Brain (1992), 115, 1045-1059
EFFECTS OF FOCAL TRANSCRANIAL MAGNETIC
STIMULATION ON SIMPLE REACTION TIME TO
ACOUSTIC, VISUAL AND SOMATOSENSORY
STIMULI
by ALVARO PASCUAL-LEONE, JOSEP VALLS-SOLE\
ERIC M. WASSERMANN, JOAQUIM BRASIL-NETO,
LEONARDO G. COHEN and MARK HALLETT
SUMMARY
In a simple reaction time (RT) paradigm, magnetic stimulation of different intensities was delivered over
different scalp positions and at variable delays before (negative) or after (positive) the go-signal. Magnetic
stimulation shortened RT to different go-signals (auditory, visual and somatosensory stimuli) by approximately 30 ms when delivered over the motor cortex contralateral to the responding arm at intensities below
motor threshold. This effect was maximal at a delay of approximately + 10 ms. A similar effect was found
with suprathreshold stimulation to the ipsilateral motor cortex. Magnetic stimulation over other scalp areas
did not affect RT regardless of the delay. No differences were found between the effects on elbow flexion
and thumb abduction. The shortening of RT was not associated with changes in the timing development
of premovement excitability increase in the motor cortex. We conclude that magnetic stimulation shortens
RT by inducing an earlier initiation of this excitability increase.
INTRODUCTION
When simple reaction time (RT) to focal transcranial stimulation was compared with
RT to acoustic, visual and somatosensory stimuli (Pascual-Leone et al., 1992), RT was
longest to a magnetic or electrical stimulus delivered over the contralateral motor cortex
at an intensity high enough to induce motor evoked potentials (MEPs) in muscles involved
in the response (suprathreshold intensity). Conversely, RT was shortest to subthreshold
transcranial stimulation over the same scalp position. This report describes the effects
of transcranial magnetic stimulation (TMS)', to which the subjects are not supposed to
respond, on RT to acoustic, visual or somatosensory stimuli (go-signal).
In RT paradigms, transcranial stimulation.properly delivered in time and space can
modify the onset latencies of the responses. Reaction time may be prolonged or shortened
depending on the intensity of the stimulus. Suprathreshold transcranial stimuli can delay
RT to an auditory go-signal if delivered over the motor cortex close to the expected
response time (Day et al., 1989). Transcranial weak direct currents applied over the
sensorimotor cortex can shorten RT to go-signals of different modalities (Elbert et al.,
1981; Jaeger et al., 1987).
Correspondence to: Dr Mark Hallett, Building 10, Room 5N226, NINDS, NIH, Bethesda, MD 20892, USA.
© Oxford University Press 1992
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
(From the Human Cortical Physiology Unit, Human Motor Control Section, Medical Neurology
Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health,
Bethesda, Maryland, USA)
1046
A. PASCUAL-LEONE AND OTHERS
Reaction time can be divided into a first period during which the motor cortex is less
excitable, and a second period during which it becomes increasingly excitable, leading
to movement onset (Starr et al., 1988). Hallett et al. (1991) have shown that the study
of the relative duration of these two periods can provide some understanding of RT
abnormalities in patients with Parkinson's disease. Similarly, the effects of transcranial
stimulation on RT to different go-signals may be clarified by comparing the duration
of these two RT periods in trials with and without transcranial stimulation.
METHODS
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
Reaction time experiments
We studied five naive, right-handed normal volunteers (three men and two women), aged 26—42 yrs.
An auditory warning signal, used to alert the subject, was followed at random intervals (foreperiod, 1 —5 s)
by an auditory, visual or somatosensory go-signal. In response to the go-signal, the subject flexed the
right elbow or abducted the right thumb as rapidly as possible. When the response was elbow flexion,
the subjects were seated comfortably on a chair with the right arm slightly abducted at the shoulder and
flexed 90° at the elbow so that the pronated forearm rested on a horizontal platform. When the response
was thumb abduction, the subjects were seated with the right hand supinated and resting on a horizontal
platform, the thumb adducted and the elbow flexed at 90°.
Each subject completed sets of trials using three different go-signals: a click, a flash and an electrical
stimulus to the left index finger. The click was generated by a Grass auditory stimulator and delivered
by a loudspeaker suspended 15—20 cm over the subject's head. The flash was generated by a Grass PS22
photic stimulator and delivered at an intensity of 100% of the stimulator's output by a lamp positioned
at eye level 30 cm in front of the subject. The electrical stimulus was generated by a Grass electric stimulator
and delivered by two surface electrodes taped 3 cm apart to the subject's left index finger. The intensity
of the electric stimulus was kept at three times sensory threshold as determined by the method of limits
(Gescheider, 1976).
Reaction time was measured from the go-signal to the onset of biceps or abductor pollicis brevis (APB)
electromyographic (EMG) activity. The EMG was recorded with two surface electrodes taped over the
muscle belly. The EMG signal was amplified and filtered (100-2000 Hz) by Grass amplifiers, digitized
with a sampling rate of 5000 Hz per channel and rectified. The device delivering the go-signal was triggered
with a 100 ms delay after EMG recording began; the total sweep time was - 1 0 0 to 400 ms. All data
were collected using an AST personal computer.
