Impaired Refractory Periods of Peripheral Sensory Nerves in Multiple Sclerosis Hanns C.Hopf, MD, and Magrit Eysholdt,
MD
Maximum conduction velocity and relative refractory period (RRP) of median nerve sensory fibers were studied in 36 patients diagnosed as having multiple sclerosis (MS) and in 31 controls. Maximum conduction velocity did not differ in the two groups, but the RRP was significantly prolonged in MS patients. Increased RRP is observed mainly when peripheral nerve myelin is abnormal or damaged. Our findings support the assumption that peripheral nerve myeliaated fibers are usually involved in MS. Hopf HC, Eysholdt M: Impaired refractory periods of peripheral sensory nerves in multiple sclerosis. Ann Neurol4:499-501, 1978
Pollock et al [23] recently reported on teased fiber experiments in multiple sclerosis (MS)that demonstrated myelin changes in the peripheral nerves. Electrophysiological studies have been reported with contradictory results: traditional maximum conduction velocity (MCV) was within normal range [3,6, 7, 281; mild disturbances were found, however, particularly involving the slow alpha motor fiber pool when the collision technique was applied [ll, 121. We undertook a different approach to the problem of peripheral nerve involvement in MS by determining refractory periods. The results are reported here.
Material and Methods The 36 patients in this study were classified as having definite or probable MS according to the criteria suggested by Bauer [ 11: (1) a clinical course with at least two bouts or chronic progressive development over one year or more; (2) symptoms and signs that are disseminated in place or time; and (3) characteristic cerebrospinal fluid changes, such as mononuclear pleocytosis of up to 50 cells per cubic millimeter and increased immunoglobulin G with ap proximately normal total protein content. The control group consisted of 31 volunteers, including medical students, staff personnel, and some patients with lumbar disc protrusion. The MCV and relative refractory period (RRP) of median nerve sensory fibers were determined by stimulating the digital nerves at the middle finger. Rectangular pulses of 0.1 msec duration and supramaximal intensity (20 to 50 times threshold) served as stimuli. They were generated by a battery operated stimulation unit and triggered by a digital clock (Digitimer, Model 4030). The stimuli were applied percutaneously with salinated ring electrodes, the
From the Department ofClinical Neurology, Johannes Gutenberg University, Maim, Federal Republic of Germany. Accepted for publication May 8, 1978.
cathode being placed close to the basal joint of the middle finger and the anode more distally, 30 to 35 mm away. The evoked sensory nerve action potential (SNAP) was picked up percutaneously by silver disc electrodes covered with saline-soaked cotton. The active electrode was placed above the median nerve at the wrist and the reference electrode 40 mm proximally. The readings from 12 to 20 single sweeps were fed into a hardware averager (Tonnis Model 2 T, 1,024 addresses, 12 bits), summated, and stored. The averaged potential was photographed. The reference points for latency measurements were the artifact of the stimulus and the first positive peak of the SNAP. The RRP was evaluated by relying on changes in latency of a second SNAP that followed the conditioning volley after a short delay. Paired stimuli were used. The intervals between the two stimuli were progressively decreased in successive tests. For each stimulus interyal, the latency of the test SNAP was compared to the resting latency by switching off the conditioning stimulus and superimposing both averaged signals on the screen (Fig 1).The amplitude of the unconditioned SNAP depends on the number of summated events and spontaneous changes of the signal amplitude. Because such changes may amount to almost 10% with unequivocally supramaximal stimulus intensities, the signal amplitude was disregarded [ 171. The shortest stimulus interval at which changes in latency could not be observed will be referred to as the RRP. It was estimated using the “Polynomanpassung”(polynomial fit) according to Marquardt [20](Fig 2). Tissue temperature was measured continuously by a thermocouple needle. Temperature was maintained at 34°C by a heat lamp. Iatencies were not accepted if temperature changes exceeded 0.5”C.
