Available online at www.sciencedirect.

com

Drug and Alcohol Dependence 92 (2008) 55–60

Prefrontal cortex modulation using transcranial DC stimulation reduces

alcohol craving: A double-blind, sham-controlled study
Paulo S. Boggio b , Natasha Sultani b , Shirley Fecteau a , Lotfi Merabet a ,
Tatiana Mecca b , Alvaro Pascual-Leone a , Aline Basaglia c , Felipe Fregni a,∗
a Center for Noninvasive Brain Stimulation, Beth Israel Deaconess Medical Center,
330 Brookline Avenue, KS 430, Harvard Medical School, Boston, MA 02215, USA
b Nucleo de Neurociencias, Mackenzie University, Sao Paulo, SP, Brazil
c Psychology Institute, University of Sao Paulo, Sao Paulo, Brazil

Received 30 January 2007; received in revised form 3 June 2007; accepted 11 June 2007
Available online 18 July 2007

Abstract
Background: Functional neuroimaging studies have shown that specific brain areas are associated with alcohol craving including the dorsolateral
prefrontal cortex (DLPFC). We tested whether modulation of DLPFC using transcranial direct current stimulation (tDCS) could alter alcohol
craving in patients with alcohol dependence while being exposed to alcohol cues.
Methods: We performed a randomized sham-controlled study in which 13 subjects received sham and active bilateral tDCS delivered to DLPFC
(anodal left/cathodal right and anodal right/cathodal left). For sham stimulation, the electrodes were placed at the same positions as in active
stimulation; however, the stimulator was turned off after 30 s of stimulation. Subjects were presented videos depicting alcohol consumption to
increase alcohol craving.
Results: Our results showed that both anodal left/cathodal right and anodal right/cathodal left significantly decreased alcohol craving compared
to sham stimulation (p < 0.0001). In addition, we found that following treatment, craving could not be further increased by alcohol cues.
Conclusions: Our findings showed that tDCS treatment to DLPFC can reduce alcohol craving. These findings extend the results of previous studies
using noninvasive brain stimulation to reduce craving in humans. Given the relatively rapid suppressive effect of tDCS and the highly fluctuating
nature of alcohol craving, this technique may prove to be a valuable treatment strategy within the clinical setting.
© 2007 Elsevier Ireland Ltd. All rights reserved.

Keywords: Alcohol craving; Brain stimulation; Transcranial direct current stimulation; Dorsolateral prefrontal cortex

1. Introduction (rTMS), a technique that can transiently modulate focal cor-

tical activity, show that high-frequency rTMS (i.e. excitability
Craving might be defined as a strong or inward desire. Dif- enhancing) delivered to DLPFC results in significant reductions
ferent types of craving have been reported including smoking, in smoking craving (Eichhammer et al., 2003), food craving
food, cocaine and alcohol craving. Recent neuroimaging stud- (Uher et al., 2005) and cocaine craving (Camprodon et al.,
ies have identified cortical areas associated with craving. Most 2007). Because alcohol craving is also associated with activ-
of these studies suggest that the prefrontal regions, particularly ity in DLPFC (Olbrich et al., 2006; George et al., 2001; Myrick
the dorsolateral prefrontal cortex (DLPFC), play a major role in et al., 2004), we decided to explore whether modulating corti-
craving associated with drugs and smoking (Olbrich et al., 2006; cal excitability in this area could change alcohol craving using
George et al., 2001; Myrick et al., 2004; Mcbride et al., 2006; another noninvasive method of brain stimulation namely, tran-
Wilson et al., 2004; Brody et al., 2002; Due et al., 2002; Grant scranial direct current stimulation (tDCS).
et al., 1996; Garavan et al., 2000; Sell et al., 2000). In addi- tDCS is a simple technique of noninvasive brain stimula-
tion, studies using repetitive transcranial magnetic stimulation tion in which a weak DC current is applied into the brain
for several minutes resulting in a polarity-dependent modula-
tion of brain activity. Studies in humans have demonstrated
∗ Corresponding author. Tel.: +1 617 667 5272; fax: +1 617 975 5322. that stimulation of the motor cortex changes its excitability
E-mail address: [email protected] (F. Fregni). depending on the stimulation polarity. Specifically, anodal stim-

