~~
Normal results of CMCT to upper extremities do not correlate with height, age, or sex. In cases of peripheral conduction slowing, upper limits of CMCT must be adapted. Cervical root excitation can be carried out by electrical or magnetic stimulation. The F wave technique also produces reliable results. Electrical stimulation of lumbar roots produces the most reliable results when measuring CMCT to leg muscles. CMCT to the tibialis anterior correlates with height, but not with age or sex. The same is true for intraspinal motor conduction time. Contraction of the target muscle enlarges CMAP amplitudes after cervical and transcranial stimulation. However, a latency shortening is only seen after brain stimulation. This is thought to be at least partially due to spatial summation of 0 and I waves. Age affects peripheral nerve conduction and central nerve conduction in different ways. Key words: Central motor conduction time magnetic stimulation normal results spinal motor conduction time MUSCLE & NERVE 13:1125-1132 1990
.
CENTRAL MOTOR CONDUCTION: METHOD AND NORMAL RESULTS DETLEF CLAUS. MD
Recently tlevelopcd nietlwtis allow noninvasive stirnulation of tlie motor cortex ant3 nerve rools. 1.17 Me;tsuring central motor conduction time ( C M C T ) ciiri be cxpccted to bccorne a routine investigation i r i c~inic:a~ ri(:iit~o~~~iysio1OSy.T).X.'" I t is, lhertforc, of interest t.o collect normal results. I Iic. influencc of' chaiigcs in technical vnriablcs on measuring results was evaluated by comparing stantiardized mntcrials and different techniques.
r 1
SUBJECTS AND METHODS
Normal values werc recorded in 54 subjects with no liistory of nervc disorders (25 female, 29 male; 1!)-59 yrurs, mean 34 2 1 I years; height 15619 1 cm). Experiments with stirnulation at C7 arid L1 w e r e carried out in 29 siib-jects (18-62, nican 38 years; height 15 1- 193, meail 175 cln) in order t o calculate spinal (:onduction time. They had all
From the Department of Neurology. University frlangen-Nurernberg, Erlaiigen, West Germany
Acknowledgment: This work was supported by the Sander Foundation. The resulls were presented in part at ltie 36th AAEE meeting. Wastiington, DC, 1989 Addrcss reprint requests to Detlef Clnus, MD, Department of Neurology, University Erlanyen-Nuremberg, Schwabacharilage 6.D 8520 Erlanyeii, West Germany. Accepted for publication November 24, 1989. CCC 0148-639X/90101201125-08 $04 00 0 1990 John Wiley & Sons. lnc.
Central Motor Conductiori
given tlicir iiifbrmed c:orisent. 'I'hc. procedures were approv(:d b y the local Medical Ethics CornII ii t t r e . 'IlIree dil'ferent magnetic stiniu1;Iiors were uscrl: tlie Digiliriier. U 190 (slimiilator A: lhgitinicr Lt.d., Herlfordshire, UK) thc Novametrix Magstiin 200 (sti rn ulator K : Novainet rix Medical Systems Inc., Whiiland, Wales), a r i d the C:adwell MES-10 (stiinulator- C;: Cadwrll Laboratories Inc., Kennewick, W A , USA). Technical data are presented in 'l'ablc I . Transcranial stimulation was carried out wiih h e stimulation coil positioned Hat on the scalp arid its center over C r (international 10/20 system) with an abductor digiti minimi target muscle (ADM). 'The coil center was over Fz for recording from the leg (tibialis anicrior, ~ T A ) .With each magnetic stimulalor, clockwise stimulation current in the coil was uscd for target muscles 011 ilie right side and counterclockwise ciirrent for those on the left (physical rurrent direction). T h e coil position had to be slightly adapted in every case to achieve optirrial excitation. 'To establish t.he threshold, st.imulus strength (given in percentage of maximum output) was increased i n 5% steps with the target muscle exerting 10% isometric cont.ractiorl"," (recorded by surface EM(;, 100 ms/D). Aclditional monitoring of muscle contraction by a miniature load cell (Kyowa Electronic I nstrumcnts Co. Ltd., Tokyo, Japan) had already proved in scveral cases that surface E M C was sufficient. 'The thr-eshold was defined as itie lowest intensity
MUSCLE & NERVE
December 1990
1125
Table 1. Technical data of the different rnagnettc stimulators
COll
field (Tesla)
Mean coil diameter (mm) (measured by X-ray)
Standard Standard
1-1.5 2 4 2-2 5
125 90 40 85
Peak magnetic
Stimulator
Number of coil
Induced voltage
windings
maximum (V)
7 19 21 14
0.7 3.1 8.1 3.4
~~
Digitimer D 190 Novametrix Magstim 200
Cadwell MES-I 0
Small
Standard
Peak voltage induced IR a sniall coil (100% output, m e winding circular copper coil. 16 mm diameter) close to the inner edge of the round stimulalior~coil and parallel to Its surface (tarryenrial) according to Claus et a/.'
