Differences in the recovery of excitability in sensory and motor axons of human median nerve

Transcription

1 Brain (1996), 119, Differences in the recovery of excitability in sensory and motor axons of human median nerve Matthew C. Kiernan, Ilona Mogyoros and David Burke Prince of Wales Medical Research Institute and the Department of Neurology, Prince of Wales Hospital, Randwick, Sydney, Australia Correspondence to: Dr Matthew Kiernan, Prince of Wales Medical Research Institute, High Street, Randwick, NSW, Australia 2031 Summary Following conduction of an action potential there is a stereotyped sequence of changes in excitability as axons are initially refractory, then superexcitable and finally subexcitable. These activity-dependent oscillations in excitability subside over 0 ms and together constitute the recovery cycle. The present study was undertaken first to document the recovery cycle of sensory and motor axons of different threshold and, secondly, to compare the changes in sensory axons with those in motor axons. A computerized thresholdtracking system was used to measure recovery cycles in six healthy subjects; stimuli were applied to the median nerve at the wrist. Changes in the threshold required to produce an antidromic compound sensory action potential (CSAP) and an orthodromic compound muscle action potential (CMAP) of fixed amplitude (30%, 50% and 70% of maximal) were recorded following a single supramaximal conditioning stimulus. Normalized recovery cycles were identical for axons of different threshold, whether sensory or motor, and were reproducible on repeat testing. However, there were significant differences between the changes in sensory and motor axons, with greater supernormality and greater late subnormality in motor axons. The greater changes in motor axons could not be explained by differences in the strengthduration properties of sensory and motor axons. There are biophysical differences in the properties of sensory and motor axons and these differences may underlie the differential susceptibility of sensory and motor axons in peripheral nerve disorders. Key words: supernormality; subnormality; sensory axons; motor axons Abbreviations: CMAP = compound muscle action potential; CSAP = compound sensory action potential Introduction A series of changes in axonal excitability follows the transmission of a single nerve impulse. These changes (the absolutely and relatively refractory periods, the supernormal period and the late subnormal period) constitute the recovery cycle and have been well documented in human sensory and mixed nerve (Gilliatt and Willison, 1963; Guisset, 1968; Stohr, 1981; Eisen et al., 1982; Ng et al., 1987; Stys and Ashby, 1990; Shefner et al., 1992; Morita et al., 1993; Potts et al., 1994). The few studies which have addressed the recovery cycle in human motor axons have been confined to the excitability changes in single motor units (Bergmans, 1970, 1973; Borg, 1983). In part, this has been due to technical difficulties associated with measuring changes in the test potential at short conditioning-test intervals because the test CMAP is then superimposed on the large CMAP evoked by the conditioning stimulus. Consequently, although there is qualitative information about the recovery cycle in Oxford University Press 1996 motor fibres derived from single motor units, there is little quantitative information. No previous study has investigated the recovery cycle in nerve fibres of different threshold, whether sensory or motor. Furthermore, there has been no direct comparison of recovery in sensory and motor axons of similar threshold. With improved experimental techniques, much interest is focusing on differences in the behaviour of sensory and motor fibres in human subjects, and on the biophysical basis for these differences (Stys and Waxman, 1994). For example, Bostock et al. (1994) demonstrated differences in the behaviour of sensory and motor axons following the release of ischaemia, perhaps associated with a quantitative difference in an inwardly rectifying conductance. Honmou et al. (1994) have demonstrated the presence of a slowly inactivating Na + conductance in mammalian cutaneous afferents but not in muscle afferents or efferents. More recently, Bostock and

