Acta Physiol Scand 1992, 146, 385-392
Non-linearity of skin resistance response t o intraneural electrical stimulation of sudomotor nerves M. KUNIMOTO", K. K I R N O t , M. ELAM", T. KARLSSON" and B. G. WALLIN" T h e Departments of Clinical Neurophysiology and Hospital, University of Goteborg, Sweden
"
t Anaesthesiology,
Sahlgren's
KUNIMOTO, M., KIRNO,K., ELAM,M., KARLSSON, T. & WALLIN,B. G. 1992. Nonlinearity of skin resistance response to intraneural electrical stimulation of sudomotor nerves. Acta Physiol Scand 146, 383-392. Received 31 January 1992, accepted 22 May 1992. ISSN 0001-6772. Departments of Clinical Neurophysiology and Anaesthesiology, University of Goteborg, Sweden. Intraneural electrical stimuli (0.3 mA, 0.2 ms) were delivered via a tungsten microelectrode inserted into a cutaneous fascicle in the median nerve at the wrist in 16 normal subjects, and the effects on the sweat glands within the innervation zone were recorded as changes of skin resistance. In order to examine the relationship between the skin resistance level and the amplitude of transient resistance responses, trains of high frequency stimulation were used to reduce the skin resistance level and then transient resistance responses were evoked by single stimuli at 0.1 Hz. Regional anaesthesia of the brachial plexus in the axilla eliminated spontaneous sympathetic activity and reflex effects. At high skin resistance levels response amplitudes to single stimuli were low but they increased successively to a maximum at intermediate levels and then decreased again at low resistance levels. Repeated stimulation sequences evoked qualitatively similar response curves but quantitatively both response amplitudes and skin resistance levels were slightly reduced upon repetition. We suggest that the changes of response amplitudes are due to variable resistivity of the corneal layer. The shifts of the response curves with repetition of stimulation may result from increased hydration of the corneum. It is concluded that the variability of response amplitudes to constant stimuli makes the amplitude of a skin resistance response unsuitable as an indicator of the strength of sympathetic sudomotor nerve traffic. Key words : Galvanic skin response, microneurography, sweat glands.
Neuro-effector transfer in human sweat glands can b e studied by intraneural electrical stimulation of sympathetic sudomotor nerve fibres and recordings of skin resistance within the innervation zone (Wallin et al. 1983). Using this technique it was found that the transient skin resistance reduction evoked by a constant stimulus varied in amplitude depending on the number Correspondence : B. Gunnar Wallin, Department of Clinical Neurophysiology, Sahlgren's Hospital, S413 45 Goteborg, Sweden.
and frequency of previous stimuli and the time elapsed from the last stimulus (Wallin et al. 1983, Kunimoto et al. 1991). Similarly, the skin resistance response to a short train of impulses varied markedly in amplitude depending on the background frequency of stimulation (Kirno
et al. 1991). T h e reason for the variability of response amplitude is largely unknown. T h e skin resistance level may, however, be a factor of importance, since at low stimulation frequencies the transient responses evoked by each stimulus
385
386
.M. Kuninioto et al.
increased in amplitude a t the same time as the skin resistance level n.as reduced ( K u n i m o t o et al. 1091). -4gainst this background the present stud!- was undertaken to test systematicall! if and how the amplitude of transient responses in skin resistance are related t o the skin resistance level. ' l o this end we delivered trains of impulses of. different duration a n d frequency to induce reductions in skin resistance and then studied the responses to single stimuli.
.Ifter approval ot'the local ethics committee and with the nritten consent o f each subject, investigations were made on 18 health!- females aged 21-38 !-ears. .\*urr ruc.nvdr,r~ urrd . s I ~ ~ N u / ~ ~ llicro-electrode ~oN. recordings and intraneural electrical stimulations I\ ere made with tungsten and micro-electrodes, insul.ited b! \ olta lacquer, with a shaft diameter of 0 . 2 mm and a tip ofa few pm, impaled in cutaneous fascicles o f the median nerve a t the wrist (for technical details see \allbo p t 111. 1979, Kunimoto et u i . 1991). For intraneural electrical stimulation the same electrodes n ere connected to the output of a constant-current stimulator. (Grass 548 v,ith constant-current unit C(.~LI.I ; Grass Instr.) deli\-ering square-n-a\-epulses (03m.1. duration 0.2 ms). T o rapidly reduce the skin resistance lewl, trains of var>-ing frequenc! and duration \\ere used (0.5-20 Hz, duration 0.2.i-184 s). .S'k//? ves/s'tnric.r i.hangt~.c. Skin resistance (gal! anic skin response, g.s.r) was measured b!- modified van Gough g.s.r modules (type IGSR/7.\) usuall!- using .ig/.IpCI electrodes (lledicotest, Olst!-ckLe, Denmark) with a rectangular area of 5 . 5 x 4 mm. One pair o f electrodes \\as used in nine esperiments and two pairs in nine. The resistance records were displd!-ed nith two filter settings, one nithout IOU cut tilter (l>(:-display) and the other nith a Ioiv cut filter at 0.7 i i z (.l(;-displa!). T h e high cut filter \\as set at .NHz in elelen subjects and 12 Hz in sel-en depending on the aniplitier gain. \\-hen skin resistance \vas measured at tbo sites the filters nere a l ~ a y sidentical for both sites. T h e electrode gel contained 0.1 \ t SaCl in I 0 and 0.05 11in one subject. In fi\-e esperiments 0.1 11 \\-as used for one pair of electrodes and 0.0.i M for the second pair. The gel had a contact area with the skin of0.5 cm2.There was no s!-stematic difference in results related to the gel concentration. In t n o experiments skin resisrance records were obtained simultaneousl!- using one pair of ('wet ') electrodes containing thc 0.05l1 S a C l gel and one pair of ( ' dr! ') electrodes made bl- a conductive semisolid acrylate polymer containing KCI and attached to a tin foil (2.32.; \-P, Littnyan, 3JICo, St Paul, \IS, LS.1). -Imz!),u.s. For determination of the levels and response amplitudes of skin resistance the analogue