which gave 1-cprodiicihle responses three times; furtlicr stimiili were applied at 20%) above 111ITS 11oiti. Four tests were carricxl o u t t o tlcfirie the shortest onset latency aiitl the highest amplitude of coriipouncl iiiuscle 1-cspoiises (measured from the I~asclinet o the first iieptivc potential peak). 'I'ht. clcctrical stiinul;ttcii- (Digitiiiicr I1 180) gcrieixtes ;I stiniuliis witli ;I s h o r t rising phase and ;I decay tiiiie of 50 or 100 p s (pcak voltage 750 V). 'llii-cc tliffciwit dcctrotles were uscd f o r electrical stimuIatioIr. For per-ipheral ~ierve stiiiiulation (ulnar i i c ~ i - v cat thc wrist, per-oiical nerve a t the head of the fil)ula) ;I staiidarcl surface stiniulating electrode was used with cathocle c1ist;illy aiid at a 25 111ni i i i tcrelect 1-ode cl i st ;iiicc. Electrical root stirnulation was perf'oriiied with $lll.f', 'I(.L .-.electrodes, the c:athotic a b o v e the spinal processus C:7 and the anode 5 t:m more cranially i i i ridline position. For the electrical stimulation of' luiribar nerve roots an electrode was used which consisted of 7 i n r n cliariieter surl'ace e1et:trodes. ( h e cathode was placed over the I , ] spinal processus ;ind thi-cc anodes wci-c positioned 1%) i i i i i i paixvertebrally and contr-alatci-allyto the target muscle in vertical orientation (interelectrode distance 40 i i i i i i ) . This method was found to produce relial.de lurnbar root excitation."3'('T h e stiriiulus sti-ength was elevated until it became suprarnaxiriial. Magnetic I-oot stirnulation was c:or~luctedwith the coil center ovcr ihe spinal p~-oc:essusC7 o r L1 (1 OOT, output). 'I'he stiiiiulatioii current (physical tiirec:~iori)was clockwise for larget rnuscles on the right and counterclockwise for those 011 the l e k 7 1: waves wei-c elicited A'ter ulnar nerve stitnulation at tlie wrist and af'ter-pcrorieal nerve stimulation at the fibula head. At least 5 E' waves were recorded a r i d the shortest latency was taken as a result. 1: ',9 According t o various autliors ',--the excitatioii of niotoiiciirons requires ;rppiwxiniately 1 111s. 'T711er-efore,per-iplieral c.oiicluction time f'rom thc
1126
Central Motor Conduction
r ~ ~ o t o i i c u r oto~ i11ie site 01' nerve excitation can he estimated by the fol-rnllla: (F - M - 1)/~.'"-'('*'(' This calculation was used to investigate peripheral triotor conduction time from C7 to the wrist arid fr-orri I , I to tlic fibula head. Compound muscle r e s p i i s e s (CMAPs) were recorded fi-orn the AI)M and the T A using siirface cup electrotlcs (7 nim cliaiiietcr-).'l'hey were filled with electi-odc jelly and adapted by the tendon I~ellyriictliod. Responses w e r e amplified (giiiii 0.5-5 rnVA), h i i d p a s s 20 Hz-2 kHz) and stor-cd digitally for I'irrtlicr ~~ieasiire~iiciits (Medelec MS20, l'oshiba m ic-rocoiiip u Lei,, 1) A'l'hM ED p r()gram). After hrain or L1 root stimulation, recordings were usually oiily made fiwn one target rnuscle. Aftel. C7 stimulation, however, recordings were made siniultaneously from both hands. The CMC'I' to ADM a n d '1-A was calculated in each subject using the rriagiietic stirniilators A, U , and C iri rniiclorii order-. Spinal root latencies were nieasur-ed for CMC'I' t o 1i;tnd niuscles using the 1' wave technique: (F + M - I)/'.'., electrical arid iiiagnetic root stirnulation (stimulator K). b'or C M C T to 'I'A, only electrical stiriiu1;itioii was used at L1 since magnetic lu~nbarroot stinirrlation had pi-oved t o be inefficient. 'I'he CMC'I' was represented h y the latency dif'lereiicc to a targct muscle after Lranscrmial stirnulatioil a r i d after stiniulation of spirial nerve roots (Pig. I). Spinal conduction tirrie (SC'17)was calculated 1.1-oiii latency diffcrcrices riieasurecl t o T A after electi-ic:al transcutaneous spinal coid stimulation at C7 (target muscle 10% contraction) aiid af'ter root stimulatioii a t I , 1 . Peripheral 1110tor conduction velocities (PM(;V) in the ulnar iicrve anti peroneal nerve were ciilculatetl from the anterior spinal r o o t LO the distal stiiiiulaLion sit.c (distances assessed with a stretched measuring '
tape).
