Journal of
J. Neurol. 216, 173--180 (1977)
Neurology © by Springer-Verlag 1977
Temperature Dependence of Normal Sensory Nerve Action Potentials* H. P. Ludin and F. Beyeler The Neurological Clinic, University of Berne, Inselspital, CH-3010 Berne, Switzerland
Summary. Sensory conduction velocities of normal subjects are increasing linearly with rising temperature. The duration of the c o m p o u n d sensory action potentials recorded from the median nerve at the wrist and elbow shows a negative temperature coefficient. The peak-to-peak amplitude of these potentials increases from 22 ° to approximately 26 ° C and then decreases progressively again up to 36 ° C. It is believed that this behavior is due to a combination of decreasing temporal dispersion, height and duration of the individual spike potentials. Key words: Conduction velocity, sensory - Temperature dependence.
Zusammenfassung. Die sensorische Leitgeschwindigkeit steigt mit zunehmender T e m p e r a t u r linear an. Die Dauer der sensiblen Nervenaktionspotentiale vom N. medianus nimmt dagegen ab wenn die Temperatur h/Sher wird. Die Amplitude dieser Potentiale nimmt zwischen 22 ° und zirka 26 ° C zu und dann k o m m t es wieder zu einem langsamen Abfall. Dieses Verhalten ist wahrscheinlich bedingt durch ein Zusammenspiel der abnehmenden zeitlichen Dispersion und der kleiner werdenden Dauer und H6he der einzelnen Spitzenpotentiale. During the last few years measurements of sensory conduction velocities have become increasingly important in clinical electromyography. Originally it was the aim of this study to establish normal values for temperature dependence, not only for sensory conduction velocity in the median nerve, but especially for the nerve action potential parameters. As the amplitude of these potentials showed a rather unexpected behavior when tissue temperature was increased from 22 ° to 36°C, the main attention of this study turned towards a possible explanation of this observation. The behavior was first thought to be due to an artefact, but now we feel that it is a real p h e n o m e n o n which may very well be explained by currently accepted theories of membrane excitation. *
Dedicated to Prof. F. Buchthal on the occasion of his 70th birthday
174
H.P. Ludin and F. Beyeler
Methods Two types of experiments were performed on 13 healthy volunteers (9 males, 4 females) with ages ranging from 23 to 41 years. The cooling and the subsequent heating procedure was kept constant for all experiments: after the application of the electrodes, the right arm of the subject was put into ice water where it was kept until the tissue temperature near the median nerve had fallen to approximately 21°C. The ice water was then removed and the arm was warmed slowly by irradiating it with an infrared lamp until a tissue temperature of 36°C was reached. The temperature was measured by a thermocouple at the end of a cannula placed in the vicinty of the median nerve in the middle of the forearm. The error of measurement was _+0.2°C. Care was taken that the position of this electrode remained unchanges during the whole experiment which normally lasted about 3h.
a) Temperature Dependence of Nerve Action Potentials Eight experiments of this type have been included in this series. The criteria for accepting or rejecting an experiment will be discussed below.
Stimulation. Sensory fibers of the index finger of the right hand were stimulated with surface electrodes (Buchthal and Rosenfalck, 1966). Rectangular stimuli of 0.2 msec duration, 60mA amplitude, were delivered by a constant current stimulator (Disa 15 E 07). This amplitude was considered to be supramaximal as it was about 15 times the mean sensory threshold current. Recording. Unipolar, teflon-coated needle electrodes (Buchthal and Rosenfalck, 1966) were used for recording the sensory nerve action potentials from the median nerve at the wrist and elbow. The position of the different electrode was adjusted by using it for stimulation and varying its position until a motor response was elicited in the thenar muscles with stimulus currents well below 1 mA. The electrode position was controlled by this motor threshold and by the sensory threshold when stimulated at the index finger and recorded from the electrodes at the wrist and elbow. Only those experiments were accepted as valid for which the electrode position did not change appreciably during the whole experiment. The signals were fed into a differential amplifier (Disa 15 C 01) and then averaged by a digital averager with 1024 addresses (Disa 15 G 07). At every temperature examined 50 potentials from each recording site were averaged. To reduce temperature changes during averaging, this was carried out in about 13 s at a rate of 4/s. In addition, the temperature was read just before and after averaging. The mean of these two values was used for the subsequent evaluations. The frequency response of the entire recording system ranged from 100 to 10 000 c/sec at 3 db down. Control measurements showed that frequencies up to 3000c/sec were reproduced without any measurable loss of amplitude. The nerve action potentials were recorded before cooling at normal body temperature and eight times after cooling at temperature steps of approximately 2°C between 22 ° and 36 °C. The evaluation of the action potentials was as follows: latency was measured from the stimulus artefact to the peak of the first positive deflection. Amplitude was measured peak-topeak. The time from the positive peak to the negative peak was taken as the measure of the potential duration. The total area under the potential curve was determined planimetrically from enlarged curves. The regression lines were calculated using the method of "least squares".
b) Measurement of Tissue Resistance An attempt was made to determine the temperature dependence of the tissue resistance near the median nerve in the forearm. Absolute resistance values could not be measured in situ, but with a "two electrode method" (Burger and van Dongen, 1961) resistance changes could be recognized easily.
Temperature Dependence of Normal Sensory Nerve Action Potentials
175
The measurements were performed on five subjects. Two electrodes, the same as those used for recording the nerve action potentials, were inserted near the median nerve in the middle of the forearm at an interelectrode distance of 5 mm. Care was taken to keep the electrode position unchanged during an experiment. A Wheatstone bridge was first balanced with a defined a.c. current (10 ~tA, 100 c/sec) with the measuring electrodes shortcircuited. Then this shortcircuit was turned off. The current which was measured now, was inversely proportional to the resistance between the electrodes.
Results
a) Temperature Dependence of Sensory Nerve Action Potentials Figure 1 shows the sensory action potentials recorded at the wrist of one subject at different temperatures. I n what follows only the values of the action p o t e n t i a l p a r a m e t e r s recorded at the wrist will be reported a n d discussed, because the p o t e n t i a l s recorded at the elbow showed essentially the same t e m p e r a t u r e d e p e n d e n c e with the only exception that the scatter o f the i n d i v i d u a l values was greater. The conduction velocity shows a linear increase with rising t e m p e r a t u r e (Fig. 2). The m e a n value of this increase between wrist a n d elbow is 1.51 m / s e c / o C. F o r the distal p a r t of the nerve (index finger to wrist) the c o r r e s p o n d i n g value is 1.96 m / s e c / ° C, but, since the t e m p e r a t u r e was controlled in the forearm, this figure is less reliable.
30.*0C ~ ~ 32.*0C ~ 25.*C~ 9
34.*5C~ , ~
5
ms~
Fig. 1. Median nerve sensory action potentials redorded at the wrist in a normal subject. Each potential has been averaged 50 times
176
H.P. Ludin and F. Beyeler
m/sec 80 60 40 20*C
212 2r/. 2'6
3'0
2'8
3T2
3'4
3'6
Fig. 2. Effect of temperature on conduction velocity in normal sensory fibers of the median nerve between wrist and elbow
m sec 1.041
~7
~7