0376-8716/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.drugalcdep.2007.06.011
56 P.S. Boggio et al. / Drug and Alcohol Dependence 92 (2008) 55–60

ulation increases cortical excitability while cathodal stimulation 2.2. Study protocol
decreases it (Nitsche and Paulus, 2000; Nitsche et al., 2003).
These excitability shifts during stimulation are believed to be due This study was a randomized, double-blind, sham-controlled, cross-over
study in which subjects received three different types of bilateral stimulation of
to subthreshold neuronal membrane depolarization (Nitsche and
DLPFC with tDCS: (1) active anodal left/cathodal right tDCS, (2) active anodal
Paulus, 2000; Liebetanz et al., 2002) as a result of the opening right/cathodal left tDCS and (3) sham tDCS. A 48-h inter-session interval was
or closing of voltage-gated ions induced by the electrical current used to avoid the potential of any carry-over effects due to stimulation. The
flow (Purpura and McMurtry, 1965; Nitsche et al., 2003) during order of stimulation was randomized and counterbalanced across subjects using
stimulation. On the other hand, the after-effects are shown to be a Latin square design. Participants and the evaluating investigators (except the
investigators that applied tDCS) were blinded to the treatment arm.
regulated by NMDA receptors (Nitsche et al., 2005; Liebetanz
All stimulation sessions were carried out by the same researchers and at
et al., 2002). the same time of the day. Demographic and alcohol habits profile data were
In this study, we stimulated the DLPFC using tDCS rather collected at baseline. The following instruments of evaluation were used:
than rTMS for several reasons. First, tDCS offers an advan-
tage in that the scalp sensation associated with stimulation only (1) For baseline assessment: (i) stages of change readiness and treatment eager-
lasts for a couple of seconds. Therefore, in a sham-controlled ness scale (SOCRATES) (Figlie et al., 2004); (ii) Short Alcohol Dependence
Data (SADD) questionnaire (Davidson and Raistrick, 1986).
trial, subjects can be adequately blinded to the condition they (2) To evaluate the effects of treatment: (i) alcohol urge questionnaire (AUQ)
are receiving (Gandiga et al., 2006). This characteristic is criti- (Drummond and Phillips, 2002); (ii) tDCS side-effect questionnaire (side-
cal given our cross-over study design. Second, tDCS is a simple, effect checklist) (Fregni et al., in press); (iii) Visual Analogue Scale for
safe and inexpensive technique and the device is highly portable. mood domains (Fregni et al., in press).
These characteristics make this treatment strategy highly appeal-
The following steps were then performed:
ing within the clinical setting.
Given that previous neuroimaging studies have demonstrated
(1) Baseline evaluation (T0): subjects were instructed to complete a visual
activity within the DLPFC in association with alcohol craving analogue scale (VAS) with 16 items evaluating mood and a questionnaire
(George et al., 2001; Wilson et al., 2004), we hypothesized that to measure alcohol craving (AUQ).
modulation of this area would alter alcohol craving. Further- (2) Cue-provoked craving: subjects were then exposed to alcohol cues to
more, a previous brain stimulation study carried out by our group increase craving that included viewing a video showing scenes of peo-
ple drinking in a pleasant way different types of alcoholic beverages, such
(Fregni et al., in press) has shown that tDCS delivered to both the
as beer, wine, vodka and ‘pinga’ (a popular local alcoholic beverage made
right and left hemisphere reduces craving for smoking. Based from sugar cane and has an ethanol concentration that ranges from 38 to
on these findings, we implemented a similar bilateral stimula- 48%). The video lasted for 5 min. Six different equivalent movies were
tion strategy in which we: (1) increased excitability in the left used and randomized across subjects. In this way, subjects were exposed to
and decreased it in the right DLPFC (anodal left/cathodal right) a different movie before and after the three conditions of treatment.
(3) Subjects were assessed again regarding their alcohol craving (T1).
and (2) increased excitability in the right and decreased it in the
(4) Subjects underwent tDCS treatment for 20 min (as detailed below).
left DLPFC (anodal right/cathodal left) following a randomized, (5) The procedure of the pre-treatment was repeated: initial craving evalua-
sham-controlled, double-blind, cross-over design. tion (T2), alcohol cues to increase craving (movie) and new assessment of
craving (T3) and also mood using VAS.
2. Methods
It is worthy of note that our urge-elicitation strategy might have increased
craving processing in a manner that differs slightly from other types of natu-
2.1. Study subjects
rally occurring craving. For instance, there is evidence showing that exposure
to drug cues in conjunction with drug availability may produce a more intense
We selected 13 subjects (mean age of 41.3 ± 5.7, two females) satisfying
craving than would occur without drug availability (Juliano and Brandon, 1998).
the diagnostic criteria for alcohol dependence defined by Diagnostic and Sta-
However, in this study, we were interested in craving associated with alcohol
tistical Manual of Mental Disorders (4th ed.) (DSM-IV) (American Psychiatric
abstinence and thus drinking after craving was never allowed. It is also important
Association, 1994). Patients were recruited from a specialized center for alcohol
to mention that we did not show a neutral cue to the subjects (such as a movie of
and substance abuse disorders in Santana do Parnaiba, outside of Sao Paulo,
non-alcoholic beverages); however it has been shown that craving as induced by
Brazil. All patients were enrolled in the clinic’s rehabilitation program and
alcoholic beverage images is effective in inducing alcohol craving as compared
were abstinent for 41.0 ± 51.3 days (a minimum of 10 days). Inclusion cri-
to neutral cues (George et al., 2001).
teria included aged between 30 and 55 years of age, alcohol dependence, and
a minimum drink history (14 drinks for females and 21 drinks for males) on
average per week for four consecutive weeks (during the period of drinking). 2.3. Transcranial direct current stimulation (tDCS)
Subjects were excluded if they had any current or past neuropsychiatric disorder,
substantial drug abuse other than alcohol, were taking any neuropsychotropic Direct current was transferred by using a saline-soaked pair of surface sponge
medications or were pregnant. electrodes and delivered by a custom developed, battery-driven, constant cur-
This study was performed at University of Mackenzie, Sao Paulo, Brazil. rent stimulator with a maximum output of 10 mA and electrode size of 35 cm2
Subjects gave written informed consent to participate in the study and approval (Nitsche and Paulus, 2000). The tDCS device has a special feature that makes
was obtained from the local research ethics committee of the institution (process it particularly reliable for double-blind study designs. We noted in our previ-
approval number 0020.0.272.000-06). The experimental protocol was designed ous trials that patients try to look at the tDCS display during stimulation and
and carried out according to the principles of the Declaration of Helsinki. encountered situations in which we had to hide the device from patients receiv-
We were aware of the possibility that exposing patients to alcohol cues may ing sham treatment. We therefore incorporated a switch in the back of the tDCS
potentially result in relapse. Given this important ethical concern, we decided, device that can be activated by the researcher to interrupt the electrical current
therefore, not to use real alcoholic beverages in a glass as cues. We chose images while maintaining the display “ON” and displaying the parameters of stimula-
of alcoholic beverages and situations that subjects would be more typically tion throughout the procedure. As described earlier, participants received three
exposed to through television commercials and advertisements. different types of treatment:
 P.S. Boggio et al. / Drug and Alcohol Dependence 92 (2008) 55–60 57

