Content uploaded by Igor B. Mekjavic
Author content
All content in this area was uploaded by Igor B. Mekjavic on Feb 27, 2014
Content may be subject to copyright.
Eur J Appl Physiol (2008) 104:257–264
DOI 10.1007/s00421-007-0645-y
123
ORIGINAL ARTICLE
Local diVerences in sweat secretion from the head
during rest and exercise in the heat
Christiano A. Machado-Moreira · Frederik Wilmink ·
Annieka Meijer · Igor B. Mekjavic · Nigel A. S. Taylor
Accepted: 3 December 2007 / Published online: 20 December 2007
© Springer-Verlag 2007
Abstract The importance of the head in dissipating body
heat under hot conditions is well recognised, although very
little is known about local diVerences in sweat secretion
across the surface of the head. In this study, we focused on
the intra-segmental distribution of head sweating. Ten
healthy males were exposed to passive heating and exercise-
induced hyperthermia (36°C, 60% relative humidity, water-
perfusion suit: 46°C), with ventilated sweat capsules
(3.16 cm
2
) used to measure sweat rates from the forehead
and nine sites inside the hairline. Sweat secretion from both
non-hairy (glabrous) and hairy areas of the head increased
linearly with increments in work rate and core temperature,
with heart rate and core temperature peaking at 175 b min
¡1
(§6) b min
¡1
and 39.2°C (§0.1). The mean sweat rate
during exercise for sites within the hairline was
1.95 mg cm
¡2
min
¡1
. However, the evolution of this secre-
tion pattern was not uniformly distributed within the head,
with the average sweat rate for the top of the head being sig-
niWcantly lower than at the anterior lateral aspect of the head
(P < 0.05), and representing only 30% of the forehead sweat
rate (P < 0.05). It is hypothesised that these intra-segmental
observations may reXect variations in the local adaptation of
eccrine glands to diVerences in local evaporation associated
either with bipedal locomotion, which will inXuence fore-
head sweating, or the hidromeiotic suppression of sweating,
which impacts upon sweat glands within the hairline.
Keywords Eccrine sweat glands · Regional sweating ·
Sudomotor · Thermoregulation
Introduction
Eccrine sweat glands are found over virtually the entire sur-
face of the human body. However, the density of these
glands is not uniform, and we know very little concerning
the number of active sweat glands, or the sweat rate per
gland within diVerent body segments. Although gross
diVerences in the sweat secretion among diVerent body seg-
ments have already been reported (Weiner 1945; Hertzman
1957; Cotter et al. 1995), there is a gap in the literature con-
cerning local diVerences in sweat secretion within most
body segments.
Many studies have evaluated forehead sweating, and it is
now widely known that this site has one of the highest sweat
gland densities, and usually has a greater sweat response
than all other body segments during thermal loading (Hertz-
man et al. 1952; Szabo 1962; Cotter et al. 1995). Neverthe-
less, one should not assume that forehead sweating
adequately represents sweat secretion from the entire head,
since this has not yet been thoroughly investigated. Further-
more, it is generally regarded that glabrous (non-hairy) sur-
faces are more responsive to psychological (mental and
emotional) stimuli, and that non-glabrous (hairy) regions are
more sensitive to thermal stress, although increases in
sweating due to thermal and non-thermal stimulation have
been observed in both types of skin (Ogawa et al. 1977;
Ogawa 1984). Thus, one might reasonably expect to observe
diVerences in sweat secretion from the hairy and non-hairy
skin surfaces of the head during thermal loading. With this
background, we undertook the measurement of sweat rates
from ten sites on the surface of the head.
