Proximal versus Distal Slowing of Motor Nerve Conduction Velocity in the Guillain-Bar& Syndrome Jun Kimura, MD
Using the F wave, a simple equation was devised to calculate the ratio between motor nerve conduction time from the spinal cord to the stimulus site and that of the remaining nerve segment to the muscle (F ratio). In 33 healthy subjects the F ratio (mean t SD) was close to unity for the median nerve (1.04 2 0.09)with stimulation at the elbow and for the peroneal(l.11 2 0.09) and tibial nerves (1.17 2 0.10) with stimulation at the knee. Thus, in these nerves, the time required for the passage of impulses from the cord to the site of stimulation is approximately the same as that from the stimulussite to the muscle. The ratio was significantly more for the ulnar nerve (1.40 ? 0.11) with stimulation below the elbow. Of 126 nerves in the upper and lower extremities from 45 patients with Guillain-Barri syndrome, the F ratio was normal in 65 (51%), increased in 30 (24%), and decreased in 31 (25%). The mean F ratios remained normal in median (1.12 0.40), ulnar (1.38 0.30),peroneal (1.07 0.25), and tibial (1.12 ? 0.20) nerves. These findings together with the results of nerve conduction studies indicate that the conduction abnormality usually affects both proximal and distal segments in the Guillain-Barre syndrome. If selective, it is distributed a t random between the two segments, but there is a tendency toward involvement of common sites of compression and the most proximal, possibly radicular, portion of the nerve.
*
Kimura J: Proximal versus distal slowing of motor nerve conduction velocity in the Gulllain-Barre syndrome. Ann Neurol 3: 144-350, 1978
may be subject to certain errors, especially if a standardized method is not used. Eisen et air63 have recently shown that the ratio between F and M latency is a useful addition in this regard for evaluating proximal entrapment syndromes. For the present study we devised a simple equation to calculate the ratio between motor conduction time from the cord to the stimulus site and that of the remaining nerve segment to the muscle ( F ratio). We have previously shown in afew patients with the Guillain-Bark syndrome that the FWCV may be slow despite normal or borderline MNCV distally [ 11, 121. King and Ashby [13] have since reported similar results. The purpose of this communication is, first, to report the normal values and variations of the F ratio and, second, to delineate whether the Guillain-BarrP syndrome affects preferentially proximal o r distal nerve segments or whether conduction abnormalities are distributed at random. Circumventing the need for the distance determination, the F ratio provides an accurate measure of conduction abnormality of the proximal nerve segment as compared with that of the distal segment. It will be shown that the F ratio is useful in assessing the relative degree of involvement
After a supramaximal electric shock is delivered to a nerve, a late response, designated the F wave [I?], often can be recorded from the innervated muscle following the direct motor potential. There has been considerable discussion in the literature as to whether the F wave represents a reflex response [I51 o r a recurrent discharge of antidromically activated motor neurons 19, 16, 17, 241 or both [71. T h e presence of the F wave in deafferentated limbs [ 16, 171 and after transverse myelotomy [IS], however, proves that it results at least in part from backfiring of motor neurons. Studies by single-fiber electromyography [24, 251 also have shown that the F wave requires direct activation of the motor axon, which can be taken as evidence of its recurrent nature. The latency of the F wave represents passage of the impulse to and from the spinal cord through the most proximal segment of the nerve [9]. Thus, it is possible to calculate the F wave conduction velocity (FWCV) of the most proximal segment analogous to the motor nerve conduction velocity (MNCV) of the more distal segments [3, 6 , 9, 11-13, 19-22]. Determination of FWCV, however, requires accurate estimation of the length of the most proximal nerve segment, which ~
~
From the Division of Clinical Electrophysiology, Department of Neurology, University of Iowa, College of Medicine, IowaCity, IA. Accepted for publication Nov 2, 1377.
344
0364-5134/78/0003-0411$01.25 @ 1978 by Jun Kimura
~~~~~~~~~~~~~~~
Address reprint requests to Dr Kimurd, Division of Clinical Electrophysiology, University Hospitals, Iowa City, IA 52242.
of the proximal nerve segment in Guillain-Bar& syndrome.
FWCV =
Methods The median, ulnar, peroneal, and tibial nerves were stimulated using supramaximai intensity at cwo or three points, and the M response and F wave were recorded with the surface electrodes from the appropriate muscles [9, 11, 121 (Fig 1). All responses were stored for later measurements using an FM tape recorder. The latencies of M response and F wave were measured from the stimulus artifact to the beginning of the evoked potential. The F wave latency was relatively constant in some individuals but in others varied by a few milliseconds from one stimulus to the next. Therefore, at least ten F waves were recorded at each stimulus point to determine the response with the shortest latency. The use of a storage oscilloscope was of particular value for this purpose. If the F wave was difficult to elicit, slight voluntary muscle contraction often enhanced the potential. The F wave represented the muscle potential evoked by discharges of antidromically activated motor neurons. Thus, the F wave first traveled in the centripetal direction toward the spinal cord before it was rransmicted distaiiy to activate the muscle. The time required for passage of impulses to and from the spinal cord was determined as the latency difference between the F wave and the M response. Considering an estimated delay of 1.0 msec for recurrent discharge of motor neurons [23] and the refractory period of motor axons [lo], the conduction time from the cord to the stimulus sire was calculated as (F-M-1)/2, where F and M were latencies of the E wave and M response, respecriveJy (Fig 2). The time required from the stimulus site to the muscle was represented by the latency of the M response. Thus, the latency ratio between the nerve segment proximal Fig I. M response (open brackets) and F waue (small arrows) vecordedfrom the hypothenar muscles (left) and thenar muscles (right) in 2 patients uiilh Guillain-Bark syndrome. The sites of.rupramaximal stimulus to the ulnar and median nerves are shoum. Three consrcutive trials are superimposedfor each tracing. Because of slowing in nerve conduction, the M response and F wave are distinctly separate even with stimulation at the axilla, which is usual4 not the case in normal persons. Rlght Ulnor Nerve
Left Median Nerve
Site of
1
Slimulatiq?
0 Stimulation
Fratio =
U
b 2 mv
5 msec
(F-M-1)/2 ~
=
M
-D
-
(F-M-1)/2
DistunccID) 4
l-(F-M-1)/2
t
F-M-1 (msec)
M x 2 (rnsec) D x 2 (rnm) F-M
- ! (rnsec)
F wove
M
+
F t g 2. Conduction pathways of the M response and F wave. A minimal delay of 1.O msec is estimated for recarrent dmharge of motor neurons and refractory period of motor axom. Thus, the conduction time from the cord t o the stimulus site is calculated as (F - M - 1)12 where F and M are latencies of the F wave and M response, respectively.