(1) Anodal stimulation of the left DLPFC and cathodal stimulation of the right and interaction condition versus time. When appropriate, post hoc comparisons
DLPFC (referred in the text as “anodal left/cathodal right”). The anode were performed using Bonferroni correction.
electrode was placed over F3 (using EEG 10/20 system) and the cathode At the completion of the study, a total of two subjects were lost to attrition
electrode over F4. (one dropout after one session and one dropout after two sessions of tDCS).
(2) Anodal stimulation of right DLPFC and cathodal stimulation of the left Data that could not be obtained were handled as missing at random.
DLPFC (referred in the text as “anodal right/cathodal left”). The anode
electrode was placed over F4 (using EEG 10/20 system) and the cathode
electrode over F3. 3. Results
(3) Sham stimulation of DLPFC (referred as “sham stimulation”). For sham
stimulation, the electrodes were placed at the same positions as in active All patients were able to well tolerate the treatment with
stimulation; however, the stimulator was turned off after 30 s of stimu- tDCS. The demographic and clinical characteristics are summa-
lation. Therefore, the subjects felt the initial itching sensation associated
rized in Table 1. Adverse effects were rare and their frequency
with turning on the device, but received no current stimulation for the rest
of the treatment period. A recent study showed that this method of sham was not significantly different across the three conditions of
stimulation is reliable (Gandiga et al., 2006). treatment (p = 0.51, fisher’s exact test). Discomfort at the site
of stimulation was the most common adverse effect (1 report
The principal target for treatment was the DLFPC cortex based on prior in the sham and 2 reports in anodal left/cathodal right and
TMS studies showing that modulation of this area results in a decrease in anodal right/cathodal left stimulation conditions) followed by
smoking and also food craving (Johann et al., 2003; Uher et al., 2005) and neu-
roimaging studies showing that the activity in this area is significantly associated
headache (1, 0 and 2 reports, respectively), mood changes (1,
with alcohol craving (Olbrich et al., 2006; George et al., 2001; Myrick et al., 0 and 0 reports, respectively) and itching on the stimulation
2004). site (1, 0 and 0 reports, respectively). Importantly, when com-
A constant current of 2 mA was applied for 20 min. These parameters have paring baseline craving across the three conditions, we found
been shown to be safe in healthy volunteers (Iyer et al., 2005). no significant difference (F(2,32) = 0.09, p = 0.91), suggesting no
carry-over effects due to stimulation.
2.4. Statistical analysis Exploratory analysis evaluating mood effects showed that a
single session of tDCS induced no significant effects in these
Analyses were done with SAS version 9.1 statistical software (Cary, NC,
USA). In this study, we had three main research questions that were addressed subjects except for the item “worried/unconcerned” (p = 0.02).
separately: Interestingly, after anodal right/cathodal left DLPFC stimula-
tion, subjects had an increased rating towards “worried” as
• Research question 1: did our cues (alcoholic beverage movie) increase crav- compared to anodal left/cathodal right and sham stimulation.
ing? To answer this question, we merged the data of the three conditions and
compared craving before (T0) and after (T1) alcohol cues, but before stimu-
lation. We then performed an analysis of variance model in which time (T0 3.1. Question 1: craving increase by alcohol cues
versus T1) was the main covariate and subsequently, we performed ANCOVA
models to adjust for the effects of potential confounders such as age, gender, In order to analyze the initial question of whether our strategy
order of stimulation and duration of smoking. to induce craving was effective (movie of alcoholic beverage
• Research question 2: did active tDCS (with either anodal left/cathodal right
or anodal right/cathodal left) change craving levels as compared to sham
consumption), we compared alcohol craving before and after
tDCS? For this analysis, we used a mixed linear model with repeated mea- alcohol cues (T0 versus T1) using an analysis of variance model
sures on time in which we modeled craving changes using the covariates of in which we include time (T0 and T1) as the main covariate
time, condition and interaction between condition and time. For the outcome and subsequently, we performed ANCOVA models to adjust the
measure, we summed items to produce a total craving intensity score for each results for potential confounders such as age, gender and order
time point. We compared two time-points: after cues/before stimulation (T1)
versus before the second cues/after stimulation (T2).
and years of drinking. We found a significant increase in craving
• Research question 3: can tDCS induce after-effects in smoking craving? In of 8.1% (±8.8%, F(1,34) = 29.26, p < 0.0001) after alcohol cues.
other words, do alcohol cues viewed after tDCS stimulation change crav- Similar results were obtained when we performed ANCOVA
ing differently when active versus sham conditions are compared? For this models including age, gender, order of stimulation and years of
assessment, we performed a similar analysis as compared to research ques- drinking as confounders (P < 0.0001 when adjusting for all these
tion number 2; however, using different time-points (before the movie/after
stimulation (T2) and after the movie/after stimulation (T3)).
potential confounders). Interestingly, age and years of drinking
were significantly associated with an increase in craving from
Given we have three different questions, we corrected the p-value for multiple T0 to T1 (p = 0.0162 and p = 0.0076, respectively) (Table 2).
comparisons using Bonferroni correction. Therefore, we used a threshold for
significance of p < 0.017 for each research question.
Table 1
We defined the time points as:
Demographic and clinical characteristics