signals were fed into a PDP 11/70 computer (sampling frequenc! 96 Hz) or a personal computer (sampling frequent!. 200 Hz). T o correlate the amplitude ofthe transient response to a single stimulus with the skin resistance level the resistance level was measured by averaging all data points during a 0.3-s period after deliver!- of the stimulus. This was justified, since the latent!- to the start of the response to a single impulse is 0.6 s (Kunimoto et a / . 1991). T h e amplitude of the transient response was measured as the difference between the minimum and maximum value after the stimulus in the X(:-curves. T o calculate the mean slope of the DC-resistance curve between responses to two consecutive stimuli the slope of the line connecting the resistance values a t the delivery of the t h o stimuli was used (and expressed as k 0 s-'). T h e slope was compared with the amplitude of the second response. \.slues are given as means & SERI. E.~.per.~nirrita/ pro~.et/irre.The procedures have been described in detail previously (Kunimoto et ul. 1991). In short the microelectrode was inserted into the median nerve a t the kvrist and the receptive field of the impaled fascicle was mapped b! mechanical stimuli. The skin resistance electrodes were positioned within the inner\-ation zone of the impaled nerve fascicle. If t n o pairs of skin resistance electrodes were used, one electrode in each pair mas put outside the receptive field. .I common reference electrode was used in two esperiments and then the polarity was reversed between the pairs which made the net current in the conimon reference electrode almost zero. T o eliminate spontaneous and reflex sympathetic activity to the hand and tn enable stimulation of C: fibres without e\-oking pain, the brachial plexus was blocked by regional anaesthesia iiith 35-4.5 ml of 1n o carbocaine (Selander 1977). After around 30 min complete anaesthesia had developed in the hand. ?'he room temperature \\-as 13-25 "C and the skin temperature of the examined fingers varied before anaesthesia (20.0-34.4 "C) but became 35.7 "C: on the average (.35.3--.36.3 "C) after the block and did not change during experiment. Additional regional anaesthesia w a s given as needed
RESULTS Efer.ts qf s m l l or moderate redzirtions of the skirr mistunre l e d -1s described previously ( K u n i m o t o et a/. 1991), the first skin resistance responses to 0.1 Hz single pulse stimulation (initiated after 5-15 m i n of inactirity) appeared after u p to 15 stimuli. K i t h continuing stimulation transient responses occurred after each stimulus; there was a slow reduction of the skin resistance level a n d
Non-linearity of skin resistance response
a
387
Skin resistance r310 kO
DC
- 340
b
320
330
340 kO
Skin resistance (DC)
Fig. 1. (a) Skin resistance responses evoked by 0.1 Hz single pulse stimulation beginning approximately 30 min after the application of local anaesthesia. At 'T' the single stimulus was replaced by a train (six impulses, 20 Hz). The first response was observed following the seventh stimulus (arrow). Note the marked potentiation of the response amplitudes after train. * Indicates sudden spontaneous reduction of resistance. Reduction of skin resistance shown as upward deflection in this and all subsequent figures. (b) The relationship between the skin resistance level (DC) and the change of skin resistance (AC amplitude) for the sequence shown in Figure la. Symbols are open before and closed after the train.
concomitantly the transient responses increased in amplitude (Fig. la). When an approximate steady state was reached the effect of reducing the skin resistance level was examined by giving a short train (six impulses, 20 Hz) between single stimuli a t 0.1 Hz in seven subjects. As illustrated in Figure 1(a) the train led to a rapid skin resistance reduction and potentiation of the responses to single stimuli. When the amplitudes of all the transient responses to single stimuli in Figure l ( a ) were plotted against the skin resistance level at each stimulus, there was a negative linear relationship with high correlation coefficient (Y = -0.96, Fig. 1b). Potentiation effects were observed also in the other six subjects but the relationship between skin resistance level and response amplitude was not always linear.
T o test the reproducibility of the relationship between the skin resistance level and amplitude of transient responses, 0.1 Hz single pulse stimulation was given continuously in two subjects but every tenth stimulus was replaced by a train (six impulses, 20 Hz). Qualitatively, the train-induced reduction of skin resistance and the potentiation of the amplitude of transient responses were reproducible after the second train (Fig. 2a). However, the response amplitudes tended to diminish slightly with repetition of the train. Figure 2(b) summarizes data for six sequences of nine responses to single stimuli, starting with the nine single stimuli immediately preceding the first train. Data points are averaged from four sites in two subjects. F o r each sequence the relationship between response amplitude and skin resistance level was approximately linear
388
M. Kunimoto et al.
a Skin resistance
Stimulation
I I- !1 I1 1
T
T
T
T-
1 min
b
k0
2'o 1.5
T
1
1.0
.5'
.o
250
270
290
kO
Skin resistance (DC) Fig. 2. (a) Skin resistance responses evoked by 0.1 Hz single pulse stimulation where every 10th stimulus was replaced by a train of six impulses at 20 Hz (indicated by T). The first response increase of response amplitudes concomitant with the gradual decrease of skin resistance and the potentiation of response amplitudes after trains 2-5. (b) Relationship between skin resistance level (DC) and change of skin resistance (AC amplitude) evoked by single pulse stimulation (0.1 Hz) during six consecutive sequences of stimuli (nine pulses in each) separated from each other by a single train (six impulses, 20 Hz) as in a. Separate symbols for each series as shown in the figure. %leandata from four sites in two subjects occurred after the third stimulus (arrow). Note the initial SEhI for the amplitude (not shown) was around 0.5 kohm during first and second series and decreased successively to around to 0.15 kohm during the sixth series.
and the slopes of the regression lines were similar. T h e data points for the first two sequences (before and after the first train) showed a reasonable fit to a common regression line. F o r response sequence 3-6 (after train 2-5) the regression lines were shifted successively to the left and slightly downwards (lines not shown in the figure) indicating that response amplitudes to single stimuli tended to diminish in spite of successive small reductions of the skin resistance level after each train.
