Clinical Science (1970) 39, 823-832.
ROLE OF CIRCULATING CATECHOLAMINES IN
HUMAN ECCRINE SWEAT GLAND CONTROL
K. G. FOSTER, JEAN GINSBURG
AND
J. S. WEINER
M.R.C. Environmental Physiology Unit, London School of Hygiene and Tropical Medicine, and
Department of Obstetrics, Royal Free Hospital Medical School and Endocrine Unit, London
(Received 19 August 1970)
SUMMARY
1. Intra-arterially as well as intradermally administered adrenaline or noradrenaline produces a sweat response.
2. The intensity of the intradermal response is always considerably less than that due
to acetylcholine and the threshold of effective concentration is considerably greater
than for acetylcholine. The glands also fail to respond to repeated small doses of
intradermal catecholamine and to continued arterial catecholamine infusion.
3. Intradermal administration of a mixture of adrenaline and acetylcholine did not
indicate that adrenaline potentiates the cholinergic response.
4. Intravenous infusions of 20 pg/min adrenaline neither produce a response by the
resting glands nor affect the response of the active glands. Whilst intra-arterial
infusions of high doses of adrenaline or noradrenaline produce a response by the
resting glands they reduce the response of the active glands.
5. The results show that even though the catecholamine levels in a discharge of a
phaeochromocytoma can be high enough to cause a discharge by resting glands, as
the glands are likely to be secreting in response to neural stimulation, the effect of the
secreted catecholamine would be inhibitory.
Human eccrine sweat glands respond to intradermal adrenaline and noradrenaline, and this
response is blocked by a-receptor blocking agents such as tolazoline hydrochloride (Chalmers
& Keele, 1951). Although there would appear to be specific adrenergic receptors in these glands
the existence of an adrenergic component in their innervation could not be established (Chalmers & Keele, 1951). These receptors could conceivably respond to circulating catecholamines
released by the adrenal medulla, particularly during the secretion of excessive amounts of
catecholamines as during the discharge of a phaeochromocytoma when sweating is a prominent feature. This possibility is, however, unlikely in view of the finding by Prout & Wardell
(1969) that hyoscine abolished the sweat response in patients with phaeochromocytomas.
Correspondence: Professor J. S. Weiner, Environmental Physiology Unit, London School of Hygiene and
Tropical Medicine, Keppel Street, London, W.C.!.
823
824
K. G. Foster, Jean Ginsburg and J. S. Weiner
Chalmers & Keele (1951) and Barcroft & Swan (1953) could not detect sweating after
intravascular adrenaline. Their methods may, however, have been too insensitive to detect
either small sweat responses or even large rapid transient responses. More recently, Kuno
(1965) has claimed that the action of adrenaline is to potentiate the sudorific action of acetylcholine, and has postulated that circulating adrenaline will potentiate the nervous response.
Using a rapid and sensitive method for monitoring the sweat response we have, therefore,
re-examined the possible role of circulating catecholamines in the control of the human eccrine
sweat glands.
METHODS
Sweat secretion was measured in glands on the flexor surface of the forearm using a ventilated
capsule in conjunction with an infra-red analyser (Foster & Weiner, 1967).
The effects of catecholamines were studied on (1) resting glands, (2) thermally stimulated
glands, and (3) glands stimulated by acetylcholine. Thermal sweating was induced either by
placing the subject's feet in hot water or by putting the subject in a suit ventilated by warm air.
The body temperature was measured by placing a thermistor in the external auditory meatus.
Catecholamines, acetylcholine, and blocking agents were given by intradermal injection
through an indwelling needle (Foster & Weiner, 1967). All solutions for intradermal injection
were at 37°, since the acetylcholine response is optimal at this temperature. Catecholamines
were also administered parenterally-intravenously by infusion into the cubital vein of the
opposite forearm, and intra-arterially into the brachial artery leading directly to the glands.
