890 Journal of Pain and Symptom Management Vol. 39 No.

5 May 2010

Original Article

Anodal Transcranial Direct Current

Stimulation of the Motor Cortex Ameliorates
Chronic Pain and Reduces Short Intracortical
Inhibition
Andrea Antal, PhD, Daniella Terney, MD, Stefanie Kühnl, and Walter Paulus, MD
Department of Clinical Neurophysiology, Georg-August University, Göttingen, Germany

Abstract
Context. Consecutive sessions of transcranial direct current stimulation (tDCS)
over the primary motor cortex (M1) may be a suitable therapy to treat chronic
pain, as it can modulate neural activities in the stimulated and interconnected
regions.
Objectives. The present study investigated the analgesic effect of five
consecutive days of anodal/sham tDCS using subjective (visual analog scale
[VAS]) and objective (cortical excitability measured by transcranial magnetic
stimulation [TMS]) measurements.
Methods. Patients with therapy-resistant chronic pain syndromes (trigeminal
neuralgia, poststroke pain syndrome, back pain, fibromyalgia) participated. As
this clinical trial was an exploratory study, statistical analyses implemented
exploratory methods. Twelve patients, who underwent both anodal and sham
tDCS, were analyzed using a crossover design. An additional nine patients had
only anodal or sham stimulation. tDCS was applied over the hand area of the M1
for 20 minutes, at 1 mA for five consecutive days, using a randomized, double-
blind design. Pain was assessed daily using a VAS rating for one month before,
during, and one month post-stimulation. M1 excitability was determined using
paired-pulse TMS.
Results. Anodal tDCS led to a greater improvement in VAS ratings than sham
tDCS, evident even three to four weeks post-treatment. Decreased intracortical
inhibition was demonstrated after anodal stimulation, indicating changes in
cortico-cortical excitability. No patient experienced severe adverse effects; seven
patients suffered from light headache after anodal and six after sham stimulation.
Conclusion. Results confirm that five daily sessions of tDCS over the hand area of
the M1 can produce long-lasting pain relief in patients with chronic pain. J Pain
Symptom Manage 2010;39:890e903. Ó 2010 U.S. Cancer Pain Relief Committee.
Published by Elsevier Inc. All rights reserved.

This study was supported by the German Ministry University, Robert Koch Strasse 40, 37075 Göttin-
for Research and Education (BMBF-01EM 0513). gen, Germany. E-mail: Aantal@gwdg.de
Address correspondence to: Andrea Antal, PhD, Depart- Accepted for publication: October 16, 2009.
ment of Clinical Neurophysiology, Georg-August

Ó 2010 U.S. Cancer Pain Relief Committee 0885-3924/$esee front matter

Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jpainsymman.2009.09.023
Vol. 39 No. 5 May 2010 Anodal Stimulation Improves Chronic Pain 891

Key Words
Chronic pain, tDCS, TMS, motor cortex stimulation, intracortical inhibition

Introduction right motor cortex (2 mA, 20 minutes for five

consecutive days). There was a significant
Despite the availability of multiple pharma-
pain perception improvement after anodal
cological approaches concerning chronic
stimulation of the M1 but not after sham stim-
pain, it is not uncommon that patients fail to
ulation. A follow-up study reported a similar
experience sufficient pain relief. This dilemma
decrease in pain perception in a patient group
establishes the need for new therapeutic inter-
with fibromyalgia.9
ventions and has prompted a renewed interest
Compared with the aforementioned stud-
in neuromodulatory approaches with brain
ies,8,9 we changed three parameters related
stimulation. Stimulation of the primary motor
to the stimulation: 1) the intensity of stimula-
cortex (M1) for the treatment of certain forms
tion was 1 mA (2 mA in the study8); 2) related
of refractory neuropathic pain has attracted
electrode size was 4  4 cm to increase the fo-
much interest in recent years. Tsubokawa
cality of the stimulation (5  7 cm in the stud-
et al.1 first showed that central poststroke
ies8,9); and 3) a crossover design was used in 13
pain could be reduced by means of chronic
participants, who received both types of stimu-
motor cortex stimulation (MCS) through im-
lations. Furthermore, we examined the possi-
planted epidural electrodes. Further studies
ble intracortical effects of repeated tDCS over
proved that MCS could also relieve trigeminal
M1 using paired-pulse stimulation. Paired-
neuropathic pain.2 Deep brain stimulation has
pulse TMS techniques include several proto-
shown promising results,3 but less invasive
cols to study the modulation of motor cortex
forms of stimulation also might be effective.
excitability by local circuits of afferent inputs
Indeed, a number of studies have shown that
from other cortical areas of the brain.15,16
both a single session and repeated sessions of
Finally, the possible side effects of the stimula-
repetitive transcranial magnetic stimulation
tion were evaluated by a questionnaire devel-
(rTMS) can relieve pain transiently in some
oped by our research group.17
patients with chronic neuropathic pain,4e6
although others have found the effect to be
small and not significant.7
Recent studies have demonstrated the effec- Methods
tiveness of transcranial direct current stimula- Subjects
tion (tDCS) on pain symptoms in patients All participants (patients with trigeminal
with central pain because of traumatic spinal neuralgia, poststroke pain syndrome, back
cord injury8 and fibromyalgia.9 Several studies pain, fibromyalgia) (Table 1) were outpatients
have shown that this technique modulates cor- at the Department of Clinical Neurophysiol-
tical excitability in the M1 (for a review, see ogy, Göttingen, Germany. They were regarded
Ref. 10), and its modulatory effect endures as suitable to participate if they fulfilled the fol-
after stimulation. tDCS not only shifts the activ- lowing criteria: 1) stable chronic pain for at
ity of cortical areas situated directly under the least the preceding six months; 2) score
electrodes but also of distant areas, probably greater than or equal to 3 (0 ¼ no pain and
through interconnections between the princi- 10 ¼ worst possible pain) on the visual analog
pal stimulated area and these structures.11 scale (VAS) for pain perception during the
The primary effect of tDCS is a neuronal last month before baseline/start of the stimu-
de- or hyperpolarization of membrane poten- lation; and 3) refractoriness to drugs for pain
tials,12,13 whereby the induced aftereffects relief, such as nonopioid analgesics, tricyclic
depend on N-methyl-D-aspartate receptor- antidepressants, antiepileptic drugs, and/or
efficacy changes.14 opioids (pain resistance to at least two of these
In a recent study,8 patients were randomized drugs supplied in adequate dosages for six
to receive sham or active tDCS over the left or months). The diagnoses were made by trained
892 Antal et al. Vol. 39 No. 5 May 2010