Results The mean values of maximum sensory fiber conduction velocity (MSCV), RRP, and age for the MS Address reprint requests to Prof Hopf, Department of Clinical Neurology. Universitiit Johanna Gutenberg, langenbeckstr. 1. D-6500 Mninz. Federal Republic of Germany.
0 3 6 4 - 5 1 3 4 / 7 8 / ~ 3 $ 0 1 . 2 5@ 1978 by
HW
C. Hopf
499
Tabk I . Comparatht Dctcnnitutionr f w the MS Patients and Contmls (Total Grorrps)
Data No. of subjects Age (yr) Nerve conduction velocity ( d s e c ) Relaave refractory period (msec) Temperature (“C)
Controls 31 28.4f 7.4 63.5 f 4.5
MS Patients 36 38:l f 9.2 61.3 f 5.1
3.12 f 0.3
3.31
34 f 0.3
34.2 2 0.26
* 0.35
Vducs are mean f SD.
Pig I . The paired stimnlns response (lower trace)and the nnconditioned response (upper trace) snperimposed to show dgfmnce in latency. (Intmtimnlns intewal: 2.5 msec.)
’-1
n
.5
Fig 3. Histograms of the RRP valnes from rontrolsnbjerts (domed lines)andMSpatients (solid lines).
Pig 2. Stimnlns intewah (x axis) phtted against SNAP latencies (yaxis)for a singkpatient. The latencies originally measnrtd art indicated by triangks. The cnwc was drawn by the computer according t o “Polynomanpassnng”(polynomialf i t ) . (rR= ralrnlated refrartoy period.)
patients and the controls are presented in Table 1. The two groups differ significantly with r e w d to MSCV and RRP (x* test: p C 0.05). Histograms of the RRP from patients and controls are plotted in Figure 3. Phenytoin, carbamazepine, and diazepam have been observed to decrease MSCV [13]. If subjects are excluded who were receiving one of these drugs (2 controls and 5 MS patients), the difference in MCV between these reduced groups is still significant (p < 0.05; Table 2). MCV decreases with advancing age 12, 221. Because the age factor might have biased our results, two subgroups comprising 16 patients and 15 controls were matched for age. When the two subgroups were compared, the average RRP values were still significantly different (p < 0.05>, but not the MSCV (see Table 2).
Discussion Our results correspond to those of previous studies in that MCV in MS patients could not be demonstrated to be different from control values [3, 6, 7, 12, 281. MCV, however, is a relatively insensitive function that is frequently unaffected in mild peripheral neuropathy, and up to now only s m d patient groups have been examined. It has been proved that slight disturbances of the excitation cycle of peripheral nerves influence RRP more than MCV [15-171. On the other hand, RRP apparently changes little, if at all, with age [14, 271, but considerably with temperature 1181 and hypokalemia [21]. Excluding such influences, RRP differed significantly for the total groups (3 1 controls and 36 MS patients), and for the reduced groups (15 controls and 16 patients) as well. This finding concerning fast fibers is in agreement with comparable observations on slow fibers [12] based on the range of conduction velocities of motor fibers supplying a single muscle, estimated by applying the collision technique [ 111. Our findings support the report of Pollock et al [23], who found an increased number of abnormal teased fibers in nerves from MS patients. Many internodes showed a 50% reduction in myelin thick-
5 0 0 Annals of Neurology Vol 4 No 6 December 1978
Table 2. Comparative Detenninations fw the Reduced Groups of MS Patients and Contmh Data
Controls
MS Patients
EXCLUDING SUBJECTS RECEIVING DRUGS
No. of subjects Nerve conduction
29 63.4 f 4.3
31 61.2 & 5.0
velocity ( d s e c ) Relative refractory period (msec)
3.12 f 0.3
3.28 2 0.33
SUBIECTS MATCHED FOR AGE
No. of subjects Age (yr) Nerve conduction velocity ( d s e c ) Relative refractory period (msec) Temperature (“C)
30.3 & 5.6 62.5 f 4.3
16 30.6 k 5.3 60.8 2 5.3
3.04 2 0.2
3.37 f 0.3
34.1 f 0.35
34.2 f 0.36
15
Values are mean f SD.