C. A. Machado-Moreira · F. Wilmink · A. Meijer ·
N. A. S. Taylor (&)
School of Health Sciences,
University of Wollongong, Wollongong,
NSW 2522, Australia
e-mail: nigel_taylor@uow.edu.au
I. B. Mekjavic
Jozef Stefan Institute, 1000 Ljubljana, Slovenia
258 Eur J Appl Physiol (2008) 104:257–264
123
The importance of the head in dissipating body heat, and
the prevention of excessive thermal strain under uncom-
pensable thermal conditions, has been well recognised
(Desruelle and Candas 2000). Despite its small surface area
(6–7% of body surface area; Hardy and DuBois 1938),
cooling the head can elicit signiWcant reductions in thermal
strain (Nunneley et al. 1982). Indeed, it has been estab-
lished that this body segment is highly responsive to locally
applied thermal stimuli (Cotter and Taylor 2005). The large
relative surface area to mass ratio of the head, in combina-
tion with its cutaneous vasculature, facilitates heat loss,
which is signiWcantly higher than reported from other body
segments when normalised to surface area (Froese and Bur-
ton 1957; Clark and Toy 1975; Rasch et al. 1991). Further-
more, the relative importance of the head for heat
dissipation increases when clothing is used. Unfortunately,
the use of headwear can dramatically reduce heat loss due
to altered dry and evaporative heat transfers.
Mathematical and manikin-based modelling of human
thermoregulation represent valuable research tools. How-
ever, the utility of such methods is entirely dependent upon
the precision of the physiological data upon which they are
based. Since we currently lack a detailed description of the
intra-segmental sudomotor responses during passively- and
exercise-induced thermal loads, then these modelling tools
remain imprecise. For example, Liu and Holmer (1995)
used a head manikin to assess the evaporative heat transfer
characteristics of Wve industrial safety helmets, but a uni-
form sweat secretion was assumed to exist over the entire
head surface. More recently, Brühwiler (2003) developed a
headform that consisted of three separately-heated areas,
that also had the capacity to mimic diVerent sweat rates.
Unfortunately, precise information about localised sweat
secretion rates within the head was not available for use
within this manikin.
In this study, we focussed on the distribution of sweat
secretion from nine non-glabrous skin surfaces of the scalp,
and the forehead, in shaved male subjects. Thermal strain
consisted of passive heating (water-perfusion garment) and
incremental cycling in a hot and humid environment. The
observations from this experiment contribute to our under-
standing of human thermoregulation in general, and sudomo-
tor control in particular. In addition, such information forms
important background knowledge for the development of
thermal (head) manikins and for the design of headwear.
Methods
Subjects
Ten healthy and physically-active males [25.5 years (SD
2.7); 74.8 kg (SD 8.4); 1.79 m (SD 0.09)] were exposed to
passive heating and exercise, consisting of incremental
cycling performed in a climate-controlled chamber (36°C,
60% relative humidity, wind speed <0.5 m s
¡1
). Subjects
wore a whole-body, water-perfusion suit provided with
water at 46°C for the entire trial. Procedures were approved
by the Human Research Ethics Committee (University of
Wollongong), and fully explained to the subjects prior to
the provision of written, informed consent.
Procedures
None of our subjects was naturally bald, and each was
shaved completely on the day of each experiment. Subjects
were then instrumented (thermistors and sweat capsules
attached) and dressed in a water-perfusion suit, before
entering a climate-controlled chamber. Exercise started
after a 30-min period of seated rest, during which heated
water (46°C: 38-L water bath; Type VFP, Grant Instru-
ments, UK) was passed through the perfusion suit (Cool
Tubesuit, Med-Eng, Ottawa, ON, Canada) at a Xow of
0.3 L min
¡1
(Delta Wing pump, Med-Eng, Ottawa, ON,
Canada). Incremental cycling was performed using an elec-
tronically-braked ergometer (Lode Excalibur Sport; Gron-
ingen, The Netherlands) with an initial work rate of 50 W
(15 min), which increased by 25 W every 15 min
(60 rev min
¡1
). Trials were terminated when core tempera-
ture exceeded 39.5°C for 2 min (N = 1) or at volitional
fatigue (N = 9). The mean exercise duration was 55.6 min
(range: 45–69 min).