2 10 M. C. Kiernan et al. Rothwell (1995) have reported greater expression of noninactivating 'threshold' Na + channels in human sensory axons than in motor axons, a difference which may underlie the difference in their strength-duration properties (Panizza et al., 1994; Mogyoros et al., 1996). The aim of the present study was to explore the recovery cycles of motor axons and to compare the findings with those for sensory axons under comparable conditions, in an attempt to explore further the molecular organization of these fibres. The recovery cycles for axons of different threshold were identical but there were hitherto undescribed differences in the degree of supernormality and late subnormality in human sensory and motor axons. Methods Four sets of experiments were conducted on six normal subjects (four male and two female, aged years) with no clinical or neurophysiological evidence of a peripheral nerve disorder. All subjects gave informed consent to the experimental procedures which had the approval of the appropriate institutional ethics committee. The recovery cycle was determined using the technique of threshold tracking, as described by Bostock and Baker (1988) and Weigl et al. (1989). The median nerve was stimulated at the wrist with 0.1 ms current pulses through surface electrodes of 1 cm diameter, taped to the skin 4 cm apart and orientated longitudinally along the course of the nerve. In sensory studies the antidromic CSAP was recorded from the index finger using ring electrodes set 3 cm apart around the proximal phalanx. In motor studies, the orthodromic CMAP was recorded using surface electrodes over abductor pollicis brevis, with the active electrode at the motor point and the reference electrode on the proximal phalanx. Skin temperature was monitored continuously at the first metacarpophalangeal joint and at the wrist, and was kept above 32 C by wrapping the limb in a blanket and by radiant heat if necessary. In two separate series of experiments, one on sensory axons and the other on motor axons, stimulus intensity was adjusted automatically in steps of 2% by computer to produce CSAPs or CMAPs which were 30%, 50% and 70% of maximum. The test stimulus consisted of a single current pulse of 0.1 ms duration, delivered alternately by itself or after a conditioning stimulus. The conditioning stimulus consisted of a single supramaximal current pulse of 0.1 ms duration. The conditioning-test interval was stepped down automatically by computer from 0 ms to 2 ms. For each size of the CSAP (or CMAP), the computer tracked the current required to produce the desired potential amplitude when the test stimulus was delivered with or without a conditioning stimulus. The potential produced by the supramaximal conditioning stimulus contaminated the conditioned potential when the conditioning-test interval was short (< ms). The conditioned potential was therefore measured after subtraction of the response evoked by the conditioning stimulus alone, the subtraction being performed on line by computer. For each curve, the conditioned trace was normalized relative to its unconditioned control. Amplitude was measured from the negative peak to positive peak (falling phase). Latency was measured to the negative peak in all studies. In a third series of experiments in the same subjects, the recovery cycles were measured for a CSAP of 30% maximum and a CMAP of 30% maximum in the same sequence; stimuli were presented at one per second alternating between the stimulus for sensory axons and that for motor axons. This provided data on reproducibility of recovery cycles on different days and allowed direct comparison of recovery cycles for comparable populations of sensory and motor axons in the same experiments. In the final series of experiments strength-duration curves for sensory and motor axons of the median nerve were measured at two conditioning-test intervals (7 and 40 ms) in the same six subjects. The test stimulus intensity was adjusted to produce a CSAP (and alternately a CMAP) which was 30% of maximum using test stimulus durations of 0.1, 0.2, 0.5, 0.7 and 1.0 ms (Mogyoros etai, 1996): the duration of the stimulus was increased every 2 min and the stimulus intensity was adjusted in steps of 2% until the amplitude of the compound potential was accurately maintained at 30% of maximum. Strength-duration curves were established for unconditioned test potentials and compared with those of conditioned responses. The conditioning-test intervals of 7 and 40 ms were chosen to sample peak supernormality and subnormality, respectively. Results The 2 ms conditioning-test interval sampled the relatively refractory period, and a higher current was then