Efeects of large reducttons lei.eI.
of
the skin resistunce
U hen 20 €32 trains were given for 25-38 s in three subjects (six skin sites), a rapid maximal
reduction of skin resistance occurred (cf Kunimoto et al. 1991). In most cases an approximately stable resistance level was reached during the stimulation but sometimes resistance started to increase slow~lyagain despite continuing stimulation (cf Fig. 3a). After termination of the high frequency train, resistance started to return towards the starting level, first rapidly, then more slowly. T h e first responses to 0.1 Hz single stimuli after the end of the train were usually small but they grew successively in amplitude during the rapid phase of skin resistance return, reached a maximum after 3-10 stimuli and then started to diminish again. Figure 3 (a) shows experimental records from the distal skin site in one subject and Figure 3(b) summarizes quantitatively the relationship between response
Non-linearity of skin resistance response
a
389
Skin resistance
[ lkn
AC
Stimulation liillilllllllllllllllilllllllillllllilil 20 HZ
OIHZ
u 1 min
h
7i 2 0 -
f
5
16-
“0 -4 0-
12-
aa 0 0 c c
.a-
22 0 .cn
4-
cn
!!
0-
16-
- -
amplitude to single stimuli and skin resistance level for both the proximal and the distal site in the same subject. Results were also similar in both records when response curves were obtained simultaneously with one pair of ‘wet’ electrodes and one pair of ‘dry’ electrodes (2 expts). I n one subject (two skin sites) the same procedure was repeated three times during the same experiment (20 H z initial train for 25 s, followed by 0.1 Hz single stimuli). Qualitatively, findings were similar every time but quantitatively the response amplitudes to the single pulses became successively smaller and the response curves were shifted to the left (i.e. towards lower skin resistance levels) and downwards (Fig. 3 c). Figure 3(b) and (c) also illustrate that although response curves from two sites in the same experiment usually were qualitatively similar, quantitative differences were nevertheless fairly 14
..>
0’
common. For example, in Figure 3(b) response amplitudes to single stimuli had their peak at the fourth response in the proximal and at the ninth in the distal site. Peaks in response amplitudes similar to those in Figure 3 were observed in all three subjects (six skin sites) in whom the initial train frequency was 20 Hz (averaged data shown in Fig. 4).T h e effect was present also if the frequency of the initial train was 10 Hz (six subjects) and the frequency of the single stimuli 0.1 H3 (four subjects, five sites) or 0.2 H 3 (two subjects, three sites), but not if the frequency of the initial train was 0.5 or 1.5 Hz. This is summarized in Figure 5 which shows the effects of varying the train frequency from 0.5 to 1.5 to 10Hz in four subjects (five sites). T h e peak in amplitude a t 10 Hz occurred when the skin resistance level was below approximately 490 kR and the A C T 146
390
M. Ktrninioto et al.
-.
I
16
-16
12
-12
-
0
@
g
-
m
0
- 8 2 -
4
- 4
0
- 0
I
X
3
> H
I
0
1
2
3
4
5min
Time
Fig. 1. Changes of response amplitude (.XC-amplitude), skin resistance and slope of the DC-curve during single pulse stiniulation a t 0.1 Hz folloaing a 20 Hz train for 25-38 s. Averaged data from six skin sites in three subjects. Yertical bars indicate SElL For clarity SEMs were omitted for skin resistance (ralues were large due to differences in skin resistance between skin sites but the range of variabilit! was small (77-82 kohm).
-
b
a
I
C -1
2
-10
w 5
:
0
- a -6
8
1
- 4
=
- 2
5
- 0
0
1
2
3
4
5
0
1
3
2
4
5
0
1
2
3
4
-
5min
Time
Fig. 5. T h e same relationship as in Figure 4 obtained after train stimulation at 0.5 (a), 1.5 (b) and 10 Hz (c) for 96-184, 35-171 and 49-90 s, respectively. Averaged data from five skin sites in four subjects. Note that onl!- in ( c ) did response amplitudes first increase and then decrease.
slope of DC-curves ahore 0.5 kR s-', Yalues m-hich were not reached at 0.3 and 1.5 Ilz.
DISCUSSION The present results confirm our previous observations (IVaIlin el 01. 1983, Kirno et n l . 1991, Kuninioto et (11. 1991) that a g i x n sudomotor
nerve stimulus may induce skin resistance responses of very variable amplitude. Although the resuonse amplitudes were related to the skin resistance level, the relationship was complex and neither linear nor constant. T o a large extent skin resistance is determined by the amount of sweat (and its concentration of electrolytes) in the sweat ducts (Fowles 1974).
Non-linearity ojskin resistance response The resistance of the corneal layer of the skin itself is high but may be reduced by hydration (e.g. infiltration of sweat) (cf Fowles 1974). During -on-going low frequency nerve stimulation the small volume of sweat produced by each impulse leads to a decrease of resistance which slowly decays due to reabsorption of electrolytes and water (Edelberg 1971, Sat0 1977) (since the recording site is covered by the skin resistance electrode, evaporation of water from the orifice cannot occur). By using an initial high frequency train as a conditioning stimulus the responses to single impulses could be studied at different resistance levels, i.e. different degrees of filling of the sweat ducts. Using this procedure we found that response amplitudes were low both at high and low resistance levels and had a clear maximum in between. Our data give no direct insight into the underlying mechanisms but several factors may contribute: (1) If the amount of sweat produced in the gland per impulse varied with the degree of duct filling (i.e. the skin resistance level) this could lead to varying response amplitudes. We cannot exclude this alternative at the beginning of a low frequency stimulation series (when response amplitudes grow slowly concomitant with a slight skin resistance reduction). It seems unlikely, however, that the sweat volume produced in the gland per impulse should fall at low resistance levels (which under physiological conditions occur only when sweating is high). If this nevertheless were the case it would constitute a local negative feedback mechanism that would limit excessive sweating; (2) another possibility would be that there is loss of water through the ductal wall which varies in speed in proportion to the degree of duct filling. Loss of water from the duct into the corneal layer of the skin does occur but the rate is fairly slow (Adams 1966, Edelberg 1971). This mechanism may be responsible for effects that occur over minutes (see below) but would probably be too slow to significantly affect peak amplitudes of our transient responses (the peak occurs after approximately 1.6 s, cf Kunimoto et al. 1991); and (3) a more likely alternative is that the longitudinal resistivity of the empty duct varies in different parts of the duct. Skin resistance to current flow occurs mainly between the germinating layer and the upper surface of the corneum (cf Fowles 1974). This means that the skin resistance can change appreciably only when the sweat moves
