Ventral Striatal Activation During Reward Anticipation
Correlates with Impulsivity in Alcoholics
Anne Beck, Florian Schlagenhauf, Torsten Wüstenberg, Jakob Hein, Thorsten Kienast, Thorsten Kahnt,
Katharina Schmack, Claudia Hägele, Brian Knutson, Andreas Heinz, and Jana Wrase
Background: Alcohol dependence is often associated with impulsivity, which may be correlated with dysfunction of the brain reward
system. We explored whether functional brain activation during anticipation of incentive stimuli is associated with impulsiveness in
detoxified alcoholics and healthy control subjects.
would either result in monetary gain, avoidance of monetary loss, or no consequence. Impulsivity was assessed with the Barratt Impulsive-
ness Scale-Version 10 (BIS-10).
Results: Detoxified alcoholics showed reduced activation of the ventral striatum during anticipation of monetary gain relative to healthy
only in alcoholics, not in control subjects.
Conclusions: This study suggests that reduced ventral striatal recruitment during anticipation of conventional rewards in alcoholics may
be related to their increased impulsivity and indicate possibilities for enhanced treatment approaches in alcohol dependence.
tem, ventral striatum
behavior seems to be associated with dysfunctions of the dopa-
minergic mesolimbic reward system (2,3), which can elicit a
conditioned attention allocation for alcohol-associated stimuli
rendering them specifically salient. In functional magnetic reso-
nance imaging (fMRI) studies with alcoholics, alcohol cues
activated the ventral striatum (4–6), whereas in healthy volun-
teers, the same area responded toward conventional reward-
indicating cues (7–10). Such an effect could describe a reorgani-
zation (“hijacking”) of the priorities of reward circuitry, such that
drug cues elicit more appetitive behavior than cues for conven-
tional rewards (11–13). Furthermore, prefrontal control seems to
be reduced in addiction (e.g., [14–17]).
It has been suggested that alcohol dependence and drug
addiction are characterized by dysfunctional preference of im-
mediate versus delayed reward, which manifests as impulsivity
and may contribute to early disease onset and increased social
problems (18–20). A series of studies suggested that alcoholics
are more impulsive than control subjects (21–24) and that
impulse control disorders (like pathological gambling and im-
pulsive violent behavior) are more common among alcoholics
lcohol dependence is one of the most devastating disor-
ders in men in industrialized nations and number one risk
factor for more than 60 chronic diseases (1). Addictive
than in healthy volunteers (23,25). In addition, the acute effect of
alcohol itself also enhances impulsive behavior (26,27).
One factor contributing to impulsivity is a reduced ability to
choose larger but delayed rewards compared with smaller but
earlier rewards (“delay discounting” as an index of impulsive
tendencies) (28–31). This dysfunctional delay gratification may
be associated with neuronal dysfunction of reward anticipation
and contribute to craving for immediate alcohol reward
(12,21,32). In fact, neuronal correlates for immediate reward bias
were observed in alcoholics (14).
Brain activation elicited by reward processing can be mea-
sured with a monetary incentive delay (MID) task and fMRI
(10,33). In accordance with the hypothesis that reward anticipa-
tion is altered in impulsive individuals, impulsivity was nega-
tively associated with functional activation of the ventral striatum
during reward anticipation in patients with attention-deficit/
hyperactivity disorder (ADHD) (34,35) but with increased acti-
vation proportional to the amount of anticipated reward as
shown in healthy control subjects (29). Attention-deficit/hyper-
activity disorder patients are a group of patients with heightened
impulsivity similar to alcohol-dependent patients and there may
be similar neurobiological mechanisms underlying these symp-
toms in both groups.
Several studies have confirmed the role of ventral and dorsal
striatum and its ascending monoaminergic (i.e., dopaminergic)
innervation from ventral tegmental area (VTA) in impulsive
behavior. In a human study, van Gaalen et al. (36) demonstrated
that tolerance to delay reinforcement was impaired after appli-
cation of a dopamine receptor D1 antagonist, indicating a close
link between dopaminergic neurotransmission and impulsivity.
Cardinal et al. (37) and Cardinal and Howes (38) demonstrated
that lesions of the nucleus accumbens provoked impulsive
choice behavior in rodents, which preferred immediate small
rewards to delayed larger ones. Increased delay discounting has
been repeatedly observed in patients with alcoholism and drug
Cloninger et al. (41) developed a typology proposing that
high impulsivity in alcoholics is associated with the personality
Charité Universitätsmedizin Berlin, Campus Mitte, and Bernstein Center
for Computational Neuroscience (Tka, AH), Charité –, Universitätsmedi-
University, Stanford, California.