required to produce the target compound potential, be it sensory or motor (Figs 1 and 2). Following this refractory period, axons entered a supernormal phase characterized by a reduction in the current required to produce the target potential of = 20%. This occurred 5-16 ms after the conditioning volley and reflected greater axonal excitability in the supernormal period. Axons then entered a phase of subnormality associated with an increase in the required current of -15%. The late subnormal period was complete by 0 ms in sensory axons but lasted slightly longer in motor axons. To compare the recovery cycles of sensory fibres obtained with test compound potentials of different size, the stimulus intensities necessary to produce compound sensory action potentials of 30%, 50% and 70% were recorded, and the results are shown in Fig. 1, normalized to the relevant control stimulus intensity. There were no significant differences between the three recovery curves at any conditioning-test interval (ANOVA, P > 0.5). At the 2 ms interval, sampling the relatively refractory period, there was a tendency for axons of lower threshold to require a smaller current increase than those of higher threshold. This might be expected as

3 Sensory and motor nerve excitability 11 ^0l 30% Sensory 50% Sensory 70% Sensory 1.8 o Sensory Motor 1.6 I Conditioning-Test Interval (ms) Fig. 1 The changes in excitability of sensory axons following a single supramaximal conditioning stimulus. The test stimulus intensity was adjusted to produce CSAPs which were 30%, 50% and 70% of maximum ('threshold tracking'). Each data point shows the mean ± SE for six subjects normalized to the current required for an unconditioned (control) potential of the same size. 30% Motor 50% Motor 70% Motor ^ b Conditioning-Test Interval (ms) Fig. 2 The recovery cycle of motor axons. As in Fig. 1 the stimulus intensity was adjusted, this time to produce CMAPs which were 30%, 50% and 70% of maximum. Each data point shows the mean ± SE for six subjects normalized to the current required for an unconditioned (control) potential of the same size. other studies have demonstrated changes in the refractory period dependent on fibre size (Brink and Mackel, 1993). The recovery cycles for motor axons are shown in Fig. 2. Again the curves were remarkably similar in the magnitude and timing of changes, with no significant differences between the three curves at any interval (ANOVA, P > 0.5). To assess reproducibility, the recovery cycles for a CSAP of 30% and a CMAP of 30% were recorded in a third experimental session in the same six subjects. These results were virtually identical with the earlier ones, there being no significant differences at any interval between the two studies, for both sensory axons and motor axons. However, there were clear differences between the recovery curves of sensory and motor axons, with greater supernormality and greater subnormality in motor axons. -02 Fig. 3 Comparison of the recovery cycles of sensory and motor axons. (A) The recovery cycle for sensory axons (open circles) and motor axons (filled squares) each tested at three threshold levels in the six subjects. Data are reproduced from Figs 1 and 2. (B) The mean of the three values for sensory axons was subtracted from that for motor axons at each conditioning-test interval to illustrate further the greater changes in supernormality and late subnormality in motor axons. Compound potentials of different threshold were analysed to ensure that this difference was not due to differences in axonal size (see Fig. 3A). It can be seen that at each conditioning-test interval during the supernormal and late subnormal periods the three data points for sensory potentials lie within the three data points for motor axons. These differences between sensory and motor fibres were significant (P = for both periods). This is illustrated further by direct subtraction of the recovery cycles shown in Fig. 3B; the mean of the three values for sensory axons was subtracted from the mean of the three for motor axons at each conditioning-test interval. Use of the mean value was justified because the recovery cycles were identical for compound potentials of different size. In addition to the changes in the degree of supernormality and late subnormality, the time course for recovery was shorter in sensory than in motor fibres: in sensory fibres the late subnormal period was complete by 80 ms but it continued beyond 0 ms in motor fibres.