to the site of stimulation and the remaining distal segment including the neuromnscular junction was obtained by the following equation:
F ratio =
(F latency - M latency) - 1 (msec) M latency x 2 (msec)
The motor nerve conduction velocities (MNCV) were calculated in the conventional manner. For assessments of FWCV, the approximate length of the median and ulnar nerves from the axilla to the spinal cord was obtained by measuring the surface distance from the stimulus point at the axilla to the C7 spinous process via the midclavicuiar point [9].Likewise, the surface distance from the stimulus site at the knee to the T12 spinous process was determined by way of the greater trochanter of the femur for the peroneal and tibial nerves [ 111. T h e FWCV in the segment to and from the spinal cord was calculated according to the foilowing equation l3.6, 9 , 11-13, 19, 21, 221: FWCV(m/sec) =
(Distance to C7 or T12) x 2 (mm) (F latency - M latency) - 1 (msec)
Materials and Results Normal Subjects Table 1 summarizes rhe normal values in our laboratory. The normal ranges of M and F latencies and conduction velocities were previously established in 33 normal subjects for median and ulnar nerves [9] and in another group of 33 subjects for peroneal and tibial nerves [ 111. The mean minus two standard deviations was taken as the lower limit of normal for MNCV and FWCV. Thus, the FWCV in the cord-toaxilla segment of median and ulnar nerves was considered slow if it was below 50 m/sec. The MNCV of median and ulnar nerves was considered slow if it was less than 50 m/sec in the axilla-to-elbow segment and 46 m/sec in the elbow-to-wrist segment. In the
Kimura: Nerve Conduction Velocity in Guillain-Barrk Syndrome 345
Tclble 1 . Norrncll ValueJ for Motor Newe Conduction (Alean
No. of Nerves Tested
Site of Stimulation
66 Median nerves ( 3 3 subjects)”
Wrist
M Latency (nisec) 3.5
Elbow
66 Ulnar nerves ( 3 3 subjects)”
* S D in 33 Subjects)
7.8
2 2
MNCV between the Two Stimulus Sites
F Ratio ( F - M - 1)/2M (rn/sec)
Stimulus Site im/sec)
29.1 i 2.3
3.64 i 0.45
59.2
5
3.9
62.2
2
5.2
64.3
i 6.4
56.7
i 2.9
0.5 0.8
24.8
* 2.0
1.04
* 0.09
11.3 t 1.0
21.7 i 2.8
0.40 i 0.07
Wrist
2.9 t 0.5
30.5 i 3.0
4.65
F
56.0 i 5.0 63.3
0.75 55.9
6.7
t
to the
F Latency (rnsec)
Axilla
Below elbow
FWCV from the Spinal Cord
0.’
26.0 i 2.0
1.40
2
0.11
2
6.0
* 5.1
58.2 t 2.9
56.9 i 4.6 9.2 i 0.9
Above elbow
23.5 i 2.0
61.1 i 5.4
0 . 7 2 t 0.07
61.3 11.2 i 1 . 0
Axilla
66 Peroneal nerves ( 3 3 subjects)
4.5
Ankle
2
0.9
1.9
0.43
i 0.06
51.3 t 4.7
5.17
-+ 0.91
21.9
i
2
6.8 63.0 t 5.9 53.3
i
3.7
56.3
i 4.9
51.3
2
2.9
54.4
jr
3.6
49.4 t 3.8
66 Tibia1 nerves ( 3 3 subjects)
Knee
12.9
2
1.4
42.7 2 4.0
1.11
Ankle
4.1
i
0.6
52.3
5.91 i 0.90
i 4.3
0.09 46.8
Knee “F wave was clicited by axillary stimulation
12.8
F
1.3
43.5 i 3.4
1.17 t o.10
i
3.4
2 1 o f 13 subjects. ”Mid& segment across elbow was tested in 17 c ) t 33 subjects. In
peroneal and tibial nerves, the lower limits of normal were 4 5 m/sec for the FWCV in the cord-to-knee segment and 4 0 m/sec for the M N C V in the knee-toankle segment. T h e F ratio was calculated in the same normal subjects for the purpose of this study. It varied little among different normal subjects when the stimulus was given at the elbow o r the knee in the upper and lower extremities, respectively. T h e F ratio was more variable with stimulation at the wrist, above the elbow, at the axilla, o r a t the ankle. Therefore, evaluation was primarily based o n the values obtained at the elbow for the median nerve, below the elbow for the ulnar nerve, and at the knee for the peroneal and tibial nerves. At these stimulus sites, the F ratio was close to unity for all bur the ulnar nerve, indicating thar the nerves were bisected into two segments of approximately equal conduction time. T h e ratio was significantly m o r e for the ulnar nerve as expected from the stimulus site below the elbow. T w o standard deviations above and below the mean was taken as the lower and upper limits of normal. Thus, the F ratio was considered abnormal if it fell outside the range of0.86 co 1.22 at the elbow for rhe median nerve and 1.18 to 1.62 below t h e elbow for the ulnar nerve. For the peroneal and tibial nerves the ratio should be between 0.93 and 1.29, and between 0.97 and 1.37, respectively, when the nerves are stimulated at the knee.
346 Annals of Neurology
Vol 3 N o 4
April 1978
GadLain-Buvk Sytidvonie Tables 2 through 4 and Figure 3 summarize the results in 45 patients with the Guillain-Bark syndrome including 14 reported previously [ 11, 121. T h e r e were 24 males and 21 females ranging in age from 7 to 80 years (average age, 41). A total of 57 median, 54 ulnar, 5 I peroneal, and 4 0 tibial nerves were tested in various combinations from these 45 patients. Of these, both M response and F wave !Fig 1) were recorded in 39 median, 33 ulnar, 3 1 peroneal, and 23 tibial nerves. Only these nerves were included in the analysis. As shown in Table 2 , the latencies of M response and F wave were greater (Fig 3 ) and MNCV and FWCV smaller in the patients with Guillain-BarrP syndrome than in the normal subjects Cj < 0.01). However, the F ratio remained normal, indicating that the nerves were affected to the same degree above and below the stimulus site at the elbow and knee, respectively. Analysis of the F ratio in the individual nerves (Table 3) showed that it was normal in 65 f5lc/c), increased in 30 (24q,), and decreased in 3 1 (25v;) of the 126 nerves in which both the M response and F wave were recorded. Thus, a majority of the nerves tested were slowed equally in the segments above and below the stimulus site, whereas the remaining nerves were predominantly slowed in either proximal or distal segments but with equal preference between the two.
Table 2. Motor Nerve Conduction i n Guillain-BarriSyndrome (Mean +- SD in 45 Patients) a
No. of Nerves Tested
Site of Stimulation
39 Median nerves (27 patients)
Wrist
M Latency (msec) 5.8
Elbow
33 Ulnar nerves (22 patients)
?
3.4
11.2 t 5.5
F Latency (msec)
(mlsec)
38.1 2 14.4
3.25 +- 1.00
47.3 t 11.0
32.5
4
Axilla
14.7
%
6.4
29.7
5
Wrist
4.4
2
2.2
37.7
2
10.8
8.6 % 2.6
to the
Stimulus Sire
49.2
k
13.1
56.2
?