Measurements
Ventilated sweat capsules (3.16 cm
2
) were used to measure
local sweat rates (mg cm
¡2
min
¡1
) from ten head sites: the
forehead and nine sites inside the hairline (Fig. 1). Capsules
were glued to the skin (Collodion U.S.P., Mavidon Medical
Products, FL, USA). Pre-capsular airXow was indepen-
dently regulated at 600 mL min
¡1
, with inlet humidity
maintained at 12% by passing room air for all capsules over
a common, saturated lithium chloride solution. The humid-
ity of the post-capsular air was measured using capacitance
hygrometers, which formed parts of a sweat data acquisi-
tion system (Clinical Engineering Solutions, NSW, Austra-
lia), with inlet and exhaust air temperatures and humidities
sampled at 5-s intervals from six channels (DAS1602,
Keithley Instruments, Inc., Cleveland, OH, USA), and used
to compute local sweat rates (Taylor et al. 1997). Hygrome-
ter calibration, using saturated salt solution standards, pre-
ceded experimentation.
Local sweat rates could only be recorded from six sites
simultaneously, with the four remaining capsules being
ventilated with room air to sustain a dry skin surface.
During rest, only six sites were investigated, while during
Eur J Appl Physiol (2008) 104:257–264 259
123
exercise, capsules were connected to the sweat system in a
rotating pattern, leaving two capsules always connected
(Fig. 1: sites 2 and 9). To control this pattern, two trial
sequences were established, with Wve subjects commencing
these trials using each sequence, as shown in Fig. 1. Subse-
quently, 5 min prior to each work rate increase, the sites of
sweat measurement were changed so that the four remain-
ing sites were now connected to the sweat system. These
changes were performed manually, taking approximately
2 min to complete. This rotation pattern was then continued
until the trial was terminated.
Auditory canal (insulated) and rectal (10 cm beyond
the anal sphincter) temperatures were recorded at 5-s
intervals (Edale instruments Ltd., Cambridge, UK), with
the former being used as the representative core temper-
ature. Skin temperatures from eight sites (forehead,
chest, scapula, upper arm, forearm, dorsal hand, thigh
and calf) were measured at the same time intervals, and
mean skin temperature was derived using an area-
weighted summation (ISO 9886: 1992). Local skin tem-
peratures were also recorded from sites adjacent to each
sweat capsule. All thermistors were calibrated against a
certiWed reference thermometer in a stirred water bath
(Dobros total immersion, Dobbie Instruments, Sydney,
NSW, Australia), and data were collected using a data
logger (1206 Series Squirrel, Grant Instruments Pty Ltd.,
Cambridge, UK). Heart rate was continuously monitored
from ventricular depolarisation, and recorded every 5 s
(Vantage NV Sports Tester, Polar Electro Oy, Kempele,
Finland).
Analysis
Data from discrete 2-min periods were used to compare
local sweat rates across sites: the 2 min immediately before
changing the sampling sites; and the 2 min immediately
prior to each increase in work rate. Local sudomotor sensi-
tivities (gain: mg cm
¡2
min
¡1
°C
¡1
) were derived from
changes in local sweat rates (mg cm
¡2
min
¡1
) and auditory
canal temperature, computed across the Wrst 45 min of
exercise for each subject (after Taylor et al. 2006), since
such analyses may be of greater utility to modellers and
clothing engineers. One-way Analyses of Variance were
used to compare between-site sudomotor diVerences for the
averaged sweat rates over the entire exercise phase, and
sudomotor sensitivities. Tukeys HSD post hoc tests were
used to isolate sources of signiWcant diVerences. Two-way
Analyses of Variance were used to evaluate localised sweat
rates during the Wrst 45 min of exercise. Alpha was set at
the 0.05 level for all analyses. Data are presented as means
with standard errors of the means (§SEM) and standard
deviations (SD).
Fig. 1 Sweat capsule place-
ment. Capsules were positioned
at the forehead, and at nine sites
within the hairline. Also shown
are the two experimental se-
quences used to ensure that these
sites were evaluated in a bal-
anced order
260 Eur J Appl Physiol (2008) 104:257–264
123
Results
Body core temperature and heart rate increased signiW-
cantly during the 30-min rest period (changes: 0.25°C and
18 b min
¡1
), and also during incremental exercise
(P < 0.05), peaking at 39.2°C (§0.1; moderately hyperther-
mic) and 175 b min
¡1
(§6; Table 1). The mean sweat rate
for sites within the hairline, when computed over the entire
exercise period, was 1.95 mg cm
¡2
min
¡1
.