4 12 M. C. Kiernan et al. C. Conditioned Threshold 3 Senson/ Motor 0.0 Q Q Stimu'us Duration (ms) Fig. 4 Strength-duration curves for unconditioned and conditioned sensory (A) and motor (B) potentials of 30% maximum for six subjects (mean ± SE). Unconditioned potentials are indicated by filled circles and conditioned potentials by open circles. The conditioning-test interval was fixed at 7 ms to study peak supemormality. (C) Normalized thresholds for the conditioned potentials (mean ± SE) show greater supemormality in motor fibres than in sensory fibres for each stimulus duration. Latency to negative peak was measured in all experiments, alternating at one per second across six recording channels using an overall sampling frequency of khz which is too slow to define the expected changes in latency clearly (Stys and Ashby, 1990; Stys and Waxman, 1994). Despite this, there were appropriate changes in latency in the majority of studies, with clear prolongation at the 2 ms conditioningtest interval (0.47 ms in motor axons and 0.11 ms in sensory axons) and subtle decreases during the supernormal period and the late subnormal period (both of the order of 0.05 ms, i.e. equivalent to <1 sampling bin) similar in magnitude to previously reported changes in latency (Stys and Ashby, 1990). To determine whether the differences in supemormality and subnormality reflected differences in the strengthduration properties of sensory and motor axons, strengthduration curves were constructed at two conditioning-test intervals (7 ms and 40 ms assessing supemormality and subnormality, respectively) and compared with curves from unconditioned controls. The difference in current between the conditioned and unconditioned test stimuli was recorded at each stimulus duration for sensory and motor axons, during both supemormality and subnormality {see Fig. 4A and B). Direct comparison of the curves reveals that the threshold reduction produced by supemormality was greater for motor axons than sensory axons at each stimulus duration, with no significant relationship to stimulus duration (Fig. 4C; P = 0.26). The strength-duration time constant (chronaxie) was determined from Weiss' formula {see Bostock, 1983; Mogyoros et al., 1996). As expected for a depolarizing afterpotential, there was a small increase in the mean strengthduration time constant during the supernormal period, and this was greater for motor fibres (from 363 u.s to 473 (is) than sensory (from 70 i is to 748 u.s), though for neither was the change outside the 95% confidence intervals for reproducibility of time constant measurements (Mogyoros et al., 1996). On the other hand, during the late subnormal period strength-duration time constants of sensory and motor axons underwent little change; average decreases were 11 (is for motor axons and 24 is for sensory axons. Desmedt (1958) demonstrated (in a single figure) that the second of two maximal CMAPs was smaller than the first when the interval between them was <20 ms. Similar studies were performed on three subjects; pairs of supramaximal studies were delivered 7 ms apart, the interval chosen to correspond to greatest supemormality. As previously, the amplitude of the "test" CMAP was measured on line after subtraction of the response to the "conditioning" stimulus given by itself. With two definitely supramaximal stimuli 7 ms apart, the amplitude of the second CMAP was indeed smaller in all three subjects by, on average, 9.2%. To determine the effect of this factor on earlier measurements of motor supemormality, comparison was made of the current required to generate a conditioned CMAP of 30% of maximum using 2 separate paradigms. In the first, the target 30% potential was referenced to the maximal unconditioned CMAP, as used in previous recovery studies. In the second paradigm, the target potential was 30% of the maximal conditioned CMAP, recorded as above by delivering two supramaximal stimulus 7 ms apart. There was a small but significant difference in current required in the two methods, but the current was less (on average by 1.2%) when the target potential was referenced to a conditioned CMAP (P = 0.007). This implies that the supemormality of motor axons may have been underestimated in the earlier studies. Discussion This study provides the first measurements in human subjects of the recovery cycle of compound muscle action potentials and the first direct comparison of recovery in sensory and motor nerve fibres after activation by a single conditioning stimulus. The recovery curves were qualitatively similar for motor and sensory axons, but there were quantitative differences in