Table 1
Clinical and Demographic Characteristics of the Patients With Regard to the Different Stimulation Conditions
Only Anodal tDCS Only Sham tDCS Active and Sham tDCS

Number (n) 6 (1 patient in 4 13 (1 patient in

the follow-up) the follow-up)
Gender (male) 2 1 3
Age range (years) 28e70 50e70 41e70
Etiology and side of the pain
Fibromyalgia 2 (both sides) 1 (both sides)
Chronic back pain 2 (both sides) 1 (both sides) 5 (both sides)
Trigeminal neuralgia 2 (right side) 1 (1 right side)
Atypical face pain 2 (both sides)
Arthrosis (finger, foot) 1 (both sides) 1 (both sides) 2 (both sides)
Phantom pain (leg) 1 (left side)
Poststroke pain (arm) 1 (right side)
Polyneuropathy 1 (both sides)
Duration of chronic pain
More than 5 years 4 3 6
Between 2 and 5 years 1 1 5
Less than 2 years 1 2
Baseline VAS scores (SD) 7.11 (1.2) 7.0 (1.5) 5.8 (2.1); 5.95 (2.2)
Present medication
Pregabalin 1 2
Nonsteroidal anti-inflammatory drug 2 2 2
Morphine 2 2
Amitriptyline 1
No medication 2 7

neurologists at least two years before the study were not included among the follow-up analy-
started. We excluded patients with any clini- ses (one from the crossover design and one
cally significant or unstable medical or psychi- who had received only anodal stimulation;
atric disorder, history of substance abuse, or see Table 1), because they did not provide us
neuropsychiatric comorbidity. In addition, we with their pain diaries in the follow-up period.
excluded patients with metallic implants/im- In the case of 13 patients, a crossover design
planted electric devices. As carbamazepine was applied; they had a second session of stim-
might decrease the effects of anodal stimula- ulation with at least a six-week break after the
tion after a single session of tDCS,14 patients first one and only if the VAS values returned
taking carbamazepine also were excluded. Fi- to the previous baseline values for at least 10
nally, 23 patients (age range 28e70 years; six days’ prior stimulation. For the remaining 10
males) participated in the study. The general patients, six had anodal and four had sham
clinical characteristics of the patients are sum- stimulations.
marized in Table 1. Table 2 shows the individ-
ual demographic and clinical parameters of Experimental Design
the patients. Patients with lumber pain had The study had three phases: 1) baseline eval-
not undergone surgery. Facial pains were con- uation consisting of a four-week period of the
tinuous in all cases. The patients were aware of daily registration of subjective, baseline pain;
the fact that they could get sham or real 2) one-week treatment, which consisted of daily
stimulation. treatment sessions with sham or active tDCS (20
The study was a randomized, double- minutes) for five consecutive days; and 3) a fol-
blinded, placebo-controlled, single-center trial low-up period of four weeks. Before the first
that was designed to evaluate the efficacy of stimulation and two to three days after the last
five daily sessions of anodal tDCS in patients stimulation, the patient underwent several mea-
with chronic pain. The study conformed to surements related to cortical excitability (see
the ethical standards of the 1964 Helsinki Dec- below). During the baseline period, patients
laration and was approved by the institutional were randomized to receive sham or active
ethics committee. All patients provided writ- tDCS. Randomization was performed using
ten, informed consent. Two of 23 patients the order of entrance into the study.
Vol. 39 No. 5 May 2010 Anodal Stimulation Improves Chronic Pain 893

Table 2
Individual Patient Parameters With Regard to the Different Stimulation Conditions
Patient Stimulation Gender Age (Years) Diagnosis Duration of Pain (years) Medication

1 Anodal-sham F 58 Chronic back pain 19 Nonsteroidal

2 Anodal-sham F 63 Arthrosis 5 Morphine
3 Anodal-sham F 70 Fibromyalgia 20 Nonsteroidal
4 Anodal-sham F 68 Chronic back pain 4 Morphine
5 Anodal-sham F 63 Arthrosis (foot, finger) 3 d
6 Anodal-sham M 70 Chronic back pain 1.5 d
7 Anodal-sham F 51 Polyneuropathy 5 Pregabalin
8 Anodal-sham F 56 Atypical face pain 12 d
9 Anodal-sham M 53 Trigeminal neuralgia 6 d
10 Anodal-sham M 61 Poststroke pain 8 d
11 Anodal-sham F 41 Atypical face pain 1.5 d
12 Anodal-sham F 54 Chronic back pain 6 Pregabalin
13 Anodal-sham F 55 Chronic back pain 25 d
14 Anodal F 70 Chronic back pain 8 Nonsteroidal
15 Anodal F 42 Arthrosis 2 d
16 Anodal F 65 Trigeminal neuralgia 3 Nonsteroidal
17 Anodal M 28 Phantom pain (leg) 7 Morphine
18 Anodal F 57 Trigeminal neuralgia 3 Morphine
19 Anodal M 59 Chronic back pain 6 d
20 Sham F 56 Fibromyalgia 6 Nonsteroidal
21 Sham M 70 Arthrosis 2 Nonsteroidal
22 Sham F 50 Chronic back pain 8 Pregabalin
23 Sham F 52 Fibromyalgia 10 Amitriptyline
F ¼ female; M ¼ male.

Transcranial Direct Current Stimulation bilateral pain. In addition, because the elec-
Subjects were seated in a comfortable reclin- trode for tDCS is large, the stimulation encom-
ing chair with a mounted headrest throughout passed a broad area of the motor cortex
the experiments. The stimulations with regard (upper limb and face). A constant current of
to one given patient were always done by the 1-mA intensity was applied for 20 minutes.
same investigator. Direct current was trans- The maximal current density was 62.5 mA/cm2
ferred by a saline-soaked pair of surface sponge over the M1 and 12 mA/cm2 at the reference
electrodes (4  4 cm over the M1 and 5  10 cm electrode. Subjects felt the current as an itching
over the contralateral orbit) and delivered by sensation at both electrodes at the beginning of
a specially developed, battery-driven, constant the stimulation. For sham stimulation, the elec-
current stimulator (NeuroConn, Ilmenau, Ger- trodes were placed in the same positions as for
many) with a maximum output of 5 mA. anodal M1 stimulation; however, the stimulator
Patients received either anodal stimulation or was turned off automatically after 30 seconds of
sham stimulation of M1. First, the hand area stimulation. Therefore, the subjects felt the ini-
over the left hemisphere was determined by tial itching sensation but received no current
single-pulse TMS. For anodal stimulation, the for the rest of the stimulation period. The stim-
active electrode was placed over the hand repre- ulators were coded using a five-letter code, pre-
sentation field of M1 at the left side and the programmed by one of the department
reference (cathode) electrode over the contra- members, who otherwise did not participate in
lateral supraorbital area, independently from the study. Therefore, neither the investigator
the lateralization of the pain. This electrode po- nor the patient knew the type (anodal or
sition has been shown to be effective to enhance sham) of the stimulation. After the five-day stim-
the excitability of the M1.18 Furthermore, most ulation period, patients were asked if they could
of the patients had pain at the right side or both guess the type of stimulation they received.
sides of the body. It was observed that M1 stimu-
lation with tDCS induces widespread changes in Single-Pulse Transcranial Magnetic
the activity of cortical areas and can indeed Stimulation
change the activity of the contralateral hemi- The patients who participated in the cross-
sphere.11 This suggests that unilateral tDCS over study (n ¼ 13) were included in this ex-
treatment might be sufficient for patients with periment. To detect current-driven changes
894 Antal et al. Vol. 39 No. 5 May 2010