Local sweat rates from all head sites increased during
passive heating and during exercise (P < 0.05). Sweat rates
from contra-lateral sites within the same lateral aspect of
the head, and along the top of the head, did not diVer sig-
niWcantly (P > 0.05), so these sites were combined for sub-
sequent analyses. Thus, sweating from six skin surfaces of
the head were compared: forehead (site 10), lateral 1 (front:
sites 1 and 3), lateral 2 (middle: sites 4 and 8), lateral 3
(rear: sites 5 and 7), top (sites 2 and 9) and rear (site 6). At
the end of passive heating, and immediately before com-
mencing exercise (28–30 min), these local sweat rates
were: forehead: 1.20 mg cm
¡2
min
¡1
(§0.40); lateral 1
(front): 0.31 mg cm
¡2
min
¡1
(§0.06); lateral 2 (middle):
0.29 mg cm
¡2
min
¡1
(§0.07); lateral 3 (rear): 0.34 mg
cm
¡2
min
¡1
(§0.08); top: 0.17 mg cm
¡2
min
¡1
(§0.03);
the rear site was not measured at rest.
Sweat rates increased in an approximately linear manner
in all sites from the beginning of exercise until the work
rate reached »100 W, with forehead sweating being the
highest, and sweating from the top of the head being the
lowest at each work rate (Fig. 2). There was a signiWcant
main eVect for measurement site when the forehead was
compared with each of the other locations (P <0.05). Simi-
lar main eVects also existed for comparisons between the
top of the head and each of the other loci (P < 0.05), except
for the rear of the head (P = 0.06). In addition, signiWcant
site by time interactions were evident for the comparisons
between the lateral 1 site and the rear of the head
(P < 0.05), and between the top of the head and each of the
other loci (P < 0.05), except for the rear of the head. That
is, the site-speciWc diVerences in sweat secretion seen
during lower thermal strain became more marked as time
and core temperature increased.
Local sweat rates were averaged over the full exercise
period to provide a global view of local sweating within the
head (Fig. 3). A greater sweat secretion was observed at the
forehead (P < 0.05), followed by the lateral, rear and top
sites, respectively. Figure 4 shows the averaged local sweat
rates after these areas were combined. In addition to higher
sweat secretion at the forehead, relative to all other sur-
faces, the averaged sweat rate for the top site was signiW-
cantly lower than that measured at the lateral 1 site
(P < 0.05), and this represented only 30% of the forehead
sweat rate. Lateral 1 sweat rate corresponded to 63% of
forehead sweating, lateral 2 was 53%, lateral 3 was 56%
and the rear of the head was 45%. Nevertheless, no statisti-
cal diVerences were observed among sweat rates from these
sites. Table 2 contains the local skin temperatures for the
exercise period from 40–45 min. The comparison between
the forehead skin temperature and that of the skin surface at
the rear of the head was the only source of a signiWcant
diVerence (P <0.05).
Sudomotor sensitivity was signiWcantly lower at the top
surfaces of the head than at the lateral 1 surface (P <0.05),
whilst no statistical diVerences were apparent among the
sensitivities for the other head sites (P >0.05; Table2). If
sudomotor sensitivity is compared across subjects, and
expressed as a ratio, such that the highest sensitivity within
each site is compared to the lowest observed for that site,
then one can obtain a measure of the variability of sensitiv-
ity across subjects. The site that had the greatest range in
sensitivity was the rear of the head, with this ratio being
35.9, which was at least three times greater than observed at
Table 1 Body core temperatures and heart rates at the end of 30 min
of passive heating (36°C, 60% relative humidity; water-perfusion suit
46°C), and during incremental cycling
Data are means with standard errors of means
Work rate (W) N Core temperature (°C) Heart rate (b min
¡1
)
Rest 10 36.9 (0.1) 90 (4)
50 10 37.1 (0.1) 110 (5)
75 10 37.6 (0.1) 135 (5)
100 10 38.3 (0.1) 160 (5)
125 6 39.2 (0.1) 175 (6)
150 1 39.7 177
Fig. 2 Local sweat rates from six head surfaces following 30 min o
f
passive heating, and then during incremental cycling in the heat (36°C,
60% relative humidity; water-perfusion suit 46°C). Data were aver-
aged over the second half (7.5 min) of each work rate, and are pre-
sented as means with standard errors of the means
Eur J Appl Physiol (2008) 104:257–264 261
123
the other areas (top: 12.1; forehead: 9.7; lateral 3: 7.5; lat-
eral 2: 7.3; and lateral 1: 3.8). One subject showed a very
low sudomotor sensitivity across sites, but especially at the
rear of the head (S9: 0.05 mg cm
¡2
min
¡1
°C
¡1
; average of
other nine subjects: 1.20 mg cm
¡2
min
¡1
°C
¡1
).
