Exp Brain Res (2005) 166: 23–30
DOI 10.1007/s00221-005-2334-6
R ES E AR C H A RT I C L E
Felipe Fregni Æ Paulo S. Boggio Æ Michael Nitsche
Felix Bermpohl Æ Andrea Antal Æ Eva Feredoes
Marco A. Marcolin Æ Sergio P. Rigonatti
Maria T.A. Silva Æ Walter Paulus
Alvaro Pascual-Leone
Anodal transcranial direct current stimulation of prefrontal cortex
enhances working memory
Received: 26 November 2004 / Accepted: 22 February 2005 / Published online: 6 July 2005
Springer-Verlag 2005
Abstract Previous studies have claimed that weak
transcranial direct current stimulation (tDCS) induces
persisting excitability changes in the human motor cortex that can be more pronounced than cortical modulation induced by transcranial magnetic stimulation, but
there are no studies that have evaluated the effects of
tDCS on working memory. Our aim was to determine
whether anodal transcranial direct current stimulation,
which enhances brain cortical excitability and activity,
would modify performance in a sequential-letter working memory task when administered to the dorsolateral
prefrontal cortex (DLPFC). Fifteen subjects underwent
a three-back working memory task based on letters. This
task was performed during sham and anodal stimulation
applied over the left DLPFC. Moreover seven of these
subjects performed the same task, but with inverse
polarity (cathodal stimulation of the left DLPFC) and
anodal stimulation of the primary motor cortex (M1).
Our results indicate that only anodal stimulation of the
left prefrontal cortex, but not cathodal stimulation of
left DLPFC or anodal stimulation of M1, increases the
accuracy of the task performance when compared to
Felipe Fregni and Paulo S. Boggio contributed equally to this
work.
F. Fregni (&) Æ F. Bermpohl Æ A. Pascual-Leone
Harvard Center for Non-invasive Brain Stimulation,
Beth Israel Deaconess Medical Center, Harvard Medical School,
330, Brookline Avenue, KS 452., Boston, MA 02215, USA
E-mail: ffregni@bidmc.harvard.edu
Tel.: +1-617-6675272
Fax: +1-617-9755322
P. S. Boggio Æ M. T. Silva
Department of Experimental Psychology, Institute of Psychology,
University of Sao Paulo, Sao Paulo, Brazil
M. Nitsche Æ A. Antal Æ E. Feredoes Æ W. Paulus
Department of Clinical Neurophysiology,
Georg-August-University, Goettingen, Germany
M. A. Marcolin Æ S. P. Rigonatti
Department of Psychiatry, University of Sao Paulo, Sao Paulo,
Brazil
sham stimulation of the same area. This accuracy
enhancement during active stimulation cannot be accounted for by slowed responses, as response times were
not changed by stimulation. Our results indicate that left
prefrontal anodal stimulation leads to an enhancement
of working memory performance. Furthermore, this
effect depends on the stimulation polarity and is specific
to the site of stimulation. This result may be helpful to
develop future interventions aiming at clinical benefits.
Keywords Electrical stimulation Æ Prefrontal cortex Æ
Transcranial magnetic stimulation Æ Working memory
Introduction
Despite it being an old technique to stimulate the brain,
not much is known about the behavioral effects of
transcranial direct current stimulation (tDCS) in
humans. Several animal studies carried out in the past
(Bindman et al. 1964; Purpura and McMurtry 1965)
showed that this method of brain stimulation has strong
effects on brain activity and excitability. The recent
development and the study of other methods of brain
stimulation, particularly transcranial magnetic stimulation (TMS), have placed the tDCS in the research agenda
of brain stimulation once more. Recently, a number of
studies using tDCS in humans have been published
(Nitsche and Paulus 2001; Nitsche et al. 2003a, b, 2004;
Antal et al. 2004a). These studies have shown that this
technique can be safely used in human beings.