(1) T0—baseline assessment. Mean/N (S.D.)

(2) T1—assessment after (first) alcohol cues and before tDCS.
Age 41.3 (5.7)
(3) T2—assessment after tDCS and before (second) alcohol cues.
Gender (F/M) 2/11
(4) T3—assessment after tDCS and (second) alcohol cues.
Age at onset of drinking (years) 15.0 (4.6)
Alcohol intake (g/l) 227.7 (123.2)
For the other endpoints such as mood changes, we used a repeated measures
Number of drinks per day 10.0 (4.6)
analysis of variance in which the dependent variable was one of these endpoints
SADD 27.4 (7.8)
and the independent variables were: condition (sham, anodal left/cathodal right
and anodal right/cathodal left tDCS), time of treatment (pre- and post-treatment) F/M—females/males; SAAD—Short Alcohol Dependence Data.
58 P.S. Boggio et al. / Drug and Alcohol Dependence 92 (2008) 55–60

Table 2
Craving levels throughout the study
T0 T1 T2 T3

Sham tDCS 37.9 (15.0) 35.8 (15.5) 35.2 (15.7) 32.5 (15.8)
Anodal left/cathodal right tDCS 35.2 (17.0) 33.0 (17.3) 39.6 (14.8) 38.7 (14.9)
Anodal right/cathodal left tDCS 37.3 (16.1) 32.7 (15.7) 41.5 (10.4) 40.0 (13.0)

Note that the scale for craving (alcohol urge questionnaire) is an instrument with eight questions measuring craving. For some of the questions, lower values means
more craving and for other items is the opposite. We therefore, made the necessary adjustments in order to have all the items going in the same direction; and
therefore, calculated a composite score for each patient that ranges from 0 to 56 in which 0 indicates maximal craving and 56 indicates no craving. Therefore, an
increase in this scale (e.g. from 33 to 37) indicates less craving and a decrease (e.g. from 40 to 37) indicates more craving. T0 corresponds to baseline assessment:
pre tDCS, pre first cue exposure. T1 corresponds to pre tDCS, post first cue exposure. T2 corresponds to post tDCS, pre second cue exposure. T3 corresponds to post
tDCS, post second cue exposure.