Dilutions of acetylcholine chloride, L-noradrenaline bitartrate, adrenaline, isoprenaline
hydrochloride, and phentolamine mesylate were made with physiological saline immediately
before each experiment. Ascorbic acid was added to the solutions of catecholamine to prevent
oxidation (Gad dum, Peart & Vogt, 1949). All subjects were healthy adult volunteers and were
fully informed of the nature and purpose of the study.
RESULTS
Effect of intradermal administration of catecholamines
As with acetylcholine (Fig. la) there was a rapid onset of the sweat response following the
intradermal injection ofO'l-lOj!g adrenaline or noradrenaline in 0·1 ml physiological saline.
There were no obvious differences between the adrenaline and noradrenaline responses.
Results obtained from five subjects showed that the maximum height of the catecholamine
response was always much lower than that due to the same dose of acetylcholine in a given
subject. Fig. 1 shows a comparison between these responses in one of these subjects. Results
obtained from three subjects also showed that no response occurred with 10- 2 j!g/O'l ml
adrenaline or noradrenaline. These three subjects always responded to doses of acetylcholine
as low as 10-2 • 5 j!g/O'l ml. The decline in the response from its peak was, however, usually
much slower than that due to a similar dose of acetylcholine (Fig. la).
On repeated intradermal administration of catecholamine (0,1 and 5 j!g/O·l ml) the glandular
response invariably declined. A similar failure occurs in the eat's pad (Foster, 1968). The response to a second dose of catecholamine was markedly decreased compared to the first dose,
and subsequent doses showed further decreases with, in some cases, eventual failure to respond
10- I,ug Adrenaline
t
0'14pL/min
t
Time (min)
IQ-Ip
g Acetylcholine
1
0'5pLlmin
I
Time (min)
FIG. la. Comparison between the adrenaline and acetylcholine responses due to intradermal injections made on the same forearm. Adrenaline and acetylcholine responses due to intradermal
injections made on the same site. The adrenaline response here followed the acetylcholine response.
Capsule diameter = 0·63 em. Injection volume = 0'1 mI. Solutions at 37°.
1,6
1·2
004
Dose
(~g/O·1
mll
FIG. lb. Maximum heights of the adrenaline and acetylcholine responses due to intradermal
injections made at different sites on the same forearm. (1). Acetylcholine responses. Each mean
±SEM was obtained from five different sites. (2) Adrenaline responses. Each response (e) was
obtained from a different site.
826
K. G. Foster, Jean Ginsburg and J. S. Weiner
(Fig. 2). This failure still occurred when subjects had been heated (active sweating having
previously been blocked by atropine). In one subject, however, during very hot weather, it
was possible to obtain repeated adrenaline responses, but the glands eventually failed to secrete.
The possibility of a potentiating effect of adrenaline on the acetylcholine response was
examined. In these experiments the effects due to a given dose of acetylcholine were observed
first; the adrenaline-acetylcholine mixture was then administered at the same site. Although
10,u9 Adrenaline
101'9 Adrenaline
~ ..-----...-
~_--"
I J
I"" I
""" I J "
I! I "
I "
I "
II J "
I I I I I I I! "
I "
0'2,uL
I "
"
"
I I "
Time (min)
FIG. 2. Responses due to repeated intradermal doses of adrenaline.
1. Comparisons between the amplitudes of the responses due to acetylcholine and
adrenaline mixtures and to acetylcholine alone, drugs administered intradermaIIy
TABLE
Dose of drugs (pg/O'1 m\)
Sweat responses (pI min- 1 cmr ')
Subject
Acetylcholine
Adrenaline
Acetylcholine
response
K.G.F.
10- 2
10- 2
10- 2
10- 2
10- 2
10- 1
10- 1
10- 1
10
10- 2
10- 1
10- 1
10- 2
10- 1
0·15
0·29
0·86
0·26
0·21
0·37
0'55
0·42
0·83
J.P.
J.S.W.
G.W.C.
A.C.