the refractory, supernormal and late subnormal periods, differences which could not be explained by strengthduration properties. This study has shown for the first time that, after normalization, there is no difference in the recovery cycles of fibres with different thresholds, be they sensory or motor. These findings suggest that there are differences in the molecular organization of sensory and motor axons, and that these differences cannot be attributed to fibre size. Mechanisms of recovery The mechanisms involved in different phases of the recovery cycle have been extensively investigated. The established interpretation of the refractory period is inactivation of sodium conductances (Hodgkin and Huxley, 1952). The refractory period has been shown to be prolonged by cooling, presumably because of slowed channel kinetics at lower temperatures. The supernormal period is characterized by an increase in excitability (Gilliatt and Willison, 1963; Guisset,

5 Sensory and motor nerve excitability ; Bergmans 1970, 1973; Ng et al., 1978; Stohr, 1981; Stys and Ashby, 1990; Morita et al, 1993; Potts et al., 1994) which is related to a depolarizing afterpotential (Barrett and Barrett, 1982; Blight and Someya, 1985; Bowe et al., 1987). In unmyelinated fibres, this results from the efflux of potassium that occurs with the repolarization phase of the action potential leading to periaxonal accumulation of potassium due to a diffusion barrier (Kocsis et al., 1983; see also Barrett and Barrett, 1982; Stys and Waxman, 1994). In myelinated fibres the depolarizing afterpotential probably results from the capacitive charging of the internode by the action potential, with subsequent discharge occurring through or under the myelin sheath (Barrett and Barrett, 1982; Blight and Someya, 1985; Bowe et al., 1987). The evidence for the association between the depolarizing afterpotential and the supernormal period is threefold (Bowe et al., 1987). First, axons exhibiting a depolarizing afterpotential also have a supernormal period. Secondly, there is a close temporal association between the two. Finally, brief subthreshold depolarizing pulses applied intra-axonally result in a passive, slowly decaying depolarization and a parallel period of increased excitability without generation of an action potential. Recovery of an axon after a single action potential terminates in a phase of subnormal ity. This phase of lowered excitability reflects a corresponding membrane hyperpolarization. The precise mechanism of this late subnormal period remains unclear. Ritchie (1973) implicated electrogenic Na + / K + pump activity. If two K + ions are exchanged for three Na + ions, a net deficit of positive ions is produced on the inner side of the membrane, with resultant hyperpolarization. However, it is unlikely that a single conditioning discharge would be sufficient to activate pump mechanisms significantly and, more recently, Stys and Waxman (1994) have suggested that activation of a slow K + conductance may be responsible for the late subnormal period. Differences between motor and sensory fibres The present study has established significant differences in both the supernormal period and late subnormal period in sensory and motor axons. It is unlikely that technical factors were responsible. The third series of experiments was conducted simultaneously on sensory and motor fibres of similar threshold so that experimental conditions would be identical. It is unlikely that the differences were due to differences in axonal size because, within each group, the recovery cycles were identical for axons of different threshold. Excitability was measured by recording changes in the threshold stimulus required to produce a potential of constant amplitude ('threshold tracking'). Unequal changes in latency for axons of slightly different threshold could result in amplitude changes merely due to changes in temporal dispersion. However temporal dispersion would cause greater changes in the amplitude of sensory potentials (Olney et al., 1987) and, accordingly, this factor cannot explain why the changes in threshold current were smaller for sensory axons in the present study. The studies on motor axons involved the recording of a CMAP. They reflect axonal properties only if neither the neuromuscular junction nor the muscle fibre itself distort the response. Limitations at these levels are probably responsible for the smaller size of a supramaximal test CMAP when the conditioning-test interval is short (<20 ms, see Desmedt, 1958). This could explain greater refractoriness for the CMAP than the CSAP, but not the greater