Transcranial magnetic stimulation was delivered with a Cadwell MES 10 magnetic stimulator equipped
with an 8-shaped coil in which each component measured 4.5 cm in diameter. The coil was held flat on
the scalp over the position at which TMS induced MEPs of maximal amplitude in the contralateral biceps
or APB. (These positions were determined, with the patient at rest, by delivering TMS at an intensity
of 100% of the stimulator's output over different scalp areas several times during the experiment.) The
handle of the coil was held parallel to the sagittal axis of the subject's head, pointing occipitally. This
technique allows relatively focal cortical stimulation (Cohen et al., 1990); the characteristics of the electric
field induced in the cortex are discussed elsewhere (Roth et al., 1991). Subthreshold intensity is the highest
intensity that did not evoke a MEP in the target muscle at rest (at a recording sensitivity of 50 ^V/division)
in five trials.
The experiments were performed in sets of 12 trials presented in random order. In each set, one-third
of the trials were control trials (go-signal only), one-third were test trials (go-signal plus TMS) and onethird were catch trials (TMS only). In the test trials, TMS was delivered before (negative) or after (positive)
the go-signal; delay ranged from - 5 0 ms to +50 ms and was randomly varied in the different trials. The
catch trials served to ensure that the subjects were responding to the go-signals and not TMS.
The study of each subject was completed in three different recording sessions of 11 sets each (132 trials)
using a single go-signal in each session. This procedure was intended to avoid fatigue. The order of the
recording sessions for the different go-signals was varied in different subjects. The first set of 12 trials
in each recording session was considered a practice set and discarded from analysis. During the practice
trials, the subjects were encouraged to perform as rapidly as possible, thus minimizing RT variability.
EFFECT OF TMS ON SIMPLE REACTION TIME
1047
For each subject, we measured RT in the control trials and calculated a mean and standard deviation of
RT for each go-signal. In the test trials, we measured RT and calculated a mean and standard deviation
of RT for each delay tested. Results across subjects were compared with one-way analysis of variance
(ANOVA) repeated for the different go-signals. Comparison of RTs to the different go-signals in control
and test trials was performed using one-way ANOVAs collapsing across subjects. Significance level, tested
with Scheffc's test, was set at P < 0.05.
We also studied the effects of variable intensities of TMS delivered concurrently with the go-signal (delay
= 0). The initial stimulus, delivered at subthreshold intensity, was randomly increased or decreased stepwise
by 5% of the stimulator's output. Finally, to evaluate the topographic specificity of the effects of TMS
on RT, we delivered TMS to F3/4, ipsilateral motor cortex, or P3/4 concurrently with a visual go-signal
(delay = 0).
RESULTS
Reaction time experiments
All subjects occasionally responded to isolated catch trials (Fig. 1). When this occurred,
the entire set of 12 trials was discarded from further analysis because of the possibility
that the subject was responding to the magnetic stimulus rather than to the go-signal
in the test trials. Errors in catch trials occurred regardless of the go-signal modality,
but were significantly more frequent when the go-signal was auditory (Fig. 1).
Reaction time with TMS (test trials) was shorter (P < 0.001) than RT without TMS
(control trials) regardless of the go-signal modality. The difference between RT in control
and test trials, considered to be the amount of RT shortening due to TMS, was
approximately 30 ms regardless of the go-signal modality. The results for elbow flexion
and for thumb abduction were similar (Table 1). The shortening of RT by TMS started
at a delay of —30 ms, was maximal at +5 ms to +10 ms, and lasted up a delay of
+30 ms. When the delay was +50 ms or longer, RT was prolonged rather than shortened.
The relationship between RT shortening and delay was the same regardless of the
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
Motor cortex excitability experiments
The experimental design was the same as for the RT experiments involving thumb abduction, and the
same subjects were studied. The study of APB allowed us to compare our results with those of Starr et al.
(1988). In half of the trials, we used the same visual go-signal as in the RT experiments (control trials).
In the other half, the go-signal was the same visual stimulus coupled with a subthreshold TMS delivered
to the ideal position for evoking MEPs in the contralateral APB (test trials). In both control and test trials,
a subthreshold TMS (probing stimulus, S) was delivered at variable times during the RT to assess the
probability of evoking MEPs in the APB as a function of the proximity of voluntary EMG onset. The
MEP amplitude was expressed as a percentage of the maximal M-response following peripheral electrical
nerve stimulation. In the test trials, S was identical in intensity and localization to the TMS coupled with
the visual stimulus as part of the go-signal.
Transcranial magnetic stimulation was delivered with a Cad well magnetic stimulator capable of delivering
single or twin pulses at intervals as short as 30 ms without changes in the amplitude of the pulse. Technical
information about this stimulator is presented elsewhere (Pascual-Leone et al., 1991, Appendix 1). The
stimulation coil and its position on the scalp were the same as in the RT experiments.