5
5
5
5
Response due Response due
to 10 -1 Ilg/0·1
to mixture
ml adrenaline*
0·10
0·48
1-19
0·20
0·31
0·37
0·48
0·47
0·74
0·30
0·04
* Determined on another forearm site.
the response due to a submaximally active dose of acetylcholine (10- 2 p.g/O·l ml) was increased
by above threshold doses of adrenaline it was reduced by subthreshold doses of adrenaline
(10- 2 and 10- 3 p.g/O·l ml), Increased responses were either slightly greater or slightly less
than that due to a simple summation of the separate acetylcholine and adrenaline responses
(Table 1). When the dose of acetylcholine produced a maximal rate of secretion, the rate was
not increased any further even by large doses ofadrenaline (Table I). The decline in the response
due to the acetylcholine-adrenaline mixture tended to be slower than that due to the acetylcholine alone. These results do not support Kuno's hypothesis.
Catecholamines and sweat gland activity
827
The loss of responsiveness of the glands to repeated doses of catecholamines did not affect
the response of the glands to a subsequent administration of acetylcholine provided this was at
a level (10- 1 J-lgjO'1 ml) capable of producing a maximal response (Table 2). The previous
administration of adrenaline, even at a subthreshold dose (10- 2 J-lgjO'1 ml) did however
reduce the response to acetylcholine given in a dose of submaximal potency.
TABLE
2. Comparisons between the amplitudes of the acetylcholine responses before
and after the adrenaline response, drugs administered intradermally
Subject
K.G.F.
A.C.
J.S.W.
Dose of
acetylcholine
(pg/O'l ml)
Dose of
adrenaline
(pg/O'l ml)
10- 2
10- 2
10- 1
10- 1
10- 1
10- 1
10- 1
10- 3
10- 2
10- 2
10- 1
10- 1
5
5
Acetylcholine
Acetylcholine
response after
response before
adrenaline
adrenaline
(pI min- 1 cmr ') (pI min" ' cmr ')
0·33
0·50
1·07
1-67
0·79
0·26
0·49
0·17
0·20
1·07
1·55
0·83
0·29
0·45
Effect ofparenteral infusions of catecholamines on resting glands
Intravenous infusions. The intravenous infusion of 20 p,gjmin adrenaline for periods of 20
min into one unheated or four heated subjects had no effect on sweat production. In the heated
subjects a high rate of thermal sweating was first induced, and then just before the start of the
adrenaline infusion the thermal response was completely blocked by the intradermal injection
of 20 J-lg atropine. The intradermal injection of adrenaline after the cessation of the infusion
showed, however, that the glands were still responsive to adrenaline.
Intra-arterial infusions. The intra-arterial infusion of adrenaline in doses ranging from 4 to
10J-lgjmin in five unheated subjects produced a definite response. This response was not blocked
by 20 J-lg atropine administered just before the start of the infusion (Fig. 3a) or during the
infusion (Fig. 3b). The response was, however, blocked by 20 p,g phentolamine (Fig. 3b). The
infusion produced intense pallor in the whole of the arm peripheral to the infusion site. A
larger sweat response was obtained by keeping the arm warm with lamps.
The response failed to be maintained with continued infusion of adrenaline (Fig. 3a); with a
subsequent infusion of adrenaline little or no response occurred.
Five J-lgjmin noradrenaline in one subject also produced a response when infused intraarterially. The arterial infusion of 2 J-lgjmin of isoprenaline in one subject on the other hand was
not effective.
Effect ofparenteral infusion of catecholamines on active glands
Intravenous infusions. During heating of the subject, the intravenous infusion of 20 J-lgjmin
adrenaline whilst the glands were moderately active and still capable of an increased response
(ear temperature about 37,2°) did not enhance the thermal response (Fig. 4).