The relative atiiplitude of' CMAPs af'ter brain stimulation aiid after root stirnulatioii (quoticnt: (;xiroot), c:al(:ulated i n pel-cent, was takcri as a pa-
MUSCLE & NERVE
December 1990
Normal rcsults and tmrtierline values of CMC'I' to ADM and PMCV are prescnted in '[able 2. 'l'he upper limit of norm of threshold stimulus strcngth (5%) for wanscranial stitnulat.iorias defined by the nincty-ninth pcrceritile is: coil A: 90%, L: 70%, (2: 70%. Comparison of latencies mcasured to ADM after clectrical root stimulation (stirnulus streri th 60%7596, decay 50 ps) with F wave Btenc:ie& shows a diffcrencc of 0.41 f 0.41 ms. Assuming that the YMCV in the proximal root. scgnient is not slower than that measured from rriotor roou to wrist., the site of root excitation after transcutaneous stimulation is asscssed to be 26 rrim distal from the alphamotoncuroii. I n cases of delayed pcripheral conduction, a cielay in the proximal cervical root scgnierit can be estirnated, assuming that the slowing down in the root is proportional to that of PMCV it1 thc whole riervc.' Sincc proxirnal root conductioii is included in CMCYI', delay tirnes have to be aclded to the irpper limits (PMCV 50 m/s: 0.1 I nis, 40 m/s: 0.24 nis, 30 m/s: 0.46 ms,20 m/s: 0.8'3 ins). The C M C I ' calculated by subt.racting F wave latcricics l'ro~nthc latencies measured to ADM after transcranial stimulation is slightly shorter than t.hat obtained aftcr elcc:r.rical root stimulation. This difference is due to the fact that the F wave rcpreseiits the latency to thc alphamotoncuron while thc site of' elcct.rical root excitation is scparate liorn the spinal root exit (Table 2). However, with F waves r i o atiiplitude can be calculated to judge CMC (Fig. 2). Magnetic cervical root stimulation was always successful irsirig the large. (:oil H. The latencies CYCT to Uppor Extremity.
FIGURE 1. Example of the investigation to ADM and TA. PMCV is calculated from the latency differences after stimulation at wrist and C7, fib., and Ll CMCT is defined by the difference of latencies to a target muscle after stimulation at C7 (Ll) and scalp.
.
rarrieter for corticospirial stimulus conduction; (CMC) target muscle at least 10% contmction. Because linear regression fits thc results, correlation was irivestigatecl by the linear correlation coellicient (R). When histograms or scattergrams proved near-normal distribution of' results, limits of nortn were calculated by nican f 2.5 SD. 'I'his provides simple 1i)rmulae for the c;ilculation of tmderline values. In other cases percentiles were used fhr defining limits ol' norm. -1'0 investigate thc significance of' diflerences the Wilcoxon's matclicd-paired signed-ranks test h r srnall groups was carricd out.