3.2. Question 2: effects of tDCS treatment

In order to analyze whether tDCS reduced craving, we com-

pared craving levels before and after tDCS treatment (T1 versus
T2), performing a mixed model with repeated measures on time.
This model revealed a significant interaction time versus condi-
tion (F(5,52) = 7.4, p < 0.0001), suggesting that craving changes
were different across the different conditions of treatment.
Indeed, when comparing the three conditions of treatment, we
found a significant difference between anodal left/cathodal right
versus sham stimulation (p = 0.02) and anodal right/cathodal left
versus sham stimulation (p < 0.0001), but not between anodal
left/cathodal right versus anodal right/cathodal left (p = 0.53). Fig. 2. Craving changes after the second alcohol cues exposure (after
The mean craving change was −2% (±6%), +20% (±29%) and treatment—from T2 to T3) in the three conditions of treatment: anodal
+27% (±20%) for sham, anodal left/cathodal right and anodal left/cathodal right DLPFC (dorsolateral prefrontal cortex) tDCS, anodal
right/cathodal left DLPFC stimulation, respectively (see Fig. 1 right/cathodal left tDCS and sham tDCS. Each time point represents mean
craving and error bar represents standard error of mean (S.E.M). *Statistically
and Table 2). Finally, we tested whether gender, order, age and significant when compared with baseline (p < 0.017). This figure shows that
years of drinking were significant confounders and found that after the treatment craving could be significantly increased only in subjects
these terms were not significant (p > 0.05) and did not change who received sham stimulation. Note that normalized craving levels increase
the results of our model. indicates craving increase.

3.3. Question 3: after-effects of tDCS patients that received active stimulation as compared with sham
stimulation. We therefore performed another model in which
We hypothesized that tDCS effects would last for several min- the dependent measure was craving level and had three inde-
utes after the end of the stimulation period. In order to investigate pendent variables: time (T2 and T3), condition (sham, anodal
this effect, we evaluated whether craving could be increased in left/cathodal right and anodal right/cathodal left) and the interac-
tion condition with time. We found a significant interaction effect
(F(5,52) = 5.7, p = 0.0003). Indeed, craving could not be increased
by alcohol cues in the two conditions that received active stim-
ulation (p = 0.12 and p = 0.64, anodal left/cathodal right and
anodal right/cathodal left stimulation conditions, respectively);
however, there was a significant craving increase in the condition
that received sham stimulation after alcohol cues (mean craving
change of 7.3% (±7.8%)—p = 0.007)—see Fig. 2 and Table 2.
When including potential confounders in the model (age, gender,
order of stimulation, years of drinking), the results for the inter-
action term did not change and these terms were not significant,
except for gender that was marginally significant (p = 0.04).

Fig. 1. Craving changes after stimulation (T2) as compared with baseline. Each 4. Discussion
column represents mean craving changes and error bar represents standard
error of mean (S.E.M). *Statistically significant when compared with baseline
(p < 0.017). This figure shows that craving was significantly changed after active Our results show that both anodal left/cathodal right and
stimulation only. Note that a positive change indicates craving reduction and a anodal right/cathodal left DLPFC stimulation significantly
negative change indicates craving increase. decreased alcohol craving as compared to sham stimulation. In
 P.S. Boggio et al. / Drug and Alcohol Dependence 92 (2008) 55–60 59