In each subject, we recorded 120 trials (60 control and 60 test trials) in a single recording session. We
compared RT in control trials with RT in test trials using two-way ANOVA (subject and trial type). To
analyse the probability of S evoking an MEP, we aligned the trials at EMG onset (response) and expressed
the probability as a function of the interval between S and EMG onset. We compared the probability curves
in the control trials with those in the test trials to assess the effect of TMS in the go-signal (test trials)
on the build-up of motor cortex excitability during RT.
1048
A. P A S C U A L L E O N E AND OTHERS
GO-SIGNAL
AUDITORY
VISUAL
SOMATOSENSORY
en en
cc
2C-
oCC
i
<
<
cc
cc
LU 1—
o
cc
SUBJECT
D 5
X
o
LU
CD LJ
• 4
IS 3
10-
m 2
• i
51
SET i
10
9
8
7
6
5
4
3
2
1
SUBJECT
HIBBH RRDH
ll
B
3
4
5
1
2
3
4
5
1
2
3
4
5
FIG. 1. Sets of arm flexion RT experiments for each subject and go-signal modality (bottom) and stacked bar-histogram
of the number of errors in catch trials according to the go-signal modality (top). The experiments were organized in
11 sets of 12 trials containing four control, four test and four catch trials. If the subject committed an error in a catch
trial, the entire set was discarded (closed squares). The first set in each experiment was considered practice and was
also discarded (stippled squares). Results are based on the remaining sets (open squares). Significantly more errors
were committed by the subjects when an auditory go-signal was used (*P < 0.05). The results were essentially the
same when the response was thumb abduction.
TABLE I. SHORTENING OF RT BY SUBTHRESHOLD TMS TO THE
CONTRALATERAL MOTOR CORTEX
Reaction time (ms)
Arm flexion
Go-signals
Auditory
Visual
Somatosensory
Thumb abduction
Control trials
Test trials
Difference
Control trials
Test trials
Difference
150.3±10.2
137.4±9.4
91.6±7.1*
121.7±9.2»
106.3±9.6»
32.5*13.2
28.6±12.3
31.1*10.2
129.2± 13.5
154.4±11.8
139.9*14.8
94.6^9.7*
123.6*17.9*
106.3*13.6*
34.6*14.1
30.8*18.1
33.6*15.2
Values are mean ±SD of the total number of trials in all five subjects. *P < 0.001 for comparison with control trials.
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
0J
EFFECT OF TMS ON SIMPLE REACTION TIME
1049
o __
— o
-10-
-20-
-40
-50 - 4 0 - 3 0 - 2 0 - 1 0 0
10 2 0
Delay (ms)
30 40 50
-20 -15 -10 - 5
i.
»5 * I O » I 5 * 2 0 »25
'SuUtnrestiold
Intensity'
TMS Intensity i% stimulator output)
FIG. 2. Effect of TMS on arm flexion RT according to the interval between go-signal and TMS (delay, left) and
according to TMS intensity (delay = 0, right). Symbols represent different go-signals: open circles = somatosensory;
open triangles = visual; open squares =• auditory. Maximal shortening of RT occurs at delays of +5 ms to +10 ms
and subthreshold intensity. Subthreshold intensity refers to the intensity at which no biceps MEP was recorded (sensitivity
50 /jV/division) in any of five trials with the subject at rest when TMS was delivered over the ideal scalp position for
biceps MEP. Note that TMS at a delay of +50 ms and at suprathreshold intensities prolonged RT regardless of gosignal modality.
go-signal modality (Figs 2, 3). With concurrent delivery of the go-signal and TMS,
the effect of TMS to shorten the RT decreased at intensities below subthreshold. At
intensities above subthreshold, TMS prolonged the RT (Figs 2, 3).
There were no significant differences for any of these effects regardless of whether
the tested movement was elbow flexion or thumb abduction. There was, however, a
tendency for the effects of TMS to be more prominent in the APB (Table 1 and Fig. 4).
Reaction time to the auditory go-signal was shortest, followed by the somatosensory
go-signal and then the visual go-signal (Table 1). These results were predicted by our
previous study (Pascual-Leone et al., 1992) and will not be further discussed here.
We found no significant differences between RT in trials without TMS and trials with
TMS over F3, F4, P3 or P4. Transcranial magnetic stimulation over the ipsilateral motor
area did not affect RT when delivered at subthreshold intensities; however, it shortened
RT when delivered at suprathreshold intensities (Fig. 5).
