.0
0.27
42
48
28
I-4
0.9
0.8
0.25
51
1.0
57
59
1.0
the last 9, those
I*3
o-33
0.26
0.28
0.37
o-37
of the winter
ratio of the apparent insulation in the cold to that in the heat is constant
(2.1 in the summer and I .9 in the winter).
The validity of this difference
will be discussed later.
DiJkrences Between the Three Periods of Exposure to Heat. Table 3
shows that the evaporative heat loss of the foot was considerably reduced by
previous exposure to cold even when the exposure was as short as the 6-day
interval between the two periods of heat in the summer. A similar but
smaller reduction can be seen in the q-hour loss of the whole body. This
decreased ability to sweat was accompanied (and probably partiallycom-
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
3393
33*3
33.3
32.8
OC.
27.8
28.0
28.6
28.6
28.6
28.0
July
BLOOD
1948
AND
HEAT
EXCHANGE
IN
THE
FEET
27
pensated) by a raised skin temperature and an increased rate of non-evaporative heat loss.
Blood flow measurements also show differences between the periods of
heat (table 2). In subject R, the difference in flow between any two of the
three periods of heat was statistically significant, as was the smaller increase
shown by subject Y for the summer periods. The amount of heat lost by
each volume of blood under warm conditions was reduced after exposure to
cold; both subjects are consistent in indicating this change.
TABLE
2.
AVERAGE
VALUES
Heat I, Summer ........
Heat II, Summer .......
Heat, Winter.
.........
Cold,
Cold,
Winter.
Summer.
.........
........
32.9
32.7
3393
36-s
36.6
20.6
20.7
24.3
25.6
36.4
41
47
43
“C.
Subject
Heat I, Summer..
. . . . . . 33 .o
Heat II, Summer .......
33.2
.........
33.3
Heat, Winter.
Cold,
Cold,
Winter.
Summer.
.........
........
36.0
36.5
36*7
49
60
80
70 of
Cal/d/hr.
total
I 25 i
Cal/cc.
61
2.0
54
37
1.6
23.2
20
24.5
29
4-4
6.0
35-6
3596
36.0
7
6
PERIODS
64
57
54
20
1
39
31
34
/
20
--
I.5
-
R
34.8
63
20
43
68
1.9
0.34
35.3
36.0
57
64
I9
33
38
31
67
48
I.4
1.1
o-49
0.30
20.6
25.2
6
33
6.8
0.56
20.4
26.1
7
3=
5-9
o-89
Evidence of Acclimatization Dwing any One Period. From the appearance and attitude of the subjects it was apparent that their condition improved during each of the longer exposures but this improvement
was not
The only gradual change which
reflected in any of the variables tabulated.
could be demonstrated in the heat was that the toe regularly became the
warmest part of the foot. In most of the exposures this rise in toe temperature relative to that of the rest of the foot took several days to develop. In
the cold the toe was regularly the coldest of the four points measured with
a single exception.
This exception occurred on the first morning of one of
the exposures to cold.
DISCUSSION
The various levels obtained can be compared with the results obtained
on the hand by Forster et al. (3). The hand experiments in which the sur-
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
cc/min.
FORENTIRE
28
LOIS
H. LOVE
Volume
1
TABLE
3.
AVERAGEVALUESFOREVAPORATIVE
SWJECTY
PERIOD
Night.
los~o~;lre
...........
Heat I, Summer.
Heat II, Summer ............
Heat, Winter.
..............
..............
Cold, Winter.
Cold, Summer.
.............
50
54
24hr.
entire
loss,
body
72
6G
WEIGHTLOSS~
SWJECTR
Foot loss
63
46
5
56
IO
53
34
8
5
II
I4
Night
entire
50
48
loss,
body
2Chr. loss,
entire body
Foot loss
47
58
55
46
IO
12
54
I9
IO
12
16
75
66
1 All figures are in grams/m2/hr.
An approximate
correction
for the loss of water vapor
from the respiratory
tract and for the difference
in weight of the respiratory
gases has been
applied to the figures for the body.
The correction
used was 6 grams/m2/hr.
in the cold, and
IO grams/m2/hr.
in the heat.