of excitability, motor-evoked potentials (MEPs) experiment. SI1 mV was determined with

of the right first dorsal interosseus (FDI) mus- single-pulse TMS first. RMT was defined as
cle were recorded before and after the stimula- the minimal output of the stimulator that in-
tion of its motor-cortical representation field duced a reliable MEP (w50 mV in amplitude)
by single-pulse TMS. TMS was performed by in at least three of six consecutive trials when
using a Magstim standard double (‘‘figure the FDI muscle was completely relaxed. AMT
eight’’) 70 mm coil (2.2 T; average inductance, was defined as the lowest stimulus intensity at
16.35 mH) connected to two monophasic Mag- which three of six consecutive stimuli elicited
stim 200 stimulators by means of a bistim mod- reliable MEP (w200 mV in amplitude) in the
ule (Magstim Co., Whiteland, Dyfed, UK). tonically contracting FDI muscle.19
Surface electromyogram (EMG) was recorded SICI/ICF, LICI, and SICF were measured
from the right FDI through a pair of Ag-AgCl with three different protocols of single- and
surface electrodes in a belly-tendon montage. paired-pulse TMS applied in random order at
Raw signals were amplified, band-pass filtered 0.25 Hz. For SICI/ICF, two magnetic stimuli
(2 Hz to 3 kHz; sampling rate, 5 kHz), digi- were given through the same stimulating coil,
tized with a micro 1401 AD converter (Cam- and the effect of the first (conditioning) stim-
bridge Electronic Design, Cambridge, UK) ulus on the second (test) stimulus was investi-
controlled by Signal Software (version 2.13; gated.15 To avoid any floor or ceiling effect,
Cambridge Electronic Design), and stored on the intensity of the conditioning stimulus was
a personal computer for offline analysis. set to a relatively low value of 80% of AMT.
Whenever necessary, complete relaxation was The test-stimulus intensity was adjusted to
controlled through auditory and visual feed- SI1 mV. SICI was measured with interstimulus
back of EMG activity. The coil was held tangen- intervals (ISI) of two and four milliseconds,
tially to the skull, with the handle pointing and ICF with ISIs of 9 and 12 milliseconds.
backward and laterally at 45 from the midline, The control condition (test pulse alone)
resulting in a posterior-anterior direction of was tested 40 times, and each of the
current flow in the brain. This orientation of conditioning-test stimuli was tested 20 times.
the induced electrical field is thought to be The mean peak-to-peak amplitude of the con-
optimal for predominantly transsynaptic ditioned MEP at each ISI was expressed as
mode of activation of the corticospinal system. a percentage of the mean peak-to-peak size
The optimum position was defined as the site of the unconditioned test pulse. SICI was
where TMS resulted consistently in the largest taken as the mean percentage inhibition at
MEP in the resting muscle. The site was ISIs of two and four milliseconds, whereas
marked with a skin marker to ensure that the ICF was taken as the mean facilitation at ISIs
coil was held in the correct position through- of 9 and 12 milliseconds.
out the experiment. The second protocol tested LICI with two
suprathreshold stimuli applied with ISIs of
Paired-Pulse Transcranial Magnetic 50, 100, 150, and 200 milliseconds.16 The in-
Stimulation tensities of both stimuli were set to 110% of
TMS measurements included resting motor RMT. Here also, the intensity was set to this rel-
threshold (RMT), active motor threshold atively low value to avoid any floor or ceiling ef-
(AMT), the intensity to evoke an MEP of w1- fect. The control condition (first pulse alone)
mV peak-to-peak amplitude (SI1 mV), short- was tested 40 times, whereas each of the paired
interval intracortical inhibition/intracortical stimuli was tested 20 times. LICI was taken as
facilitation (SICI/ICF), long-interval intracort- the mean percentage inhibition of condi-
ical inhibition (LICI), short-interval intracorti- tioned MEP at ISIs of 50, 100, 150, and 200
cal facilitation (SICF), recruitment curves and milliseconds. SICF also was measured in
cortical silent period (CSP). a paired-pulse TMS protocol, but here, the first
Patients participated in four experimental pulse was suprathreshold, and the second
conditions (before and after sham and anodal pulse was subthreshold.20,21 The intensity of
tDCS). Stimulus intensities (in percentage of the first (test) pulse was set to SI1 mV, and
maximal stimulator output) of TMS were de- the intensity of the second (conditioning)
termined at the beginning of each pulse to 90% of RMT. In this protocol, the
Vol. 39 No. 5 May 2010 Anodal Stimulation Improves Chronic Pain 895