ness, and the number of myelin lamellae was reduced compared to the axonal area. Morphological studies had previously suggested peripheral nerve involvement in MS [4, 5, 8, 9, 24, 25, 291; particularly, the papers of Strauber [26],Margulis [19], ikd Hasson et al [lo] should be mentioned in this context. An increased RRP has been found mainly with conditions affecting the myelin of peripheral nemes [IS, 17, 271. We conclude from our results that the basic process in MS is not confined to the central nervous system, but also extends to the myelinated peripheral nerves. The patients in this study were not examined during the initial stages or acute exacerbations of their disease. Thus the peripheral nerve involvement, if brought about during relapses, results at the very least in a mild but permanent deficit. References 1. Bauer HJ: Judgment of the validity of a clinical M S d i q n d s . Acta Neurol Scand 50:171-173, 1974 2. Buchthal P, Rosenfalck A, Trojaborg W Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J Neurol Neurosurg Psychiatry 37:340-360, 1974 3. Conrad B, Bechinger D: Sensorische and motorische Nervenleitgeschwindigkeit and distde Latenz bei Multipler Sklerose. Arch Psychiatr Nervenkr 2 12:140-149, 1969 4. Dawson JW:The histology of disseminated sclerosis. Trans R SOCEdinb 50:517-725, 1916 5. Dinkler P: Zur Kasuistik der Multiplen Sklerose des Gehirns und Ruckenmarks. Dtsch 2 Nervenheilkd 26:233-237,1904 6. Gilliatt RW, Goodman HV, Willison RG: The recording of lateral popliteal nerve action potentials in man. J Neurol Neurosurg Psychiatry 24:305-318, 1961 7. Ginzburg M, Lee M, Ginzburg J: Correlation between muscle
strength and peripheral nerve conduction in multiple sclerosis. Electromyography 11:491-5 13, 1971 8. HallervordenJ: Die zentralen EntmarkuagJkrankheiten.Dtsch Z Nervenheilkd 150201-304, 1940 9. Hassin G B Zwei ungewohnliche F a e von Multipler Sklerose. 2 Ges Neurol Psychiatr 128:99-106, 1930 10. Hasson J, Terry RD, Zimmermann HM: Peripheral neuropathy in multiple sclerosis. Neurology (Minneap) 8:503-5 10, 1958 11. Hopf HC: Electromyographic study on so-called mononeuritis. Arch Neurol 9:307-312, 1963 12. Hopf HC: Leitgeschwindigkeit motorischer Nerven bei der Multiplen Sklerose und unter dem Einfluss hoher CortisonMedikation. Dtsch Z Nervenheilkd 187:522-526, 1965 13. Hopf HC: Anticonvulsant drugs and spike propagation of motor nerves and skeletal muscle. J Neurol Neurosurg Psychiatry 36574-580, 1973 14. Hopf HC. LeQuesne P,Willison RG.Refractory periods and lower limiting frequencies of sensory fibers of the hand, in Kunze K, Desmedt JE (eds): Studies on Neuromuscular Diseases. Basel, Karger, 1975 15. Lehrnann HJ, Lehmann G, Tackmann W: Refraktiirperiode und Ubermialung von Serienimpulsen im N. tibialis des Meerschweinchens. Z Neurol 199:67-85, 1971 16. Low PA, McLeod JG: Refractory period, conduction of trains of impulses, and effect of temperature on conduction in chronic hypertrophic neuropathy. J Neurol Neurosurg Psychiatry 40:434-447, 1977 17. Lowitzsch K, Hopf HC: Refraktkperioden und frequente Impulsfortleitung im gemischten N. ulnaris des Menschen bei Polyneuropathien. Z Neurol 205: 123-144, 1973 18. Lowitzsch K, Hopf HC, Galland J: Changes of sensory conduction velocity and refractory periods with decreasing temperature in man. J N e w 1 216:181-188, 1977 19. Margulis MS: iiber die pathologische Anatomic und Pathogenese der Multiplen Sklerose. Dtsch Z Nervenheilkd 131:125-126, 1933 20. Marquardt DW: An algorithm for least squares estimation of nonlinear parameters. J SOCIndust Appl Math 11:431-441, 1963 21. Maurer K, Hopf