14.4
1.12 2 0.40
46.7 2 10.0
10.5 0.52 +- 0.15
9.5
44.2 t 12.4
4.22 +- 1.32 51.3
Below elbow
FWCV from the Spinal Cord
MNCV between the Two Stimulus F Ratio Sites (F - M - 1)/2M (mlsec)
32.9 7t 7.9
-C
46.4
2
9.7
45.5
rt_
9.6
45.0
rt
9.7
11.8
1.38 f 0.30 46.2 t 12.1
Above elbow
11.8 4 3.7
31.0
4
8.0
0.78
If:
0.14 56.6
31 Peroneal nerves ( 1 9 patients)
Axilla
14.7
2
4.8
27.9 t 6.7
0.43 +- 0.11
Ankle
7.8
2
5.3
60.9 2 12.9
4.21 +- 1.96
23 Tibial nerves (15 patients)
17.1 % 6.5
Ankle
6.1
?
2.7
50.5
2
12.5
1.07 +- 0.25
60.6
2
12.1
5.08
5
41.2 t 7.3
15.6 5 4.4
49.5 2 9.7
?
8.5 43.6 c 13.2
41.4 t 9.3
1.69 41.1
Knee
16.3 46.6 c 11.2
42.4
Knee
&
1.12 +- 0.20
?
10.0 41.9 t 11.2
“Only those nerves for which both F wave and M response were elicited are included.
Table 3. F Ratio wtth Stirnwlation at the Elbou o r Knee i92 the Guzlluin-Burr>Syndrome
No. Nerve
Tested
Normal Decreased
Increased
Median Ulnar Peroneal Tibial All nerves combined
39 33 31 23 126
16 20 13 16
12 10 6 2
11 3 12 5 31
65
30
Table 4. MNCV in the Most Distal Segment versus FWCV in the Most Proximal Segment in the Guillain-Bur& Syndrome ~~
Nerve Median Ulnar Peroneal Tibial All nerves combined
No. Tested 39 33 31 23 126
MNCV in Elbow-toWrist or Knee-toAnkle Segmen t
FWCV in the Cord-to-Axilla or Cord-toKnee Segment
Normal Slow
Normal Slow
25 25 21 16 87
14 12 16
14 8 10 7 39
9 51
25 21 15 14 75
Although the F ratio remained normal in t h e patients, as discussed, it was not necessarily associated with uniformly slow conduction along the entire length of the peripheral nerve (see Table 2). In the median and ulnar nerves, the FWCV in the cord-toaxilla segment was significantly (p < 0.01) slower than the M N C V in the axilla-to-elbow segment. The average value of FWCV in the cord-to-axilla segment was also less than that of M N C V in the elbow-to-wrist segment although the difference was not statistically significant. In both nerves, conduction delay in the terminal segment from the wrist to the muscle was out of proportion to the slowing of M N C V in the elbowto-wrist segment. In the ulnar nerve, the FWCV in the cord-to-axilla segment was similar to the MNCV in the above-to-below elbow segment. In the lower extremities the FWCV of the proximal segment and the MNCV of the distal segment were equally slow for both peroneal and tibia1 nerves. In calculating the F ratio, all the factors mentioned were cumulative, but in general a marked increase in terminal latency tended to compensate for an equally prominent slowing of nerve conduction of the most proximal segment. When the incidence of abnormality was anaIyted in individual nerves (Table 4 ) , slowing in FWCV of the cord-to-axilla segment was more frequent than that in M N C V of the elbow-to-wrist segment for both median (p < 0.05) and ulnar nerves ( p < 0.01). Indeed,
Kimura: Nerve Conduction Velocity i n Guillain-Barr&Syndrome 347
7
'OOr
[I
- ..
40
8,
I
10 0'
A
Normol Control ( E l subleclsl
Guilloin-Barre' Syndrome
0
(27 patients)
B
the FWCV in the most proximal segment was slow when the M N C V in the more distal segments was normal or borderfine in 15 median and 14 ulnar nerves. In the lower extremities there was no significant difference in the incidence of conduction abnormality between the proximal and distal segments for both tibial and peroneal nerves. In 7 peroneal and 7 tibial nerves, however, the FWCV of the proximal segment was slow when the M N C V of the distal segment was normal.
Discussion Since the F wave presumably represents the conduction time of the motor impulse to and from the spinal cord [16-18, 24, 251, F wave conduction in this segment can be directly compared to the motor conduction in the more distal segments [3, 6, 9, 11-13, 19-22]. In the present study, we used a simple equation to calculate the ratio of the motor nerve conduction time from the cord to the stimulus site and that of the remaining distal nerve segment to the muscle. Designated the F ratio, it provided an accurate electrophysiological means to assess the conduction characteristics of the proximal segment relative to those of the distal segment. Unlike FWCV, the F ratio is entirely independent of the total length of the nerve but is based on the assumption that the proportion between the proximal and distal segments is the same among different subjects. In an experiment using a baboon, it has been shown
348 Annals of Neurology
Vol 3 No
4 April
1978
Normol Control (33 subleclsl
Gu~lloin-BorreSyndrome (22 patients)
Fig 3 . Latencies of F and M response of (A, the t i i t d i d i f n e w t and t B ) the ulnur nerve in norrrialsu~j~t.land in Guiilazri-Bark syndronie. Responxs on the right and ieji iidei ure combined in each category. Only those nerz!esfor tchirh both F i c m e and M response u'ere elicited a r e inrhded. Sitri ofnrrtme stimulation are indicated in the key. Thr diflerence in lutency between F waoe and M response (triangles) is the time interi'ul requiredfor passage of the impulse.c t o and from the Jpinul mrd. w m l e
that the conduction velocity of single motor axons is slower in the brachial plexus than in the nerves [Z]. In man, nerve conduction as determined by FWCV and M N C V is normally faster in the proximal segment above the elbow or knee than in the remaining distal segment [3,6,9,11-13,19-221. However, the Fratio is close to unity for the median nerve with stimulation at the elbow and for the peroneal and tibia1 nerves with stimulation at the knee. Thus, the time required for the passage ofimpulses from the cord to the site of stimulation is approximately the same as that from the stimulus site to the muscle. This is expected since surface determination indicates that the cord-toelbow segment is longer than the elbow-to-muscle segment. Likewise, the cord-to-knee segment is substantially longer than the knee-to-muscle segment whether estimated by surface determination o r measured directly in cadavers [ l l]. Clinical values of the F wave have previously been demonstrated in various diseases 13, 6, 9, 11-13, 19-22]. In Guillain-Bar& syndrome it has been