In tDCS, the cerebral cortex is stimulated through a
weak constant electric current in a non-invasive and
painless manner. This weak current can induce focal
changes of cortical excitability—increase or decrease
depending on the electrode polarity—that lasts beyond
the period of stimulation. Several studies have shown
that this technique might modulate cortical excitability
in the human motor cortex (Nitsche and Paulus 2000;
24
Rosenkranz et al. 2000; Baudewig et al. 2001) and visual
cortex (Antal et al. 2001, 2004a). Recent studies have
demonstrated a beneficial effect of excitability-enhancing
anodal DC stimulation on simple reaction times and
implicit motor learning when the primary motor cortex
was stimulated (Nitsche et al. 2003c), as well as
improved learning of a visuo-motor coordination task
by stimulation of the primary motor area or the visual
area V5 (Antal et al. 2004b). Moreover, frontopolar
stimulation enhanced probabilistic classification learning (Kincses et al. 2004). Thus anodal tDCS appears to
improve cognitive functions in humans, and it has been
proposed that this cognition enhancement might be
accomplished by its strengthening effects on glutamatergic synapses. The effects are particularly intriguing,
given that subjects can indeed be blinded as to the nature
of the stimulation, anodal, cathodal or sham, given the
lack of associated perceptions. Therefore, the aim of the
present investigation was to study the effects of tDCS on
working memory, which can be considered a paradigmatic case of cognitive functioning.
Working memory refers to temporary storage and
manipulation of the information necessary for complex
tasks such as language comprehension, learning and
reasoning. Neuroimaging studies have shown that prefrontal cortex, particularly the dorsolateral prefrontal
cortex (DLPFC) (Brodmann areas 9 and 46) plays a
crucial role during working memory tasks (D’Esposito
et al. 1998; Mottaghy et al. 2000). Studies using electroencephalogram (EEG) have demonstrated a theta
coupling in the DLPFC during working memory tasks
(Sauseng et al. 2004) and temporary disruption of the
activity of the DLPFC by TMS that can lead to performance deterioration in different working memory
tasks (Grafman et al. 1994; Pascual-Leone and Hallett
1994; Jahanshahi et al. 1998; Mottaghy et al. 2000; Mull
and Seyal 2001). However, although these studies deliver
convincing evidence that the DLPFC is involved in
working memory, these techniques, in a strict sense, allow no definite conclusion about the specific involvement
of this cortical area in these processes. For example,
changes of brain activation and EEG modifications
could be an epiphenomena and a disruption of cortical
processing, as delivered by TMS, could diminish performance by disturbing working memory storage or just
performance. Therefore, tDCS has an advantage over
these techniques, as this method demonstrates a causal
link between the stimulated area and behavior—which is
deficient in neuroimaging studies—and does not disrupt
cortical processing. Although tDCS does not have the
same spatial resolution as TMS, the potential enhancement of cortical function by tDCS may provide further
evidence of the association between the DLPFC and
working memory, thus strengthening this relationship.
In addition, an enhancement of working memory, although only short-lived and on-line, might provide insights that may lead to further studies of this technique
exploring working memory function in healthy subjects
and patients with disturbed working memory.
The aim of this study was to investigate the effects of
anodal stimulation of the DLPFC on working memory.
We postulated that the stimulation would improve task
performance if the DLPFC is critically involved in
working memory formation and a cortical activity
enhancement is important for this process, as suggested
by neuroimaging studies. Moreover, this study will be
important to increase our knowledge about the behavioral effects induced by tDCS because this is the first
study to test the effects of this stimulation technique on
DLPFC function.
Materials and methods
Subjects
Fifteen healthy human subjects (11 females) were tested.
The age range was 19–22 years (mean 20.2 years). All
participants were right-handed. All subjects were college
students, thus all shared the same level of education.
Seven (six females) out of these 15 subjects participated
in an additional control experiment. Subjects gave
informed consent and the local Human Subjects Review
Committee approved the study, which was conducted in
strict adherence to the Declaration of Helsinki.
Direct current stimulation
Direct current was transferred by a saline-soaked pair of
surface sponge electrodes (35 cm2) and delivered by a
specially developed, battery-driven, constant current
stimulator (Schneider Electronic, Gleichen, Germany)
with a maximum output of 10 mA. To stimulate the
DLPFC, the anode electrode was placed over F3
according to the 10–20 international system for EEG
electrode placement. This method of DLPFC localization has been used before in TMS studies (Gerloff et al.
1997; Rossi et al. 2001), and has been confirmed as a
relatively accurate method of localization by neuronavigation techniques (Herwig et al. 2003). The cathode was
placed over the contralateral supraorbital area.
Although neuroimaging (D’Esposito et al. 1998; Smith
and Jonides 1999) and TMS studies (Mottaghy et al.