HC, Lowitzsch K Hypokalemia shortens relative refractory period of peripheral sensory nerves in man. J Neurol216:67-71, 1977 22. Mulder DW, Lambert EH, Bastron JA, et al: The neuropathies associated with diabetes mellitus. Neurology (Minneap) 11:275-284, 1961 23. Pollock M, CPLder C, Allpress S Peripheral nerve abnormality in multiple sclerosis. Ann Neurol 2:41-48, 1977 24. Schaltenbrand G. Die Mulriple Sklerose des Menschen. Stuttgart, Thieme, 1943 25. Schob F Ein Beitrag fur pathologischen Anatomie der Multiplen Sklerose. Monatsschr Psychiatr Neurol 22:62-73, 1907 2 6. Striihuber A: iibet Regenerations- und Proliferationsvorg h g e bei multipler Sklerose des Nervensysterns. Beitr Pathol Anat 33:409-412, 1903 Refractory period in human sen27. Tackmann W, Lehmann sory nerve fibers. Eur Neurol 12277-292, 1974 28. Taraschi G, Lanzi G ktude de la vitesse de conduction du nerf cubital dans sclerose en plaque. Electroencephalogr Clin Neurophysiol Suppl 22:54-55, 1962 29. Taylor E W Zur pathologischen Anatomie der Multiplen Sklerose. Dtsch Z Nervenheilkd 5:l-2, 1894
w:
H o p f and Eysholdt: Impulse Conduction in MS
501
MD
Maximum conduction velocity and relative refractory period (RRP) of median nerve sensory fibers were studied in 36 patients diagnosed as having multiple sclerosis (MS) and in 31 controls. Maximum conduction velocity did not differ in the two groups, but the RRP was significantly prolonged in MS patients. Increased RRP is observed mainly when peripheral nerve myelin is abnormal or damaged. Our findings support the assumption that peripheral nerve myeliaated fibers are usually involved in MS. Hopf HC, Eysholdt M: Impaired refractory periods of peripheral sensory nerves in multiple sclerosis. Ann Neurol4:499-501, 1978
Pollock et al [23] recently reported on teased fiber experiments in multiple sclerosis (MS)that demonstrated myelin changes in the peripheral nerves. Electrophysiological studies have been reported with contradictory results: traditional maximum conduction velocity (MCV) was within normal range [3,6, 7, 281; mild disturbances were found, however, particularly involving the slow alpha motor fiber pool when the collision technique was applied [ll, 121. We undertook a different approach to the problem of peripheral nerve involvement in MS by determining refractory periods. The results are reported here.
Material and Methods The 36 patients in this study were classified as having definite or probable MS according to the criteria suggested by Bauer [ 11: (1) a clinical course with at least two bouts or chronic progressive development over one year or more; (2) symptoms and signs that are disseminated in place or time; and (3) characteristic cerebrospinal fluid changes, such as mononuclear pleocytosis of up to 50 cells per cubic millimeter and increased immunoglobulin G with ap proximately normal total protein content. The control group consisted of 31 volunteers, including medical students, staff personnel, and some patients with lumbar disc protrusion. The MCV and relative refractory period (RRP) of median nerve sensory fibers were determined by stimulating the digital nerves at the middle finger. Rectangular pulses of 0.1 msec duration and supramaximal intensity (20 to 50 times threshold) served as stimuli. They were generated by a battery operated stimulation unit and triggered by a digital clock (Digitimer, Model 4030). The stimuli were applied percutaneously with salinated ring electrodes, the
From the Department ofClinical Neurology, Johannes Gutenberg University, Maim, Federal Republic of Germany. Accepted for publication May 8, 1978.