addition, alcohol craving could not be increased by alcohol cues areas associated with craving such as the orbitofrontal cortex
in the active stimulation conditions after treatment. (London et al., 2000; Fowler and Volkow, 1998) that also has
Because we targeted the dorsolateral prefrontal cortex, it is extensive connections to other brain areas such as the striatum
conceivable that stimulation was associated with a change in and amygdala; therefore, integrating the cortical and subcortical
the activity of this area. Indeed, other tDCS studies targeting processing of motivational behavior and reward. This hypothe-
stimulation of DLPFC showed that modulation of this area is sis is also supported by recent data showing that tDCS results
associated with behavioral changes, such as cognitive changes in widespread changes in regional brain activity in this same
in healthy subjects (Fregni et al., 2005; Kincses et al., 2004) and network (Lang et al., 2005).
patients with major depression (Fregni et al., 2006b) and also Although we showed that DC stimulation of the prefrontal
mood changes in patients with depression (Fregni et al., 2006a). cortex reduces alcohol craving, further studies are needed to
Several neuroimaging studies have correlated activity in the establish the use of tDCS as a viable clinical and therapeutic
dorsolateral prefrontal cortex with alcohol craving. In a recent application. One potential advantage of developing tDCS as an
study, George et al. (2001) reported that alcoholic subjects exhib- alternative therapeutic strategy is the fact that the effects of tDCS
ited increased activity in the left dorsolateral prefrontal cortex are immediate. Because craving levels are highly variable and
while viewing alcohol cues presented within a scanner environ- dependent on social and even circadian influences, a single treat-
ment. Furthermore, this change was not observed when alcoholic ment that can transiently block craving levels quickly would be
patients viewed control pictures (such as non-alcoholic bever- highly desirable compared to drug treatment therapies that are
ages) and in social drinkers (George et al., 2001). Activity in typically more long-lasting and lead to tonic effects and thus can-
the dorsolateral prefrontal cortex area is also associated with not capture craving variations. In addition, if tDCS does prove
craving for other substances, such as smoking and cocaine. to have clinical value, it has additional advantage because it is
Previous studies investigating neural responses to cues in nico- safe and has a low incidence of only very mild adverse effects.
tine abusers demonstrated that the anterior cingulate, amygdala, Some potential limitations of this study should be discussed.
insula, orbitofrontal and dorsolateral prefrontal cortex are asso- First, we studied a relatively small sample size of 13 patients.
ciated with craving (Wilson et al., 2004). However, this study used a cross-over design and therefore the
As the effects of alcohol and other drugs might be related results would be comparable to a parallel design study with
to activity in mesolimbic dopamine pathways (Berridge and 39 subjects (13 subjects times 3 different conditions). Second,
Robinson, 1998), this might explain the involvement of the though not formally assessed in our study, the use of a 48-h
dorsolateral prefrontal cortex (through the mesofrontolimbic washout period appears to be sufficient given that we showed that
connections). Several dopaminergic antagonists have demon- baseline craving values were not significantly different across
strated a significant effect in reducing alcohol craving such as different days of testing. Third, there was no neutral cue condi-
clozapine and olanzapine (Green et al., 1999; Hutchison et al., tion; thus, it is difficult to interpret conclusively whether or not
2006). Given that the mesolimbic pathways are associated with tDCS reduced cue-specific craving or whether the effects were a
reward behavior, this might explain the addiction behavior as result of a general attenuation in craving report. As shown by our
alcohol consumption plays a critical role in incentive sensitiza- data, tDCS had an effect on both types of craving as it reduced
tion of dopamine receptors. Therefore, with a lack of alcohol or cue-provoked craving and also baseline craving. We decided not
when there is a cue associated with alcohol such as images of to implement a neutral cue condition given that several studies
the alcoholic beverages or people drinking alcohol, this creates have reported that images of non-alcoholic beverages as com-
a response in the mesolimbic pathways generating an increase pared with alcoholic beverages are not effective to elicit craving
in DLPFC activity that in part may be responsible for the drug- (George et al., 2001). Finally, we cannot determine, based on our
seeking behavior. It would follow that if the activity of DLPFC is findings, whether the effects of tDCS on alcohol craving were
modulated externally by tDCS, this might block this cascade of due to anodal or cathodal stimulation (or combination of both).
events due to the competition with the input coming from tDCS Future studies using different electrode montages and sizes are
that can ultimately decrease the signal to noise in the neural critical to explore this matter further.
system associated with reward. All together, this evidence suggests that noninvasive brain
One aspect that warrants further discussion is the fact stimulation might be an efficacious method to reduce differ-
that both strategies of DLPFC stimulation, that is, anodal ent types of craving and thus further investigation with studies
left/cathodal right and anodal right/cathodal left stimulation, including larger sample sizes and also evaluating the clinical
resulted in craving reduction. Indeed, neuroimaging studies benefits of this treatment are warranted.
showed that either the right (Olbrich et al., 2006) or left (George
et al., 2001) DLPFC are involved in alcohol craving. There- Acknowledgements
fore, it might be conjectured that anodal stimulation or cathodal
inhibition of either right or left DLPFC ruptured the balance This work was supported in part by a grant from Harvard
between the right and left DLPFC activity that might be nor- University David Rockefeller Center—Jorge Paulo Lemann to
mally necessary for craving states. Support for this notion of F.F. A.P.-L. is supported by a NIH grant K24 RR018875. N.S.
balanced bilateral activation of DLPFC during craving states is supported by a research grant from Mackenzie University
has been shown in neuroimaging studies (Wilson et al., 2004). (Mack-pesquisa). Sincere appreciation is expressed to Paola
Finally, modulation of the DLPFC might have modulated other Liguori for her assistance in the study and to the patients for
60 P.S. Boggio et al. / Drug and Alcohol Dependence 92 (2008) 55–60