Authors AB and FS contributed equally to this work.
Address correspondence to Jana Wrase, Ph.D., Department of Psychiatry,
Charité Campus Mitte, Charitéplatz 1, 10117 Berlin, Germany; E-mail:
Received Sep 17, 2008; revised Apr 29, 2009; accepted Apr 30, 2009.
BIOL PSYCHIATRY 2009;66:734–742
© 2009 Society of Biological Psychiatry
traits of high novelty seeking and low harm avoidance and
suggested a modulation of these traits by dopaminergic and ser-
otonergic neurotransmission, respectively (19). Human and ani-
mal studies partially confirmed this hypothesis by suggesting that
in alcohol dependence, serotonin dysfunction is associated with
negative mood states, which may trigger impulsive behavior in
individuals who feel anxious or threatened (20,42). Heinz et al.
(20) showed that nonhuman primates who were isolated in their
early childhood had a reduced serotonin turnover rate and were
more anxious. After adolescence, these primates were especially
aggressive and impulsive and displayed enhanced self-adminis-
tered alcohol consumption. Based on rodent studies, King et al.
(43) suggested that impulsivity is reduced when stimulation of
serotonin 1A (5-HT1A) receptors increases brain activation in a
corticostriatal circuitry including components of the ventral
To further explore the correlation between impulsivity and
striatal activation elicited during the processing of gains and
losses, especially during the anticipation period, we examined
recently detoxified male alcoholics and healthy men with a
monetary incentive delay (MID) task and fMRI and assessed
impulsivity with the Barratt Impulsiveness Scale-Version 10
(BIS-10) (44), as well as negative mood states. To the best of our
knowledge, our study is one of the first directly investigating
reward anticipation and impulsivity in a homogeneous sample of
alcohol-dependent patients, helping to better understand the
neural substrates underlying this clinically important trait.
Based on the studies by Scheres et al. (34) and Ströhle et al.
(35), we hypothesized that 1) the previously shown reduced
ventral striatal activation for conventional rewards (12) would
correlate inversely with impulsivity, and 2) that this correlation
would be particularly strong in alcoholics compared with control
Methods and Materials
Nineteen alcohol-dependent right-handed male patients and
19 age-matched healthy subjects were included in the study.
Written informed consent was obtained from all participants. The
study was approved by the Ethics Committee of the Charité
Universitätsmedizin Berlin. All patients were diagnosed as alco-
hol-dependent according to ICD-10 and DSM-IV criteria and had
no other neurological and psychiatric Axis I disorders and no
past history of dependency or current abuse of other drugs than
alcohol and nicotine as verified by random urine drug testing and
Structured Clinical Interview for DSM-IV Axis I Disorders
(SCID-I) interview (47). The severity of alcoholism was assessed
with the Alcohol Dependence Scale (48) and severity of alcohol
craving was assessed with the Obsessive Compulsive Drinking
Scale (OCDS) (49) (Table 1).
Healthy control subjects had no neurological and psychiatric
Axis I or Axis II disorders (SCID-I and Structured Clinical
Interview for DSM-IV Axis II Disorders [SCID-II] interviews)
(47,50) and no history of psychiatric disorder in first-degree
relatives. Before the fMRI experiment, patients had abstained
from alcohol in an inpatient detoxification treatment program for
at least 7 days (14 subjects ? 14 days; 9 subjects ? 10 days of
sobriety) as verified by random administration of alcohol breath
test. All patients were free of benzodiazepine or clomethiazole
medication for at least 4 days and had no bodily withdra-
wal symptoms. Thirty-seven alcohol-dependent patients were
screened: 19 were excluded due to comorbidities or magnetic
resonance imaging (MRI) contraindications and 2 were excluded
due to movement during scanning. Three patients with com-
pleted BIS-10 from a previous sample (12) were included.
Severity of anxiety was assessed with Spielberger’s State Trait
Anxiety Inventory (STAI) (46) and depression was assessed with
the Hamilton Depression Rating Scale (45). All participants were
right-handed as confirmed by the Edinburgh Handedness Inven-
tory (51). Years of education, socioeconomic status measured
with the Hollingshead Index of Social Status (52), and verbal IQ
measured with a German vocabulary test (53) were collected. All
patients and 13 of 19 control subjects were smokers. Subjects last
smoked about 45 minutes before scanning to avoid withdrawal.