control condition (test pulse alone) was given contained rating scales for the presence and
40 times, and each of the conditioning-test severity of headache; difficulties in con-
stimuli with the ISIs 1.3, 2.1, 2.5, 3.5, and 4.1 centrating; acute mood changes; visual per-
milliseconds, was tested 20 times. The mean ceptual changes; fatigue; and discomforting
peak-to-peak amplitude of the conditioned sensations, such as pain, tingling, itching, or
MEP at each ISI was expressed as a percentage burning under the electrodes during and af-
of the mean peak-to-peak size of the uncondi- ter tDCS.
tioned test pulse.
Recruitment curves (RECR) were measured Statistical Analysis
with four different and increasing stimulus in- Pain Perception. Two kinds of analyses were
tensities (100%, 110%, 130%, and 150% of done: 1) all of the patients (n ¼ 21) were in-
RMT), each with 10 pulses. Mean amplitude cluded, and the anodal and sham conditions
was calculated for each intensity. Finally, 10 were compared (17 anodal vs. 16 sham stimula-
pulses with SI1 mV and 10 pulses with 120% tions); and 2) the data from 12 patients were
RMT were applied under tonic contraction of analyzed, using a crossover analysis (12 anodal
the right FDI muscle. CSPs were separately de- vs. 12 sham stimulations). We normalized the
termined in rectified and averaged EMG traces post-stimulation VAS values to baseline
with a prestimulus period of 100 milliseconds. (mean of the last 10 days prestimulation).
CSP (in milliseconds) was measured from the The data were treated using a between-
TMS stimulus to the point where the signal subject design because of the partly different
reached the amplitude of the mean prestimu- groups. Additionally, we compared the presti-
lus EMG activity again for more than 5 mulation VAS values with Student’s t-test for
milliseconds. both groups of patients to rule out that differ-
ences between sham and active stimulation
Pain Measurement: Visual Analog Scale for might have been because of a priori differences
Pain between these two stimulation groups.
This self-evaluation scale ranges from 0 to 10 For the analysis for VAS scores, we used
as visually described in centimeter units: 0 cm repeated-measures analysis of variance (AN-
indicates no pain and 10 cm means the worst OVA), in which the dependent variable was
pain possible. This scale has been widely used the VAS score and the factors were STIMULA-
in studies that evaluate pain as an outcome; TION (sham and anodal) and TIME of treat-
both validity and reproducibility have been ment (baseline: mean of the last 10 days
demonstrated.22 Participants were asked to before stimulation, during stimulation,1e5
rate their pains three times a day at the same and after stimulation [Days 7, 14, 21, and
time point at least 30 days before, during, 28]). When appropriate, post hoc comparisons
and 30 days after the end of the stimulation. (between sham and anodal stimulation and be-
These values were averaged, giving one VAS tween post-treatment evaluations and baseline
value per day. We instructed the participants evaluations) were carried out using Bonferroni
to continue their routine medication regimen. correction for multiple comparisons. In addi-
Herbal remedies and alternative therapies, tion, for all endpoints, we also ran a Student’s
such as massage or acupuncture, were allowed t-test, including only the time factor as a re-
if they had been used for at least four weeks be- peated measurement.
fore randomization, and the regimen was
maintained constant throughout the study. Transcranial Magnetic Stimulation Measurements.
For each measurement (SI1 mV, RMT, AMT,
Adverse Effects of Transcranial Direct Current SICI, ICF, SICF, LICI, CSP), we performed sepa-
Stimulation rate ANOVAs for repeated measurements by us-
The electric current was applied continu- ing the mean values from each subject as the
ously for 20 minutes/day in this protocol. dependent variable to compare the parameters
Because potential adverse effects of this tech- of cortical excitability. In addition to the factor
nique are not fully known yet, the patients com- ‘‘STIMULATION type’’ (anodal vs. sham) and
pleted a questionnaire17 after anodal and sham factor ‘‘TIME’’ (baseline, post-stimulation), the
stimulation, separately. The questionnaire ANOVA model included the factor ‘‘ISI’’ (2, 4,
896 Antal et al. Vol. 39 No. 5 May 2010

7, 9, and 12 milliseconds) when SICI and ICF

were analyzed or the factor ‘‘intensity’’ (100%,
110%, 130%, and 150% of RMT) for recruitment
curves, or the factor ‘‘INTENSITY’’ (120% RMT
and SI1 mV) for CSP. Here, only the data of the
patients who participated in the crossover design
were included. Unless stated otherwise, all re-
sults are presented as means and standard errors,
and statistical significance refers to a two-tailed P-
value less than 0.05.

Adverse Effects of Transcranial Direct Current Stim-

ulation. The incidences of side effects were
coded using a binary system (no ¼ 0, yes ¼ 1),
and the severities of the side effects were rated
using a numerical analog scale from 1 to 5; 1
being very mild and 5 indicating an extremely
strong intensity of any given side effect. From
these values, a mean intensity was calculated.
The incidences and severities of the adverse
effects were separately calculated during and
after stimulation. The percent of affected pa-
tients in the sham and anodal conditions was
calculated independently.
Fig. 1. Pain scores as indexed by VAS throughout
and after the stimulation (a) with regard to all of
Results the patients and (b) the patients who participated
in the crossover stimulation study. The VAS scores
Pain ControldVisual Analog Scale for each time point were normalized to the
There was no significant difference concern- before-stimulation values (10 days’ mean VAS be-
ing the baseline VAS values (t ¼ 1.8, df ¼ 32, fore the first day of stimulation). Bars show standard
error of mean.
P ¼ 0.1). To analyze whether tDCS treatment
was associated with pain improvement, we per-
formed a repeated-measures ANOVA, in which
the dependent variable was change of normal- (F(8,176) ¼ 1.24, P ¼ 0.27) and the interaction
ized VAS pain scores, and the independent vari- between TIME and STIMULATION
able was TIME of evaluation (baseline; Days 1, 2, (F(8,176) ¼ 0.35; P ¼ 0.9) were not significant
3, 4, 5; and follow-up) and type of STIMULA- (Fig. 1b). Student’s t-test revealed a significant
TION (anodal vs. sham tDCS). Concerning the difference on Days 3, 4, 5, 7, 14, and 28 between
whole group, this analysis revealed a significant anodal and sham stimulations (P < 0.05; VAS
main effect of STIMULATION (F(1,32) ¼ 5.32, change: 36.5% decrease vs. 9% increase, 33.5%
P < 0.05) and TIME (F(8,256) ¼ 2.14, P < 0.03); decrease vs. 7% increase, 35% decrease vs. 1%
however, the interaction between them was not decrease, 11% decrease vs. 23% increase,
significant: F(8,256) ¼ 0.28 and P ¼ 0.9 (Fig. 1a). 27.5% decrease vs. 5% increase, 26.5% decrease
Student’s t-test revealed a significant difference vs. 6% increase, respectively).
on Days 3 and 7 between anodal and sham stim- Finally, we compared the baseline VAS
ulations (P < 0.05; VAS change: 27.3% decrease values with all of the time points (the factor
[anodal] vs. 2.7% [sham] and 8.9% decrease vs. TIME was significant using ANOVA) for each
11% increase, respectively). Concerning the 12 condition separately, using Student’s t-test.
patients who participated in both stimulation Concerning the full group, this analysis
sessions, the ANOVA revealed a significant showed a significant difference during the an-
main effect of the group of STIMULATION odal stimulation (baseline comparison with
(F(1,22) ¼ 14.3, P < 0.005). However, the TIME first stimulation [P ¼ 0.03], second stimulation
Vol. 39 No. 5 May 2010 Anodal Stimulation Improves Chronic Pain 897