shown that the FWCV in the most proximal segment may be abnormal even when the M N C V in the more distal segments is normal or borderline [ 11-13]. The present findings confirm these observations and further document that slowing of conduction tends to be localized to the most proximal segment and to the terminal portion of motor fibers. Although the FWCV in the cord-to-axilla segment is in general slower than the MNCV in the more distal segments in both median and ulnar nerves, terminal latencies are also greatly increased in these nerves. Thus, the F ratio remains normal i n a majority of the affected nerves, indicating rhat slowing of nerve conduction is romparable between the cord-to-elbow and elbow-to-muscle segments. Similarly, slowing of nerve conduction is the same between the cord-to-knee and knee-tomuscle segments for both peroneal and tibia1 nerves in the lower extremities. A localized conduction abnormality can occur in any of these nerves but, if selective, is distributed at random without particular preference between proximal and distal segmenrs. From the data presented here it may be concluded that conduction abnormalities in the Guillain-Barre syndrome are usually diffuse, involving any segment of the peripheral nerve. There is, however, a tendency toward involvement of the terminal portion of the motor fibers and the most proximal, possibly radicular, portions of the nerve, whereas the main nerve trunk is relatively less affecred. The distal slowing may be secondary to superimposed compression, as previously suggested by Lambert and Mulder [14] and others [ 5 , 8 ] .Prominent slowing of the ulnar nerve in the segment across the elbow supports such a contention. Additionally, the terminal segment may be peculiarly vulnerable because of its distance from the cell body or for y e t undetermined reasons. Obviously, these findings do not necessarily indicate the location of the initial abnormality. It may be that the original lesions center around the radicular portion, which can cause a secondary conduction abnormality in the distal segments. Nonetheless, it is clear from our data that demyelination or some other pathological process responsible for slowing of conduction in this syndrome affecrs rhe entire length of the peripheral nerve either continuously o r discontinuously. This is consistent with the histological findings of Asbury et a1 [ l ] that lesions exist throughout the peripheral nervous system although the segment of maximal involvement varies. In a series of patients with chronic inflammatory polyradiculoneuropathy, Dyck e t a1 141 reported diffusely slow conduction velocity of peripheral nerves although the most marked slowing was often very proximal. Electrophysiological study of the most proximal segment of the peripheral nerve is desirable for overall appraisal of the affected nerves to delineate the under-
lying pathophysiology of various polyneuropathies. In addition to the F wave latency and FWCV, the F ratio appears promising in providing R simple means to compare the motor nerve conduction of the proximal segment to that of the distal segment. For calculation of FWCV, it is assumed rhat surface determinations provide the length of the most proximal nerve segment with fair accuracy. For the F ratio to be meaningful, it is essential that the proximal and distal portions of the extremities be of the same proportion among different subjects. Although both prerequisites are reasonable, they are subject to certain errors. Since these assumptions are of a totally different nature, it is udikely both would be incorrect at the same time. When the FWCV and F ratio are both altered, therefore, the validity of the abnormality under consideration is strengthened considerably. Eisen et al [b] have reported that the ratio between F and M latency is of practical clinical value in differentiating radiculopathp from distal compression syndromes. Since the rightto-left difference of the F ratio described here is very small in individual sublecrs, this comparison might also provide a reliable means of assessing unilateral root lesions. 1 wish to thank Mr David Walker, MSEE, for electrical engineering assistance, Dr Leon F. Burmrister for providing statistical analysis, and Sheila Mennen, Joanne Colter, and Karen Thompson for technical assistance.
References 1. Asbury AK, Arnason BG, Adarns RD: The inflammatory lesion in idiopathic polyneuritis: its role in pathogenesis. Medicine 48:173-215, 1969 2. Clough JFM, Kernell D, Phillips CG: Conduction velocity in proximal and distal porrions offorelimb axons in the baboon. J Physiol (Lond) 198:16?-178, 1968 3 . Conrad 8,Aschoff JC, Fischler M: Der diagncisiische Werr Jer F-Wellen-Latenz. J Neurol 210:151-159, 1975 4. Dyck PJ, Lais AC, Ohta M, et al: Chronic inflammatory polyradiculoneuropathy. Mayo Clin Proc 50:621-637, 1975 5 . Eisen A, Humphreys P: T h e Guillain Bar& syndrome: aclinical and electrodiagnostic study of 25 cases. Arch Neurol30:438-
441, 1974
6. Eisen A, Schorner D, Melrned C: The application of F wave measurements in the differentiation of proximal aod distal upper limb entrapments. Nwrology (Minneap) 2?:662-668, 1977 7 . Gassel MM, Wiesendanger M: Recurrent and reflex discharges in plantar muscles of the cat. Acta Physiol Scand 63.138-142, 1965 8. Humphrey Jd: Motor nerve conducrion studies in the Landry-Guillain-BarrP syndrome (acute ascending polyneuropathy). Electroencephalogr Clin Neurophysiol 17:96, 1964 9. Klmura J: F wave veloc~ryin the centrkl segmenr of the median and uInar nerves: a study in normal subjects and in parienrs wirh Charcot-Marie-Tooth disease. Neurology (Minneap) 24: 539-546, 1974 1 0 ~Kimura J: A method for estimating the refractory period of motor fibers in the human peripheral nerve. J Neurol Sci 28:485-490, I976
Kimura: Nerve Conduction Velocity in Guillain-Barre Syndrome
349
11. Kimura J, Bosch P, Lindsay GM: F wave conduction velocity in the central segment of the peroneal and tibial nerves. Arch Phys Med Rehab 56:492-497, 1975 12. Kimura J, Butzer JF: F wave conduction velocity in Guillain Bar& syndrome: assessment of nerve segment between axilla and spinal cord. Arch Neurol 33:524-529, 1075 13. King D , Ashby P: Conduction velocity in the proximal segments of a motor nerve in the Guillain Bar& syndrome. J Neurol Neurosurg Psychiatry 39:538-544, 1976 14. Lambert EH, Mulder DW: Nerve conduction in the Guillain Bar& syndrome. Electroencephalogr Clin Neurophysiol 17:86, 1964 15. Magladery JW, McDougal DB Jr: Electrophysiological studies of nerve and reflex activity in normal man: identification of certain reflexes in the electromyogram and the conduction velocity of peripheral nerve fibers. Johns Hopkins Bull 86:265-290, 1950 16. Mayer RF, Feldman RG: Observations on the nature of the F wave in man. Neurology (Minneap) 17:147-156, 1967 17. McLeod JG, Wray SH: An experimental study of the F wave in the baboon. J Neurol Neurosurg Psychiatry 29: 196-200.1966 18. Miglietta OE: The F-response after transverse myelotomy, in Desmedt JE (ed): New Developments in Electromyography
3 5 0 Annals of Neurology Vol 3 No 4 April 1978
19.
20. 2 1.
22.
23.
24.
25.