Eliminating this subject dramatically altered these sensi-
tivity ratios. For the rear of the head, the ratio became 4.6,
and ranged from 3.8 to 9.7 for the other skin sites. This is
consistent with ratios previously reported for dorsal sur-
faces of the foot and forehead (Taylor et al. 2006). In those
experiments, large inter-subject variability was also
observed for sweating at the plantar surface, but, in that
case, hyperhidrotic responses were found in two subjects.
The very low head sweat gland responsiveness to increases
in core temperature, as observed in S9, appears not to have
been described in the literature. This response was not evi-
dent at the forehead, but was restricted just to sites within
the hairline.
Discussion
In the current study, sweat secretion rates from the forehead
and nine non-glabrous (hairy) skin surfaces of the head
(within the hairline) were measured during passive and
exercise-induced heating in a hot-humid environment. To
the best of our knowledge, such a broad description of head
sweating has not previously been reported, and three sig-
niWcant observations arise from these experiments. First,
each of the sites that were investigated responded to the
thermal loading, with the hairy sites of the head secreting
less sweat than the forehead. Second, the sites at the top of
the head secreted the least sweat, and generally displayed a
lower sudomotor sensitivity to changes in core temperature.
Third, although the forehead sweat rate was the highest, its
sudomotor sensitivity did not diVer signiWcantly from that
observed at the lateral aspect, rear or top of the head.
Those familiar with the literature may recognise the sim-
ilarity of the Wrst two primary observations with those
reported by Cabanac and Brinnel (1988). These authors
investigated sweating from the head in both bald (N = 10)
and hairy men (N = 10). However, they studied just three
sites [forehead, temple (side of face) and the top of the head
(which they called the calvaria)], and did not explore sweat
secretion within the hairline. For their hairy subjects, they
did observe a lower sweat rate for the top of the head
Fig. 3 Mean sweat rates for ten head skin surfaces. Data are means
with standard errors of the means collected during incremental cycling
in the heat (36°C, 60% relative humidity; water-perfusion suit 46°C)
and averaged across the entire exercise phase, the mean duration o
f
which was 55.6 min (range: 45–69 min). *SigniWcantly diVerent from
all sites below the horizontal line (P <0.05)
Fig. 4 Mean sweat rates for six head sites. Data are means with stan-
dard errors of the means collected during incremental cycling in the
heat (36°C, 60% relative humidity; water-perfusion suit 46°C) and
averaged across the entire exercise phase. *SigniWcantly diVerent from
all sites (P <0.05); H, signiWcantly diVerent from the sites identiWed by
either end of the horizontal line (P <0.05)
Table 2 Intra-segmental diVerences in sudomotor sensitivity and lo-
cal skin temperatures of the head during incremental cycling in the heat
(36°C, 60% relative humidity; water-perfusion suit 46°C)
Sudomotor sensitivities were derived from changes in local sweat rates
and auditory canal temperature, computed across 45 min of exercise.
Skin temperatures were derived from the period 40–45 min of exercise.
Data are means with standard errors of the means.