2000) have demonstrated that right and left DLPFC are
involved in working memory paradigms, we decided to
focus our investigation on the left DLPFC, as the
modulation of this area by rapid repetitive transcranial
magnetic stimulation (rTMS) (off-line rTMS) can cause
an improvement in some aspects of the cognitive function in patients with major depression (Padberg et al.
1999; Moser et al. 2002; Martis et al. 2003) and Parkinson’s disease (Boggio et al. 2005). Therefore, we
planned to test if on-line tDCS can also improve one
aspect of the cognition, working memory, in normal
subjects. For the control experiment, the position of
electrodes was changed (see ‘‘Control experiment’’). A
constant current of 1 mA intensity was applied for
25
10 min. Subjects felt the current as an itching sensation
at both electrodes at the beginning of the stimulation.
For sham stimulation, the electrodes were placed in the
same position; however, the stimulator was turned off
after 5 s as previously described (Siebner et al. 2004).
Therefore, the subjects felt the initial itching sensation in
the beginning, but received no current for the rest of the
stimulation period. This procedure allowed to blind
subjects for the respective stimulation condition (Nitsche
et al. 2003a).
Working memory assessment
We used the three-back letter working memory paradigm described elsewhere (Mull and Seyal 2001). Subjects were presented with a pseudo-random set of ten
letters (A J). The stimuli were generated using the
Superlab pro v2.0 software (Cedrus Corporation, San
Pedro, Calif., USA). Each letter was displayed on
computer monitor for 30 ms. A different letter was displayed every 2 s. Black letters were presented on a white
background and subtended 2.4 cm (when viewed at
50 cm). Subjects were required to respond (key press) if
the presented letter was the same as the letter presented
three stimuli previously (Fig. 1). In this test, a total of 30
correct responses were possible. In each set of this task,
the targets could be separated by three to five letters.
Subjects were allowed to practice the task for 20 min or
until they obtained an accuracy of ‡50%.
Experimental protocol (main experiment)
Following a first practice run, subjects were tested during sham and active stimulation. Since the test run lasted
5 min, it was delivered during the last 5 min of active
and sham stimulation (Fig. 2). The two test runs differed
in the order of the letters and were randomized across
subjects to avoid difficulty bias. To avoid carryover
effects, the order of active versus sham stimulation was
Fig. 1 The sequence of the
3-back letter working memory
paradigm. Note that subjects
were required to respond (key
press) if the presented letter was
the same as the letter presented
three stimuli previously
Fig. 2 The experimental protocol design. Each subject was tested
during sham and active stimulation. The two tests runs were
randomized within subject and the order (active versus sham
stimulation) was counterbalanced across subjects
fully counterbalanced across subjects, such that seven
subjects received first active stimulation and eight subjects received first sham stimulation. In addition, each
condition was separated by at least 1 h to washout the
effects of the previous run. Subjects could not distinguish between real and sham stimulation as they felt the
initial itching in both conditions.
Control experiment
In order to test if the anodal stimulation of the left
DLPFC was indeed responsible for the observed effects,
seven out of the 15 subjects that participated in the main
experiment were enrolled in a control experiment. This
control experiment was carried out 6 months after the
main experiment. In this control experiment, we tested: (1)
whether the effects of the tDCS on DLPFC were focal and
(2) whether the effects of tDCS on DLPFC were dependent on polarity (anodal versus cathodal stimulation).
To test aim (1) (focality of tDCS), subjects underwent
an identical study protocol; however, with the anodal
electrode placed over the primary motor cortex (M1)
26
rather than left DLPFC. The cathodal electrode was
again placed on the right supraorbital area. To test aim
(2) (polarity of tDCS), the same experimental design as
in the main experiment was used, however, with inverted
electrode polarity. The anode was placed over the right
supraorbital area and the cathode over the left DLPFC.
These subjects also underwent sham tDCS. Therefore, in the control experiment, three different types of
stimulation (anodal M1 stimulation, cathodal left
DLPFC stimulation and sham stimulation) were
applied. We used the same experimental design: 10 min
of 1 mA of tDCS (on-line test in the final 5 min of
stimulation). The order of these three conditions was
randomized and counterbalanced across subjects. The
washout period was 1 h.