[P ¼ 0.03], third stimulation [P ¼ 0.003], started during the stimulation. In these cases,
fourth stimulation [P ¼ 0.001], fifth stimula- only threshold measurements were done.
tion [P ¼ 0.0006]) and after the stimulation First, RMT, AMT, and SI1 mV baseline values
(Day 14 [P ¼ 0.03] and Day 28 [P ¼ 0.03]). were compared between active and sham stim-
The largest pain reduction was achieved after ulation conditions. There was no significant
the fifth session of stimulation (30%; mean difference between anodal and sham stimula-
VAS pain scores: 4.22 [2.23]). Applying tion in any of the measurements (Table 3).
sham stimulation, there was no significant dif- Anodal stimulation had no effect on SICF,
ference between baseline and during/post- ICF, LICI, CSP, or motor-evoked recruitment
stimulation VAS values. curves, as revealed by repeated-measures
Concerning the 12 patients, the analysis ANOVA (Table 3). However, Student’s t-tests
showed a similar tendency: significant differ- showed a significant difference in the case of
ence was found during the anodal stimulation SICI (P < 0.05), showing decreased intracorti-
(baseline comparison with first stimulation cal inhibition after anodal stimulation
[P ¼ 0.02], second stimulation [P ¼ 0.05], (Fig. 2). Eight of the patients showed a clear
third stimulation [P ¼ 0.004], fourth stimula- trend: the increase in SICI correlated with
tion [P ¼ 0.01], fifth stimulation [P ¼ 0.008]), the decrease in VAS after anodal but not after
and after the stimulation (Day 14 [P ¼ 0.03] sham stimulation. However, because of the low
and Day 28 [P ¼ 0.02]). The largest pain re- number of subjects, this was not significant.
duction was achieved after the third session
of stimulation (37%; mean VAS pain scores: Adverse Effects of Transcranial Direct Current
3.54 [2.29]). Applying sham stimulation, Stimulation
there was no significant decrease with regard Neither of the subjects terminated the stim-
to baseline and during/post-stimulation VAS ulation or needed any medical intervention
values except the first-day values (P ¼ 0.05). during or after the end of tDCS. Tables 4
There were eight responders (reduction of and 5 summarize the adverse effects during
30% or more in the VAS after five-day stimula- and after tDCS, including all of the sham
tion) in the anodal tDCS group (63% of and anodal stimulation conditions. During
patients) and only four responders in the the stimulation, mild tingling sensation was
sham tDCS group (16%). In the follow-up the most common adverse effect; it was re-
evaluation, four patients in the active tDCS ported by 66.6% of the subjects during anodal
group were still considered as responders, and 52.9% during sham stimulation. Moderate
and three responders in the sham tDCS group fatigue was the second most frequent conse-
reported no more pain improvement. Impor- quence; it was reported by 44.4% of the partic-
tantly, active tDCS did not worsen pain in ipants during anodal stimulation and,
any of the patients who received this interestingly, 64.7% of the subjects during
treatment. sham stimulation. Similarly, after the stimula-
Because we had a small sample size and a het- tion, 33.3% of the patients felt tired in the an-
erogeneous group with regard to the origin of odal group, and 70.6% reported this symptom
the chronic pain, we did not perform any cor- in the sham group. Headache occurred in
relation analysis for the characteristics of the 38.9% in the anodal group and 35.3% in the
disease. However, larger pain improvement sham group. Four patients in the anodal and
was observed among patients with arthrosis two in the sham groups reported acute sleep-
than back pain. None of the patients with fi- ing disturbances for two to five days after
bromyalgia showed a different pattern of tDCS. Concerning the patients who partici-
pain reduction after the five-day treatment pated in the crossover study, only one reported
compared with sham stimulation. that there was a difference between the two
types of stimulations when it was explicitly
asked.
Single- and Paired-Pulse Transcranial There were two dropouts in the follow-up
Magnetic Stimulation period (one anodal and one sham). Both the
For two subjects, the experimental sessions patients showed no improvement that could
were interrupted because of headache that have contributed to their decision to abandon
898 Antal et al. Vol. 39 No. 5 May 2010

Table 3
Statistical Analysis of the Single- and Paired-Pulse TMS Experiments
TMS measurement Factor df F P

ANOVA SI1 mV Stimulation 1 0.08 0.77

Time 1 0.0 0.99
Stimulation  time 25 0.03 0.85
AMT Stimulation 1 0.2 0.66
Time 1 0.68 0.42
Stimulation  time 25 0.98 0.20
RMT Stimulation 1 0.22 0.56
Time 1 0.33 0.57
Stimulation  time 25 0.09 0.77
RECR Stimulation 1 0.36 0.55
Time 1 1.67 0.21
Stimulation  time 22 0.08 0.78
Intensity 3 39.26 <0.01
Stimulation  intensity 3 0.18 0.90
Time  intensity 3 1.52 0.22
Stimulation  time  intensity 66 0.17 0.92
CSP Stimulation 1 0.25 0.61
Time 1 0.43 0.52
Stimulation  time 22 0.002 0.95
Intensity 1 2.66 0.11
Stimulation  intensity 1 0.76 0.39
Time  intensity 1 0.002 0.96
Stimulation  time  intensity 22 0.26 0.61
SICI/ICF Stimulation 1 0.01 0.9
Time 1 3.32 0.08
Stimulation  time 20 0.69 0.41
ISI 4 88.96 <0.001
Stimulation  ISI 4 0.39 0.81
Time  ISI 4 0.56 0.69
Stimulation  time  ISI 80 3.68 0.05
LICI Stimulation 1 0.05 0.99
Time 1 4.19 0.05
Stimulation  time 23 0.08 0.77
ISI 3 2.84 0.04
Stimulation  ISI 3 0.33 0.80
Time  ISI 3 4.77 0.004
Stimulation  time  ISI 69 0.6 0.59
Student’s t-test Stimulation Type df t P