RESULTS
Results nicasurccl to right and left wcrc treat.erl as one, sincc no relevant dil'li.rence was found between them. This iriay lle illustr;ttcd by separately calculated limits 0 1 norm (riglit/lcft): latency wristADM 4.0/3.9 ins, el. stini. (3-ADM 1'7.2/17.1 ms, magstiin. CV-ADM 1 7.71 17.5 ins, PMGV arm 5 % 53 m/s, CMCT (el. stim. C7) 8.U7.9 ms, CMCT (F wave) 7.8/7.I) 111s; latency libula hcad-'rA 5.0/5.1 ms, latcncy Ll-'rA IS.Y/IY.5 I ~ S , l'M(:V leg 49/50 tll/S, CMC'T ICg (Coil u) 16.6/16.8Ills, (
Normal results of CMCT to upper extremities do not correlate with height, age, or sex. In cases of peripheral conduction slowing, upper limits of CMCT must be adapted. Cervical root excitation can be carried out by electrical or magnetic stimulation. The F wave technique also produces reliable results. Electrical stimulation of lumbar roots produces the most reliable results when measuring CMCT to leg muscles. CMCT to the tibialis anterior correlates with height, but not with age or sex. The same is true for intraspinal motor conduction time. Contraction of the target muscle enlarges CMAP amplitudes after cervical and transcranial stimulation. However, a latency shortening is only seen after brain stimulation. This is thought to be at least partially due to spatial summation of 0 and I waves. Age affects peripheral nerve conduction and central nerve conduction in different ways. Key words: Central motor conduction time magnetic stimulation normal results spinal motor conduction time MUSCLE & NERVE 13:1125-1132 1990
.
CENTRAL MOTOR CONDUCTION: METHOD AND NORMAL RESULTS DETLEF CLAUS. MD
Recently tlevelopcd nietlwtis allow noninvasive stirnulation of tlie motor cortex ant3 nerve rools. 1.17 Me;tsuring central motor conduction time ( C M C T ) ciiri be cxpccted to bccorne a routine investigation i r i c~inic:a~ ri(:iit~o~~~iysio1OSy.T).X.'" I t is, lhertforc, of interest t.o collect normal results. I Iic. influencc of' chaiigcs in technical vnriablcs on measuring results was evaluated by comparing stantiardized mntcrials and different techniques.
r 1
SUBJECTS AND METHODS
Normal values werc recorded in 54 subjects with no liistory of nervc disorders (25 female, 29 male; 1!)-59 yrurs, mean 34 2 1 I years; height 15619 1 cm). Experiments with stirnulation at C7 arid L1 w e r e carried out in 29 siib-jects (18-62, nican 38 years; height 15 1- 193, meail 175 cln) in order t o calculate spinal (:onduction time. They had all
From the Department of Neurology. University frlangen-Nurernberg, Erlaiigen, West Germany
Acknowledgment: This work was supported by the Sander Foundation. The resulls were presented in part at ltie 36th AAEE meeting. Wastiington, DC, 1989 Addrcss reprint requests to Detlef Clnus, MD, Department of Neurology, University Erlanyen-Nuremberg, Schwabacharilage 6.D 8520 Erlanyeii, West Germany. Accepted for publication November 24, 1989. CCC 0148-639X/90101201125-08 $04 00 0 1990 John Wiley & Sons. lnc.
Central Motor Conductiori
given tlicir iiifbrmed c:orisent. 'I'hc. procedures were approv(:d b y the local Medical Ethics CornII ii t t r e . 'IlIree dil'ferent magnetic stiniu1;Iiors were uscrl: tlie Digiliriier. U 190 (slimiilator A: lhgitinicr Lt.d., Herlfordshire, UK) thc Novametrix Magstiin 200 (sti rn ulator K : Novainet rix Medical Systems Inc., Whiiland, Wales), a r i d the C:adwell MES-10 (stiinulator- C;: Cadwrll Laboratories Inc., Kennewick, W A , USA). Technical data are presented in 'l'ablc I . Transcranial stimulation was carried out wiih h e stimulation coil positioned Hat on the scalp arid its center over C r (international 10/20 system) with an abductor digiti minimi target muscle (ADM). 'The coil center was over Fz for recording from the leg (tibialis anicrior, ~ T A ) .With each magnetic stimulalor, clockwise stimulation current in the coil was uscd for target muscles 011 ilie right side and counterclockwise ciirrent for those on the left (physical rurrent direction). T h e coil position had to be slightly adapted in every case to achieve optirrial excitation. 'To establish t.he threshold, st.imulus strength (given in percentage of maximum output) was increased i n 5% steps with the target muscle exerting 10% isometric cont.ractiorl"," (recorded by surface EM(;, 100 ms/D). Aclditional monitoring of muscle contraction by a miniature load cell (Kyowa Electronic I nstrumcnts Co. Ltd., Tokyo, Japan) had already proved in scveral cases that surface E M C was sufficient. 'The thr-eshold was defined as itie lowest intensity
MUSCLE & NERVE
December 1990
1125
Table 1. Technical data of the different rnagnettc stimulators
COll
field (Tesla)
Mean coil diameter (mm) (measured by X-ray)
Standard Standard
1-1.5 2 4 2-2 5
125 90 40 85
Peak magnetic
Stimulator
Number of coil
Induced voltage
windings
maximum (V)
7 19 21 14
0.7 3.1 8.1 3.4
~~
Digitimer D 190 Novametrix Magstim 200
Cadwell MES-I 0
Small
Standard
Peak voltage induced IR a sniall coil (100% output, m e winding circular copper coil. 16 mm diameter) close to the inner edge of the round stimulalior~coil and parallel to Its surface (tarryenrial) according to Claus et a/.'