committing the time for this study. We are also grateful to the circuits during cue-elicited cocaine craving. Proc. Natl. Acad. Sci. U.S.A.
rehabilitation center in the city of Santana do Parnaiba for the 93, 12040–12045.
collaboration in this study. Green, A.I., Zimmet, S.V., Strous, R.D., Schildkraut, J.J., 1999. Clozapine
for comorbid substance use disorder and schizophrenia: do patients with
schizophrenia have a reward-deficiency syndrome that can be ameliorated
References by clozapine? Harv. Rev. Psychiatr. 6, 287–296.
Hutchison, K.E., Ray, L., Sandman, E., Rutter, M.C., Peters, A., Davidson, D.,
American Psychiatric Association, 1994. Diagnostic Criteria from DSM-IV, Swift, R., 2006. The effect of olanzapine on craving and alcohol consump-
Washington D.C., American Psychiatric Association. tion. Neuropsychopharmacology 31, 1310–1317.
Berridge, K.C., Robinson, T.E., 1998. What is the role of dopamine in reward: Iyer, M.B., Mattu, U., Grafman, J., Lomarev, M., Sato, S., Wassermann, E.M.,
hedonic impact, reward learning, or incentive salience? Brain Res. Brain 2005. Safety and cognitive effect of frontal DC brain polarization in healthy
Res. Rev. 28, 309–369. individuals. Neurology 64, 872–875.
Brody, A.L., Mandelkern, M.A., London, E.D., Childress, A.R., Lee, G.S., Bota, Johann, M., Wiegand, R., Kharraz, A., Bobbe, G., Sommer, G., Hajak, G.,
R.G., Ho, M.L., Saxena, S., Baxter Jr., L.R., Madsen, D., Jarvik, M.E., 2002. Wodarz, N., Eichhammer, P., 2003. Transcranial magnetic stimulation for
Brain metabolic changes during cigarette craving. Arch. Gen. Psychiatr. 59, nicotine dependence. Psychiatr. Prax. 30 (Suppl. 2), S129–S131.
1162–1172. Juliano, L.M., Brandon, T.H., 1998. Reactivity to instructed smoking availability
Camprodon, J.A., Martinez-Raga, J., Alonso-Alonso, M., Shih, M.C., Pascual- and environmental cues: evidence with urge and reaction time. Exp. Clin.
Leone, A., 2007. One session of high frequency repetitive transcranial Psychopharmacol. 6, 45–53.
magnetic stimulation (rTMS) to the right prefrontal cortex transiently Kincses, T.Z., Antal, A., Nitsche, M.A., Bartfai, O., Paulus, W., 2004. Facili-
reduces cocaine craving. Drug Alcohol Depend. 86, 91–94. tation of probabilistic classification learning by transcranial direct current
Davidson, R., Raistrick, D., 1986. The validity of the Short Alcohol Depen- stimulation of the prefrontal cortex in the human. Neuropsychologia 42,
dence Data (SADD) Questionnaire: a short self-report questionnaire for the 113–117.
assessment of alcohol dependence. Br. J. Addict 81, 217–222. Lang, N., Siebner, H.R., Ward, N.S., Lee, L., Nitsche, M.A., Paulus, W., Roth-
Drummond, D.C., Phillips, T.S., 2002. Alcohol urges in alcohol-dependent well, J.C., Lemon, R.N., Frackowiak, R.S., 2005. How does transcranial DC
drinkers: further validation of the Alcohol Urge Questionnaire in an untreated stimulation of the primary motor cortex alter regional neuronal activity in
community clinical population. Addiction 97, 1465–1472. the human brain? Eur. J. Neurosci. 22, 495–504.
Due, D.L., Huettel, S.A., Hall, W.G., Rubin, D.C., 2002. Activation in mesolim- Liebetanz, D., Nitsche, M.A., Tergau, F., Paulus, W., 2002. Pharmacological
bic and visuospatial neural circuits elicited by smoking cues: evidence approach to the mechanisms of transcranial DC-stimulation-induced after-
from functional magnetic resonance imaging. Am. J. Psychiatr. 159, 954– effects of human motor cortex excitability. Brain 125, 2238–2247.
960. London, E.D., Ernst, M., Grant, S., Bonson, K., Weinstein, A., 2000.
Eichhammer, P., Johann, M., Kharraz, A., Binder, H., Pittrow, D., Wodarz, N., Orbitofrontal cortex and human drug abuse: functional imaging. Cereb.
Hajak, G., 2003. High-frequency repetitive transcranial magnetic stimulation Cortex 10, 334–342.
decreases cigarette smoking. J. Clin. Psychiatr. 64, 951–953. Mcbride, D., Barrett, S.P., Kelly, J.T., Aw, A., Dagher, A., 2006. Effects of
Figlie, N.B., Dunn, J., Laranjeira, R., 2004. Factor structure of the Stages of expectancy and abstinence on the neural response to smoking cues in