cathode being placed close to the basal joint of the middle finger and the anode more distally, 30 to 35 mm away. The evoked sensory nerve action potential (SNAP) was picked up percutaneously by silver disc electrodes covered with saline-soaked cotton. The active electrode was placed above the median nerve at the wrist and the reference electrode 40 mm proximally. The readings from 12 to 20 single sweeps were fed into a hardware averager (Tonnis Model 2 T, 1,024 addresses, 12 bits), summated, and stored. The averaged potential was photographed. The reference points for latency measurements were the artifact of the stimulus and the first positive peak of the SNAP. The RRP was evaluated by relying on changes in latency of a second SNAP that followed the conditioning volley after a short delay. Paired stimuli were used. The intervals between the two stimuli were progressively decreased in successive tests. For each stimulus interyal, the latency of the test SNAP was compared to the resting latency by switching off the conditioning stimulus and superimposing both averaged signals on the screen (Fig 1).The amplitude of the unconditioned SNAP depends on the number of summated events and spontaneous changes of the signal amplitude. Because such changes may amount to almost 10% with unequivocally supramaximal stimulus intensities, the signal amplitude was disregarded [ 171. The shortest stimulus interval at which changes in latency could not be observed will be referred to as the RRP. It was estimated using the “Polynomanpassung”(polynomial fit) according to Marquardt [20](Fig 2). Tissue temperature was measured continuously by a thermocouple needle. Temperature was maintained at 34°C by a heat lamp. Iatencies were not accepted if temperature changes exceeded 0.5”C.
Results The mean values of maximum sensory fiber conduction velocity (MSCV), RRP, and age for the MS Address reprint requests to Prof Hopf, Department of Clinical Neurology. Universitiit Johanna Gutenberg, langenbeckstr. 1. D-6500 Mninz. Federal Republic of Germany.
0 3 6 4 - 5 1 3 4 / 7 8 / ~ 3 $ 0 1 . 2 5@ 1978 by
HW
C. Hopf
499
Tabk I . Comparatht Dctcnnitutionr f w the MS Patients and Contmls (Total Grorrps)
Data No. of subjects Age (yr) Nerve conduction velocity ( d s e c ) Relaave refractory period (msec) Temperature (“C)
Controls 31 28.4f 7.4 63.5 f 4.5
MS Patients 36 38:l f 9.2 61.3 f 5.1
3.12 f 0.3
3.31
34 f 0.3
34.2 2 0.26
* 0.35
Vducs are mean f SD.
Pig I . The paired stimnlns response (lower trace)and the nnconditioned response (upper trace) snperimposed to show dgfmnce in latency. (Intmtimnlns intewal: 2.5 msec.)
’-1
n
.5
Fig 3. Histograms of the RRP valnes from rontrolsnbjerts (domed lines)andMSpatients (solid lines).
Pig 2. Stimnlns intewah (x axis) phtted against SNAP latencies (yaxis)for a singkpatient. The latencies originally measnrtd art indicated by triangks. The cnwc was drawn by the computer according t o “Polynomanpassnng”(polynomialf i t ) . (rR= ralrnlated refrartoy period.)
patients and the controls are presented in Table 1. The two groups differ significantly with r e w d to MSCV and RRP (x* test: p C 0.05). Histograms of the RRP from patients and controls are plotted in Figure 3. Phenytoin, carbamazepine, and diazepam have been observed to decrease MSCV [13]. If subjects are excluded who were receiving one of these drugs (2 controls and 5 MS patients), the difference in MCV between these reduced groups is still significant (p < 0.05; Table 2). MCV decreases with advancing age 12, 221. Because the age factor might have biased our results, two subgroups comprising 16 patients and 15 controls were matched for age. When the two subgroups were compared, the average RRP values were still significantly different (p < 0.05>, but not the MSCV (see Table 2).