Data analysis
The primary outcomes for this study were number of
correct responses, false alarms (errors) and response time
during active compared to sham stimulation. Analyses
were done with SAS statistical software (version 8.0,
Cary, N.C., USA). We used the Shapiro-Wilk test to
evaluate whether the data were normally distributed. The
results from this test for the data from reaction time
(W=0.94, P=0.39), correct answers (W=0.89, P=0.08)
and errors (W=0.91, P=0.11) show that the null
hypothesis (sample is taken from a population with normal distribution) should not be rejected; therefore these
data are normally distributed. Assuming normal distribution, paired Student’s t-test was used to compare each
pair of results (response time, number of errors and correct answers). Paired t-test, rather than two independent
samples t-test, was used as data are dependent—each
subject was measured after two different interventions
(active and sham stimulation). Repeated measures of
analysis of variance (ANOVA) was performed to investigate if there was an order effect between sham and active
stimulation. This two-way ANOVA assessed the main
effect of type of stimulation (active versus sham) and
order of stimulation (active first or sham first). Statistical
significance refers to a two-tailed P-value <0.05.
All subjects completed the entire experiment. One
important observation is that there were no side effects
observed throughout the experiment. All subjects tolerated the treatment well, and there was no complaint of
pain or any uncomfortable symptom during the stimulation. All subjects reported that they could not feel the
difference between the active and sham conditions and
forced guessing was at chance level.
Correct responses (main experiment)
Subjects had significantly more correct answers during
active condition when compared to the sham stimulation
(t=3.4, df=14, P=0.0042). The mean number of correct
responses during the sham stimulation was 19.8 (±5.8
SD) whereas the mean number of correct answers during
active stimulation was 21.7 (±5.0 SD), and the mean
difference between these two types of stimulation was 1.92
(±2.18 SD) (Fig. 3). In order to test if the order effect was
significant, a two-way ANOVA (stimulation type versus
order) was performed. This analysis disclosed that there
was no order effect (F=1.77, df=1,13, P=0.21), but only
a stimulation effect (F=12.85, df=1,13; P=0.003). This
finding confirmed that the order of stimulation did not
influence our results.
Errors (main experiment)
Subjects could make two types of errors when performing this task. They could either omit the correct
response or press the response key at a wrong time. We
used only this last variable—designated as false alarms—to compute total errors as omissions are implicitly
analyzed under correct responses. This analysis showed
that subjects made significantly more errors during sham
condition when compared to the active stimulation
(t=2.77, df=14, P=0.015). The mean number of errors
during the sham stimulation was 6.9 (±6.1 SD), whereas
Correct responses
25
N u m be r of cor r e ct r e spon se s
Fig. 3 Number of correct
responses during each
stimulation condition (active
and sham). Dark bar indicates
mean number of correct
responses during sham
stimulation. White bar
represents mean number of
correct responses during active
stimulation. There was a
significant difference in the
mean number of correct
response between sham and
active stimulation. Error bars
indicate ±SEM (standard error
of the mean)
Results
24
23
22
21
20
19
18
17
16
15
Sham
Active
27
Errors
9
Sham
8
Active
7
Number of errors
Fig. 4 Number of errors during
each stimulation condition
(active and sham). Dark bar
indicates mean number of
errors during sham stimulation.
White bar represents mean
number of errors during active
stimulation. There was a
significant difference in the
mean number of errors between
sham and active stimulation.
Error bars indicate ±SEM
(standard error of the mean)
6
5
4
3
2
1
0
the mean number of errors during active stimulation was
4.7 (±4.7 SD) and the mean difference between these
two types of stimulation was 2.2 (±3.1 SD) (Fig. 4). In
order to test if the order effect was significant, a two-way
ANOVA (stimulation versus order) was performed. This
analysis showed that there was no order effect (F=0.88,
df=1,13, P=0.36), but only a stimulation effect
(F=7.93, df=1,13; P=0.014). This finding confirmed
that the order of stimulation did not influence our results. Indeed, Fig. 5 shows that subjects that received
active stimulation first had larger improvement than
subjects that received active stimulation second
(although this difference was not significant). This result
speaks against a carryover or learning effect.
Response time (main experiment)
There was no significant difference in the mean response
time between the active and sham conditions (t=0.04;
df=14; P=0.97). The mean response time during the
active condition was 573.0 ms (±160.3 SD) whereas the
mean response time during the sham condition was
572.4 ms (±126.7 SD) (Fig. 6).