and Clinical Neurophysiology. Basel, Karger, 1973, vol 3, pp 323-327 Muller D: Die Bestimmung der F-Wellengeschwindigkeit am N. ulnaris Gesunder. Psychiatr Neurol Med Psycho1 (Leipz) 27:619-623, 1975 Panayiotopoulos CP, Scarpalezos S: Neural and muscular involvement in dystrophia myotonica. Lancet 2:329, 1975 Panayiotopoulos CP, Scarpalezos S: F-wave studies on the deep peroneal nerve. Part 2, 1. Chronic renal failure; 2. Limb-girdle muscular dystrophy. J Neurol Sci 31:331-341, 1977 Panayiotopoulos CP, Scarpalezos S, Nasras PE: F-wave studies on the deep peroneal nerve. Part 1, Conrrol subjects. J Neurol Sci 31:319-329, 1977 Renshaw B: Influence of discharge of motoneurons upon excitation of neighboring motoneurons. J Neurophysiol 4: 167183, 1941 Thorne J: Central responses to electrical activation of the peripheral nerves supplying the intrinsic hand muscles. J Neurol Neurosurg Psychiatry 28:482-495, 1965 Trontelj JV: A study of the F-response by single fibre electromyography, in Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology. Basel, Karger, 1973, vol 3, pp 318-322
Using the F wave, a simple equation was devised to calculate the ratio between motor nerve conduction time from the spinal cord to the stimulus site and that of the remaining nerve segment to the muscle (F ratio). In 33 healthy subjects the F ratio (mean t SD) was close to unity for the median nerve (1.04 2 0.09)with stimulation at the elbow and for the peroneal(l.11 2 0.09) and tibial nerves (1.17 2 0.10) with stimulation at the knee. Thus, in these nerves, the time required for the passage of impulses from the cord to the site of stimulation is approximately the same as that from the stimulussite to the muscle. The ratio was significantly more for the ulnar nerve (1.40 ? 0.11) with stimulation below the elbow. Of 126 nerves in the upper and lower extremities from 45 patients with Guillain-Barri syndrome, the F ratio was normal in 65 (51%), increased in 30 (24%), and decreased in 31 (25%). The mean F ratios remained normal in median (1.12 0.40), ulnar (1.38 0.30),peroneal (1.07 0.25), and tibial (1.12 ? 0.20) nerves. These findings together with the results of nerve conduction studies indicate that the conduction abnormality usually affects both proximal and distal segments in the Guillain-Barre syndrome. If selective, it is distributed a t random between the two segments, but there is a tendency toward involvement of common sites of compression and the most proximal, possibly radicular, portion of the nerve.
*
Kimura J: Proximal versus distal slowing of motor nerve conduction velocity in the Gulllain-Barre syndrome. Ann Neurol 3: 144-350, 1978
may be subject to certain errors, especially if a standardized method is not used. Eisen et air63 have recently shown that the ratio between F and M latency is a useful addition in this regard for evaluating proximal entrapment syndromes. For the present study we devised a simple equation to calculate the ratio between motor conduction time from the cord to the stimulus site and that of the remaining nerve segment to the muscle ( F ratio). We have previously shown in afew patients with the Guillain-Bark syndrome that the FWCV may be slow despite normal or borderline MNCV distally [ 11, 121. King and Ashby [13] have since reported similar results. The purpose of this communication is, first, to report the normal values and variations of the F ratio and, second, to delineate whether the Guillain-BarrP syndrome affects preferentially proximal o r distal nerve segments or whether conduction abnormalities are distributed at random. Circumventing the need for the distance determination, the F ratio provides an accurate measure of conduction abnormality of the proximal nerve segment as compared with that of the distal segment. It will be shown that the F ratio is useful in assessing the relative degree of involvement
After a supramaximal electric shock is delivered to a nerve, a late response, designated the F wave [I?], often can be recorded from the innervated muscle following the direct motor potential. There has been considerable discussion in the literature as to whether the F wave represents a reflex response [I51 o r a recurrent discharge of antidromically activated motor neurons 19, 16, 17, 241 or both [71. T h e presence of the F wave in deafferentated limbs [ 16, 171 and after transverse myelotomy [IS], however, proves that it results at least in part from backfiring of motor neurons. Studies by single-fiber electromyography [24, 251 also have shown that the F wave requires direct activation of the motor axon, which can be taken as evidence of its recurrent nature. The latency of the F wave represents passage of the impulse to and from the spinal cord through the most proximal segment of the nerve [9]. Thus, it is possible to calculate the F wave conduction velocity (FWCV) of the most proximal segment analogous to the motor nerve conduction velocity (MNCV) of the more distal segments [3, 6 , 9, 11-13, 19-22]. Determination of FWCV, however, requires accurate estimation of the length of the most proximal nerve segment, which ~
~
From the Division of Clinical Electrophysiology, Department of Neurology, University of Iowa, College of Medicine, IowaCity, IA. Accepted for publication Nov 2, 1377.
344
0364-5134/78/0003-0411$01.25 @ 1978 by Jun Kimura
~~~~~~~~~~~~~~~
Address reprint requests to Dr Kimurd, Division of Clinical Electrophysiology, University Hospitals, Iowa City, IA 52242.
of the proximal nerve segment in Guillain-Bar& syndrome.
FWCV =
Methods The median, ulnar, peroneal, and tibial nerves were stimulated using supramaximai intensity at cwo or three points, and the M response and F wave were recorded with the surface electrodes from the appropriate muscles [9, 11, 121 (Fig 1). All responses were stored for later measurements using an FM tape recorder. The latencies of M response and F wave were measured from the stimulus artifact to the beginning of the evoked potential. The F wave latency was relatively constant in some individuals but in others varied by a few milliseconds from one stimulus to the next. Therefore, at least ten F waves were recorded at each stimulus point to determine the response with the shortest latency. The use of a storage oscilloscope was of particular value for this purpose. If the F wave was difficult to elicit, slight voluntary muscle contraction often enhanced the potential. The F wave represented the muscle potential evoked by discharges of antidromically activated motor neurons. Thus, the F wave first traveled in the centripetal direction toward the spinal cord before it was rransmicted distaiiy to activate the muscle. The time required for passage of impulses to and from the spinal cord was determined as the latency difference between the F wave and the M response. Considering an estimated delay of 1.0 msec for recurrent discharge of motor neurons [23] and the refractory period of motor axons [lo], the conduction time from the cord to the stimulus sire was calculated as (F-M-1)/2, where F and M were latencies of the E wave and M response, respecriveJy (Fig 2). The time required from the stimulus site to the muscle was represented by the latency of the M response. Thus, the latency ratio between the nerve segment proximal Fig I. M response (open brackets) and F waue (small arrows) vecordedfrom the hypothenar muscles (left) and thenar muscles (right) in 2 patients uiilh Guillain-Bark syndrome. The sites of.rupramaximal stimulus to the ulnar and median nerves are shoum. Three consrcutive trials are superimposedfor each tracing. Because of slowing in nerve conduction, the M response and F wave are distinctly separate even with stimulation at the axilla, which is usual4 not the case in normal persons. Rlght Ulnor Nerve
Left Median Nerve
Site of
1
Slimulatiq?
0 Stimulation
Fratio =
U
b 2 mv
5 msec
(F-M-1)/2 ~
=
M
-D
-
(F-M-1)/2
DistunccID) 4
l-(F-M-1)/2
t
F-M-1 (msec)
M x 2 (rnsec) D x 2 (rnm) F-M
- ! (rnsec)
F wove
M
+
F t g 2. Conduction pathways of the M response and F wave. A minimal delay of 1.O msec is estimated for recarrent dmharge of motor neurons and refractory period of motor axom. Thus, the conduction time from the cord t o the stimulus site is calculated as (F - M - 1)12 where F and M are latencies of the F wave and M response, respectively.