*SigniWcantly diVerent from the lateral 1 site; ** signiWcantly diVerent
from the forehead (P <0.05)
Head sites Sudomotor sensitivity
(mg cm
¡2
min
¡1
°C
¡1
)
Skin
temperature (°C)
Forehead 1.89 (0.28) 37.5 (0.1)
Lateral 1 2.10 (0.22) 37.3 (0.1)
Lateral 2 1.85 (0.27) 37.2 (0.1)
Lateral 3 1.85 (0.27) 37.2 (0.1)
Top 1.03 (0.23)* 36.9 (0.1)
Rear 1.09 (0.19) 36.8 (0.2)**
262 Eur J Appl Physiol (2008) 104:257–264
123
relative to the forehead and temple sites, with temple
sweating representing approximately 60%, and sweating
from the top of the head was about 30% of that observed at
the forehead. In the current investigation, sweating from the
temples was not examined. However, the relative diVer-
ences in sweat secretion from the forehead and top of the
head are in close agreement for both investigations. Follow-
ing comparisons with bald subjects, Cabanac and Brinnel
(1988) concluded that sweating correlated with the pres-
ence of hair, with hair acting in some way to keep sweating
lower, relative to that observed without hair. The authors
suggested this was related to an altered local cholinergic
sensitivity (which was not currently evident; Table 2), and
was perhaps modiWed by the plasma concentration of
androgens, with male-pattern baldness developing in
response to the growth of facial hair, and thereby facilitat-
ing a constant rate of evaporation and brain cooling.
While data from the top of the head in our hairy subjects
are consistent with the observation of Cabanac and Brinnel
(1988), a closer inspection of our data shows that the distri-
bution of sweating appears to be more closely related to
location than to the presence of hair. One could presume
the presence of hair would inXuence sweating, but one
could also assume that its inXuence would probably have
equally aVected sweat secretion from all sites below the
hairline. We cannot thoroughly test this hypothesis, since
we did not study bald subjects. However, since we have
known for many years that most skin surfaces sweat less
than the forehead (Weiner 1945; Hertzman 1957; Cotter
et al. 1995), with some of these also being hairless (gla-
brous) skin areas, then the observation of lower sweat rates
within the hairline is not a phenomenon peculiar to the
head. In addition, six of the nine scalp sites currently tested
(Fig. 3), displayed sweat rates approximately twice that
observed for the top of the head. We, therefore, interpret
these data as being inconsistent with the hypothesis of Cab-
anac and Brinnel (1988). Thus, it seems more likely that, in
addition to the diVerences in sudomotor function that exist
between hairy and non-hairy skin surfaces (Saad et al.
2001), there are sweat secretion diVerences even within the
hairy skin surfaces of the head. Indeed, we suspect that if
Cabanac and Brinnel (1988) had investigated more sites
within the hairline, then they may perhaps have arrived at a
diVerent conclusion.
To explain these intra-segmental diVerences in sweating,
one should look Wrst for diVerences in local skin tempera-
tures, sweat gland densities and individual sweat gland
Xows. Since sweat rate can be modiWed by local skin tem-
perature (Nadel et al. 1971), there is the possibility that
these diVerences were due to variations in local skin tem-
peratures. This possibility was anticipated, and skin tem-
peratures were recorded adjacent to each sweat capsule so
that this eVect could be evaluated. However, these data
revealed a uniformly high skin temperature distribution
across the forehead and scalp (Table 2), and cannot account
for the variance observed in sweating among these sites.
Unfortunately, there is a lack of information regarding
the sweat gland density for the scalp, and we know of no
data that describe intra-segmental diVerences in either
sweat gland
Xow or recruitment patterns within the hairline
of the head. However, we know that the forehead sweat
gland density is one of the highest of all body surfaces
(»226 glands cm
¡2
; mean from 14 available studies), and
is only lower than the densities for sites within the hands
and feet. We are aware of only two (cadaver) studies report-
ing glandular densities for areas within the hairline (Szabo
1967; Hwang and Baik 1997). In the former case, data
came from the scalp area of only one individual, with one
count of 100 glands cm
¡2
being obtained from skin within
the hairline, and another of 70 glands cm
¡2
from a bald
area of the scalp. In the latter study, counts were obtained
from the parietal surfaces (dome of the skull »3–4 cm
above the eyebrows) of 27 male cadavers, revealing a mean
glandular density of 194 glands cm
¡2
. Of these two data
sets, the current authors have greater faith in the latter,
since the often-cited observations of Szabo (1962, 1967)
were derived from a very small number of cadavers, and
these data, whilst being quantitatively similar, are often at
variance with data obtained by other groups using larger
sample sizes.