Alertness during the task was assessed with the Stanford Sleep-
iness Scale (54) and self-reported effort for gain, loss, and neutral
cues was assessed with visual analog scales (VAS) (55) (Table 1).
Assessment of Impulsiveness
Impulsivity was assessed with a German version of the Barratt
Impulsiveness Scale-Version 10 (44), which contains 34 items
overall subdivided into three different subscores of impulsivity:
nonplanning, motor, and cognitive impulsiveness.
Monetary Incentive Delay Task
We used the monetary incentive delay task as described by
Knutson et al. (10) to examine neural responses in volunteers
during trials in which they anticipated potential monetary gain,
loss, or no consequences. Participants’ monetary gain depended
on their performance in a simple reaction time (RT) task at the
Table 1. Clinical Data
Age in Years
Obsessive Compulsive Drinking Scale
VAS Effort to Obtain Gain (1–10)
VAS Effort to Avoid Loss (1–10)
Mean Hit Rate in % Gain
Mean Hit Rate in % Neutral
Mean Hit Rate in % Loss
Total Gain in Euros per Run
Stanford Sleepiness Scale
Edinburgh Handedness Inventory
Socioeconomic Status (Hollingshead
Index of Social Status)a
IQ Estimates (WST)a
Number of Cigarettes per Daya
Severity of Alcohol Dependence (ADS)
Duration of Alcohol Dependence (in
Alcohol Consumption During the Last
Year in kg (Pure Alcohol)
Number of Professional Detoxifications
Length of Abstinence in Days
ADS, Alcohol Dependence Scale; HAMD, Hamilton Depression Rating
Scale; IQ, intelligence quotient; STAI, State Trait Anxiety Inventory; VAS,
visual analog scale; WST, Wortschatztest.
aStudent t test: p ? .05.
A. Beck et al.
BIOL PSYCHIATRY 2009;66:734–742 735
end of each trial, which required pressing a button upon the brief
presentation of a visual target (Figure 1, Supplement 1).
Money was shown in cash to the subjects before entering the
scanner and they were informed that they would receive the
money that they earned immediately after the scanning session.
During structural scans, participants completed a short practice
version of the task to minimize later learning effects and to
ensure that participants had learned the association between
cues and corresponding euro value (the latter was not displayed
during the actual task).
In each trial, participants saw one of seven geometric figures
(cue) for 250 msec, which indicated that they could either gain or
avoid losing different amounts of money (€3.00, €.60, or €.10) if
they pressed the response button during target presentation
(white square presented for 200 msec up to maximum 1000
msec). Participants were instructed to respond as fast as possible
during target presentation. The different cues are shown at the
bottom of Figure 1. Between cues and target, a variable delay of
2250, 2500, or 2750 msec was inserted. After responding, a
feedback was given for 1650 msec. Due to application of an
adaptive algorithm for target duration, subjects succeeded on
about 67% of the trials. Hits (? success) were defined as button
presses within the time frame of the target presentation (maxi-
mum 1 sec), including wins as well as no-losses. Subjects
performed two sessions consisting of 54 gain, 54 loss, and 36
neutral trials, which were presented in a random sequence. Each
run lasted about 14 min with a mean trial duration of approxi-
mately 7.69 sec and a mean intertrial interval of 3.53 sec.
Functional Magnetic Resonance Imaging
Functional magnetic resonance imaging was performed on a
1.5 Tesla scanner (Magnetom VISION, Siemens, Erlangen, Ger-
many) equipped with a standard circularly polarized (CP) head
coil using gradient-echo echo-planar imaging (GE-EPI). For the
acquisition of functional images, the following parameters were
used: repetition time (TR) ? 1870 msec, echo time (TE) ? 40
msec, flip ? 90°, matrix ? 64 ? 64, field of view (FOV) ? 256,
voxel size ? 4 ? 4 ? 3.3 mm3. Eighteen slices were collected
approximately parallel to the bicommissural plane, covering the
inferior part of the frontal lobe (superior border above the cau-
date nucleus), the whole temporal lobe, and large parts of the
occipital region. Approximately 6 fMRI volumes were acquired
per trial, resulting in 900 volumes total.