To test if the effects of left DLPFC tDCS were focal,
anodal stimulation of the primary motor cortex (M1)
was performed. The results showed that there was no
significant difference between anodal stimulation of the
primary motor cortex and sham stimulation regarding
number of correct response (P=0.70), errors (P=0.46)
and reaction time (P=0.71) (Fig. 7).
In order to test if the effects of left DLPFC tDCS
depend on polarity, we inverted the electrodes polarity:
cathode electrode was placed over left DLPFC and
anode electrode over the right supraorbital area. Similarly to the other control experiment, results from this
stimulation revealed that there was no significant difference between cathodal stimulation of the left DLPFC
and sham stimulation regarding number of correct response (P=0.67), errors (P=0.64) and reaction time
(P=0.72) (Fig. 7).
Discussion
Our results indicate that anodal stimulation of the left
prefrontal area increases the accuracy of the task
Active stimulation first
Active stimulation second
0.45
0.4
compared to sham stimulation
Improvement rate (%) during active stimulation
Fig. 5 This chart shows that
subjects that received active
stimulation first (black column)
had larger improvement
(compared to sham stimulation)
than subjects that received
active stimulation second (white
column). However, this
difference was not statistically
significant
Control experiment
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Correct responses
Errors
28
Response time
700
Sham
600
Active
500
RT (msec)
Fig. 6 Mean response time
during each stimulation
condition (active and sham).
Dark bar indicates mean
response time during sham
stimulation. White bar
represents mean response time
during active stimulation. There
was no significant difference in
the mean response time between
sham and active stimulation.
Error bars indicate ±SEM
(standard error of the mean)
400
300
200
100
0
performance when compared with sham stimulation of
the same area. Furthermore, our control experiment
showed that the effect of the anodal stimulation of the
left DLPFC was relatively focal and depends on the
polarity of stimulation. This accuracy enhancement
during active stimulation cannot be accounted for by
slowed responses, as response times were not changed by
stimulation. These results showed that left prefrontal
anodal stimulation leads to an enhancement of working
memory performance.
It is interesting to note that opposite effects were
demonstrated when the stimulation was performed with
TMS. Although TMS can modulate the activity of a
given cortical area, this technique transiently disrupts
brain activity during the period of stimulation and hence
creates a temporary ‘‘virtual lesion’’ (Pascual-Leone
et al. 1999). Mull and Seyal (2001) and Mottaghy et al.
(2000) showed that on-line single pulse TMS and 1 Hzrepetitive TMS, respectively, applied over the left
DLPFC resulted in an increase of task errors compared
to the control condition (Mottaghy et al. 2000; Mull and
Seyal 2001). This degradation in task performance is
likely related to transient disruption of the left DLPFC
information processing caused by TMS. Opposed to
Sham tDCS
Performance improvement (%) from sham tDCS
Fig. 7 Results from the two
control experiments.
Normalized average
performance of working
memory during sham, anodal
primary motor cortex (M1) and
cathodal left dorsolateral
prefrontal cortex (DLPFC)
stimulation with tDCS.
Normalized values were
obtained setting the
performance during sham tDCS
to 100( (black column). There
was no significant difference in
working memory performance
during either anodal M1 (white
column) or cathodal DLPFC
(gray column) stimulation when
compared with sham
stimulation (black column).
Error bars indicate SEM
(standard error of the mean)
these findings, the present study showed that tDCS
causes no degradation in task performance during anodal stimulation. On the contrary, on-line anodal tDCS
causes an enhancement in working memory. This
observation is important as it might indicate that tDCS
can stimulate the brain in a different way compared to
TMS. The main characteristic that could underlay this
difference is the amount of electric current involved in
these two techniques. While TMS is likely to elicit neuronal depolarizations and induction of action potentials,
tDCS as applied here only causes a slight change in the
resting potential of the stimulated cells (Creutzfeldt et al.
1962), and thus may not disrupt but rather improve
information processing by bringing neurons closer to
depolarization thresholds in response to appropriate
inputs.
The results of this study underscore the importance of
cortical excitability and activity enhancements for
working memory function, as suggested by previous
functional imaging studies. Moreover, since accuracy,
but not response times, differed between the stimulation
conditions, the results are in accordance with a critical
role of the DLPFC in working memory formation rather
than simply task execution.