The total heat loss for the cold foot averaged 28 Cal/m2/hr. as compared to 21 for the hand (assumed hand area 0.05 m2). Corresponding
values in the heat were 55 and 61. However, there is a considerable difference in the amount of heat lost over the various pathways.
The foot lost
54 per cent of its heat by vaporization in the heat and 30 per cent in the cold.
In contrast to this the evaporative heat loss from the warm hand was 72 per
cent of the total and 74 per cent for the cold hand. The probable reason
for this difference is that the volume rate of air flow in the hand calorimeter
was only one-fifth of that used on the foot. This probably would limit the
convective heat loss from the hand more than the evaporative loss and so
alter the ratio.
Benedict and Wardlaw (13) have reported that in comfortable conditions the rate of water loss from a unit area of the feet is greater than that
for the entire body. In contrast, during profuse sweating the evaporative
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
face temperature fell within the limits found for the foot were used for this
cc/min. for
comparison.
In the cold the average blood flow was 1.5 cc/100
the hand and 0.5 for the foot. The flow in the warm foot (5 cc/100 cc/mm)
was considerably less than that of the warm hand (16 cc/100 cc/min.).
The
figures for the heat lost by each volume of blood are in good agreement with
the results on the hand, as are the probable temperatures of the blood entering the arteries. These have been estimated (as were those of Forster et al.)
on the assumption that the blood leaves the foot at a temperature equal to
the average surface temperature.
It must be concluded that in the heat
the arterial blood entered the foot at a temperature only slightly below that
of the rectum, while in the cold the incoming blood must have been in the
neighborhood of 3oOC.
J&Y w8
BLOOD
AND
HEAT
EXCHANGE
IN
THE
FEET
29
I. Air insulations
for an electrically heated boot were determined at different room
temperatures.
The insulation was found to be at most IO per cent greater in the cold
than in the heat. These results show that the phenomenon was not dependent on physical factors in the calorimeter nor to errors in the estimation of calorimeter temperature
from measurements of room temperature.
2. The observed differences cannot be due to errors in calculation
of the evaporative
heat loss. Both corrections for the heat absorbed during the expansion of the water
vapor, as previously described, have been calculated. The non-evaporative
heat recovered has been corrected by this amount on the assumption that all of the heat absorbed
by the vapor comes from the air. Both corrections lower the air insulation but do not
decrease the ratio of the air insulation in the cold to that in the heat. The uncorrected
ratio is 1.9. With the minimum correction it becomes 2.0 and with the maximum correction, 2.3.
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
loss of the feet was found to be less than that of other areas of the body
(14, IS).
The figures given in table 3 show that under the conditions of these
experiments, in which the foot was exposed to dry air and the rest of the
body to moist air, such differences were not apparent.
An essential factor was the reduction of the ability of the foot to sweat
in the heat after exposure to cold. This is in accord with the results of
Adolph (16) for the entire body. The more marked reduction in the loss
from the foot than from the entire body may indicate that the regional differences described by Kuno and Weiner are more marked, or appear at a
lower temperature, in the winter than in the summer.
The other differences between the three periods of heat are all probably
related to this decreased ability to sweat following an exposure to cold.
Thus, the raised skin temperature would result from the decreased evaporative heat loss and would lead to an increased rate of non-evaporative heat
loss. The decreased cooling of the blood would result from the decreased
temperature gradient from blood to skin. The increase in blood flow, which
was conspicuous in one subject and partially present in the other, could
be adequately explained as a compensatory mechanism used to keep up
the nonevaporative
heat loss in spite of the decreased cooling of each
volume of blood.
Although no consistent changes in blood flow could be demonstrated
during any one exposure to heat or cold, the increase in toe temperature
relative to that of the rest of the foot in the heat strongly suggests that the
toes have a gradual increase in flow similar to that found by Burton (I) for
the finger.
The marked difference in the apparent air insulation for the hot and
cold conditions was unexpected, for, as is pointed out by Burton (I ;I), the
air insulation should change only a very small amount with temperature.
However, the following considerations suggest that the difference is real.
30
LOIS
H. LOVE
Volume r
Since it has not been possible to find an error in measurement or a change
in the physical properties of the system which would account for the difference in the apparent air insulation between the hot and cold conditions it is
suggested that the change is real and has a physiological basis.