These inter-site comparisons in eccrine gland density fail
to provide a simple explanation for the greater forehead
sweat secretion, in the presence of a presumably equivalent
thermoeVerent drive. Alternatively, we have recently
shown that the forehead sweat glands appear to be much
less responsive to heat adaptation than are glands from
other body segments (Patterson et al. 2004). This may indi-
cate that the eccrine glands of the forehead operate closer to
their maximal secretion capacity, even in the unacclima-
tised state. That is, glandular Xows from this site do not
generally increase as much as Xows from other sites follow-
ing heat adaptation. Accordingly, explanations other than
diVerences in sweat gland density may help to account for
the diVerences in sweat secretion from sites within the hair-
line.
Since humans are bipedal, then evaporative cooling dur-
ing locomotion will be greatest from wetted surfaces that
face directly into the wind. Since the forehead lacks hair,
and is rarely covered during exercise in the heat, then one
would predict that evaporation from the forehead would be
very eYcient. As it is well established that the forehead also
has possibly the highest skin temperature (Werner and
Reents 1980), then these factors may combine to ensure
that, during hyperthermic states, sweat secretion from
glands on the forehead is generally profuse and well evapo-
rated. This state increases the probability of sweat gland
Eur J Appl Physiol (2008) 104:257–264 263
123
adaptation. Conversely, the skin surfaces below the hair-
line, which also have high skin temperatures (Table 2), will
not experience high evaporation due to the presence of hair,
and this results in the partial trapping of a boundary layer of
moist air close to the skin surface. This layer reduces evap-
oration, leading to moisture accumulation on the skin, and a
possible hidromeiotic suppression of sweating. Thus, it is
probable that, without hair, these glands may also have
adapted to produce a higher sweat rate, just as observed by
Cabanac and Brinnel (1988), though not for reasons associ-
ated with brain cooling, but due to diVerences in local adap-
tation state. Furthermore, one wonders whether the lower
sweat rate at the top of the head may also be partially
explained on the same basis, but now augmented by the
behavioural practice of wearing a hat during outdoor exer-
cise in hot conditions, with greater heat and moisture accu-
mulating over the parietal areas of the head than over its
lateral aspects.
In addition to providing useful information regarding
human sudomotor function, the current observations should
be considered by clothing design engineers. In the case of
headwear and headbands, heat loss from the head may be
impaired, thus increasing heat storage, particularly in those
who are obliged to wear protective helmets. Two outcomes
from this experiment should be considered in this regard.
First, given the higher sweat rates on the lateral sites of the
head, headwear must be designed to optimise evaporation
from the sides of the head. Most existing headwear can sat-
isfy this physiological speciWcation. Second, if the lower
sweat rate on the top of the head is the result of a poor local
thermal adaptation due to the wearing of hats, then head-
wear should be modiWed to improve ventilation, and there-
fore evaporation, for the parietal area. This will have two
positive eVects: increased evaporation and improved sweat
gland performance.
Conclusion
Both non-hairy (glabrous) and hairy areas of the head
increased sweat secretion in response to passively- and
exercise-induced thermal strain. However, the evolution of
this secretion pattern was not uniformly distributed across
the surface of the head, being less evident in sites within the
hairline, and least apparent at the top of the head, which had
a sweat rate that was only 30% of that secreted from the
forehead. It is hypothesised that these intra-segmental
observations may reXect variations in the local adaptation
of eccrine glands to diVerences in local evaporation associ-
ated either with bipedal locomotion, which will inXuence
forehead sweating, or the hidromeiotic suppression of
sweating, which impacts upon sweat glands within the hair-
line.
Acknowledgments This project was supported, in part, by a grant
from the Ministry of Defence (Republic of Slovenia). It was also sup-
ported by a Doctoral scholarship from Coordenação de Aperfeiçoa-
mento de Pessoal de Nível Superior––CAPES (Ministry of Education,
Brazil).
References
Brühwiler PA (2003) Heated, perspiring manikin headform for the
measurement of headgear ventilation characteristics. Meas Sci
Technol 14:217–227
Cabanac M, Brinnel H (1988) Beards, baldness, and sweat secretion.