supernormality. Following an action potential, muscle fibres undergo comparable phases of refractoriness and supernormality (Stalberg, 1966). These fluctuations in excitability would have contributed to the recorded changes in the CMAP only if they resulted in a change in the size of the action potentials of single muscle fibres. However, if such a change occurred, one would expect hyperpolarization to result in a greater single fibre action potential and depolarization to result in a smaller action potential. Again these changes are the opposite of those necessary to explain the greater supernormality and greater subnormality of the CMAP. Conventional outwardly rectifying fast and slow K + conductances are unlikely to contribute to the differences seen between motor and sensory fibres because accommodation to prolonged subthreshold depolarizing currents is identical in human sensory and motor axons, suggesting that outward rectification due to K + conductances is similar in these fibres (Bostock et al., 1994). It must be conceded that our conditioning stimuli were clearly suprathreshold and might therefore have activated K + conductances not activated by subthreshold stimuli. However, these would be 'fast' K + conductances, and there is little evidence for them at the mammalian node of Ranvier (Vogel and Schwarz, 1995). There is greater inward rectification in sensory axons than in motor axons (Bostock et al., 1994). This difference could contribute to the lesser degree and shorter duration of late subnormality seen in sensory axons, but it is unlikely that inwardly rectifying channels are activated with the degrees of hyperpolarization accompanying the late subnormal period: Bostock et al. (1994) required hyperpolarizing threshold changes of 40% and 80% to demonstrate a significant difference whereas in this study the late subnormal period was associated with reductions of threshold of only -20%. Na + channels and strength-duration properties Strength-duration time constants are longer for sensory axons than for motor axons (Panizza et al., 1994; Bostock and Rothwell, 1995; Mogyoros et al., 1996). The electrical properties of nodal membrane are different from those at the paranodal and internodal regions, and it is the nodal membrane that determines the axon's strength-duration properties. The difference between the time constants of sensory and motor nerves probably reflects greater expression of non-inactivating 'threshold' Na + channels in sensory axons (Bostock and Rothwell, 1995). However, the differences in excitability

6 14 M. C. Kiernan et al. seen between sensory and motor fibres in the present study do not result from differences in their strength-duration properties. As expected, time constants were longer for sensory axons than motor axons (701 is for sensory; 363 fis for motor) but there were similar differences between sensory and motor axons in supernormality and late subnormality regardless of stimulus duration. Honmou et al. (1994) have demonstrated the existence of slowly inactivating Na + channels in cutaneous afferents of the rat. If they were activated following conduction of an action potential then a depolarizing afterpotential and hence supernormality would result (Stys and Waxman, 1994). These kinetically slower conductances exist in greater numbers in sensory than motor axons but it is unlikely that these channels contribute to the difference in supernormality seen in the present study because supernormality would then be greater in sensory than motor fibres, not smaller. Conclusions The mechanisms responsible for the differences in the recovery cycles of human sensory and motor axons remain uncertain but these differences constitute further evidence that there are biophysical differences between afferent and efferent axons. The differences probably represent an adaptation to normal activity patterns and they presumably contribute to the differential susceptibility of sensory and motor axons in peripheral nerve lesions. In addition, the present findings may contribute to further understanding of disorders of peripheral nerve function. Two previous studies have established abnormalities in supernormality in the recovery cycle of sensory fibres in patients with multiple sclerosis (Eisene/ al., 1982; Shefner etal., 1992), traditionally considered to be a disease of the CNS. Acknowledgements This work was supported by the National Health and Medical Research Council of Australia. The authors thank Dr Hugh Bostock for helpful comments on the studies. References Barrett EF, Barrett JN. Intracellular recording from vertebrate myelinated axons: mechanism