39 1
within this layer, i.e. mainly within the corneum. If then the resistivity were low in the bottom part of the layer, high in the middle part and then lower again towards the top, the addition of a constant volume of sweat could explain the variations of response amplitudes found in the present study. A variable resistivity is likely to exist for structural reasons. T h e corneal layer contains dead cells which become older and drier the closer to the skin surface they are situated (cf Peiss et al. 1956) and Suchi (1955) found an increasing resistivity of the corneal layer and the empty duct all the way up to the skin surface. This does not agree with our finding of lower response amplitudes at minimal resistance levels (i.e. when the ducts are full or almost full). Theoretically, there may be an artefactual gradient of reduced resistivity near the skin surface due to the electrode gel hydrating the upper corneum. In addition, excessive hydration of the corneal layer may result in closure of the sweat duct opening on the skin (Randall et a/. 1957, Sarkany et al. 1965) and Fowles (1974) has suggested that this may reduce amplitudes of resistance responses. It is, however, unlikely that excessive hydration from the electrode gel can explain our findings since the changes of response amplitudes were similar with both ‘wet’ and ‘dry’ resistance electrodes. Another structural factor which could contribute to a variable resistivity would be variations of the ductal anatomy. For example, the sweat duct has a spiral pathway in the corneum and towards the surface there is a widening of the lumen (Hambrick & Blank, 1954, Montagna 1956) which may tend to reduce resistivity in that part. In addition to being non-linear the relationship between skin resistance levels and response amplitudes to single stimuli was shifted successively to the left and downwards with repetition of the stimulation sequences (Figs 2 & 3 ) . This finding indicates that response amplitudes diminished slightly between sequences in spite of a decrease of the skin resistance level. 4 t the beginning of stimulation (i.e. about 30 min after local anaesthesia) the basal epidermal layer probably was relatively dry and therefore it can be expected that the sweat produced by the stimulation caused a slow successive increase of the hydration of those structures (Adams 1966). This probably explains the shift of the response curves, since an increased hydration would reduce skin resistance (see above). 162
392
‘21. Kirnimoto et al.
The present findings have important practical consequences. ,Measurements of skin resistance changes are often used as indicators of the strength of sympathetic sudomotor nerve traflic in phJ-siological, psychological or clinical studies. The marked variabilitl- of the responses to a constant stimulus makes such data difficult to interpret. When skin resistance levels are reasonably constant and stimuli delivered with fairl!long intervals the errors may be limited (Bini et ul. 1980, Lidberg et a/. 1981). If, however, the background sudomotor nerve traffic changes during a study (e.g., with changing environmental temperature, stress or anaesthesia) the concomitant changes of skin resistance level will make measurements of skin resistance response amplitudes almost meaningless. Supported b! the Swedirh Medical Research Council Grant N o B91-04X-03546-20A and b! traxel grants to Llasanari Kunimoto from Japan Societ! for the Pi omotion of Science
REFERENCES AXIS IS, T. 1966. Characteristics of eccrine sweat gland actkit!- in the footpad of the cat. J .Jpp/ Ph,Jsrol 2 1 1004-1 0 12. B IX I , G., HAGBARTH, K.-E., HTSNINES, P. & i v A L I . I S , H . G . 1980. Thermoregulator! and rh!-thm-generating mechanisms governing the sudomotor and vasoconstrictor outflow in human cutaneous nerves. 1 Pliysiol (Lond) 306, 537-552. EDELBERG, R. 1971. Electrical properties of skin. In: H.R. Weiss (ed.) Biopli,)tsical Properties of the Skin, 1.01. I , pp. 713-550. LViley, Yew Yo& FOWLES. D.C. 1974. Mechanisms of electrodermal activitl-. I n : R.F. Thompson 8r X I . X l . Patterson (eds) Biodectric Recording Techniques, Part C, ~
Receptor and E#ector Processes, pp. 231-27 1. Academic Press, New York. HAMBRICK, G.W. JR. & BLANK,H. 1954. Whole mounts for the study of skin and its appendages. 3 Inrest Drrmatol 23, 437-453. KIRNO,K., KUNIMOTO, M., LUNDIN, S., ELAM,M. & WALLIN,B.G. 1991. Can galvanic skin response be used as a quantitative estimate of sympathetic nerve activity in regional anesthesia? Anesth Analg 73, 138-142. KCNIMOTO, M ,KIRYO,K., ELAM,M. & WALLIN, B.G. 1991. Neuro-effector characteristics of sweat glands in the human hand activated by regular stimuli. 3 Ph)Jsio/(Lond) 442, 391-411. LIDBERG, L. & WALLIN,B.G. 1981. Sympathetic skin nerve discharges in relation to amplitude of skin resistance responses. Psychoph.ysio/ 18, 268-270. MOSTAGNA, W.1956. T h e eccrine sweat glands. I n : The Structure and Function of Skin, pp. 89-127. Academic Press, New York. PEtss, C.N., RANDALL, W.C. & HERTZMAN, A.B. 1956. Hydration of the skin and its effect on sweating and evaporative water loss. 3 Invest Dermatol 26, 459-470. RINDALL,W.C. & PEISS,C.N. 1957. T h e relationship between skin hj-dration and the suppression of sweating. 3 Inrest Dermatol 28, 435-441. SQRKASY, I, SHLISTER, s. & STAMMERS, M.C. 1965. Occlusion of the sweat pore by hydration. Brit 3 Dermatol 77, 101-104. S4T0, K. 1977. T h e physiology, pharmacology, and biochemistry of the eccrine sweat gland. Rev P h j d Biorheni Pharnrarol 79, 5 1-131. SELANDER, D. 1977. Catheter technique in axillary plexus block. Actri Anaesthesiol Scand 21, 321t-329. SLXHI, T . 1955. Experiments on electrical resistance of the human epidermis. 3ap 3 Pitysio/ 5, 75-80. \-.ALLBO, A.B., HAGBARTH, K.-E., TOREBJORK, H.E. & WALLIN, B.G. 1979. Somatosensory, proprioceptive and sympathetic activity in human peripheral nerves. Phjlsiol R e v 59, 919-957. \VALLI’V,B.G., BLUMBERG, H. & HYNNINEN, P. 1983. Intraneural stimulation as a method to study synpathetic function in the human skin. Neurosci Lett 36, 189-194.