to the site of stimulation and the remaining distal segment including the neuromnscular junction was obtained by the following equation:
F ratio =
(F latency - M latency) - 1 (msec) M latency x 2 (msec)
The motor nerve conduction velocities (MNCV) were calculated in the conventional manner. For assessments of FWCV, the approximate length of the median and ulnar nerves from the axilla to the spinal cord was obtained by measuring the surface distance from the stimulus point at the axilla to the C7 spinous process via the midclavicuiar point [9].Likewise, the surface distance from the stimulus site at the knee to the T12 spinous process was determined by way of the greater trochanter of the femur for the peroneal and tibial nerves [ 111. T h e FWCV in the segment to and from the spinal cord was calculated according to the foilowing equation l3.6, 9 , 11-13, 19, 21, 221: FWCV(m/sec) =
(Distance to C7 or T12) x 2 (mm) (F latency - M latency) - 1 (msec)
Materials and Results Normal Subjects Table 1 summarizes rhe normal values in our laboratory. The normal ranges of M and F latencies and conduction velocities were previously established in 33 normal subjects for median and ulnar nerves [9] and in another group of 33 subjects for peroneal and tibial nerves [ 111. The mean minus two standard deviations was taken as the lower limit of normal for MNCV and FWCV. Thus, the FWCV in the cord-toaxilla segment of median and ulnar nerves was considered slow if it was below 50 m/sec. The MNCV of median and ulnar nerves was considered slow if it was less than 50 m/sec in the axilla-to-elbow segment and 46 m/sec in the elbow-to-wrist segment. In the
Kimura: Nerve Conduction Velocity in Guillain-Barrk Syndrome 345
Tclble 1 . Norrncll ValueJ for Motor Newe Conduction (Alean
No. of Nerves Tested
Site of Stimulation
66 Median nerves ( 3 3 subjects)”
Wrist
M Latency (nisec) 3.5
Elbow
66 Ulnar nerves ( 3 3 subjects)”
* S D in 33 Subjects)
7.8
2 2
MNCV between the Two Stimulus Sites
F Ratio ( F - M - 1)/2M (rn/sec)
Stimulus Site im/sec)
29.1 i 2.3
3.64 i 0.45
59.2
5
3.9
62.2
2
5.2
64.3
i 6.4
56.7
i 2.9
0.5 0.8
24.8
* 2.0
1.04
* 0.09
11.3 t 1.0
21.7 i 2.8
0.40 i 0.07
Wrist
2.9 t 0.5
30.5 i 3.0
4.65
F
56.0 i 5.0 63.3
0.75 55.9
6.7
t
to the
F Latency (rnsec)
Axilla
Below elbow
FWCV from the Spinal Cord
0.’
26.0 i 2.0
1.40
2
0.11
2
6.0
* 5.1
58.2 t 2.9
56.9 i 4.6 9.2 i 0.9
Above elbow
23.5 i 2.0
61.1 i 5.4
0 . 7 2 t 0.07
61.3 11.2 i 1 . 0
Axilla
66 Peroneal nerves ( 3 3 subjects)
4.5
Ankle
2
0.9
1.9
0.43
i 0.06
51.3 t 4.7
5.17
-+ 0.91
21.9
i
2
6.8 63.0 t 5.9 53.3
i
3.7
56.3
i 4.9
51.3
2
2.9
54.4
jr
3.6
49.4 t 3.8
66 Tibia1 nerves ( 3 3 subjects)
Knee
12.9
2
1.4
42.7 2 4.0
1.11
Ankle
4.1
i
0.6
52.3
5.91 i 0.90
i 4.3
0.09 46.8
Knee “F wave was clicited by axillary stimulation
12.8
F
1.3
43.5 i 3.4
1.17 t o.10
i
3.4
2 1 o f 13 subjects. ”Mid& segment across elbow was tested in 17 c ) t 33 subjects. In
peroneal and tibial nerves, the lower limits of normal were 4 5 m/sec for the FWCV in the cord-to-knee segment and 4 0 m/sec for the M N C V in the knee-toankle segment. T h e F ratio was calculated in the same normal subjects for the purpose of this study. It varied little among different normal subjects when the stimulus was given at the elbow o r the knee in the upper and lower extremities, respectively. T h e F ratio was more variable with stimulation at the wrist, above the elbow, at the axilla, o r a t the ankle. Therefore, evaluation was primarily based o n the values obtained at the elbow for the median nerve, below the elbow for the ulnar nerve, and at the knee for the peroneal and tibial nerves. At these stimulus sites, the F ratio was close to unity for all bur the ulnar nerve, indicating thar the nerves were bisected into two segments of approximately equal conduction time. T h e ratio was significantly m o r e for the ulnar nerve as expected from the stimulus site below the elbow. T w o standard deviations above and below the mean was taken as the lower and upper limits of normal. Thus, the F ratio was considered abnormal if it fell outside the range of0.86 co 1.22 at the elbow for rhe median nerve and 1.18 to 1.62 below t h e elbow for the ulnar nerve. For the peroneal and tibial nerves the ratio should be between 0.93 and 1.29, and between 0.97 and 1.37, respectively, when the nerves are stimulated at the knee.
346 Annals of Neurology
Vol 3 N o 4
April 1978
GadLain-Buvk Sytidvonie Tables 2 through 4 and Figure 3 summarize the results in 45 patients with the Guillain-Bark syndrome including 14 reported previously [ 11, 121. T h e r e were 24 males and 21 females ranging in age from 7 to 80 years (average age, 41). A total of 57 median, 54 ulnar, 5 I peroneal, and 4 0 tibial nerves were tested in various combinations from these 45 patients. Of these, both M response and F wave !Fig 1) were recorded in 39 median, 33 ulnar, 3 1 peroneal, and 23 tibial nerves. Only these nerves were included in the analysis. As shown in Table 2 , the latencies of M response and F wave were greater (Fig 3 ) and MNCV and FWCV smaller in the patients with Guillain-BarrP syndrome than in the normal subjects Cj < 0.01). However, the F ratio remained normal, indicating that the nerves were affected to the same degree above and below the stimulus site at the elbow and knee, respectively. Analysis of the F ratio in the individual nerves (Table 3) showed that it was normal in 65 f5lc/c), increased in 30 (24q,), and decreased in 3 1 (25v;) of the 126 nerves in which both the M response and F wave were recorded. Thus, a majority of the nerves tested were slowed equally in the segments above and below the stimulus site, whereas the remaining nerves were predominantly slowed in either proximal or distal segments but with equal preference between the two.
Table 2. Motor Nerve Conduction i n Guillain-BarriSyndrome (Mean +- SD in 45 Patients) a
No. of Nerves Tested
Site of Stimulation
39 Median nerves (27 patients)
Wrist
M Latency (msec) 5.8
Elbow
33 Ulnar nerves (22 patients)
?
3.4
11.2 t 5.5
F Latency (msec)
(mlsec)
38.1 2 14.4
3.25 +- 1.00
47.3 t 11.0
32.5
4
Axilla
14.7
%
6.4
29.7
5
Wrist
4.4
2
2.2
37.7
2
10.8
8.6 % 2.6
to the
Stimulus Sire
49.2
k
13.1
56.2
?