K. G. Foster, Jean Ginsburg and J. S. Weiner
828
Infusion on
==-1
Infusion off
5" Ad,,",I;~"";"
0·25,uL/min
Infusion
Infusion
off
on
[-5/'-9 /min---j
,
Adrenaline
I
20/,-9
Atropine
I
-{
t
11111111111111 II I 11I1 II !! III I I! 1111 II III1I III III II 1111 111111 I 1111111
Time (min)
Fig.3a
20/,-9 Phentolamine
(0'2mLl
t
t
0·06/,-Llmin
~
20/,-9 Atropine
(0·2mLl
i
20/,-9 Phentolamine
(0'2mLl
~
t-
5fl9/min Adrenaline
Infusion on
I
I II I
-t
Infusion off
I
I II
I
II
I
I
!
I
!
Time (min)
Fig.3b
FIG. 3. Forearm sweat gland response due to the intra-arterial infusions of adrenaline. Infusion
via the brachial artery. Capsule diameter = 1·0 em.
Intra-arterial infusions. The intra-arterial infusion of 5 flgfmin adrenaline during thermal
sweating caused the response to decrease (Fig. 5). The arterial infusion of2 flgfmin isoprenaline
in one subject did not, however, affect the response.
DISCUSSION
Acetylcholine and catecholamines compared
Even though noradrenaline and adrenaline can stimulate the glands either by intradermal
injection or intra-arterial infusion, the sweat output is small compared to that due to intradermal acetylcholine. The amplitude of the response to intradermal catecholamine was always
considerably less and the concentration of the effective threshold dose higher than that for
acetylcholine (Fig. lb). These results contrast sharply with those obtained by Chalmers &
Keele (1951), who found no difference between the adrenaline and acetylcholine thresholds.
829
Catecholamines and sweat gland activity
O·Ip.L/min
I
t
Infusion off
20,ug/min
Adrenaline
I
I
I
I
I
I
I
I
I
Time (min)
FIG.4. Effect of the intravenous infusion of adrenaline on the forearm sweat gland response due to
thermal stimulation. Infusion via the cubital vein of the opposite arm. Capsule diameter = 1·0
em.
O'16,uLlmin
Saline
2,ug/min
Isoprenaline
5,ug/min
Adrenaline
t
Infusion
off
I
I
I
I
I
I
I
I
I
J
I
!
J
J
I!!
I
I!'
,
J
J
Time (min)
5. Effect of the intra-arterial infusion of adrenaline and isoprenaline on the forearm sweat
gland response due to thermal stimulation. Infusion via the brachial artery leading directly to the
glands. Capsule diameter = 1·0 em.
FIG.
830
K. G. Foster, Jean Ginsburg and J. S. Weiner
It is very likely that their methods did not permit the detection of the more ephemeral responses
due to very low doses of acetylcholine.
The poorer response to adrenaline could be due to concomitant cutaneous vasoconstriction
depriving the gland of its oxygen supply (van Heyningen & Weiner, 1952; Hubbard & Weiner,
1969).
Acetylcholine and adrenaline differ also in their effects when administered repeatedly.
Whilst the sweat glands fail to respond to repeated very high doses of intradermal acetylcholine
(2-5 mgjO'05 ml) or methylcholine (400Jlg to 1 mgjO·05 ml) (Collins, Sargent & Weiner, 1959)
they will respond to repeated small doses (1-10 JlgjO·1 ml) with little or no diminution in
output. However, repeated small doses of intradermal catecholamine bring about a rapid
failure of response as does continued arterial infusion of adrenaline. Collins et al. (1959)
ascribed the reduction in sweat gland activity with repeated high doses of cholinomimetics to
desensitization of the receptors. Whether this accounts for the fall-off in output on repeated
catecholamine administration has not been explained. Thecausemay, however, be in the intense
vasoconstriction induced by repeated injection of catecholamine rendering the glands progressively more ischaemic. Since acetylcholine has a local vasodilator action this would explain
the finding that the response to maximal doses of acetylcholine is not affected by the previous
injection of catecholamine. Nevertheless, a reduction in the response due to submaximal doses
of acetylcholine occurs with the simultaneous or previous injection of subthreshold doses of
catecholamine.