Motor cortex excitability experiments
Reaction time in control trials was longer than in test trials (154.4 ± 19.7 ms versus
124.1 db 13.2 ms) (P < 0.001). The probing stimulus S never elicited a MEP when
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
-30-
1050
A
SO-SIGNAL
PASCUAL-LEONE AND OTHERS
AUDITORY
VISUAL
SOMATOSENSORY
TEST TRIALS
i Ifk
TMS
5UBTHRE5HOLD
_J200uV L i l i i
J iJL_J
TMS 25X ABOVE
SUBTHRE5HOLD
Go-signal
FIG 3. Representative EMG recorded during RT for the different go-signals without TMS and with TMS at subthreshold
and suprathreshold intensities (delay = 5 ms).
is—i
o __
~ o
in
-20
-30
-50 - 4 0 - 3 0 - 2 0 - 1 0 0
10 2 0
30
40 50
-20 - 1 5 - 1 0 - 5
A
-5 -10 -15 -20 *25
'SuDthresnoid
intensity'
Delay (ms)
TMS intensity (.% stimulator output)
FIG. 4. Comparison of the effects of TMS on RT to a visual go-signal for arm flexion (open triangles) and thumb
abduction (closed triangles) according to the interval between go-signal and TMS (delay, left) and according to TMS
intensity (delay = 0, right). The effects on RTs to auditory and somatosensory go-signals were equally similar.
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
CONTROL TRIALS
Jffllu
EFFECT OF TMS ON SIMPLE REACTION TIME
1051
APB
180-
160-
I2O J
Biceps
200 n
I 40-
I2O J •—'
a> ^
—
o
u
—. O
C *-'
o
o
FIG. 5. Effect of TMS to different scalp areas on RT (mean±SD) to visual go-signal. Stippled bars represent RT
in trials without TMS (control), closed bars represent trials with subthreshold TMS, hatched bars represent trials with
TMS at 25% suprathreshold intensity. Note the similarity of the results for elbow flexion (biceps) and thumb abduction
(APB). *P < 0.01.
the interval between S and EMG onset was > 80 ms in either control or test trials (Fig. 6).
With intervals of <80 ms, the probability of inducing a MEP gradually increased and
reached 1.0 at intervals ^ 3 0 ms in both control and test trials (Fig. 7).
The amplitude of the evoked MEPs increased rapidly with shorter S-EMG onset
intervals in control and test trials (Figs 6, 8).
DISCUSSION
Our results expand the findings of Day et al. (1989) on the effects of TMS on RT.
We have confirmed that prolongation of RT will result from stimuli of suprathreshold
intensity and stimuli delivered close to voluntary EMG onset. In addition, we have found
shortening of RT by stimuli of subthreshold intensity to the contralateral motor cortex
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
160
1052
A. PASCUAL-LEONE AND OTHERS
200
30 ms
Onset of
voluntary
movement
FIG. 6. Representative examples of test trials in motor cortex excitability experiments. Note the relationship between
the amplitude of the MEP and the interval between probing stimulus S and EMG onset.
delivered early during the RT, and by suprathreshold stimuli to the ipsilateral motor
cortex.
Theoretical model of response preparation and RT
The effects of TMS on RT can be explained using a model of response preparation.
We divide the processes required for response preparation and execution into a stimulus
evaluation system, a task-specific circuitry and a response channel. The stimulus
evaluation system has to detect, process and interpret the go-signal. The task-specific
circuitry prepares the motor program for the required response. The response channel
(Gratton et al., 1988) includes all the necessary structures to execute the response as
rapidly as possible. In a simple RT paradigm, the subject is given all the necessary
information to plan the appropriate response before presentation of the go-signal. The
task-specific circuitry can be completed well in advance (intentional aspects of motor
set) and thereafter the prepared motor program has to be held in memory until transferred
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
J
E F F E C T OF TMS ON S I M P L E R E A C T I O N TIME
1053
9
Reaction Time
180 170 160 150 140 130 120 1 10 100
90
80
70
60
50
40
ms before
30
20
10
ms after
Onset
FIG. 7. Probability of evoking an MEP in the APB as a function of the interval between probing stimulus S and
EMG onset. Symbols for go-signals: closed circles = visual; hatched squares = visual and TMS. The bar graph shows
RT (mean ± SD) plotted from EMG onset for control trials with visual go-signal (closed) and test trials with visual
plus TMS go-signal (hatched). Note the significant difference in RT between the two conditions (P < 0.001), and
the absence of differences in the probability curves (excitability increase of the motor cortex).
80
70
60
50
40
30
20
Time before EMG onset (ms)
FIG. 8. Amplitudes of MEPs (mean ±SD) evoked in the APB by the probing stimulus S as a function of the interval
between S and EMG onset. When the S-EMG onset interval was <20 ms the MEPs were evoked after onset of the
voluntary movement and coincided with ongoing EMG activity. Therefore, the exact amplitude of the MEPs could
not be determined. Columns represent trials with visual go-signal (closed) and trials with visual plus TMS go-signal
(hatched).
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
-a—a
1054
A. PASCUAL-LEONE AND OTHERS
Shortening RT by TMS: time for recognition
We define time for recognition as the time necessary for identification of a stimulus
as the go-signal. This process may not require conscious recognition or perception of
the go-signal (Taylor and McCloskey, 1990). The duration of the time for recognition
will vary for different go-signal modalities. Therefore, this process is the most likely
to be responsible for the different RTs to visual, somatosensory and auditory stimuli
(Luce, 1986; Pascual-Leone etal., 1992).