Anodal M1
Cathodal DLPFC
160
140
120
100
80
60
40
20
0
Correct responses
Errors
Reaction time
29
The results of this study extend the findings of a
previous investigation that showed a modification of
implicit probabilistic classification learning by weak
anodal tDCS (Kincses et al. 2004). It appears that tDCS
may be an interesting tool to enhance some aspects of
cognition. One important issue to consider is a possible
carryover effect. It has been demonstrated that the
effects of 11 min of 1.0 mA tDCS of the motor cortex
can shift excitability for up to 60 min (Nitsche and Paulus 2001). We allowed for at least 1 h of washout period between test conditions, but the effects of tDCS on
the prefrontal cortex might last longer than those on
motor cortex. However, we counterbalanced stimulation
conditions across subjects, and failed to observe order
effects. In addition, subjects that received tDCS first had
larger improvement, though not significant, compared
to the group of subjects that received tDCS second,
indicating that carryover and learning effects were not
likely responsible for the results. If there was carryover
or learning effect, we would expect an opposite effect:
the task performance during sham tDCS when performed after active tDCS (active tDCS first) would have
been similar or even larger than during active tDCS.
Another important concern is a potential learning effect
as the task was given repeatedly and a learning curve
could distort the results. We addressed this issue allowing the subjects to practice the test. In addition, we
compared the difference in the performance between
sham and active stimulation. As these two conditions
were counterbalanced, a learning effect would have
affected both groups to a similar extent.
It has been shown that the effect of tDCS on brain
activity seems to depend on the stimulation polarity, i.e.
whereas anodal stimulation hyperpolarizes brain tissue,
cathodal stimulation has the opposite effect (Nitsche
et al. 2003a). Our control experiment confirms behaviorally that the effects of tDCS depend on polarity.
While anodal stimulation of the DLPFC enhanced
working memory, cathodal stimulation of the same area
caused no changes. However, according to the anodal/
cathodal opposite effects, we would expect worsening of
working memory after cathodal stimulation. Nevertheless, we only observed a significant effect related to
anodal stimulation. We speculate that the lack of effects
of cathodal stimulation may be explained by two factors. First, the sample size of the control experiment was
small, although there was not even a statistical trend
that could suggest a possible effect masked by the small
sample size. Second, our test may not have been adequate to detect behavior worsening.
Because the technique of tDCS uses large electrodes
(35 cm2) and the electric current passes throughout the
brain, one can argue that the tDCS effects on the left
DLPFC might be diffuse and involve a larger area of the
left hemisphere. However, our control experiment
revealed that the effects of tDCS were relatively focal.
Anodal stimulation of the primary motor cortex did not
cause any significant effect on working memory. Due to
the fact that we maintained the cathodal electrode over
the right supraorbital area, this finding also ruled out the
alternative explanation that cathodal stimulation to the
right frontopolar cortex accounts for the working
memory improvement, unless the anodal stimulation of
the left primary motor cortex was worsening working
memory (and thus counterbalancing the effects of cathodal stimulation of supraorbital area); however, this
explanation is improbable, because it is unclear how
such an inhibition of neural activity of the motor area
would lead to working memory impairment. This finding is also in accordance with the study of Uy and
Ridding (2003). In this study, the effects of tDCS on the
motor cortex had a remarkably good spatial resolution:
anodal stimulation to the motor cortex representation of
the FDI (first dorsal interosseous) muscle did not produce any effect in nearby muscles such as ADM
(abductor Digiti Minimi) and FCU (Flexor Carpi Ulnaris) (Uy and Ridding 2003). However, this study shows
the effects for the motor cortex, and thus the same
specificity may not automatically be transferred to the
dorsolateral prefrontal cortex.
One important methodological consideration should
be entertained. Our findings could have been further
validated by varying the working memory load (modifying the n in the n-back task). We decided to use a more
challenging three-back task (Mull and Seyal 2001),
because the degree of difficulty of a test is related to the
likelihood to detect degradation or improvement in the
brain function following tDCS. An easier task might not
have detected subtle behavioral effects due to ‘‘ceiling’’
effect, while a more difficult version may have obscured
performance disruption due to the ‘‘floor’’ effect. Testing
different working memory load in the same task (as in a
traditional Sternberg paradigm) would have increased
the duration of the stimulation, possibly raising safety
concerns. Safety studies to date have only evaluated the
effects of less than 10 min of tDCS.