Non-linearity of skin resistance response t o intraneural electrical stimulation of sudomotor nerves M. KUNIMOTO", K. K I R N O t , M. ELAM", T. KARLSSON" and B. G. WALLIN" T h e Departments of Clinical Neurophysiology and Hospital, University of Goteborg, Sweden
"
t Anaesthesiology,
Sahlgren's
KUNIMOTO, M., KIRNO,K., ELAM,M., KARLSSON, T. & WALLIN,B. G. 1992. Nonlinearity of skin resistance response to intraneural electrical stimulation of sudomotor nerves. Acta Physiol Scand 146, 383-392. Received 31 January 1992, accepted 22 May 1992. ISSN 0001-6772. Departments of Clinical Neurophysiology and Anaesthesiology, University of Goteborg, Sweden. Intraneural electrical stimuli (0.3 mA, 0.2 ms) were delivered via a tungsten microelectrode inserted into a cutaneous fascicle in the median nerve at the wrist in 16 normal subjects, and the effects on the sweat glands within the innervation zone were recorded as changes of skin resistance. In order to examine the relationship between the skin resistance level and the amplitude of transient resistance responses, trains of high frequency stimulation were used to reduce the skin resistance level and then transient resistance responses were evoked by single stimuli at 0.1 Hz. Regional anaesthesia of the brachial plexus in the axilla eliminated spontaneous sympathetic activity and reflex effects. At high skin resistance levels response amplitudes to single stimuli were low but they increased successively to a maximum at intermediate levels and then decreased again at low resistance levels. Repeated stimulation sequences evoked qualitatively similar response curves but quantitatively both response amplitudes and skin resistance levels were slightly reduced upon repetition. We suggest that the changes of response amplitudes are due to variable resistivity of the corneal layer. The shifts of the response curves with repetition of stimulation may result from increased hydration of the corneum. It is concluded that the variability of response amplitudes to constant stimuli makes the amplitude of a skin resistance response unsuitable as an indicator of the strength of sympathetic sudomotor nerve traffic. Key words : Galvanic skin response, microneurography, sweat glands.
Neuro-effector transfer in human sweat glands can b e studied by intraneural electrical stimulation of sympathetic sudomotor nerve fibres and recordings of skin resistance within the innervation zone (Wallin et al. 1983). Using this technique it was found that the transient skin resistance reduction evoked by a constant stimulus varied in amplitude depending on the number Correspondence : B. Gunnar Wallin, Department of Clinical Neurophysiology, Sahlgren's Hospital, S413 45 Goteborg, Sweden.
and frequency of previous stimuli and the time elapsed from the last stimulus (Wallin et al. 1983, Kunimoto et al. 1991). Similarly, the skin resistance response to a short train of impulses varied markedly in amplitude depending on the background frequency of stimulation (Kirno
et al. 1991). T h e reason for the variability of response amplitude is largely unknown. T h e skin resistance level may, however, be a factor of importance, since at low stimulation frequencies the transient responses evoked by each stimulus
385
386
.M. Kuninioto et al.
increased in amplitude a t the same time as the skin resistance level n.as reduced ( K u n i m o t o et al. 1091). -4gainst this background the present stud!- was undertaken to test systematicall! if and how the amplitude of transient responses in skin resistance are related t o the skin resistance level. ' l o this end we delivered trains of impulses of. different duration a n d frequency to induce reductions in skin resistance and then studied the responses to single stimuli.
.Ifter approval ot'the local ethics committee and with the nritten consent o f each subject, investigations were made on 18 health!- females aged 21-38 !-ears. .\*urr ruc.nvdr,r~ urrd . s I ~ ~ N u / ~ ~ llicro-electrode ~oN. recordings and intraneural electrical stimulations I\ ere made with tungsten and micro-electrodes, insul.ited b! \ olta lacquer, with a shaft diameter of 0 . 2 mm and a tip ofa few pm, impaled in cutaneous fascicles o f the median nerve a t the wrist (for technical details see \allbo p t 111. 1979, Kunimoto et u i . 1991). For intraneural electrical stimulation the same electrodes n ere connected to the output of a constant-current stimulator. (Grass 548 v,ith constant-current unit C(.~LI.I ; Grass Instr.) deli\-ering square-n-a\-epulses (03m.1. duration 0.2 ms). T o rapidly reduce the skin resistance lewl, trains of var>-ing frequenc! and duration \\ere used (0.5-20 Hz, duration 0.2.i-184 s). .S'k//? ves/s'tnric.r i.hangt~.c. Skin resistance (gal! anic skin response, g.s.r) was measured b!- modified van Gough g.s.r modules (type IGSR/7.\) usuall!- using .ig/.IpCI electrodes (lledicotest, Olst!-ckLe, Denmark) with a rectangular area of 5 . 5 x 4 mm. One pair o f electrodes \\as used in nine esperiments and two pairs in nine. The resistance records were displd!-ed nith two filter settings, one nithout IOU cut tilter (l>(:-display) and the other nith a Ioiv cut filter at 0.7 i i z (.l(;-displa!). T h e high cut filter \\as set at .NHz in elelen subjects and 12 Hz in sel-en depending on the aniplitier gain. \\-hen skin resistance \vas measured at tbo sites the filters nere a l ~ a y sidentical for both sites. T h e electrode gel contained 0.1 \ t SaCl in I 0 and 0.05 11in one subject. In fi\-e esperiments 0.1 11 \\-as used for one pair of electrodes and 0.0.i M for the second pair. The gel had a contact area with the skin of0.5 cm2.There was no s!-stematic difference in results related to the gel concentration. In t n o experiments skin resisrance records were obtained simultaneousl!- using one pair of ('wet ') electrodes containing thc 0.05l1 S a C l gel and one pair of ( ' dr! ') electrodes made bl- a conductive semisolid acrylate polymer containing KCI and attached to a tin foil (2.32.; \-P, Littnyan, 3JICo, St Paul, \IS, LS.1). -Imz!),u.s. For determination of the levels and response amplitudes of skin resistance the analogue