The shortening of RT by TMS is unlikely to be due to influences on the time for
recognition. The introduction of TMS added complexity to the task of the stimulus
evaluation system since the subjects had to differentiate the proper go-signal from the
irrelevant TMS stimulus; failure to do so was penalized in the catch trials. Therefore,
the prediction would be that the time for recognition and thus RT should have been
prolonged rather than shortened (Luce, 1986).
Transcranial magnetic stimulation shortened RT by the same amount of time
(approximately 30 ms) regardless of the go-signal modality, while, as mentioned above,
the duration of the time for recognition depends on the go-signal modality. However,
TMS could have an alerting or attention-focusing effect on the stimulus evaluation system.
The role for such an effect would be limited with a warning signal, as used in our study.
Nevertheless, such effects have been proposed to explain the phenomenon of 'intersensory
facilitation' (Bernstein el al., 1969; Nickerson, 1973). Intersensory facilitation refers
to shortening of RT when a second signal to which the subject is not supposed to react
is given in close proximity to the go-signal, and may shorten RT by 20 — 50 ms
(Nickerson, 1973). However, TMS shortened RT only when it was delivered over the
motor cortex, and not over frontal or parietal areas. Perception of TMS is similar
regardless of the exact site of stimulation. Therefore, this topographic difference would
argue against intersensory facilitation. Furthermore, effects of intersensory facilitation
are more prominent if the second signal is of high intensity and of a different modality
than the go-signal. Transcranial magnetic stimulation is associated with an auditory click
and somatosensory scalp stimulation, both of which are proportional to the stimulus
intensity. Therefore, if due to intersensory facilitation, the effects of TMS on RT should
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
to the response channel for response execution. In a warned RT paradigm, the stimulus
evaluation system can prepare to detect the go-signal (attentional aspects of set) in response
to the warning signal. Therefore, in a warned, simple RT paradigm, such as the one
used in this study, the stimulus evaluation system, the task-specific circuitry and the
response channel are unlikely to be activated serially (Coles etal., 1985). Rather,
activation of the task-specific circuitry may begin even before the warning signal, and
the task-specific circuitry and the stimulus evaluation system may be active in parallel
during the foreperiod (Fig. 9). Identification of a stimulus as the go-signal by the stimulus
evaluation system is completed in the 'time for recognition' and sets in motion a process
by which the motor program prepared and held in memory by the task-specific circuitry,
is transferred to the response channel ('time for initiation'). Thereafter, the response
channel needs a period of time ('time of development') to execute the response (Fig. 9).
Reaction time can therefore be divided into three periods: time for recognition, time
for initiation and time of development. We think of these periods as being completed
serially (Fig. 9); shortening of any of them could lead to a shorter RT.
EFFECT OF TMS ON SIMPLE REACTION TIME
1055
SHORT RT
TO TMS
SES
WITHOUT
TMS
!
TSC
iRC
^V\
TR
'
TI
i
TD
REACTION TIME
WARNING
SIGNAL
COSIGNAL
FIG. 9. Theoretical model of the processes completed during response preparation and effects of TMS on RT. SES
" stimulus evaluation system, TSC = task specific circuitry, RC = response channel, R « response, TR = time
for recognition, TI = time for initiation, TD = time of development, a and b represent two possible mechanisms (see
text for details). The stimulus evaluation system has to identify the stimulus as the go-signal and is divided into attentional
aspects of set (during the foreperiod) and time for recognition. The task-specific circuitry prepares the motor program
for the required response and holds it in memory (dashed line) until transfer to the response channel during the time
for initiation. In a simple RT paradigm, preparation of the motor program may begin or even be completed well in
advance of the warning signal (•). The response channel includes all the necessary structures to execute the response
as rapidly as possible as soon as the go-signal is identified (time for recognition) and the motor program is transferred
from the task specific circuitry (time for initiation). Thereafter, the response channel needs a period of time (time of
development) to execute the response.
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
EFFECT OF TMS
ONRT
1056
A. PASCUAL-LEONE AND OTHERS
Shortening RT by TMS: time for initiation
We define time for initiation as the time necessary for the transfer of the motor program
generated and held in memory by the task-specific circuitry to the response channel.
We hypothesize that TMS shortens RT primarily by influencing this period. Since the
RT is shortened by the same 30 ms regardless of modality, the effect is likely to be
on a process separate from those that are modality dependent. Transcranial magnetic
stimulation could induce an earlier transfer of the motor program from the task specific
circuitry to the response channel or it could speed up the transfer (Fig. 9). In the former
mechanism, the transfer process would have to begin prior to stimulus recognition. In
the latter mechanism, the two processes could remain serial. We favour the shift in
timing of the transfer process since it provides a simpler explanation for the effects
of TMS on choice paradigms in which focal subthreshold TMS to the motor cortex
leads to faster response times at the cost of biasing the responses towards the contralateral
hand (Brasil-Neto et al., 1992). In such a condition, the completion of the transfer process
prior to stimulus recognition would account for the inappropriate responses.