Several explanations are possible. First, there is the familiar roughening of the skin, particularly of the exposed areas, which occurs in cold
weather. This would increase the still air trapped around the skin and thus
decrease the convective heat loss. The idea that such roughening is an
active process occurring at temperatures below that at which ‘goose flesh’
becomes obvious is supported by the observations on the unusual smoothness of the skin after sectioning a cutaneous nerve (IS>, and after the injection of novocaine into the region of such a nerve (19).
An additional explanation is provided by the effects of curvature on
heat loss. This has been considered by Van Dilla (20) with respect to the
problem of clothing insulation and the treatment was extended by Burton
(21)
to include the air insulation around curved, insulated surfaces. Such
a factor is well known to engineers. Heilman (22) gives the heat loss from
bare iron pipes of different diameters, and equations relating the heat loss
to the diameter of curved surfaces are given by Rice (23), who has summarized the various experimental results.
The foot and toes can be treated very roughly as a series of cylinders
with the toes having approximately 15 per cent of the entire surface area.
The effect of the curvature factor is such that with an equal temperature
gradient from the skin to the air more heat must be lost from the toes than
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
3. A possible source of error is the use of average air temperature
for ambient
temperature instead of a suitable compromise between air and wall temperatures.
Although the inner wall temperature was not measured, the thermal gradient across the
walls happened to be the same under both conditions, so that the heat transfer and the
thermal gradient from the outer wall to the room air must also have been identical.
Analysis along these lines shows that the difference between the average air temperature
in the calorime ter and the wall temperature in the two conditions could not explain the
results.
4. The skin temperature
measurements were subject to a known source of error
produced by covering the thermocouples and thus reducing the heat loss from the covered
area. This error should be greater in the heat and, therefore, the eliminationof
the error
would increase the difference between the two conditions.
5. Since the temperature distribution
on the foot varied greatly with the temperature level, the thermocouples might have been on representative areas in one condition
but not in the other. For this reason the resistance thermometer previously described
was used. Two new subjects spent a single night at each of the temperatures used before.
The procedure was the same except for additional precautions in the measurement of air
In the cold the apparent air insulation was 0.75 for one subject and 1.33
temperature.
for the other. Corresponding
values in the heat were 0.42 and 0.24.
July
BLOOD
1948
AND
HEAT
EXCHANGE
IN
THE
FEET
31
from a comparable area of the rest of the foot. As has been mentioned
before, the toes were the coldest part of the foot during the exposure to cold,
so that little of the non-evaporative heat lost could have come from the toes.
In the heat the gradient along the foot was reversed, and the toe temperature
became the highest of the four measured.
In this case a much greater proTABLE
4.
CONDITION
DATAFROMTABLES
SUBJ.
I AND 2 OF GAGGE (26),
SHOWINGCHANGE
LOSS WITHCHANGING
SURFACETEMPERATURE
SKIN TEMP.
___-
---
R I H I
C
-.
Cal/mPlhr.
I
28.9
11.8
-53
3s
+=s
2
91
3
P-9
12.0
-18
32.3
34.3
34.8
35.3
II.2
16
II.2
--I7
-50
-60
4-S
5.4
10.3
43
86
35
34
17
-21
11.1
120
-45
-65
-7s
6.3
6.8
4l
!?
6’
--
--
7
8’
33.5
34.1
9l
6.1
-30
IO1
34.7
34.4
34
71
II
35.3
s-8
96
I
29.5
30.0
2
3’
4’
5’
6l
----
--
12.4
II
.6
-57
-=7
33.3
34.4
35.1
11.9
15
11.1
35.4
11.4
49
84
I24
11.0
30
25
5
6.3
56
4.8
0
-30
-35
-42
I
- 29
--57
-2
-33
-57
-66
-78
-41
32.9
34.4
9*
IO1
34.2
34.8
37
72
-21
II
35.6
98
-58
-27
-83
37
29
5
9
6.3
8.8
-39
59
50
42
I7
-6
7
8’
4.8
I
-10
-38
-42
-51
-40
E is used by Gagge to describe the algebraic
R represents
the heat exchange by radiation.
sun-r of metabolism,
storage and evaporation.