Change Readiness and Treatment Eagerness Scale (SOCRATES) in alcohol cigarette smokers: an fMRI study. Neuropsychopharmacology.
dependent outpatients. Rev. Bras. Psiquiatr 26, 91–99. Myrick, H., Anton, R.F., Li, X., Henderson, S., Drobes, D., Voronin, K., George,
Fowler, J.S., Volkow, N.D., 1998. PET imaging studies in drug abuse. J. Toxicol. M.S., 2004. Differential brain activity in alcoholics and social drinkers
Clin. Toxicol. 36, 163–174. to alcohol cues: relationship to craving. Neuropsychopharmacology 29,
Fregni, F., Boggio, P.S., Nitsche, M., Bermpohl, F., Antal, A., Feredoes, E., 393–402.
Marcolin, M.A., Rigonatti, S.P., Silva, M.T., Paulus, W., Pascual-Leone, Nitsche, M.A., Liebetanz, D., Antal, A., Lang, N., Tergau, F., Paulus,
A., 2005. Anodal transcranial direct current stimulation of prefrontal cortex W., 2003. Modulation of cortical excitability by weak direct current
enhances working memory. Exp. Brain Res. 166, 23–30. stimulation—technical, safety and functional aspects. Suppl. Clin. Neuro-
Fregni, F., Boggio, P.S., Nitsche, M.A., Marcolin, M.A., Rigonatti, S.P., Pascual- physiol. 56, 255–276.
Leone, A., 2006a. Treatment of major depression with transcranial direct Nitsche, M.A., Paulus, W., 2000. Excitability changes induced in the human
current stimulation. Bipolar Disord. 8, 203–204. motor cortex by weak transcranial direct current stimulation. J. Physiol. 527
Fregni, F., Boggio, P.S., Nitsche, M.A., Rigonatti, S.P., Pascual-Leone, A., (Pt 3), 633–639.
2006b. Cognitive effects of repeated sessions of transcranial direct current Nitsche, M.A., Seeber, A., Frommann, K., Klein, C.C., Rochford, C., Nitsche,
stimulation in patients with depression. Depress Anxiety 23, 482–484. M.S., Fricke, K., Liebetanz, D., Lang, N., Antal, A., Paulus, W., Tergau,
Fregni, F., Liguori, P., Fecteau, S., Nitsche, M., Pascual-Leone, A., Boggio, P.S., F., 2005. Modulating parameters of excitability during and after transcra-
in press. Cortical stimulation of the prefrontal cortex with transcranial DC nial direct current stimulation of the human motor cortex. J. Physiol. 568,
stimulation reduces smoking cue-provoked craving: a randomized, sham- 291–303.
controlled study. J. Clin. Psychiatr. Olbrich, H.M., Valerius, G., Paris, C., Hagenbuch, F., Ebert, D., Juengling,
Gandiga, P.C., Hummel, F.C., Cohen, L.G., 2006. Transcranial DC stimulation F.D., 2006. Brain activation during craving for alcohol measured by positron
(tDCS): a tool for double-blind sham-controlled clinical studies in brain emission tomography. Aust. N.Z. J. Psychiatr. 40, 171–178.
stimulation. Clin. Neurophysiol. 117, 845–850. Purpura, D.P., McMurtry, J.G., 1965. Intracellular activities and evoked poten-
Garavan, H., Pankiewicz, J., Bloom, A., Cho, J.K., Sperry, L., Ross, T.J., tial changes during polarization of motor cortex. J. Neurophysiol. 28,
Salmeron, B.J., Risinger, R., Kelley, D., Stein, E.A., 2000. Cue-induced 166–185.
cocaine craving: neuroanatomical specificity for drug users and drug stimuli. Sell, L.A., Morris, J.S., Bearn, J., Frackowiak, R.S., Friston, K.J., Dolan, R.J.,
Am. J. Psychiatr. 157, 1789–1798. 2000. Neural responses associated with cue evoked emotional states and
George, M.S., Anton, R.F., Bloomer, C., Teneback, C., Drobes, D.J., Lorber- heroin in opiate addicts. Drug Alcohol Depend. 60, 207–216.
baum, J.P., Nahas, Z., Vincent, D.J., 2001. Activation of prefrontal cortex Uher, R., Yoganathan, D., Mogg, A., Eranti, S.V., Treasure, J., Campbell, I.C.,
and anterior thalamus in alcoholic subjects on exposure to alcohol-specific Mcloughlin, D.M., Schmidt, U., 2005. Effect of left prefrontal repetitive tran-
cues. Arch. Gen. Psychiatr. 58, 345–352. scranial magnetic stimulation on food craving. Biol. Psychiatr. 58, 840–842.
Grant, S., London, E.D., Newlin, D.B., Villemagne, V.L., Liu, X., Contoreggi, Wilson, S.J., Sayette, M.A., Fiez, J.A., 2004. Prefrontal responses to drug cues:
C., Phillips, R.L., Kimes, A.S., Margolin, A., 1996. Activation of memory a neurocognitive analysis. Nat. Neurosci. 7, 211–214.