Eur J Appl Physiol 58:39–46
Clark RP, Toy N (1975) Forced convection around the human head.
J Physiol 244:295–302
Cotter JD, Taylor NAS (2005) The distribution of cutaneous sudomo-
tor and alliesthesial thermosensitivity in mildly heat-stressed hu-
mans: an open-loop approach. J Physiol 565:335–345
Cotter JD, Patterson MJ, Taylor NAS (1995) Topography of eccrine
sweating in humans during exercise. Eur J Appl Physiol
71(6):549–554
Desruelle AV, Candas V (2000) Thermoregulatory eVects of three
diVerent types of head cooling in humans during a mild hyperther-
mia. Eur J Appl Physiol 81:33–39
Froese G, Burton AC (1957) Heat losses from the human head. J Appl
Physiol 10:235–241
Hardy JD, DuBois EF (1938) The technic of measuring radiation and
convection. J Nutr 15:461–475
Hertzman AB (1957) Individual diVerences in regional sweating.
J Appl Physiol 10:242–248
Hertzman AB, Randall WC, Peiss CN, Seckendorf R (1952) Regional
rates of evaporation from the skin at various environmental tem-
peraures. J Appl Physiol 5:153–161
Hwang K, Baik SH (1997) Distribution of hairs and sweat glands on
the bodies of Korean adults: a morphometric study. Acta Anat
158:112–120
ISO 9886 (1992) Evaluation of thermal strain by physiological mea-
surements. International Standard Organisation, Geneva
Liu X, Holmér I (1995) Evaporative heat transfer characteristics of
industrial safety helmets. Appl Ergon 26:135–140
Nadel ER, Bullard RW, Stolwijki JAJ (1971) Importance of skin
temperature in the regulation of sweating. J Appl Physiol
31(l):80–87
Nunneley SA, Reader DC, Maldonado RJ (1982) Head-temperature
eVects on physiology, comfort and performance during hyperther-
mia. Aviat Space Environ Med 53:623–628
Ogawa T (1984) Regional diVerences in sweating activity. In: Hales
JRS (eds) Thermal physiology. Raven Press, New York, pp 229–
234
Ogawa T, Asayama M, Ito M (1977) Comparison of sudomotor neural
activities between palmar and non palmar sweating. In: Proceed-
ings of XVIII international congress of neurovegetative research,
Tokyo, Japan, 4–6 Nov, pp 236–238
Patterson MJ, Stocks JM, Taylor NAS (2004) Humid heat acclimation
does not elicit a preferential sweat redistribution towards the
limbs. Am J Physiol 286(3):R512–R518
Rasch W, Samsom P, Cote J, Cabanac M (1991) Heat loss from the hu-
man head during exercise. J Appl Physiol 71(2):590–595
Saad AR, Stephens DP, Bennett LAT, Charkoudian N, Kosiba W,
Johnson J (2001) InXuence of isometric exercise on blood Xow
and sweating in glabrous and nonglabrous human skin. J Appl
Physiol 91(6):2487–2492
Szabo G (1962) The number of eccrine sweat glands in human skin.
Adv Biol Skin 3:1–5
264 Eur J Appl Physiol (2008) 104:257–264
123
Szabo G (1967) The regional anatomy of the human integument
with special reference to the distribution of hair follicles,
sweat glands and melanocytes. Phil Trans R Soc Lond B
252:447–485
Taylor NAS, Patterson MJ, Cotter JD, Macfarlane DJ (1997) EVects of
artiWcially-induced anaemia on sudomotor and cutaneous blood
Xow responses to heat stress. Eur J Appl Physiol 76:380–386
Taylor NAS, Caldwell JN, Mekjavic IB (2006) The sweating foot: lo-
cal diVerences in sweat secretion during exercise-induced hyper-
thermia. Aviat Space Environ Med 77:1020–1027
Weiner JS (1945) The regional distribution of sweating. J Physiol
104:32–40
Werner J, Reents T (1980) A contribution to the topography of temper-
ature regulation in man. Eur J Appl Physiol 45:87–94