which gave 1-cprodiicihle responses three times; furtlicr stimiili were applied at 20%) above 111ITS 11oiti. Four tests were carricxl o u t t o tlcfirie the shortest onset latency aiitl the highest amplitude of coriipouncl iiiuscle 1-cspoiises (measured from the I~asclinet o the first iieptivc potential peak). 'I'ht. clcctrical stiinul;ttcii- (Digitiiiicr I1 180) gcrieixtes ;I stiniuliis witli ;I s h o r t rising phase and ;I decay tiiiie of 50 or 100 p s (pcak voltage 750 V). 'llii-cc tliffciwit dcctrotles were uscd f o r electrical stimuIatioIr. For per-ipheral ~ierve stiiiiulation (ulnar i i c ~ i - v cat thc wrist, per-oiical nerve a t the head of the fil)ula) ;I staiidarcl surface stiniulating electrode was used with cathocle c1ist;illy aiid at a 25 111ni i i i tcrelect 1-ode cl i st ;iiicc. Electrical root stirnulation was perf'oriiied with $lll.f', 'I(.L .-.electrodes, the c:athotic a b o v e the spinal processus C:7 and the anode 5 t:m more cranially i i i ridline position. For the electrical stimulation of' luiribar nerve roots an electrode was used which consisted of 7 i n r n cliariieter surl'ace e1et:trodes. ( h e cathode was placed over the I , ] spinal processus ;ind thi-cc anodes wci-c positioned 1%) i i i i i i paixvertebrally and contr-alatci-allyto the target muscle in vertical orientation (interelectrode distance 40 i i i i i i ) . This method was found to produce relial.de lurnbar root excitation."3'('T h e stiriiulus sti-ength was elevated until it became suprarnaxiriial. Magnetic I-oot stirnulation was c:or~luctedwith the coil center ovcr ihe spinal p~-oc:essusC7 o r L1 (1 OOT, output). 'I'he stiiiiulatioii current (physical tiirec:~iori)was clockwise for larget rnuscles on the right and counterclockwise for those 011 the l e k 7 1: waves wei-c elicited A'ter ulnar nerve stitnulation at tlie wrist and af'ter-pcrorieal nerve stimulation at the fibula head. At least 5 E' waves were recorded a r i d the shortest latency was taken as a result. 1: ',9 According t o various autliors ',--the excitatioii of niotoiiciirons requires ;rppiwxiniately 1 111s. 'T711er-efore,per-iplieral c.oiicluction time f'rom thc
1126
Central Motor Conduction
r ~ ~ o t o i i c u r oto~ i11ie site 01' nerve excitation can he estimated by the fol-rnllla: (F - M - 1)/~.'"-'('*'(' This calculation was used to investigate peripheral triotor conduction time from C7 to the wrist arid fr-orri I , I to tlic fibula head. Compound muscle r e s p i i s e s (CMAPs) were recorded fi-orn the AI)M and the T A using siirface cup electrotlcs (7 nim cliaiiietcr-).'l'hey were filled with electi-odc jelly and adapted by the tendon I~ellyriictliod. Responses w e r e amplified (giiiii 0.5-5 rnVA), h i i d p a s s 20 Hz-2 kHz) and stor-cd digitally for I'irrtlicr ~~ieasiire~iiciits (Medelec MS20, l'oshiba m ic-rocoiiip u Lei,, 1) A'l'hM ED p r()gram). After hrain or L1 root stimulation, recordings were usually oiily made fiwn one target rnuscle. Aftel. C7 stimulation, however, recordings were made siniultaneously from both hands. The CMC'I' to ADM a n d '1-A was calculated in each subject using the rriagiietic stirniilators A, U , and C iri rniiclorii order-. Spinal root latencies were nieasur-ed for CMC'I' t o 1i;tnd niuscles using the 1' wave technique: (F + M - I)/'.'., electrical arid iiiagnetic root stirnulation (stimulator K). b'or C M C T to 'I'A, only electrical stiriiu1;itioii was used at L1 since magnetic lu~nbarroot stinirrlation had pi-oved t o be inefficient. 'I'he CMC'I' was represented h y the latency dif'lereiicc to a targct muscle after Lranscrmial stirnulatioil a r i d after stiniulation of spirial nerve roots (Pig. I). Spinal conduction tirrie (SC'17)was calculated 1.1-oiii latency diffcrcrices riieasurecl t o T A after electi-ic:al transcutaneous spinal coid stimulation at C7 (target muscle 10% contraction) aiid af'ter root stimulatioii a t I , 1 . Peripheral 1110tor conduction velocities (PM(;V) in the ulnar iicrve anti peroneal nerve were ciilculatetl from the anterior spinal r o o t LO the distal stiiiiulaLion sit.c (distances assessed with a stretched measuring '
tape).