signals were fed into a PDP 11/70 computer (sampling frequenc! 96 Hz) or a personal computer (sampling frequent!. 200 Hz). T o correlate the amplitude ofthe transient response to a single stimulus with the skin resistance level the resistance level was measured by averaging all data points during a 0.3-s period after deliver!- of the stimulus. This was justified, since the latent!- to the start of the response to a single impulse is 0.6 s (Kunimoto et a / . 1991). T h e amplitude of the transient response was measured as the difference between the minimum and maximum value after the stimulus in the X(:-curves. T o calculate the mean slope of the DC-resistance curve between responses to two consecutive stimuli the slope of the line connecting the resistance values a t the delivery of the t h o stimuli was used (and expressed as k 0 s-'). T h e slope was compared with the amplitude of the second response. \.slues are given as means & SERI. E.~.per.~nirrita/ pro~.et/irre.The procedures have been described in detail previously (Kunimoto et ul. 1991). In short the microelectrode was inserted into the median nerve a t the kvrist and the receptive field of the impaled fascicle was mapped b! mechanical stimuli. The skin resistance electrodes were positioned within the inner\-ation zone of the impaled nerve fascicle. If t n o pairs of skin resistance electrodes were used, one electrode in each pair mas put outside the receptive field. .I common reference electrode was used in two esperiments and then the polarity was reversed between the pairs which made the net current in the conimon reference electrode almost zero. T o eliminate spontaneous and reflex sympathetic activity to the hand and tn enable stimulation of C: fibres without e\-oking pain, the brachial plexus was blocked by regional anaesthesia iiith 35-4.5 ml of 1n o carbocaine (Selander 1977). After around 30 min complete anaesthesia had developed in the hand. ?'he room temperature \\-as 13-25 "C and the skin temperature of the examined fingers varied before anaesthesia (20.0-34.4 "C) but became 35.7 "C: on the average (.35.3--.36.3 "C) after the block and did not change during experiment. Additional regional anaesthesia w a s given as needed
RESULTS Efer.ts qf s m l l or moderate redzirtions of the skirr mistunre l e d -1s described previously ( K u n i m o t o et a/. 1991), the first skin resistance responses to 0.1 Hz single pulse stimulation (initiated after 5-15 m i n of inactirity) appeared after u p to 15 stimuli. K i t h continuing stimulation transient responses occurred after each stimulus; there was a slow reduction of the skin resistance level a n d
Non-linearity of skin resistance response
a
387
Skin resistance r310 kO
DC
- 340
b
320
330
340 kO
Skin resistance (DC)
Fig. 1. (a) Skin resistance responses evoked by 0.1 Hz single pulse stimulation beginning approximately 30 min after the application of local anaesthesia. At 'T' the single stimulus was replaced by a train (six impulses, 20 Hz). The first response was observed following the seventh stimulus (arrow). Note the marked potentiation of the response amplitudes after train. * Indicates sudden spontaneous reduction of resistance. Reduction of skin resistance shown as upward deflection in this and all subsequent figures. (b) The relationship between the skin resistance level (DC) and the change of skin resistance (AC amplitude) for the sequence shown in Figure la. Symbols are open before and closed after the train.
concomitantly the transient responses increased in amplitude (Fig. la). When an approximate steady state was reached the effect of reducing the skin resistance level was examined by giving a short train (six impulses, 20 Hz) between single stimuli a t 0.1 Hz in seven subjects. As illustrated in Figure 1(a) the train led to a rapid skin resistance reduction and potentiation of the responses to single stimuli. When the amplitudes of all the transient responses to single stimuli in Figure l ( a ) were plotted against the skin resistance level at each stimulus, there was a negative linear relationship with high correlation coefficient (Y = -0.96, Fig. 1b). Potentiation effects were observed also in the other six subjects but the relationship between skin resistance level and response amplitude was not always linear.
T o test the reproducibility of the relationship between the skin resistance level and amplitude of transient responses, 0.1 Hz single pulse stimulation was given continuously in two subjects but every tenth stimulus was replaced by a train (six impulses, 20 Hz). Qualitatively, the train-induced reduction of skin resistance and the potentiation of the amplitude of transient responses were reproducible after the second train (Fig. 2a). However, the response amplitudes tended to diminish slightly with repetition of the train. Figure 2(b) summarizes data for six sequences of nine responses to single stimuli, starting with the nine single stimuli immediately preceding the first train. Data points are averaged from four sites in two subjects. F o r each sequence the relationship between response amplitude and skin resistance level was approximately linear
388
M. Kunimoto et al.
a Skin resistance
Stimulation
I I- !1 I1 1
T
T
T
T-
1 min
b
k0
2'o 1.5
T
1
1.0
.5'
.o
250
270
290
kO
Skin resistance (DC) Fig. 2. (a) Skin resistance responses evoked by 0.1 Hz single pulse stimulation where every 10th stimulus was replaced by a train of six impulses at 20 Hz (indicated by T). The first response increase of response amplitudes concomitant with the gradual decrease of skin resistance and the potentiation of response amplitudes after trains 2-5. (b) Relationship between skin resistance level (DC) and change of skin resistance (AC amplitude) evoked by single pulse stimulation (0.1 Hz) during six consecutive sequences of stimuli (nine pulses in each) separated from each other by a single train (six impulses, 20 Hz) as in a. Separate symbols for each series as shown in the figure. %leandata from four sites in two subjects occurred after the third stimulus (arrow). Note the initial SEhI for the amplitude (not shown) was around 0.5 kohm during first and second series and decreased successively to around to 0.15 kohm during the sixth series.
and the slopes of the regression lines were similar. T h e data points for the first two sequences (before and after the first train) showed a reasonable fit to a common regression line. F o r response sequence 3-6 (after train 2-5) the regression lines were shifted successively to the left and slightly downwards (lines not shown in the figure) indicating that response amplitudes to single stimuli tended to diminish in spite of successive small reductions of the skin resistance level after each train.