The topographic specificity of the effects of TMS on RT suggest that during the time
for initiation the motor program is transferred to the Ml. Drawing from their study
on the activity of Ml neurons during foreperiod and RT prior to movement onset, Lecas
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
be most prominent in trials using a visual go-signal and high TMS intensity. In fact,
TMS at high intensities prolonged rather than shortened RT, and the effects were the
same regardless of the go-signal modality.
Shortening RT by TMS: time of development
We think of time of development as an 'energizing' phenomenon (Requin, 1985) by
which the motor system excitability is increased and response execution occurs when
a particular threshold level is reached (Gratton et al., 1988). Time of development is
marked by an increase in firing rate of movement onset-related neurons in primary motor
cortex (Ml) which begins approximately 70—100 ms before movement onset and is
directly linked to the EMG activity (Evarts, 1966, 1986; Luschei et al., 1968; Fetz
and Baker, 1973; Evarts and Tanji, 1976; Fetz and Cheney, 1980; Godschalk et al.,
1981). Transcranial motor cortex stimulation can be used to probe this period of premovement excitability increase. Starr et al. (1988) demonstrated that subthreshold stimuli,
unable to evoke a MEP at rest, will produce a response when delivered < 80 ms before
EMG onset. There is an increase in the probability of evoking an MEP and in the
probability of increasing the MEP amplitude when the interval is shorter between
transcranial stimulus and EMG onset. This facilitation seems temporally related to the
negative motor potential that precedes movement onset by approximately 2 0 - 5 0 ms
and seems to originate from the motor cortex contralateral to the movement (Arezzo
and Vaughan, 1980; Barrett et al., 1985; Tarkka and Hallett, 1990), as well as to the
facilitation of the H-reflex 50—100 ms before voluntary movement (Gurfinkel and
Pal'tsev, 1965).
Our motor cortex excitability experiments demonstrate the absence of effects of TMS
on the time of development. In accordance with Starr et al. (1988) and Tomberg and
Caramia (1991) we found a pre-movement motor cortex excitability build-up beginning
approximately 80 ms before voluntary EMG onset. The presence or absence of TMS
in the go-signal affected RT, but it did not change the course of pre-movement facilitation.
EFFECT OF TMS ON SIMPLE REACTION TIME
1057
Shortening RT by TMS versus short RT to TMS
The effect of TMS to shorten RT to different go-signals must be differentiated from
a short RT to TMS. When subjects are asked to flex their arm rapidly in response to
go-signals of different modalities, RT is shortest (87.8 ±4.1 ms) when the go-signal
is a transcranial stimulus (electrical or magnetic) of subthreshold intensity delivered
over the motor cortex contralateral to the responding arm (Pascual-Leone et al., 1992).
In the present study, TMS shortened auditory RT to 91.6±7.1 ms, which is not
substantially different from the short RT to TMS (Pascual-Leone et al., 1992). It is
conceivable, that the short RT to TMS is a response to the auditory artefact of the
discharging stimulation coil, the contraction of the scalp musculature and the stimulation
of the sensory scalp receptors, accelerated by the stimulation of motor cortex. In other
words, the short RT to TMS may be due to shortening of the time for recognition by
intersensory facilitation and shortening of the time for initiation by the direct effect on
the brain (Fig. 9).
The short RT to TMS (Pascual-Leone et al, 1992) is only minimally longer than
the time of development (approximately 87 ms versus 80 ms). This reinforces the notion
that 'the transcranial stimulus could directly activate the neuronal pool responsible for
response initiation in the motor cortex, thus leading to a "reflex-like" response' (PascualLeone et al., 1992). The short RT to TMS may represent a situation where the time
for recognition is bypassed. This situation may be similar to the catch trials with errors
in the present study. The stimulus evaluation system had to differentiate between the
go-signal and the TMS stimulus to which the subject was not supposed to respond, thus
demanding a particularly long time for recognition. The discharge of the stimulation
coil during TMS is associated with a loud click (Counter et al., 1990). Therefore,
differentiation of TMS from the auditory go-signal is particularly difficult, the time
for recognition takes longer and the potential for TMS-triggered time of development
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
et al. (1986) concluded: 'The features and behavioral significance of the preparationrelated neuronal activity . . . suggest that Ml is only a target for a highly integrated
process that progressively develops elsewhere and is triggered by the stimulus after
acquiring its warning significance.' Several animal studies (Evarts, 1966, 1986; Luschei
et al, 1968; Fetz and Baker, 1973; Evarts and Tanji, 1976; Fetz and Cheney, 1980;
Godschalk et al., 1981) and human studies (Arezzo and Vaughan, 1980; Tarkka and
Hallett, 1990) support this notion. The time for initiation may then be characterized
by the switch of activity from set-related neurons to movement onset-related neurons.