14.4
1.12 2 0.40
46.7 2 10.0
10.5 0.52 +- 0.15
9.5
44.2 t 12.4
4.22 +- 1.32 51.3
Below elbow
FWCV from the Spinal Cord
MNCV between the Two Stimulus F Ratio Sites (F - M - 1)/2M (mlsec)
32.9 7t 7.9
-C
46.4
2
9.7
45.5
rt_
9.6
45.0
rt
9.7
11.8
1.38 f 0.30 46.2 t 12.1
Above elbow
11.8 4 3.7
31.0
4
8.0
0.78
If:
0.14 56.6
31 Peroneal nerves ( 1 9 patients)
Axilla
14.7
2
4.8
27.9 t 6.7
0.43 +- 0.11
Ankle
7.8
2
5.3
60.9 2 12.9
4.21 +- 1.96
23 Tibial nerves (15 patients)
17.1 % 6.5
Ankle
6.1
?
2.7
50.5
2
12.5
1.07 +- 0.25
60.6
2
12.1
5.08
5
41.2 t 7.3
15.6 5 4.4
49.5 2 9.7
?
8.5 43.6 c 13.2
41.4 t 9.3
1.69 41.1
Knee
16.3 46.6 c 11.2
42.4
Knee
&
1.12 +- 0.20
?
10.0 41.9 t 11.2
“Only those nerves for which both F wave and M response were elicited are included.
Table 3. F Ratio wtth Stirnwlation at the Elbou o r Knee i92 the Guzlluin-Burr>Syndrome
No. Nerve
Tested
Normal Decreased
Increased
Median Ulnar Peroneal Tibial All nerves combined
39 33 31 23 126
16 20 13 16
12 10 6 2
11 3 12 5 31
65
30
Table 4. MNCV in the Most Distal Segment versus FWCV in the Most Proximal Segment in the Guillain-Bur& Syndrome ~~
Nerve Median Ulnar Peroneal Tibial All nerves combined
No. Tested 39 33 31 23 126
MNCV in Elbow-toWrist or Knee-toAnkle Segmen t
FWCV in the Cord-to-Axilla or Cord-toKnee Segment
Normal Slow
Normal Slow
25 25 21 16 87
14 12 16
14 8 10 7 39
9 51
25 21 15 14 75
Although the F ratio remained normal in t h e patients, as discussed, it was not necessarily associated with uniformly slow conduction along the entire length of the peripheral nerve (see Table 2). In the median and ulnar nerves, the FWCV in the cord-toaxilla segment was significantly (p < 0.01) slower than the M N C V in the axilla-to-elbow segment. The average value of FWCV in the cord-to-axilla segment was also less than that of M N C V in the elbow-to-wrist segment although the difference was not statistically significant. In both nerves, conduction delay in the terminal segment from the wrist to the muscle was out of proportion to the slowing of M N C V in the elbowto-wrist segment. In the ulnar nerve, the FWCV in the cord-to-axilla segment was similar to the MNCV in the above-to-below elbow segment. In the lower extremities the FWCV of the proximal segment and the MNCV of the distal segment were equally slow for both peroneal and tibia1 nerves. In calculating the F ratio, all the factors mentioned were cumulative, but in general a marked increase in terminal latency tended to compensate for an equally prominent slowing of nerve conduction of the most proximal segment. When the incidence of abnormality was anaIyted in individual nerves (Table 4 ) , slowing in FWCV of the cord-to-axilla segment was more frequent than that in M N C V of the elbow-to-wrist segment for both median (p < 0.05) and ulnar nerves ( p < 0.01). Indeed,
Kimura: Nerve Conduction Velocity i n Guillain-Barr&Syndrome 347
7
'OOr
[I
- ..
40
8,
I
10 0'
A
Normol Control ( E l subleclsl
Guilloin-Barre' Syndrome
0
(27 patients)
B
the FWCV in the most proximal segment was slow when the M N C V in the more distal segments was normal or borderfine in 15 median and 14 ulnar nerves. In the lower extremities there was no significant difference in the incidence of conduction abnormality between the proximal and distal segments for both tibial and peroneal nerves. In 7 peroneal and 7 tibial nerves, however, the FWCV of the proximal segment was slow when the M N C V of the distal segment was normal.
Discussion Since the F wave presumably represents the conduction time of the motor impulse to and from the spinal cord [16-18, 24, 251, F wave conduction in this segment can be directly compared to the motor conduction in the more distal segments [3, 6, 9, 11-13, 19-22]. In the present study, we used a simple equation to calculate the ratio of the motor nerve conduction time from the cord to the stimulus site and that of the remaining distal nerve segment to the muscle. Designated the F ratio, it provided an accurate electrophysiological means to assess the conduction characteristics of the proximal segment relative to those of the distal segment. Unlike FWCV, the F ratio is entirely independent of the total length of the nerve but is based on the assumption that the proportion between the proximal and distal segments is the same among different subjects. In an experiment using a baboon, it has been shown
348 Annals of Neurology
Vol 3 No
4 April
1978
Normol Control (33 subleclsl
Gu~lloin-BorreSyndrome (22 patients)
Fig 3 . Latencies of F and M response of (A, the t i i t d i d i f n e w t and t B ) the ulnur nerve in norrrialsu~j~t.land in Guiilazri-Bark syndronie. Responxs on the right and ieji iidei ure combined in each category. Only those nerz!esfor tchirh both F i c m e and M response u'ere elicited a r e inrhded. Sitri ofnrrtme stimulation are indicated in the key. Thr diflerence in lutency between F waoe and M response (triangles) is the time interi'ul requiredfor passage of the impulse.c t o and from the Jpinul mrd. w m l e
that the conduction velocity of single motor axons is slower in the brachial plexus than in the nerves [Z]. In man, nerve conduction as determined by FWCV and M N C V is normally faster in the proximal segment above the elbow or knee than in the remaining distal segment [3,6,9,11-13,19-221. However, the Fratio is close to unity for the median nerve with stimulation at the elbow and for the peroneal and tibia1 nerves with stimulation at the knee. Thus, the time required for the passage ofimpulses from the cord to the site of stimulation is approximately the same as that from the stimulus site to the muscle. This is expected since surface determination indicates that the cord-toelbow segment is longer than the elbow-to-muscle segment. Likewise, the cord-to-knee segment is substantially longer than the knee-to-muscle segment whether estimated by surface determination o r measured directly in cadavers [ l l]. Clinical values of the F wave have previously been demonstrated in various diseases 13, 6, 9, 11-13, 19-22]. In Guillain-Bar& syndrome it has been
shown that the FWCV in the most proximal segment may be abnormal even when the M N C V in the more distal segments is normal or borderline [ 11-13]. The present findings confirm these observations and further document that slowing of conduction tends to be localized to the most proximal segment and to the terminal portion of motor fibers. Although the FWCV in the cord-to-axilla segment is in general slower than the MNCV in the more distal segments in both median and ulnar nerves, terminal latencies are also greatly increased in these nerves. Thus, the F ratio remains normal i n a majority of the affected nerves, indicating rhat slowing of nerve conduction is romparable between the cord-to-elbow and elbow-to-muscle segments. Similarly, slowing of nerve conduction is the same between the cord-to-knee and knee-tomuscle segments for both peroneal and tibia1 nerves in the lower extremities. A localized conduction abnormality can occur in any of these nerves but, if selective, is distributed at random without particular preference between proximal and distal segmenrs. From the data