Efeects of large reducttons lei.eI.
of
the skin resistunce
U hen 20 €32 trains were given for 25-38 s in three subjects (six skin sites), a rapid maximal
reduction of skin resistance occurred (cf Kunimoto et al. 1991). In most cases an approximately stable resistance level was reached during the stimulation but sometimes resistance started to increase slow~lyagain despite continuing stimulation (cf Fig. 3a). After termination of the high frequency train, resistance started to return towards the starting level, first rapidly, then more slowly. T h e first responses to 0.1 Hz single stimuli after the end of the train were usually small but they grew successively in amplitude during the rapid phase of skin resistance return, reached a maximum after 3-10 stimuli and then started to diminish again. Figure 3 (a) shows experimental records from the distal skin site in one subject and Figure 3(b) summarizes quantitatively the relationship between response
Non-linearity of skin resistance response
a
389
Skin resistance
[ lkn
AC
Stimulation liillilllllllllllllllilllllllillllllilil 20 HZ
OIHZ
u 1 min
h
7i 2 0 -
f
5
16-
“0 -4 0-
12-
aa 0 0 c c
.a-
22 0 .cn
4-
cn
!!
0-
16-
- -
amplitude to single stimuli and skin resistance level for both the proximal and the distal site in the same subject. Results were also similar in both records when response curves were obtained simultaneously with one pair of ‘wet’ electrodes and one pair of ‘dry’ electrodes (2 expts). I n one subject (two skin sites) the same procedure was repeated three times during the same experiment (20 H z initial train for 25 s, followed by 0.1 Hz single stimuli). Qualitatively, findings were similar every time but quantitatively the response amplitudes to the single pulses became successively smaller and the response curves were shifted to the left (i.e. towards lower skin resistance levels) and downwards (Fig. 3 c). Figure 3(b) and (c) also illustrate that although response curves from two sites in the same experiment usually were qualitatively similar, quantitative differences were nevertheless fairly 14
..>
0’
common. For example, in Figure 3(b) response amplitudes to single stimuli had their peak at the fourth response in the proximal and at the ninth in the distal site. Peaks in response amplitudes similar to those in Figure 3 were observed in all three subjects (six skin sites) in whom the initial train frequency was 20 Hz (averaged data shown in Fig. 4).T h e effect was present also if the frequency of the initial train was 10 Hz (six subjects) and the frequency of the single stimuli 0.1 H3 (four subjects, five sites) or 0.2 H 3 (two subjects, three sites), but not if the frequency of the initial train was 0.5 or 1.5 Hz. This is summarized in Figure 5 which shows the effects of varying the train frequency from 0.5 to 1.5 to 10Hz in four subjects (five sites). T h e peak in amplitude a t 10 Hz occurred when the skin resistance level was below approximately 490 kR and the A C T 146
390
M. Ktrninioto et al.
-.
I
16
-16
12
-12
-
0
@
g
-
m
0
- 8 2 -
4
- 4
0
- 0
I
X
3
> H
I
0
1
2
3
4
5min
Time
Fig. 1. Changes of response amplitude (.XC-amplitude), skin resistance and slope of the DC-curve during single pulse stiniulation a t 0.1 Hz folloaing a 20 Hz train for 25-38 s. Averaged data from six skin sites in three subjects. Yertical bars indicate SElL For clarity SEMs were omitted for skin resistance (ralues were large due to differences in skin resistance between skin sites but the range of variabilit! was small (77-82 kohm).
-
b
a
I
C -1
2
-10
w 5
:
0
- a -6
8
1
- 4
=
- 2
5
- 0
0
1
2
3
4
5
0
1
3
2
4
5
0
1
2
3
4
-
5min
Time
Fig. 5. T h e same relationship as in Figure 4 obtained after train stimulation at 0.5 (a), 1.5 (b) and 10 Hz (c) for 96-184, 35-171 and 49-90 s, respectively. Averaged data from five skin sites in four subjects. Note that onl!- in ( c ) did response amplitudes first increase and then decrease.
slope of DC-curves ahore 0.5 kR s-', Yalues m-hich were not reached at 0.3 and 1.5 Ilz.
DISCUSSION The present results confirm our previous observations (IVaIlin el 01. 1983, Kirno et n l . 1991, Kuninioto et (11. 1991) that a g i x n sudomotor
nerve stimulus may induce skin resistance responses of very variable amplitude. Although the resuonse amplitudes were related to the skin resistance level, the relationship was complex and neither linear nor constant. T o a large extent skin resistance is determined by the amount of sweat (and its concentration of electrolytes) in the sweat ducts (Fowles 1974).
Non-linearity ojskin resistance response The resistance of the corneal layer of the skin itself is high but may be reduced by hydration (e.g. infiltration of sweat) (cf Fowles 1974). During -on-going low frequency nerve stimulation the small volume of sweat produced by each impulse leads to a decrease of resistance which slowly decays due to reabsorption of electrolytes and water (Edelberg 1971, Sat0 1977) (since the recording site is covered by the skin resistance electrode, evaporation of water from the orifice cannot occur). By using an initial high frequency train as a conditioning stimulus the responses to single impulses could be studied at different resistance levels, i.e. different degrees of filling of the sweat ducts. Using this procedure we found that response amplitudes were low both at high and low resistance levels and had a clear maximum in between. Our data give no direct insight into the underlying mechanisms but several factors may contribute: (1) If the amount of sweat produced in the gland per impulse varied with the degree of duct filling (i.e. the skin resistance level) this could lead to varying response amplitudes. We cannot exclude this alternative at the beginning of a low frequency stimulation series (when response amplitudes grow slowly concomitant with a slight skin resistance reduction). It seems unlikely, however, that the sweat volume produced in the gland per impulse should fall at low resistance levels (which under physiological conditions occur only when sweating is high). If this nevertheless were the case it would constitute a local negative feedback mechanism that would limit excessive sweating; (2) another possibility would be that there is loss of water through the ductal wall which varies in speed in proportion to the degree of duct filling. Loss of water from the duct into the corneal layer of the skin does occur but the rate is fairly slow (Adams 1966, Edelberg 1971). This mechanism may be responsible for effects that occur over minutes (see below) but would probably be too slow to significantly affect peak amplitudes of our transient responses (the peak occurs after approximately 1.6 s, cf Kunimoto et al. 1991); and (3) a more likely alternative is that the longitudinal resistivity of the empty duct varies in different parts of the duct. Skin resistance to current flow occurs mainly between the germinating layer and the upper surface of the corneum (cf Fowles 1974). This means that the skin resistance can change appreciably only when the sweat moves