The results of the present study indicate that working
memory performance, in a sequential-letter-matching
task, is enhanced by anodal tDCS of the left prefrontal
cortex. Although the aim of this study was not to
explore the therapeutic effects of tDCS on working
memory, these results should encourage further investigations for the use of tDCS in clinical applications.
Acknowledgements This work was supported by a grant within the
Harvard Medical School Scholars in Clinical Science Program
(NIH K30 HL04095-03) to F.F.; a grant within the Postdoc-Programme of the German Academic Exchange Service (DAAD, D/
02/46858) to F.B.; and by K24 RR018875 to A.P.-L. The authors
would like to thank Barbara Bonetti for the invaluable administrative support and to Adriana L. Vieira, Elizabeth M. Saade,
Carolina R.B. Souza and Patricia Otachi for the help on data
acquisition.
References
Antal A, Nitsche MA, Paulus W (2001) External modulation of
visual perception in humans. Neuroreport 12:3553–3555
30
Antal A, Kincses TZ, Nitsche MA, Bartfai O, Paulus W (2004a)
Excitability changes induced in the human primary visual cortex by transcranial direct current stimulation: direct electrophysiological evidence. Invest Ophthalmol Vis Sci 45:702–707
Antal A, Nitsche MA, Kruse W, Kincses TZ, Hoffmann KP,
Paulus W (2004b) Direct current stimulation over V5 enhances
visuomotor coordination by improving motion perception in
humans. J Cogn Neurosci 16:521–527
Baudewig J, Nitsche MA, Paulus W, Frahm J (2001) Regional
modulation of BOLD MRI responses to human sensorimotor
activation by transcranial direct current stimulation. Magn
Reson Med 45:196–201
Bindman LJ, Lippold OC, Redfearn JW (1964) The action of brief
polarizing currents on the cerebral cortex of the rat (1) during
current flow and (2) in the production of long-lasting aftereffects. J Physiol 172:369–382
Boggio PS, Fregni F, Bermpohl F, Mansur CG, Rosa M, Rumi
DO, Barbosa E, Odebrecht-Rosa M, Marcolin MA, Silva MTA
(2005) The effect of repetitive TMS and fluoxetine on cognitive
function in patients with Parkinson’s disease and concurrent
depression. Mov Disord (in press). DOI 10.1002/mds.20508
Creutzfeldt OD, Fromm GH, Kapp H (1962) Influence of transcortical d-c currents on cortical neuronal activity. Exp Neurol
5:436–452
D’Esposito M, Aguirre GK, Zarahn E, Ballard D, Shin RK, Lease
J (1998) Functional MRI studies of spatial and nonspatial
working memory. Brain Res Cogn Brain Res 7:1–13
Gerloff C, Corwell B, Chen R, Hallett M, Cohen LG (1997)
Stimulation over the human supplementary motor area interferes with the organization of future elements in complex motor
sequences. Brain 120(Pt 9):1587–1602
Grafman J, Pascual-Leone A, Alway D, Nichelli P, Gomez-Tortosa E, Hallett M (1994) Induction of a recall deficit by rapidrate transcranial magnetic stimulation. Neuroreport 5:1157–
1160
Herwig U, Satrapi P, Schonfeldt-Lecuona C (2003) Using the
international 10–20 EEG system for positioning of transcranial
magnetic stimulation. Brain Topogr 16:95–99
Jahanshahi M, Profice P, Brown RG, Ridding MC, Dirnberger G,
Rothwell JC (1998) The effects of transcranial magnetic stimulation over the dorsolateral prefrontal cortex on suppression
of habitual counting during random number generation. Brain
121(Pt 8):1533–1544
Kincses TZ, Antal A, Nitsche MA, Bartfai O, Paulus W (2004)
Facilitation of probabilistic classification learning by transcranial direct current stimulation of the prefrontal cortex in the
human. Neuropsychologia 42:113–117
Martis B, Alam D, Dowd SM, Hill SK, Sharma RP, Rosen C,
Pliskin N, Martin E, Carson V, Janicak PG (2003) Neurocognitive effects of repetitive transcranial magnetic stimulation in
severe major depression. Clin Neurophysiol 114:1125–1132
Moser DJ, Jorge RE, Manes F, Paradiso S, Benjamin ML, Robinson RG (2002) Improved executive functioning following
repetitive transcranial magnetic stimulation. Neurology
58:1288–1290
Mottaghy FM, Krause BJ, Kemna LJ, Topper R, Tellmann L, Beu
M, Pascual-Leone A, Muller-Gartner HW (2000) Modulation
of the neuronal circuitry subserving working memory in healthy
human subjects by repetitive transcranial magnetic stimulation.