presented here it may be concluded that conduction abnormalities in the Guillain-Barre syndrome are usually diffuse, involving any segment of the peripheral nerve. There is, however, a tendency toward involvement of the terminal portion of the motor fibers and the most proximal, possibly radicular, portions of the nerve, whereas the main nerve trunk is relatively less affecred. The distal slowing may be secondary to superimposed compression, as previously suggested by Lambert and Mulder [14] and others [ 5 , 8 ] .Prominent slowing of the ulnar nerve in the segment across the elbow supports such a contention. Additionally, the terminal segment may be peculiarly vulnerable because of its distance from the cell body or for y e t undetermined reasons. Obviously, these findings do not necessarily indicate the location of the initial abnormality. It may be that the original lesions center around the radicular portion, which can cause a secondary conduction abnormality in the distal segments. Nonetheless, it is clear from our data that demyelination or some other pathological process responsible for slowing of conduction in this syndrome affecrs rhe entire length of the peripheral nerve either continuously o r discontinuously. This is consistent with the histological findings of Asbury et a1 [ l ] that lesions exist throughout the peripheral nervous system although the segment of maximal involvement varies. In a series of patients with chronic inflammatory polyradiculoneuropathy, Dyck e t a1 141 reported diffusely slow conduction velocity of peripheral nerves although the most marked slowing was often very proximal. Electrophysiological study of the most proximal segment of the peripheral nerve is desirable for overall appraisal of the affected nerves to delineate the under-
lying pathophysiology of various polyneuropathies. In addition to the F wave latency and FWCV, the F ratio appears promising in providing R simple means to compare the motor nerve conduction of the proximal segment to that of the distal segment. For calculation of FWCV, it is assumed rhat surface determinations provide the length of the most proximal nerve segment with fair accuracy. For the F ratio to be meaningful, it is essential that the proximal and distal portions of the extremities be of the same proportion among different subjects. Although both prerequisites are reasonable, they are subject to certain errors. Since these assumptions are of a totally different nature, it is udikely both would be incorrect at the same time. When the FWCV and F ratio are both altered, therefore, the validity of the abnormality under consideration is strengthened considerably. Eisen et al [b] have reported that the ratio between F and M latency is of practical clinical value in differentiating radiculopathp from distal compression syndromes. Since the rightto-left difference of the F ratio described here is very small in individual sublecrs, this comparison might also provide a reliable means of assessing unilateral root lesions. 1 wish to thank Mr David Walker, MSEE, for electrical engineering assistance, Dr Leon F. Burmrister for providing statistical analysis, and Sheila Mennen, Joanne Colter, and Karen Thompson for technical assistance.
References 1. Asbury AK, Arnason BG, Adarns RD: The inflammatory lesion in idiopathic polyneuritis: its role in pathogenesis. Medicine 48:173-215, 1969 2. Clough JFM, Kernell D, Phillips CG: Conduction velocity in proximal and distal porrions offorelimb axons in the baboon. J Physiol (Lond) 198:16?-178, 1968 3 . Conrad 8,Aschoff JC, Fischler M: Der diagncisiische Werr Jer F-Wellen-Latenz. J Neurol 210:151-159, 1975 4. Dyck PJ, Lais AC, Ohta M, et al: Chronic inflammatory polyradiculoneuropathy. Mayo Clin Proc 50:621-637, 1975 5 . Eisen A, Humphreys P: T h e Guillain Bar& syndrome: aclinical and electrodiagnostic study of 25 cases. Arch Neurol30:438-
441, 1974
6. Eisen A, Schorner D, Melrned C: The application of F wave measurements in the differentiation of proximal aod distal upper limb entrapments. Nwrology (Minneap) 2?:662-668, 1977 7 . Gassel MM, Wiesendanger M: Recurrent and reflex discharges in plantar muscles of the cat. Acta Physiol Scand 63.138-142, 1965 8. Humphrey Jd: Motor nerve conducrion studies in the Landry-Guillain-BarrP syndrome (acute ascending polyneuropathy). Electroencephalogr Clin Neurophysiol 17:96, 1964 9. Klmura J: F wave veloc~ryin the centrkl segmenr of the median and uInar nerves: a study in normal subjects and in parienrs wirh Charcot-Marie-Tooth disease. Neurology (Minneap) 24: 539-546, 1974 1 0 ~Kimura J: A method for estimating the refractory period of motor fibers in the human peripheral nerve. J Neurol Sci 28:485-490, I976
Kimura: Nerve Conduction Velocity in Guillain-Barre Syndrome
349
11. Kimura J, Bosch P, Lindsay GM: F wave conduction velocity in the central segment of the peroneal and tibial nerves. Arch Phys Med Rehab 56:492-497, 1975 12. Kimura J, Butzer JF: F wave conduction velocity in Guillain Bar& syndrome: assessment of nerve segment between axilla and spinal cord. Arch Neurol 33:524-529, 1075 13. King D , Ashby P: Conduction velocity in the proximal segments of a motor nerve in the Guillain Bar& syndrome. J Neurol Neurosurg Psychiatry 39:538-544, 1976 14. Lambert EH, Mulder DW: Nerve conduction in the Guillain Bar& syndrome. Electroencephalogr Clin Neurophysiol 17:86, 1964 15. Magladery JW, McDougal DB Jr: Electrophysiological studies of nerve and reflex activity in normal man: identification of certain reflexes in the electromyogram and the conduction velocity of peripheral nerve fibers. Johns Hopkins Bull 86:265-290, 1950 16. Mayer RF, Feldman RG: Observations on the nature of the F wave in man. Neurology (Minneap) 17:147-156, 1967 17. McLeod JG, Wray SH: An experimental study of the F wave in the baboon. J Neurol Neurosurg Psychiatry 29: 196-200.1966 18. Miglietta OE: The F-response after transverse myelotomy, in Desmedt JE (ed): New Developments in Electromyography
3 5 0 Annals of Neurology Vol 3 No 4 April 1978
19.
20. 2 1.
22.
23.
24.
25.
and Clinical Neurophysiology. Basel, Karger, 1973, vol 3, pp 323-327 Muller D: Die Bestimmung der F-Wellengeschwindigkeit am N. ulnaris Gesunder. Psychiatr Neurol Med Psycho1 (Leipz) 27:619-623, 1975 Panayiotopoulos CP, Scarpalezos S: Neural and muscular involvement in dystrophia myotonica. Lancet 2:329, 1975 Panayiotopoulos CP, Scarpalezos S: F-wave studies on the deep peroneal nerve. Part 2, 1. Chronic renal failure; 2. Limb-girdle muscular dystrophy. J Neurol Sci 31:331-341, 1977 Panayiotopoulos CP, Scarpalezos S, Nasras PE: F-wave studies on the deep peroneal nerve. Part 1, Conrrol subjects. J Neurol Sci 31:319-329, 1977 Renshaw B: Influence of discharge of motoneurons upon excitation of neighboring motoneurons. J Neurophysiol 4: 167183, 1941 Thorne J: Central responses to electrical activation of the peripheral nerves supplying the intrinsic hand muscles. J Neurol Neurosurg Psychiatry 28:482-495, 1965 Trontelj JV: A study of the F-response by single fibre electromyography, in Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology. Basel, Karger, 1973, vol 3, pp 318-322