ICI Anodal 10 3.53 <0.005

7 milliseconds 10 1.5 0.16
ICF 10 0.29 0.77
ICI Sham 10 0.044 0.96
7 milliseconds 10 0.13 0.89
ICF 10 2.0 0.07

their participation in the remainder of the adverse events in both treatment groups, as
study. in previous studies including patients with
chronic pain8,9 or healthy subjects.17 These ad-
verse events consisted of mild headache,
fatigue, and itching under the electrodes.
Discussion Compared with previous studies that used an-
Our results demonstrate that five daily ses- odal tDCS for the treatment of chronic pain, the
sions of anodal tDCS using a relatively small magnitude of our results (mean pain decrease
(4  4 cm) stimulation electrode over the of 37%) is somewhat lower (58% in the study
hand area of the M1 can produce long- by Fregni et al.,8 which included patients with
lasting pain relief in patients experiencing dif- low back pain), possibly because of our hetero-
ferent types of chronic pain, probably by geneous patient group and may be a result of
decreasing the level of intracortical inhibition. the different criteria used to designate pain re-
However, there was a higher incidence of sponders. Nevertheless, our result is similar to
Vol. 39 No. 5 May 2010 Anodal Stimulation Improves Chronic Pain 899

Fig. 2. The main results of the TMS experiment. After anodal stimulation decreased, intracortical inhibition
(ISIs: 2, 4 milliseconds) was observed. Bars show standard error of mean.

those of high-frequency rTMS studies that re- the long-lasting effect of anodal stimulation,
port mean pain relief that ranges from 20% to our results are similar to those published by
45%.4,5 Furthermore, the reduced tDCS inten- Fregni et al.8,9 However, compared with epidu-
sity here may be better suited for blinding. As re- ral stimulation, tDCS may result in a smaller
ported by Furubayashi et al.,23 an intensity of treatment effect: the surgical approach re-
3 mA starts to be painful, 1 mA certainly has sulted in a 28%e70% mean pain reduction
a higher chance of going unnoticed compared in 50%e80% of responders;1,24e27 neverthe-
with placebo conditions than 2 mA. Because less, this may include a substantially stronger
there was no difference with regard to the occur- placebo effect. The mean pain relief varied
rence of itching/tingling/burning sensations from 28% to 47% in the largest series,27,28
and headache between patients after sham and between 50% and 70% in the smallest se-
and verum stimulation, and fatigue was experi- ries25,26 of patients. Nevertheless, the effects
enced by an even higher percentage of sham pa- of five daily sessions of anodal tDCS are likely
tients compared with anodally stimulated to have a shorter impact on brain activity
patients, we are quite sure at having provided than what can be achieved using MCS with epi-
an optimal placebo condition. dural electrodes for several months.
An interesting question to be solved in the The magnitude of the placebo effect was
near future is the optimal duration and repeti- lower in our study (varied from 18.9% to
tion rate of tDCS sessions. In the present study, þ9.8% during and after treatment) compared
the effects built up during the first three stim- with rTMS studies. Here, we expect a higher
ulation sessions, being mild in many patients placebo response to rTMS, simply because of
immediately after the initial stimulation but the higher technical effort and the remote
quite clear after the fourth or fifth days. Effects risk of seizure induction, for which patients ob-
were also separated from the placebo effect of taining tDCS do not have to be informed. Fur-
sham stimulation after this duration. Accord- thermore, it has been documented that the
ing to the original observation by Lefaucheur severity and therapeutic refractoriness of
et al.,5 pain relief after a single session was op- symptoms can correlate negatively with pla-
timal two to four days after rTMS. Regarding cebo response.29 Therefore, a lower placebo
900 Antal et al. Vol. 39 No. 5 May 2010

Table 4
Adverse Effects of tDCS During Stimulation
Pain Tingling Itching
Type of
Stimulation n % MI n % MI n % MI

Sham 2 11.8 10 9 52.9 1.1  0.11 0 0 0

Anodal 2 11.1 21 12 66.6 1.5  0.15 6 33.3 2.0  0.15
Burning Fatigue Nervousness

n % MI n % MI n % MI

Sham 4 23.5 10 11 64.7 1.7  0.27 1 5.9 10

Anodal 2 11.1 10 8 44.4 1.8  0.46 0 0 0
Trouble in Concentrating Changes in Visual Perception Headache

n % MI n % MI n % MI

Sham 1 5.9 20 0 0 0 5 29.4 1.4  0.24

Anodal 1 5.5 20 2 11.1 10 5 27.8 1.8  0.75
Unpleasantness Other

n % MI n

Sham 0 0 0 1 Tingling in the tongue

Anodal 2 11.1 21 1 Warm tingling with regard
to the leg
1 Involuntary muscle contraction
MI ¼ mean intensity (0e5).

response may indicate the therapeutic refrac- the thalamic and subthalamic nuclei. In fact,
toriness of our patient population. a change in the activity of these nuclei is asso-
It was suggested by previous studies that the ciated with rTMS30 and tDCS.11 Sequentially,
best stimulation site using rTMS for pain con- the activation of thalamic nuclei can modify
trol is not the area corresponding to the pain- the activity of other pain-related structures
ful zone but the adjacent one.6 However, with (e.g., anterior cingulate, periaqueductal gray)
regard to epidural stimulation, another study and can also inhibit pain of spinal origin.
showed that the most efficacious position of The higher incidence of adverse effects
the stimulating electrodes was the one corre- (mainly headache and fatigue) during and af-
sponding to the cortical somatotopic represen- ter tDCS compared with those reported among
tation of pain perception.27 We have used healthy subjects in a previous study17 might be
a smaller stimulation electrode (4  4 cm) related to the type of disorder investigated in
compared with those of previous clinical stud- our study. Similarly, a higher proportion of mi-
ies8,9 to increase the focality of the stimulation. graineurs (55.6%) reported headache after
However, because of the still large electrode tDCS compared with healthy subjects
size applied in our study, it is likely that the (7.8%).17 Therefore, it is not astonishing that
stimulation induced a modulatory effect over 35%e39% of the subjects reported headache
a large area of M1 that encompassed both in our study. However, we have not observed
the area corresponding to the pain and the ad- any serious complications, such as seizures, in
jacent areas. Further studies are needed to connection with the application of tDCS. As
explore whether somatotopically guided stimu- tDCS has been tested only recently for clinical
lation might increase the analgesic effect of applications, the impact of consecutive ses-
tDCS and to pursue the concept of Lefaucheur sions of cortical stimulations on different cog-
et al.6 nitive, motor, and other functions in different
The mechanisms responsible for the long- patient populations is not yet fully known.
lasting effect of anodal tDCS on pain are still Therefore, our data concerning the side
unknown. Upregulation of M1 excitability effects of the stimulation could be interpreted
could modulate pain perception through indi- as preliminary safety evidence with regard to
rect effects on pain-modulating areas, such as tDCS in chronic pain.
Vol. 39 No. 5 May 2010 Anodal Stimulation Improves Chronic Pain 901