Neurosci Lett 280:167–170
Mull BR, Seyal M (2001) Transcranial magnetic stimulation of left
prefrontal cortex impairs working memory. Clin Neurophysiol
112:1672–1675
Nitsche MA, Paulus W (2000) Excitability changes induced in the
human motor cortex by weak transcranial direct current stimulation. J Physiol 527(Pt 3):633–639
Nitsche MA, Paulus W (2001) Sustained excitability elevations
induced by transcranial DC motor cortex stimulation in
humans. Neurology 57:1899–1901
Nitsche MA, Liebetanz D, Antal A, Lang N, Tergau F, Paulus W
(2003a) Modulation of cortical excitability by weak direct current stimulation—technical, safety and functional aspects.
Suppl Clin Neurophysiol 56:255–276
Nitsche MA, Liebetanz D, Lang N, Antal A, Tergau F, Paulus W
(2003b) Safety criteria for transcranial direct current stimulation (tDCS) in humans. Clin Neurophysiol 114:2220–2222
Nitsche MA, Schauenburg A, Lang N, Liebetanz D, Exner C,
Paulus W, Tergau F (2003c) Facilitation of implicit motor
learning by weak transcranial direct current stimulation of the
primary motor cortex in the human. J Cogn Neurosci 15:619–
626
Nitsche MA, Liebetanz D, Schlitterlau A, Henschke U, Fricke K,
Frommann K, Lang N, Henning S, Paulus W, Tergau F (2004)
GABAergic modulation of DC stimulation-induced motor
cortex excitability shifts in humans. Eur J Neurosci 19:2720–
2726
Padberg F, Zwanzger P, Thoma H, Kathmann N, Haag C,
Greenberg BD, Hampel H, Moller HJ (1999) Repetitive
transcranial magnetic stimulation (rTMS) in pharmacotherapyrefractory major depression: comparative study of fast, slow
and sham rTMS. Psychiatry Res 88:163–171
Pascual-Leone A, Hallett M (1994) Induction of errors in a delayed
response task by repetitive transcranial magnetic stimulation of
the dorsolateral prefrontal cortex. Neuroreport 5:2517–2520
Pascual-Leone A, Bartres-Faz D, Keenan J (1999) Transcranial
magnetic stimulation: studying the brain–behaviour relationship by induction of ‘‘virtual lesions’’. Philos Trans R Soc Lond
B Biol Sci 354:1229–1238
Purpura DP, McMurtry JG (1965) Intracellular activities and
evoked potential changes during polarization of motor cortex.
J Neurophysiol 28:166–185
Rosenkranz K, Nitsche MA, Tergau F, Paulus W (2000) Diminution of training-induced transient motor cortex plasticity by
weak transcranial direct current stimulation in the human.
Neurosci Lett 296:61–63
Rossi S, Cappa SF, Babiloni C, Pasqualetti P, Miniussi C, Carducci
F, Babiloni F, Rossini PM (2001) Prefrontal [correction of
Prefontal] cortex in long-term memory: an ‘‘interference’’ approach using magnetic stimulation. Nat Neurosci 4:948–952
Sauseng P, Klimesch W, Doppelmayr M, Hanslmayr S, Schabus
M, Gruber WR (2004) Theta coupling in the human electroencephalogram during a working memory task. Neurosci Lett
354:123–126
Siebner HR, Lang N, Rizzo V, Nitsche MA, Paulus W, Lemon
RN, Rothwell JC (2004) Preconditioning of low-frequency
repetitive transcranial magnetic stimulation with transcranial
direct current stimulation: evidence for homeostatic plasticity in
the human motor cortex. J Neurosci 24:3379–3385
Smith EE, Jonides J (1999) Storage and executive processes in the
frontal lobes. Science 283:1657–1661
Uy J, Ridding MC (2003) Increased cortical excitability induced by
transcranial DC and peripheral nerve stimulation. J Neurosci
Methods 127(24):193–197