For anatomical reference, a three-dimensional (3-D) magne-
tization prepared rapid gradient echo (MPRAGE) was acquired
(TR ? 9.7 msec; TE ? 4 msec; flip ? 12°; matrix ? 256 ? 256;
FOV ? 256, voxel size ? 1 ? 1 ? 1 mm3). We minimized head
movement using a vacuum pad.
fMRI Data Analysis
Functional magnetic resonance imaging data were analyzed
using SPM5 (Wellcome Department of Neuroimaging, London,
United Kingdom, http://www.fil.ion.ucl.ac.uk/spm).
After temporal (correction for slice acquisition delay) and
spatial (movement correction, spatial normalization and smooth-
ing with 8-mm full width at half maximum [FWHM]) preprocess-
ing (for details, see Supplement 1), fMRI data were analyzed as
an event-related design in the context of the general linear model
(GLM) approach in a two-level procedure. On the first level in
the single subjects SPM models, the seven different cue condi-
tions (3? anticipation of gain, 3? anticipation of loss, and 1?
anticipation of no outcome), the target, and five feedback
conditions were modeled as events and convolved with the
canonical hemodynamic response function (HRF) to account for
the lag between event onset and expected increase of the blood
oxygenation level-dependent (BOLD) signal. The five feedback
conditions were successful reward feedback, failure to receive
monetary reward, successful loss-avoidance, failure to avoid loss,
and feedback of nonincentive trials. To account for variance
caused by head movement, realignment parameters were also
included as additional regressors in the model. For each subject,
the linear contrast images for “gain cues ? neutral cues” and
“loss cues ? neutral cues” were computed and taken to the
second level. For the feedback phase, the contrasts “successful ?
nonsuccessful gain trials” and “successful ? nonsuccessful loss-
avoidance trials” were computed.
To detect group differences, a second level random effects
analysis using a two-sample t test was conducted. Within-group
activation was assessed with one-sample t tests.
The relationship between impulsivity and regional neural
response during gain and loss anticipation (gain cues ? neutral
cues and loss cues ? neutral cues) as well as during gain and loss
feedback (successful ? nonsuccessful gain trials and success-
ful ? nonsuccessful loss-avoidance trials) was assessed using the
total BIS-10 score and the subscores (cognitive, motor, nonplan-
ning) as covariates in multiple regression analyses using SPM5
for the whole sample. To further clarify the results, group-
specific multiple regression analyses were conducted and re-
stricted to voxels showing a significant main effect over all
subjects. To reveal if correlations are specific for alcoholics
compared with control subjects, additional SPM analyses tested
group-by-covariate interactions in separate multiple regression
analyses including the covariate (BIS-10 scale), the group-by-
covariate interaction term (covariate ? group), and smoking
Figure 1. (Top) Task structure for a representative trial in the monetary
incentive delay task. The procedure is described in the text. (Bottom) Cues
used in the different trails, indicating whether different amounts of money
no consequences depending on reaction time (circle, square, triangle).
736 BIOL PSYCHIATRY 2009;66:734–742
A. Beck et al.
Since smoking may modulate neural activity in dopaminer-
gic brain regions (56–58), smoking status was used as a
covariate in all SPM analyses (t tests and multiple regression
analyses). To further ensure that findings were not confounded
by smoking, number of cigarettes smoked per day was used as
an alternative covariate instead of smoking status in additional
analyses (Supplement 1).
Given our strong a priori hypotheses regarding the ventral
striatum, we adjusted the results for false-positive findings
applying a small volume correction (SVC) as implemented in
SPM5 using binary masks from the publication-based proba-
bilistic Montreal Neurological Institute (MNI) atlas (59) at a
threshold of .75 probability (please refer to http://hendrix.
cess date September 3, 2007) and a significance level of p ? .05
familywise error (FWE)-corrected for the volume of interest
(VOI) (left and right ventral striatum). All other results are
reported at p ? .001 uncorrected with a minimum cluster size of
10 voxels. Corresponding brain regions were identified with
reference to the stereotaxic brain atlas provided by Talairach and
Behavioral Data Analysis
Behavioral data were analyzed with SPSS (SPSS Inc., Chicago,
Illinois) using two-sample t tests for clinical data and repeated
measures analysis of variance (ANOVA) for performance (i.e., hit
rate and reaction time).