39 1
within this layer, i.e. mainly within the corneum. If then the resistivity were low in the bottom part of the layer, high in the middle part and then lower again towards the top, the addition of a constant volume of sweat could explain the variations of response amplitudes found in the present study. A variable resistivity is likely to exist for structural reasons. T h e corneal layer contains dead cells which become older and drier the closer to the skin surface they are situated (cf Peiss et al. 1956) and Suchi (1955) found an increasing resistivity of the corneal layer and the empty duct all the way up to the skin surface. This does not agree with our finding of lower response amplitudes at minimal resistance levels (i.e. when the ducts are full or almost full). Theoretically, there may be an artefactual gradient of reduced resistivity near the skin surface due to the electrode gel hydrating the upper corneum. In addition, excessive hydration of the corneal layer may result in closure of the sweat duct opening on the skin (Randall et a/. 1957, Sarkany et al. 1965) and Fowles (1974) has suggested that this may reduce amplitudes of resistance responses. It is, however, unlikely that excessive hydration from the electrode gel can explain our findings since the changes of response amplitudes were similar with both ‘wet’ and ‘dry’ resistance electrodes. Another structural factor which could contribute to a variable resistivity would be variations of the ductal anatomy. For example, the sweat duct has a spiral pathway in the corneum and towards the surface there is a widening of the lumen (Hambrick & Blank, 1954, Montagna 1956) which may tend to reduce resistivity in that part. In addition to being non-linear the relationship between skin resistance levels and response amplitudes to single stimuli was shifted successively to the left and downwards with repetition of the stimulation sequences (Figs 2 & 3 ) . This finding indicates that response amplitudes diminished slightly between sequences in spite of a decrease of the skin resistance level. 4 t the beginning of stimulation (i.e. about 30 min after local anaesthesia) the basal epidermal layer probably was relatively dry and therefore it can be expected that the sweat produced by the stimulation caused a slow successive increase of the hydration of those structures (Adams 1966). This probably explains the shift of the response curves, since an increased hydration would reduce skin resistance (see above). 162
392
‘21. Kirnimoto et al.
The present findings have important practical consequences. ,Measurements of skin resistance changes are often used as indicators of the strength of sympathetic sudomotor nerve traflic in phJ-siological, psychological or clinical studies. The marked variabilitl- of the responses to a constant stimulus makes such data difficult to interpret. When skin resistance levels are reasonably constant and stimuli delivered with fairl!long intervals the errors may be limited (Bini et ul. 1980, Lidberg et a/. 1981). If, however, the background sudomotor nerve traffic changes during a study (e.g., with changing environmental temperature, stress or anaesthesia) the concomitant changes of skin resistance level will make measurements of skin resistance response amplitudes almost meaningless. Supported b! the Swedirh Medical Research Council Grant N o B91-04X-03546-20A and b! traxel grants to Llasanari Kunimoto from Japan Societ! for the Pi omotion of Science
REFERENCES AXIS IS, T. 1966. Characteristics of eccrine sweat gland actkit!- in the footpad of the cat. J .Jpp/ Ph,Jsrol 2 1 1004-1 0 12. B IX I , G., HAGBARTH, K.-E., HTSNINES, P. & i v A L I . I S , H . G . 1980. Thermoregulator! and rh!-thm-generating mechanisms governing the sudomotor and vasoconstrictor outflow in human cutaneous nerves. 1 Pliysiol (Lond) 306, 537-552. EDELBERG, R. 1971. Electrical properties of skin. In: H.R. Weiss (ed.) Biopli,)tsical Properties of the Skin, 1.01. I , pp. 713-550. LViley, Yew Yo& FOWLES. D.C. 1974. Mechanisms of electrodermal activitl-. I n : R.F. Thompson 8r X I . X l . Patterson (eds) Biodectric Recording Techniques, Part C, ~
Receptor and E#ector Processes, pp. 231-27 1. Academic Press, New York. HAMBRICK, G.W. JR. & BLANK,H. 1954. Whole mounts for the study of skin and its appendages. 3 Inrest Drrmatol 23, 437-453. KIRNO,K., KUNIMOTO, M., LUNDIN, S., ELAM,M. & WALLIN,B.G. 1991. Can galvanic skin response be used as a quantitative estimate of sympathetic nerve activity in regional anesthesia? Anesth Analg 73, 138-142. KCNIMOTO, M ,KIRYO,K., ELAM,M. & WALLIN, B.G. 1991. Neuro-effector characteristics of sweat glands in the human hand activated by regular stimuli. 3 Ph)Jsio/(Lond) 442, 391-411. LIDBERG, L. & WALLIN,B.G. 1981. Sympathetic skin nerve discharges in relation to amplitude of skin resistance responses. Psychoph.ysio/ 18, 268-270. MOSTAGNA, W.1956. T h e eccrine sweat glands. I n : The Structure and Function of Skin, pp. 89-127. Academic Press, New York. PEtss, C.N., RANDALL, W.C. & HERTZMAN, A.B. 1956. Hydration of the skin and its effect on sweating and evaporative water loss. 3 Invest Dermatol 26, 459-470. RINDALL,W.C. & PEISS,C.N. 1957. T h e relationship between skin hj-dration and the suppression of sweating. 3 Inrest Dermatol 28, 435-441. SQRKASY, I, SHLISTER, s. & STAMMERS, M.C. 1965. Occlusion of the sweat pore by hydration. Brit 3 Dermatol 77, 101-104. S4T0, K. 1977. T h e physiology, pharmacology, and biochemistry of the eccrine sweat gland. Rev P h j d Biorheni Pharnrarol 79, 5 1-131. SELANDER, D. 1977. Catheter technique in axillary plexus block. Actri Anaesthesiol Scand 21, 321t-329. SLXHI, T . 1955. Experiments on electrical resistance of the human epidermis. 3ap 3 Pitysio/ 5, 75-80. \-.ALLBO, A.B., HAGBARTH, K.-E., TOREBJORK, H.E. & WALLIN, B.G. 1979. Somatosensory, proprioceptive and sympathetic activity in human peripheral nerves. Phjlsiol R e v 59, 919-957. \VALLI’V,B.G., BLUMBERG, H. & HYNNINEN, P. 1983. Intraneural stimulation as a method to study synpathetic function in the human skin. Neurosci Lett 36, 189-194.