The relative atiiplitude of' CMAPs af'ter brain stimulation aiid after root stirnulatioii (quoticnt: (;xiroot), c:al(:ulated i n pel-cent, was takcri as a pa-
MUSCLE & NERVE
December 1990
Normal rcsults and tmrtierline values of CMC'I' to ADM and PMCV are prescnted in '[able 2. 'l'he upper limit of norm of threshold stimulus strcngth (5%) for wanscranial stitnulat.iorias defined by the nincty-ninth pcrceritile is: coil A: 90%, L: 70%, (2: 70%. Comparison of latencies mcasured to ADM after clectrical root stimulation (stirnulus streri th 60%7596, decay 50 ps) with F wave Btenc:ie& shows a diffcrencc of 0.41 f 0.41 ms. Assuming that the YMCV in the proximal root. scgnient is not slower than that measured from rriotor roou to wrist., the site of root excitation after transcutaneous stimulation is asscssed to be 26 rrim distal from the alphamotoncuroii. I n cases of delayed pcripheral conduction, a cielay in the proximal cervical root scgnierit can be estirnated, assuming that the slowing down in the root is proportional to that of PMCV it1 thc whole riervc.' Sincc proxirnal root conductioii is included in CMCYI', delay tirnes have to be aclded to the irpper limits (PMCV 50 m/s: 0.1 I nis, 40 m/s: 0.24 nis, 30 m/s: 0.46 ms,20 m/s: 0.8'3 ins). The C M C I ' calculated by subt.racting F wave latcricics l'ro~nthc latencies measured to ADM after transcranial stimulation is slightly shorter than t.hat obtained aftcr elcc:r.rical root stimulation. This difference is due to the fact that the F wave rcpreseiits the latency to thc alphamotoncuron while thc site of' elcct.rical root excitation is scparate liorn the spinal root exit (Table 2). However, with F waves r i o atiiplitude can be calculated to judge CMC (Fig. 2). Magnetic cervical root stimulation was always successful irsirig the large. (:oil H. The latencies CYCT to Uppor Extremity.
FIGURE 1. Example of the investigation to ADM and TA. PMCV is calculated from the latency differences after stimulation at wrist and C7, fib., and Ll CMCT is defined by the difference of latencies to a target muscle after stimulation at C7 (Ll) and scalp.
.
rarrieter for corticospirial stimulus conduction; (CMC) target muscle at least 10% contmction. Because linear regression fits thc results, correlation was irivestigatecl by the linear correlation coellicient (R). When histograms or scattergrams proved near-normal distribution of' results, limits of nortn were calculated by nican f 2.5 SD. 'I'his provides simple 1i)rmulae for the c;ilculation of tmderline values. In other cases percentiles were used fhr defining limits ol' norm. -1'0 investigate thc significance of' diflerences the Wilcoxon's matclicd-paired signed-ranks test h r srnall groups was carricd out.
RESULTS
Results nicasurccl to right and left wcrc treat.erl as one, sincc no relevant dil'li.rence was found between them. This iriay lle illustr;ttcd by separately calculated limits 0 1 norm (riglit/lcft): latency wristADM 4.0/3.9 ins, el. stini. (3-ADM 1'7.2/17.1 ms, magstiin. CV-ADM 1 7.71 17.5 ins, PMGV arm 5 % 53 m/s, CMCT (el. stim. C7) 8.U7.9 ms, CMCT (F wave) 7.8/7.I) 111s; latency libula hcad-'rA 5.0/5.1 ms, latcncy Ll-'rA IS.Y/IY.5 I ~ S , l'M(:V leg 49/50 tll/S, CMC'T ICg (Coil u) 16.6/16.8Ills, (