It has been converted
from Cal/hr.
to Cal/mz/hr.
C, the convective
heat loss, was not
by the use of the surface area given for these subjects.
given but can be obtained
by the difference
between R and EI.
1 Conditions
considered
by Gagge to be suitable for partitional
calorimetry.
portion of the heat must have been lost from the toes. Due to this change
in temperature distribution
the air insulation must change in the direction
which has been found.
It should be emphasized that the curvature factor does not affect the
The observed changes in air insulation are attributed
heat loss by radiation.
If this be true, such changes should
to changes in the convective fraction.
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
"C.
SKIN - AIR
TEMP.
_____"C.
INCON~ECTIVEHEAT
32
LOIS
H. LOVE
Volume
I
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
only be marked when the convective heat loss is high relative to the radiant
heat loss. Insulation values have been calculated from the results of Forster
et al. (3). For the experiments in which the hand temperature was the same
as the foot temperatures in these experiments the air insulation is 1.5 Clo
for the warm hand and 1.7 for the cold hand. This difference is much smaller
than has been found for the foot and this may be due to the fact that the
volume air flow in the hand calorimeter was only one-fifth of that used on the
foot. This would limit the convective heat loss, the only fraction which
could be affected by the factors which have been considered as possible
causes for this change in air insulation. The convective heat loss was also
very low in the experiments of Hardy and Soderstrom (24) in which the air
insulation for the entire body remained constant over a wide temperature
range. Hardy and DuBois (25) state that when the air movement in the
calorimeter was increased that this was no longer true. However, the direction and magnitude of the changes were not given.
Some of the data of the method of partitional calorimetry indicates
that the convective heat loss does not depend only on the temperature difference between skin and air and on the air movement, as is usually assumed.
Gagge (26) has summarized a large number of experiments with constant air
velocity in which the temperature gradient between the skin and air was
maintained at either 6’C. or II’C. while the wall temperature was varied.
The data given include estimates of temperatures and amounts of heat exchange so that the convective heat loss can be calculated. Figures taken
from Gagge’s tables are shown in table 4. It is apparent that the convective
heat lossper degree difference between skin and air is not constant but shows
a marked trend which is correlated with the skin temperature. This is in
the direction which would be expected if the factor influencing the heat loss
from the foot were also important in regulating the heat loss from the entire
body. If the factor responsible for these deviations is one of curvature, it
might be expected that the change in air insulation of the entire body might
be less than for the foot, for the data on the heat loss from pipes of different
sizes show that a 50 per cent reduction in diameter is more effective in increasing the heat loss when the diameter is initially small than when it is
initially large. For this reason the change in distribution of heat loss from
the torso and limbs might be lesseffective in changing the air insulation than
the change in distribution along the foot.
The curvature factor can also be used to explain the fact that in the heat
the foot lost a greater amount of heat on an area basis than the entire body.
While the low level of heat loss from the cold foot is that to be expected from
the greater reduction of foot temperature than of general surface temperature, it is not likely that the average surface temperature of the foot was
July
1948
BLOOD
AND
HEAT
EXCHANGE
IN
THE
FEET
33
significantly higher than the rest of the body in the heat. Since the foot
has a smaller effective diameter than the entire body the increased heat loss
from the foot can be assigned plausibly to the effects of curvature.
However, no positive conclusions can be drawn, since the relative velocities of
the air inside and outside the calorimeter were not known.
SUMMARY
I am deeply indebted to Dr. H. C. Bazett for valuable advice on all phases of this
work. I would also like to thank Dr. C. R. Spealman and all of the others who parThe corrections for the
ticipated in these experiments for their unfailing cooperation.
heat absorbed during the expansion of water vapor were made with the assistance of
Dr. John G. Miller of the Department of Chemistry of this university and Drs. A. C.
Burton and J. D. Hardy to whom I am also indebted.
REFERENCES
J. C., H. C. BAZETT AND G. C. MACKIE.
Am-l.
Physiol. 129: 102, 1940.
A. C., J. C. SCOTT, B. MCGLONE
AND 1-I. C. BAZETT.
Am. J. Physiol.