The former are most abundant in the premotor area (Wise et al., 1983; Tanji and Kurata,
1989), have many intracortical connections (Wise, 1989) and may underly the shortterm storage of a planned motor output until the go-signal is perceived (Wise et al.,
1983; Evarts et al., 1984). In contrast, movement onset-related neurons are most abundant
in Ml (Tanji and Kurata, 1989) and exhibit significant correlation to movement
parameters (Tanji and Kurata, 1982).
Subthreshold TMS to the motor cortex may activate cortico-cortical connections and
thus enhance the information transfer between set-related neurons in the premotor cortex
and movement-onset related neurons (cortico-spinal neurons) in Ml. The shortening
of RT by suprathreshold TMS to the motor cortex ipsilateral to the response suggests
a similar effect of transcallosal connections.
1058
A. PASCUAL-LEONE AND OTHERS
without completion or the stimulus evaluation system is greater. This may explain the
significantly higher number of errors in catch trials when the auditory go-signal was used.
The normal duration of the time for initiation must be approximately 30 ms. The
time of development is approximately 70-80 ms long. Together they take 100— 110 ms.
The shortest RT without TMS is that to an auditory go-signal and is approximately
120 ms. This allows only 10 ms for the time for recognition. Reaction times to other
go-signals are longer by virtue of longer time for recognition. Thus the processes that
occur during the RT seem fully accounted for.
ACKNOWLEDGEMENTS
The authors wish to thank B. J. Hessie for skilful editorial assistance and Nguyet Dang for technical
help. Dr Valls-Sole' supported by a grant from the Hospital Clinici Provincial de Barcelona, Barcelona, Spain.
AREZZO J, VAUGHAN HG (1980) Intracortical sources and surface topography of the motor potential and
somatosensory evoked potential in the monkey. Progress in Brain Research, 54, 77 — 83.
BARRETT G, SHIBASAKI H, NESHIGE R (1985) A computer-assisted method for averaging movement-related
cortical potentials with respect to EMG onset. Electroencephalography and Clinical Neurophysiology,
60, 2 7 6 - 2 8 1 .
BERNSTEIN IH, CLARK MH, EDELSTEIN BA (1969) Effects of an auditory signal on visual reaction time.
Journal of Experimental Psychology, 80, 567—569.
BRASIL-NETO J, PASCUAL-LEONE A, VALLS-SOL£ J, COHEN LG, HALLETT M (1992) Transcranial magnetic
stimulation of the human motor cortex induces response bias in a forced-choice task. Journal of
Neurology, Neurosurgery, and Psychiatry. In press.
COHEN LG, ROTH BJ, NILSSON J, DANG N, PANIZZA M, BANDINELLJ S et al. (1990) Effects of coil design
on delivery of focal magnetic stimulation. Technical considerations. Electroencephalography and
Clinical Neurophysiology, 75, 350-357.
COLES MGH, GRATTON G, BASHORE TR, ERIKSEN CW, DONCHIN E (1985) A psychophysiological
investigation of the continuous flow model of human information processing. Journal of Experimental
Psychology: Human Perception and Performance, 11, 529—553.
COUNTER SA, BORG E, LOFQVIST L, BRISMAR T (1990) Hearing loss from the acoustic artifact of the coil
used in extracranial magnetic stimulation. Neurology, Cleveland, 40, 1159—1162.
DAY BL, ROTHWELL JC, THOMPSON PD, MAERTENS DE NOORDHOUT A, NAKASHIMA K, SHANNON K
et al. (1989) Delay in the execution of voluntary movement by electrical or magnetic brain stimulation
in intact man. Brain, 112, 649—663.
ELBERT T, LUTZENBERGER W, ROCKSTROH B, BIRBAUMER N (1981) The influence of low-level transcortical
DC -currents on response speed in humans. International Journal of Neuroscience, 14, 101 — 114.
EVARTS EV (1966) Pyramidal tract activity associated with a conditional hand movement in the monkey.
Journal of Neurophysiology, 29, 1011-1027.
EVARTS EV (1986) Motor cortex output in primates. In: Cerebral Cortex, Volume 5, Sensory-Motor Areas
and Aspects of Cortical Connectivity. Edited by E. G. Jones and A. Peters. New York: Plenum Press,
pp. 2 1 7 - 2 4 1 .
EVARTS EV, TANJI J (1976) Reflex and intended responses in motor cortex pyramidal tract neurons of
monkey. Journal of Neurophysiology, 39, 1069—1080.
EVARTS EV, SHINODA Y, WISE SP (1984) Neurophysiological Approaches to Higher Brain Functions.
New York: John Wiley.
FETZ EE, BAKER MA (1973) Operantly conditioned patterns of precentral unit activity and correlated
responses in adjacent cells and contralateral muscles. Journal of Neurophysiology, 36, 179—204.