Table 5
Adverse Effects of tDCS After Stimulation
Pain Tingling Itching
Type of
Stimulation n % MI n % MI n % MI

Sham 1 5.9 10 1 5.9 10 1 5.9 10

Anodal 1 5.5 10 1 5.5 10 1 5.5 10
Burning Fatigue Nervousness

n % MI n % MI n % MI

Sham 1 5.9 10 12 70.6 2.25  0.30 1 5.9 10

Anodal 1 5.5 10 6 33.3 2.4  0.51 1 5.5 20
Trouble in Concentrating Changes in Visual Perception Headache

n % MI n % MI n % MI

Sham 2 11.8 2.5  0.5 0 0 0 6 35.3 2.2  0.54

Anodal 0 0 0 2 11.1 1.0  0 7 38.9 2  0.38
Nausea Vomiting Acute Sleeping Problems

n % MI (h) n % MI n % MI (d)

Sham 1 5.9 20 0 0 0 2 11.8 21

Anodal 0 0 0 0 0 0 4 22.2 3  0.71
Elevated Mood Feeling Cold Feeling Warm

n % MI (h) n % MI n % MI

Sham 2 11.8 3.5  1.5 1 5.9 0.25  0 2 11.8 1.5  0.5

Anodal 3 16.6 1.3  0.33 1 5.5 10 1 5.5 20
Others

1 Increment in the subjective good feeling

MI ¼ mean intensity (0e5); h ¼ post-tDCS hours; d ¼ post-tDCS days.

In terms of commonly used noninvasive excit- intracortical inhibition and enhanced facilita-
ability parameters, we have found a decreased tion.34 However, this effect was explained by
SICI after anodal tDCS compared with sham the fact that the aftereffects of tDCS, as well as
tDCS over M1 using the paired-pulse paradigm. intracortical inhibition and facilitation, are at
TDCS application had no significant effect on least partly modulated by NMDA receptor activ-
SICF, ICF, LICI, CSP, or motor-evoked recruit- ity.35 Interestingly, a previous study36 in which 1-
ment curves (for an overview of methods used or 10-Hz rTMS was applied over the M1 corre-
to study the modulation of human motor cortex sponding to the painful hand of patients with
excitability in local circuits, see Refs. 21,31). chronic neuropathic pain showed a significant
SICI is a low-threshold inhibition that can be eli- increase in ICI after 10-Hz rTMS, which was cor-
cited using paired-pulse TMS delivered with related with pain relief. However, in these pa-
short ISIs.12 Because the conditioning sub- tients, ICI was reduced at baseline compared
threshold stimulus during paired-pulse TMS with the measurements of healthy subjects.
neither recruits descending volleys in the corti- The authors concluded that chronic neuro-
cospinal tract32 nor alters spinal reflexes,15 it is pathic pain was associated with M1 disinhibi-
suggested that the inhibitory phenomena are tion, suggesting impaired GABAergic
produced by mechanisms acting only at the cor- neurotransmission related to underlying sen-
tical level. SICI can be mediated by gamma ami- sory or motor disturbances. From our present
nobutyric acid (GABAA) receptors. It was found study, it is difficult to conclude which neuro-
that tiagabine (a GABA-reuptake inhibitor) de- transmitter system was involved in the reduction
creases SICI.33 In a previous study, a single ses- of the SICI in our patient population; hence,
sion of anodal tDCS over the M1 reduced further studies are necessary to clarify this point.
902 Antal et al. Vol. 39 No. 5 May 2010

The limitations of this study need to be dis- stimulation in patients with thalamic pain. J Neuro-
cussed. First, this clinical trial was an exploratory surg 1993;78:393e401.
study; we had a heterogeneous patient group, 2. Meyerson BA, Lindblom U, Linderoth B,
the symptoms and the etiology of the chronic Lind G, Herregodts P. Motor cortex stimulation as
pain syndromes differed across patients, and treatment of trigeminal neuropathic pain. Acta
Neurochir Suppl (Wien) 1993;58:150e153.
our sample size might not have been large
enough to detect some characteristics associated 3. Wallace BA, Ashkan K, Benabid AL. Deep brain
with a positive effect of anodal tDCS. Nonethe- stimulation for the treatment of chronic, intractable
pain. Neurosurg Clin N Am 2004;15:343e357.
less, many of the previous rTMS studies have
investigated patients with different types of 4. Khedr EM, Kotb H, Kamel NF, et al. Long last-
ing antalgic effects of daily sessions of repetitive
chronic pain.5,6 Furthermore, in practice, it is transcranial magnetic stimulation in central and
the general pain symptoms of the patients that peripheral neuropathic pain. J Neurol Neurosurg
are targeted and treated, and not specific symp- Psychiatr 2005;76:833e838.
toms of pain, which are dependent upon the se- 5. Lefaucheur JP, Drouot X, Keravel Y, Nguyen JP.
lected groups of patients themselves, with Pain relief induced by repetitive transcranial mag-
symptoms often varying within a wide range netic stimulation of precentral cortex. Neuroreport
with regard to the intensity, quality and localiza- 2001;12:2963e2965.
tion. Second, our patients were allowed to ad- 6. Lefaucheur JP, Drouot X, Menard-Lefaucheur I,
here to their medication regimens throughout et al. Neurogenic pain relief by repetitive transcrani-
the trial; nevertheless, none of the patients re- al magnetic cortical stimulation depends on the
origin and the site of pain. J Neurol Neurosurg
quested to increase the dosage of their medica- Psychiatr 2004;75:612e616.
tion. Furthermore, the patients were taking
different types of medications in varying dos- 7. Rollnik JD, Wüstefeld S, Däuper J, et al. Repeti-
tive transcranial magnetic stimulation for the treat-
ages. Therefore, we have to conclude that med- ment of chronic painda pilot study. Eur Neurol
ication might confound our results. Because of 2002;48:6e10.
the exploratory nature of the study, the statistical 8. Fregni F, Boggio PS, Lima MC, et al. A
analysis used also implemented exploratory sham-controlled, phase II trial of transcranial direct
methods. current stimulation for the treatment of central
In summary, previous studies showed that pain in traumatic spinal cord injury. Pain 2006;
high-frequency rTMS is associated with a signifi- 122:197e209.
cant pain improvement compared with sham 9. Fregni F, Gimenes R, Valle AC, et al. A random-
rTMS.4e6,36 TDCS is a relatively new technology ized sham-controlled proof-of principle study of
that appears to stimulate the M1 in a way similar transcranial direct current stimulation for the treat-
ment of pain in fibromyalgia. Arthritis Rheum 2006;
to that of rTMS and epidural stimulation and 54:3988e3998.
can transiently reduce pain in some groups of
10. Nitsche MA, Cohen LG, Wassermann EM, et al.
patients with neuropathic pain.8,9 Although Transcranial direct current stimulation: state of the
the basic neuronal mechanisms of tDCS are art 2008. Brain Stimulat 2008;1:206e223.
probably different from those of rTMS, both
11. Lang N, Siebner HR, Ward NS, et al. How does
techniques might lead to a similar indirect transcranial DC stimulation of the primary motor
change of activity in connected areas and, cortex alter regional neuronal activity in the human
thus, result in similar effects on chronic pain. brain? Eur J Neurosci 2005;22:495e504.
12. Bindman LJ, Lippold OCJ, Redfearn JWT. The
action of brief polarizing currents on the cerebral
cortex of the rat (1) during current flow and (2)
Acknowledgment in the production of long-lasting after-effects.
The authors thank Leila Chaieb for her assis- J Physiol 1964;172:369e382.
tance with the English translation of this 13. Creutzfeldt OD, From GH, Kapp H. Influence
article. of transcortical DC currents on cortical neuronal ac-
tivity. Exp Neurol 1962;5:436e452.
14. Liebetanz D, Nitsche MA, Tergau F, Paulus W.
Pharmacological approach to the mechanisms of
References transcranial DC-stimulation-induced after-effects of
1. Tsubokawa T, Katayama Y, Yamamoto T, human motor cortex excitability. Brain 2002;125:
Hirayama T, Koyama S. Chronic motor cortex 2238e2247.
Vol. 39 No. 5 May 2010 Anodal Stimulation Improves Chronic Pain 903