Discussion Our results correspond to those of previous studies in that MCV in MS patients could not be demonstrated to be different from control values [3, 6, 7, 12, 281. MCV, however, is a relatively insensitive function that is frequently unaffected in mild peripheral neuropathy, and up to now only s m d patient groups have been examined. It has been proved that slight disturbances of the excitation cycle of peripheral nerves influence RRP more than MCV [15-171. On the other hand, RRP apparently changes little, if at all, with age [14, 271, but considerably with temperature 1181 and hypokalemia [21]. Excluding such influences, RRP differed significantly for the total groups (3 1 controls and 36 MS patients), and for the reduced groups (15 controls and 16 patients) as well. This finding concerning fast fibers is in agreement with comparable observations on slow fibers [12] based on the range of conduction velocities of motor fibers supplying a single muscle, estimated by applying the collision technique [ 111. Our findings support the report of Pollock et al [23], who found an increased number of abnormal teased fibers in nerves from MS patients. Many internodes showed a 50% reduction in myelin thick-
5 0 0 Annals of Neurology Vol 4 No 6 December 1978
Table 2. Comparative Detenninations fw the Reduced Groups of MS Patients and Contmh Data
Controls
MS Patients
EXCLUDING SUBJECTS RECEIVING DRUGS
No. of subjects Nerve conduction
29 63.4 f 4.3
31 61.2 & 5.0
velocity ( d s e c ) Relative refractory period (msec)
3.12 f 0.3
3.28 2 0.33
SUBIECTS MATCHED FOR AGE
No. of subjects Age (yr) Nerve conduction velocity ( d s e c ) Relative refractory period (msec) Temperature (“C)
30.3 & 5.6 62.5 f 4.3
16 30.6 k 5.3 60.8 2 5.3
3.04 2 0.2
3.37 f 0.3
34.1 f 0.35
34.2 f 0.36
15
Values are mean f SD.
ness, and the number of myelin lamellae was reduced compared to the axonal area. Morphological studies had previously suggested peripheral nerve involvement in MS [4, 5, 8, 9, 24, 25, 291; particularly, the papers of Strauber [26],Margulis [19], ikd Hasson et al [lo] should be mentioned in this context. An increased RRP has been found mainly with conditions affecting the myelin of peripheral nemes [IS, 17, 271. We conclude from our results that the basic process in MS is not confined to the central nervous system, but also extends to the myelinated peripheral nerves. The patients in this study were not examined during the initial stages or acute exacerbations of their disease. Thus the peripheral nerve involvement, if brought about during relapses, results at the very least in a mild but permanent deficit. References 1. Bauer HJ: Judgment of the validity of a clinical M S d i q n d s . Acta Neurol Scand 50:171-173, 1974 2. Buchthal P, Rosenfalck A, Trojaborg W Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J Neurol Neurosurg Psychiatry 37:340-360, 1974 3. Conrad B, Bechinger D: Sensorische and motorische Nervenleitgeschwindigkeit and distde Latenz bei Multipler Sklerose. Arch Psychiatr Nervenkr 2 12:140-149, 1969 4. Dawson JW:The histology of disseminated sclerosis. Trans R SOCEdinb 50:517-725, 1916 5. Dinkler P: Zur Kasuistik der Multiplen Sklerose des Gehirns und Ruckenmarks. Dtsch 2 Nervenheilkd 26:233-237,1904 6. Gilliatt RW, Goodman HV, Willison RG: The recording of lateral popliteal nerve action potentials in man. J Neurol Neurosurg Psychiatry 24:305-318, 1961 7. Ginzburg M, Lee M, Ginzburg J: Correlation between muscle