Role of circulating catecholamines
No evidence was obtained indicating that adrenaline can potentiate the cholinergic response
as postulated by Kuno (1965). The injection of mixtures of near maximally active doses of
catecholamine and submaximally active doses of acetylcholine produced responses greater
than those due to acetylcholine alone, but the response for the mixture merely approximated
to the summation of the separate responses due to the two drugs. Nor was a maximal acetylcholine response increased by adrenaline. In fact, submaximal acetylcholine responses were
reduced by subthreshold doses of catecholamine administered simultaneously or before the
acetylcholine.
We may now examine the conditions under which the sweat glands might be stimulated by
circulating catecholamine in the intact organism. From our results this would require the
circulation of a very high level of catecholamine acting on glands which are not at the same
time undergoing nervous stimulation. When glands are thermally stimulated these high levels
of catecholamine have in fact an inhibitory effect (Fig. 5).
Sweat secretion did not occur with the intravenous infusion of 20 Jlgjmin adrenaline, which
would produce a concentration of approximately 2 Jlgjl adrenaline in the peripheral venous
blood (Vendsalu, 1960). As normal resting concentrations of catecholamines in the venous
plasma are only of the order of 0·4 Jlgjlitre (price & Price, 1957; Vendsa1u, 1960) even the
4-6-fold increases in the catecholamine level occurring in severe and prolonged exercise
(Haggendal, 1963) are unlikely to induce sweating. Furthermore, the thermal sweat response
was not increased by the intravenous infusion of 20 Jlgjmin adrenaline (Fig. 4) and so the
sweating induced by neural mechanisms during exercise is unlikely to be enhanced by secreted
catecholamine as suggested by Kuno (1965).
High plasma concentrations of catecholamines are produced in asphyxia; the output of
Catecholamines and sweat gland activity
831
catecholamines from the adrenal medulla can increase in the cat to 1 J1g kg- 1 min- 1 (von
Euler & Folkow, 1953) or 1·3 J1g kg- 1 min"? (Celander, 1954). When the asphyxia is so severe
that artificial respiration must be applied this rises to 2-2'5 J1g kg- 1 min -1 (Celander, 1954). A
level of 1-1'3 J1g kg- 1 min"! is equivalent to an output of 70-90 J1g/min for a 70 kg man. A
hot sweating subject has a forearm skin blood flow five times the resting level (Barcroft &
Hamilton, 1948), and the high catecholamine levels occurring during asphyxia are likely to
reduce the skin blood flow (Celander, 1954). An infusion of 20 J1g/min in a hot sweating subject
should be as effective in its action on the glands as an output of at least 100 J1g/min in an
asphyxiated individual, but this dose did not stimulate the glands.
In our experiments doses of adrenaline and noradrenaline (4-10 J1g/min) were found to be
definitely effective in inducing a response by the resting glands when infused arterially. Resting
forearm blood flow is of the order of 1-3 mll00 g tissue"? min- 1 (Hellon, Lind & Weiner,
1956), and as the forearm volumes are approximately 1·5 I, the concentration of catecholamine
reaching the glands would in these experiments have been of the order of 90-270 J1g/l. As the
level of catecholamine in the plasma during a phaeochromocytoma crisis can be as high as
98 J1g/1 (Lund, 1952) or 85 J1g/1 (Hermann & Mornex, 1964) it seems possible that under such
conditions a sweat secretion due to direct stimulation of sweat gland receptors might be induced
by endogenous catecholamine. These levels of catecholamine, however, reduce the response of
glands which are already active (Fig. 5). Our results show, therefore, that catecholamine.levels
in phaeochromocytoma crises may in some patient's be high enough directly to stimulate glands
which are in the resting state. However, probably due to an elevated body temperature (Pickering, 1955), the glands are likely to be under neural stimulation (Prout & Wardell, 1969),
and the vasoconstrictor effect of the catecholamine would reduce rather than elevate the response.