FETZ EE, CHENEY PD (1980) Postspike facilitation of forelimb muscle activity by primate corticomotoneuronal cells. Journal of Neurophysiology, 44, 751—772.
GESCHEIDER GA (1976) Psychophysics. Method and Theory. Chichester: John Wiley.
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
REFERENCES
EFFECT OF TMS ON SIMPLE REACTION TIME
1059
GODSCHALK M, LEMON RN, NUS HGT, KUYPERS HGJM (1981) Behaviour of neurons in monkey periarcuate and precentral cortex before and during visually guided arm and hand movements. Experimental
Brain Research, 44, 113-116.
GRATTON G, COLES MGH, SIREVAAG EJ, ERIKSEN CW, DONCHIN E (1988) Pre- and poststimulus activation
of response channels: a psychophysiological analysis. Journal of Experimental Psychology: Human
Perception and Performance, 14, 331-344.
GURFINKEL VS, PAL'TSEV El (1965) Effect of the segmentary apparatus of the spinal cord on the execution
of simple movement reactions. Biofizika, 10, 855 — 860 (in Russian).
HALLETT M, COHEN LG, BIERNER SM (1991) Studies of sensory and motor cortex physiology.
Electroencephalography and Clinical Neurophysiology. Supplement 43, 76—85.
JAEGER D, ELBERT T, LUTZENBERGER W, BIRBAUMER N (1987) The effects of externally applied
PASCUAL-LEONE A, BRASIL-NETO J. VALLS-SOLE J, COHEN LG, HALLETT M (1992) Simple reaction time
to focal transcranial magnetic stimulation: comparison with reaction time to acoustic, visual, and
somatosensory stimuli. Brain, 115, 109 — 122.
REQUIN J (1985) Looking forward to moving soon: ante factum selective processes in motor control. In:
Attention and Performance XI. Edited by M. I. Posner and O. S. M. Marin. Hillsdale, NJ: Lawrence
Erlbaum, pp. 147-167.
ROTH BJ, SAYPOL JM, HALLETT M, COHEN LG (1991) A theoretical calculation of the electric field induced
in the cortex during magnetic stimulation. Electroencephalography and Clinical Neurophysiology,
81, 4 7 - 5 6 .
STARR A, CARAMIA M, ZAROLA F, ROSSINI PM (1988) Enhancement of motor cortical excitability in humans
by non-invasive electrical stimulation appears prior to voluntary movement. Electroencephalography
and Clinical Neurophysiology, 70, 26—32.
TANJI J, KURATA K (1982) Comparison of movement-related activity in two cortical motor areas of primates.
Journal of Neurophysiology, 48, 633—653.
TANJI J, KURATA K (1989) Changing concepts of motor areas of the cerebral cortex. Brain and Development,
I I , 374-377.
TARKKA IM, HALLETT M (1990) Cortical topography of premotor and motor potentials preceding selfpaced, voluntary movement of dominant and non-dominant hands. Electroencephalography and Clinical
Neurophysiology, 75, 36—43.
TAYLOR JL, MCCLOSKEY DI (1990) Triggering of preprogrammed movements as reactions to masked stimuli.
Journal of Neurophysiology, 63, 439—446.
TOMBERG C, CARAMIA MD (1991) Prime mover muscle in finger lift or finger flexion reaction times:
identification with transcranial magnetic stimulation. Electroencephalography and Clinical Neurophysiology, 81, 319-322.
WISE SP (1989) Frontal cortex activity and motor set. In: Neural Programming, Taniguchi Symposia on
the Brain Sciences. No. 12. Edited by M. Ito. Tokyo: Japan Scientific Societies Press and Basel:
Karger, pp. 2 5 - 3 8 .
WISE, SP, WEINRICH M, MAURITZ KH (1983) Motor aspects of cue-related neuronal activity in premotor
cortex of the rhesus monkey. Brain Research, Amsterdam, 260, 301—305.
(Received January 2, 1992. Accepted March 30, 1992)
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016
transcephalic weak direct currents on lateralization in choice reaction tasks. Journal ofPsychophysiology,
I, 127-133.
LECAS JC, REQUIN J, ANGER C, VITTON N (1986) Changes in neuronal activity of the monkey precentral
cortex during preparation for movement. Journal of Neurophysiology, 56, 1680-1702.
LUCE RD (1986) Response Times. Their Role in Inferring Elementary Mental Organization. New York:
Oxford University Press.
LUSCHEI ES, JOHNSON RA, GLICKSTEIN M (1968) Response of neurones in the motor cortex during
performance of a simple repetitive arm movement. Nature, London, 217, 190—191.
NICKERSON RS (1973) Intersensory facilitation of reaction time: energy summation or preparation
enhancement? Psychological Review, 80, 489-509.
PASCUAL-LEONE A, GATES JR, DHUNA A (1991) Induction of speech arrest and counting errors with rapidrate transcranial magnetic stimulation. Neurology, Cleveland, 41, 697-702.
Downloaded from http://brain.oxfordjournals.org/ by guest on May 10, 2016