129: 84, 1940.
3. FORSTER, R. E., B. G. FERRIS AXD R. DAY. Am. J. Physiol. 146: 600, 1946.
I.
SCOTT,
2. BURTON,
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
I. The heat loss, blood flow and skin temperature
of the feet were measured in two subjects living at 33OC. and at 21OC. Exposures to the two
conditions lasted 2 to 7 days. Both temperatures were used in the summer
and again during the winter.
2. During
any one experimental period of heat the only progressive
change which could be demonstrated was a gradual increase in the temperature of the toe relative to that of the rest of the foot, presumably indicating
an increased blood flow in the toe. In the cold no progressive changes were
found.
3. All other evidence of acclimatization
appeared only as differences
between the various periods of exposure to the same temperature.
Exposure
to cold reduced the ability of the foot to lose heat by vaporization during a
subsequent period in the heat. This was associated with a decrease in the
amount of heat lost by each volume of blood flowing through the foot. In
some cases absence of acclimatization
was associated with an increased rate
of blood flow through the foot.
4. It is estimated that the foot, which has about 5 per cent of the total
body area, lost 6.5 per cent of the total heat in the heat and only 3.5 per cent
in the cold.
5. The heat loss per degree difference between the skin and air in the
heat was about twice that in the cold. As possible explanations roughening
of the skin in the cold and the effect of curvature on heat loss are suggested.
Both would affect only the convective fraction of the heat loss.
LOIS H. LOVE
34
J. Nutritioft g: 233, 1935.
J. R. AND A. C. BURTON.
J. Ntitrition 15: 461, 1938.
J. D. ANI E. F. DUBOIL
Am. Heart J. 17: 206, 1939.
ABRAMSON,
D. I., H. ZAZEELA AND J. MARRUS.
J. C&z. Invest. Ig : 273,Ig4o.
WRIGHT,
G. W. AND K. PHELPS.
Acta Physiol. ScancL 4: 149, 1942.
CHRISTENSEN,
E. H. AFJD M. NIELSEN.
Report to Comm. Av. Med.
MENDELSON,
E. S., L. HOSBACH AND H. C. BAZETT.
under O.S.R.D., July 1945.
DUBOIS, D. AND E. F. DUBOIS. Arch. Int. Med. 15: 868, IgIg.
Am. J. PhysioZ. ISO: 628, 1947.
SPEALMAN,
C. R., M. NEWTON AND R. L. POST.
Science 94: 428, 1941.
GAGGE, A. P., A. C. BURTON AND H. C. BAZETT.
Arch. Int. Med. 49: 1019, 1932.
BENEDICT,
F. G. AND H. S. H. WARDLAW.
The Physiology of Human Perspiration.
London: Churchill, 1934.
KUNO, Y.
WEINER,
J. S. J. Physiol. 104: 32, 1945.
ADOLPH,
E. F. Am. J. Physiol. 123: 486,Ig38.
Comm. Aviation Med. O.S.R.D. NO.
BURTON,
A. C. Clothing Test Methods.
390, 1945
TROTTER,
W. AND H. H. DAVIS.
J. Physiol. 38: 134, Igog.
J. Anat. 67: 18, 1932.
WOOLARD,
H. H. AND R. PHILLIPS.
VANDILLA,
M.
Report 76A, Climatic Research Lab., Army Service Forces, Q. M.
Corps, 19440
BURTON,
A. C. Canadian Comm. Aviation Medical Research, Report # C2725,
19440
HEILMAN,
R. H. lnd. and Eng. Chem. 16: 451,Igzq.
RICE, C. W.
Iniernational
critical Tables 5: 234, 1929.
J. Nutrition 16: 493, 1938.
HARDY,
J. D. AND G. F. SODERSTROM.
Proc. Nat. Acad. Sci. 23: 624, 1937.
HARDY, J. D. AND E. F. DUBOIS.
GAGGE,
A. P. Am. J. Physiol. 116: 656, 1936.
4. MURLIN,
5. HARDY,
6.
7.
8.
g.
IO.
II.
12.
13.
14.
16.
17.
18.
Ig.
20.
21.
22.
23.
24.
25.
26.
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on July 31, 2017
15.