Many of the symptoms occurring in patients with hyperthyroidism closely resemble the
effects of overactivity of the sympathetic nervous system, and although the precise role of
catecholamines in determining the various manifestations of sympathetic overactivity (such as
tachycardia, tremor and sweating) in thyrotoxicosis is not fully understood, it is unlikely that
the sweat glands are responding to catecholamines released by the adrenal medulla as the
sweating is blocked by f3-receptor blocking agents such as propranolol.
REFERENCES
BARCROFT, H. & HAMILTON, G.T:C. (1948) Results of sympathectomy of the upper limb. Lancet, i, 441-444.
BARCROFT, H. & SWAN, H.J.C. (1953) Sympathetic Control of Human Blood Vessels, p. 118. Arnold, London.
CELANDER, O. (1954) The range of control exercised by the sympathico-adrenal system. Acta Physiologica
Scandinavica, 32, Suppl, 116.
CHALMERS, T.M. & KEELE, C.A. (1951) Physiological significance of the sweat response to adrenaline in man.
Journal ofPhysiology, 114, 510-514.
COLLiNS, K.J., SARGENT, F. & WEINER, J.S. (1959) Excitation and depression of eccrine sweat glands by acetylcholine, acetyl-ji-methylcholine and adrenaline. Journal ofPhysiology, 148, 592-614.
FOSTER, K.G. (1968) Response of the eat's pad eccrine sweat glands to intravascular injections of catecholamines.
Journal ofPhysiology, 195, 331-337.
FOSTER, K.G. & WEINER, J.S. (1967) The continuous monitoring of sweat secretion of man and cat using a
ventilated capsule and an infra red analyser. Journal ofPhysiology, 191, Ip-2p.
GADDUM, I.H., PEART, W.S. & VOGT, M. (1949) The estimation of adrenaline and alIied substances in blood.
Journal ofPhysiology, 108,467-481.
832
K. G. Foster, Jean Ginsburg and J. S. Weiner
HAGGENDAL, J. (1963) The presence of conjugated adrenaline and noradrenaline in human blood plasma. Acta
Physiologica Scandinavica, 59, 255-260.
HELLON, R.F., LIND, A.R. & WEINER, J.S. (1956) The physiological reactions of men of two age groups to a hot
environment. Journal ofPhysiology, 133, 118-131.
HERMANN, H. & MORNEX, R. (1964) Human Tumours Secreting Catecholamines, Pergamon Press, London and
New York.
HUBBARD, J.L. & WEINER, J.S. (1969) Effect of hyperbaric oxygen on human sweat with particular reference to
lactate content. Journal of Applied Physiology, 27, 715-720.
KUNO, Y. (1965) The mechanism of human sweat secretion. Proceedings of the International Union ofPhysiological Sciences, 4, 3-16.
LUND, A. (1952) Adrenaline and noradrenaline in blood and urine in cases ofphaeochromocytoma. Scandinavian
Journal of Clinical and Laboratory Investigation, 4, 263-265.
PICKERING, G.W. (1955) High Blood Pressure, p, 404. Churchill, London.
PRICE, H.L. & PRICE, M.L. (1957) The chemical estimation of epinephrine and norepinephrine in human and
canine plasma. Journal ofLaboratory and Clinical Medicine, 50, 769-777.
PROUT, RJ. & WARDELL, W.M. (1969) Sweating and peripheral blood flow in patients with phaeochromocytoma, Clinical Science, 36, 109-117.
VAN HEYNINGEN, R.E. & WEINER, J.S. (1952) The effect of arterial occlusion on sweat composition. Journal
ofPhysiology, 116,404-413.
VENDSALU, A. (1960) Studies on adrenaline and noradrenaline in human plasma. Acta Physiologica Scandinavica,
49, Suppl. 173.
VON EULER, U.S. & VOLKOW, B. (1953) Einfluss verschiedener afferenter Nervenreize aug die Zusammensetzung des Nebennierenmarkinkretes bei der Katze. Archiv fiir experlmentelle Pathologie und Pharmakologie,
219,242-247.