strength and peripheral nerve conduction in multiple sclerosis. Electromyography 11:491-5 13, 1971 8. HallervordenJ: Die zentralen EntmarkuagJkrankheiten.Dtsch Z Nervenheilkd 150201-304, 1940 9. Hassin G B Zwei ungewohnliche F a e von Multipler Sklerose. 2 Ges Neurol Psychiatr 128:99-106, 1930 10. Hasson J, Terry RD, Zimmermann HM: Peripheral neuropathy in multiple sclerosis. Neurology (Minneap) 8:503-5 10, 1958 11. Hopf HC: Electromyographic study on so-called mononeuritis. Arch Neurol 9:307-312, 1963 12. Hopf HC: Leitgeschwindigkeit motorischer Nerven bei der Multiplen Sklerose und unter dem Einfluss hoher CortisonMedikation. Dtsch Z Nervenheilkd 187:522-526, 1965 13. Hopf HC: Anticonvulsant drugs and spike propagation of motor nerves and skeletal muscle. J Neurol Neurosurg Psychiatry 36574-580, 1973 14. Hopf HC. LeQuesne P,Willison RG.Refractory periods and lower limiting frequencies of sensory fibers of the hand, in Kunze K, Desmedt JE (eds): Studies on Neuromuscular Diseases. Basel, Karger, 1975 15. Lehrnann HJ, Lehmann G, Tackmann W: Refraktiirperiode und Ubermialung von Serienimpulsen im N. tibialis des Meerschweinchens. Z Neurol 199:67-85, 1971 16. Low PA, McLeod JG: Refractory period, conduction of trains of impulses, and effect of temperature on conduction in chronic hypertrophic neuropathy. J Neurol Neurosurg Psychiatry 40:434-447, 1977 17. Lowitzsch K, Hopf HC: Refraktkperioden und frequente Impulsfortleitung im gemischten N. ulnaris des Menschen bei Polyneuropathien. Z Neurol 205: 123-144, 1973 18. Lowitzsch K, Hopf HC, Galland J: Changes of sensory conduction velocity and refractory periods with decreasing temperature in man. J N e w 1 216:181-188, 1977 19. Margulis MS: iiber die pathologische Anatomic und Pathogenese der Multiplen Sklerose. Dtsch Z Nervenheilkd 131:125-126, 1933 20. Marquardt DW: An algorithm for least squares estimation of nonlinear parameters. J SOCIndust Appl Math 11:431-441, 1963 21. Maurer K, Hopf HC, Lowitzsch K Hypokalemia shortens relative refractory period of peripheral sensory nerves in man. J Neurol216:67-71, 1977 22. Mulder DW, Lambert EH, Bastron JA, et al: The neuropathies associated with diabetes mellitus. Neurology (Minneap) 11:275-284, 1961 23. Pollock M, CPLder C, Allpress S Peripheral nerve abnormality in multiple sclerosis. Ann Neurol 2:41-48, 1977 24. Schaltenbrand G. Die Mulriple Sklerose des Menschen. Stuttgart, Thieme, 1943 25. Schob F Ein Beitrag fur pathologischen Anatomie der Multiplen Sklerose. Monatsschr Psychiatr Neurol 22:62-73, 1907 2 6. Striihuber A: iibet Regenerations- und Proliferationsvorg h g e bei multipler Sklerose des Nervensysterns. Beitr Pathol Anat 33:409-412, 1903 Refractory period in human sen27. Tackmann W, Lehmann sory nerve fibers. Eur Neurol 12277-292, 1974 28. Taraschi G, Lanzi G ktude de la vitesse de conduction du nerf cubital dans sclerose en plaque. Electroencephalogr Clin Neurophysiol Suppl 22:54-55, 1962 29. Taylor E W Zur pathologischen Anatomie der Multiplen Sklerose. Dtsch Z Nervenheilkd 5:l-2, 1894
w:
H o p f and Eysholdt: Impulse Conduction in MS
501