of the depolarizing afterpotential. J Physiol (Lond) 1982: 323: Bergmans J. The physiology of single human nerve fibres. Louvain: Vander, Bergmans J. Physiological observations on single human nerve fibres. In: Desmedt JE, editor. New developments in electromyography and clinical neurophysiology, Vol. 2. Basel: Karger 1973: Blight AR, Someya S. Depolarizing afterpotentials in myelinated axons of mammalian spinal cord. Neuroscience 1985; 15: Borg J. Effects of prior activity on the conduction in single motor units in man. J Neurol Neurosurg Psychiatry 1983; 46: Bostock H. The strength-duration relationship for excitation of myelinated nerve: computed dependence on membrane parameters. J Physiol (Lond) 1983; 341: Bostock H, Baker M. Evidence for two types of potassium channel in human motor nerve in vivo. Brain Res 1988; 462: Bostock H, Rothwell JC. The time constants of motor and sensory axons in human peripheral nerve [abstract]. J Physiol (Lond) 1995; 487: 47P. Bostock H, Burke D, Hales JP. Differences in behaviour of sensory and motor nerve axons following release of ischaemia. Brain 1994; 117: Bowe CM, Kocsis JD, Waxman SG. The association of the supernormal period and the depolarizing afterpotential in myelinated frog and rat sciatic nerve. Neuroscience 1987: 21: Brink EE, Mackel RG. Time course of action potentials recorded from single human afferents. Brain 1993; 116: Desmedt JE. Methodes d'etude de la fonction neuromusculaire chez l'homme. Myogramme isometrique, electromyogramme d'excitation et topographie de l'innervation terminale. Acta Neurol Psychiat Belg 1958; 58: Eisen A, Paty D, Hoirch M. Altered supernormality in multiple sclerosis peripheral nerve. Muscle Nerve 1982; 5: 411^4. Gilliatt RW, Willison RG. The refractory and supernormal periods of the human median nerve. J Neurol Neurosurg Psychiatry 1963; 26: Guisset M. The recovery of excitability of human sensory nerve fibres following activity. Arch Int Physiol Biochim 1968; 76: 139^11. Hodgkin AL, Huxley AF. A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol (Lond) 1952; 117: Honmou O, Utzschneider DA, Rizzo MA, Bowe CM, Waxman SG, Kocsis JD. Delayed depolarization and slow sodium currents in cutaneous afferents. J Neurophysiol 1994; 71: Kocsis JD, Malenka RC, Waxman SG. Effects of extracellular potassium concentration on the excitability of the parallel fibres of the rat cerebellum. J Physiol (Lond) 1983; 334: Mogyoros I, Kiernan MC, Burke D. Strength-duration properties of human peripheral nerve. Brain 1996; 119: Morita H, Shindo M, Yanagawa S, Yanagisawa N. Neuromuscular response in man to repetitive nerve stimulation. Muscle Nerve 1993; 16: Ng A, Burke D, Al-Shehab A. Hyperexcitability of cutaneous afferents during the supernormal period: relevance to paraesthesiae. Brain 1987; 1: Olney RK, Budingen HJ, Miller RG. The effect of temporal dispersion on compound action potential area in human peripheral nerve. Muscle Nerve 1987; : Panizza M, Nilsson J, Roth BJ, Rothwell J, Hallett M. The time constants of motor and sensory peripheral nerve fibers measured with

7 Sensory and motor nerve excitability 15 the method of latent addition. Electroencephalogr Clin Neurophysiol 1994; 93: Potts F, Young RR, Shefner JM. Long lasting excitability changes in human peripheral nerve. Muscle Nerve 1994; 17: Ritchie JM. Energetic aspects of nerve conduction: the relationships between heat production, electrical activity, and metabolism. [Review]. Progr Biophys Mol Biol 1973; 26: Shefner JM, Carter JL, Krarup C. Peripheral sensory abnormalities in patients with multiple sclerosis. Muscle Nerve 1992; 15: Stalberg E. Propagation velocity in human muscle fibers in situ. Acta Physiol Scand 1966; 70 Suppl 287: Stohr M. Activity-dependent variations in threshold and conduction velocity of human sensory fibers. J Neurol Sci 1981; 49: Stys PK, Ashby P. An automated technique for measuring the recovery cycle of human nerves. Muscle Nerve 1990; 13: Stys PK, Waxman SG. Activity-dependent modulation of excitability: implications for axonal physiology and pathophysiology [editorial] [see comments]. [Review]. Muscle Nerve 1994; 17: Comment in: Muscle Nerve 1995; 18: Vogel W, Schwarz JR. Voltage-clamp studies in axons: macroscopic and single-channel currents. In: Waxman SG, Kocsis JD, Stys PK, editors. The axon. New York: Oxford University Press, 1995: Weigl P, Bostock H, Franz P, Martius P, Miiller W, Grafe P. Threshold tracking provides a rapid indication of ischaemic resistance in motor axons of diabetic subjects. Electroencephalogr Clin Neurophysiol 1989; 73: Received December 29, Revised March 14, 1996 Accepted April 1, 1996.