15. Kujirai T, Caramia MD, Rotwell JC, et al. Corti- of the motor cortex. Acta Neurochir Suppl 1995;64:
cocortical inhibition in human motor cortex. J Phys- 132e135.
iol (Lond) 1993;471:501e519.
27. Nguyen JP, Lefaucheur JP, Decq P, et al. Chronic
16. Valls-Sole J, Pascual-Leone A, Wassermann EM, motor cortex stimulation in the treatment of central
Hallet M. Human motor evoked responses to paired and neuropathic pain. Correlations between clini-
transcranial magnetic stimuli. Electroencephalogr cal, electrophysiological and anatomical data. Pain
Clin Neurophysiol 1992;85:355e364. 1999;82:245e251.
17. Poreisz C, Boros K, Antal A, Paulus W. Safety as- 28. Nuti C, Peyron R, Garcia-Larrea L, et al. Motor
pects of transcranial direct current stimulation con- cortex stimulation for refractory neuropathic pain:
cerning healthy subjects and patients. Brain Res four year outcome and predictors of efficacy. Pain
Bull 2007;72:208e214. 2005;118:43e52.
18. Nitsche MA, Paulus W. Sustained excitability 29. Ruck A, Sylven C. ‘‘Improvement’’ in the pla-
elevations induced by transcranial DC motor cortex cebo group could be due to regression to the
stimulation in humans. Neurology 2001;57: mean as well as to sociobiologic factors. Am J Cardi-
1899e1901. ol 2006;97:152e153.
19. Rothwell JC, Hallett M, Berardelli A, et al. Mag-
netic stimulation: motor evoked potentials: the 30. Strafella AP, Vanderwerf Y, Sadikot AF. Transcra-
International Federation of Clinical Neurophysiol- nial magnetic stimulation of the human motor cor-
ogy. Electroencephalogr Clin Neurophysiol Suppl tex influences the neuronal activity of subthalamic
1999;52:97e103. nucleus. Eur J Neurosci 2004;20:2245e2249.

20. Ziemann U, Tergau F, Wassermann EM, et al. 31. Paulus W, Classen J, Cohen LG, et al. State of
Demonstration of facilitatory I wave interaction in the art: pharmacologic effects on cortical excitabil-
the human motor cortex by paired transcranial ity measures tested by transcranial magnetic stimula-
magnetic stimulation. J Physiol 1998;511:181e190. tion. Brain Stimulat 2008;1:151e163.
21. Ziemann U, Paulus W, Nitsche MA, et al. Con- 32. Di Lazzaro V, Rothwell JC, Oliviero A, et al. In-
sensus: motor cortex plasticity protocols. Brain tracortical origin of the short latency facilitation
Stimulat 2008;1:164e182. produced by pairs of threshold magnetic stimuli ap-
plied to human motor cortex. Exp Brain Res 1999;
22. Bolton JE, Wilkinson RC. Responsiveness of 129:494e499.
pain scales: a comparison of three pain intensity
measures in chiropractic patients. J Manipulative 33. Werhahn KJ, Kunesch E, Noachtar S,
Physiol Ther 1998;21:1e7. Benecke R, Classen J. Differential effects on motor-
23. Furubayashi T, Terao Y, Arai N, et al. Short and cortical inhibition induced by blockade of GABA
long duration transcranial direct current stimula- uptake in humans. J Physiol 1999;517:591e597.
tion (tDCS) over the human hand motor area. 34. Nitsche MA, Seeber A, Frommann K, et al. Mod-
Exp Brain Res 2008;185:279e286. ulating parameters of excitability during and after
24. Carroll D, Joint C, Maartens N, et al. Motor cor- transcranial direct current stimulation of the hu-
tex stimulation for chronic neuropathic pain: a pre- man motor cortex. J Physiol 2005;568:291e303.
liminary study of 10 cases. Pain 2000;84:431e437. 35. Ziemann U, Chen R, Cohen LG, Hallett M. Dex-
25. Ebel H, Rust D, Tronnier V, Boker D, Kunze S. tromethorphan decreases the excitability of the hu-
Chronic precentral stimulation in trigeminal neuro- man motor cortex. Neurology 1998;51:1320e1324.
pathic pain. Acta Neurochir 1996;138:1300e1306. 36. Lefaucheur JP, Drouot X, Ménard-Lefaucheur I,
26. Herregodts P, Stadnik T, De Ridder F, Keravel Y, Nguyen JP. Motor cortex rTMS restores
D’Haens J. Cortical stimulation for central neuro- defective intracortical inhibition in chronic neuro-
pathic pain: 3-D surface MRI for easy determination pathic pain. Neurology 2006;67:1568e1574.