Healthy subjects succeeded (i.e., responded during target
presentation) on 65.98% (SEM ? 2.84) of gain trials, on 67.25%
(SEM ? 1.99) of loss trials, and on 45.47% (SEM ? 3.04) of neutral
trials and earned €18.82 ? €6.82, on average. Alcohol-dependent
patients succeeded on 67.84% (SEM ? 2.41) of gain trials, on
62.28% (SEM ? 2.75) of loss trials, and on 49.71% (SEM ? 4.70)
of neutral trials and earned €17.75 ? €5.69, on average. Total
average earnings did not differ between the two groups (t ?
.524; p ? .603). There was a significant main effect of cue (F ?
29.07, p ? .001), indicating more hit responses during gain (t ?
7.55, p ? .001) and loss (t ? 6.77, p ? .001) compared wit neutral
trials (post hoc paired t tests) in both groups. There was neither
a significant main effect of group (F ? .12, p ? .91) nor a
significant group-by-cue interaction (F ? 2.85; p ? .07). Among
the 144 trials, there were, on average, 7.84 misses (SD 13.81).
Mean reaction times revealed a significant main effect for cue
(F ? 28.63, p ? .001), indicating faster responses during both
gain and loss trials (RT gain ? neutral: t ? 6.070, p ? .001 and RT
loss ? neutral: t ? 5.62, p ? .001) but no main effect for group
(F ? .003, p ? .960) nor group-by-cue interaction (F ? .495, p ?
.558) (Figure 2).
There were no significant differences between alcoholics and
control subjects in mean self-reported alertness during the task
assessed with the Stanford Sleepiness Scale (54) and no signifi-
cant differences between groups in self-reported effort to achieve
monetary gains or effort to prevent losses, as assessed with visual
analog scales (all t ? 1.85, all p ? .08) (Table 1).
Impulsiveness, Mood States and Other Sample
The total, cognitive, and motor scores of the BIS-10 were
significantly higher in alcohol-dependent patients (79.46 ?
15.85) than in control subjects (69.13 ? 7.79; t ? ?2.296, p ?
.030). The cognitive and motor scores of the BIS-10 differed
significantly between the two groups as well (cognitive: patients:
28.00 ? 5.40; control subjects: 23.81 ? 3.66; t ? ?2.48, p ? .020;
motor: patients: 24.69 ? 6.01; control subjects: 20.94 ? 2.29; t ?
2.31, p ? .050; nonplanning: patients: 26.77 ? 6.07; control
subjects: 24.83 ? 4.37; t ? 1.24, p ? .228).
Alcoholics reported stronger alcohol craving than healthy
control subjects (OCDS; t ? ?11.66, p ? .001) and higher
severity of depression (HAMD; t ? ?2.92, p ? .010) and anxiety
(STAI; t ? ?2.07, p ? .049). In addition, we observed a
significant negative correlation between impulsivity and depres-
sion (r ? ?.662, p ? .014), but not anxiety, in alcoholics. Control
subjects did not show any such significant correlations.
Figure 2. Hit rate (A) and reaction times (B) during MID
task: Box plots show the distribution of subjects’ mean
effect sizes within hit rate and reaction time during MID
task (left side: healthy control subjects; right side alcohol-
dependent patients). The boxes have lines at the lower,
median, and upper quartile values. Whiskers extend from
the interquartile range from the ends of the box. Notches
display the variability of the median between samples.
notches do not overlap have different medians at the 5%
significance level. In addition, single subject data are dis-
played as dots. Healthy control subjects are displayed on
left sides, patients on right sides; (l) indicates loss, (n)
neutral, and (g) gain trails. g, gain; l, loss; MID, monetary
incentive delay; n, neutral.
A. Beck et al.
BIOL PSYCHIATRY 2009;66:734–742 737
As expected, groups differed significantly in years of educa-
tion (t ? 4.49, p ? .002) and socioeconomic status (t ? 4.54, p ?
.001), as well as in IQ estimates (t ? 4.89, p ? .001). Alcoholics
smoked significantly more cigarettes per day than healthy control
subject (t ? ?5.39, p ? .001) (Table 1).
Neural Activity During Anticipation of Potential Monetary
Gain and Loss
During anticipation of monetary gain (contrast: gain ? neutral
cues), healthy control subjects showed a significant activation in
the bilateral ventral striatum, right caudate tail extending into
bilateral thalamus, and right insula. Alcoholics also displayed a
significant activation of bilateral ventral striatum, as well as of
right lateral globus pallidus, bilateral middle frontal gyrus (Brod-
mann area [BA] 10), right thalamus, and left superior temporal
gyrus (BA 38) (Table 1 in Supplement 1).
Comparing alcoholics directly with healthy control subjects, a
two-sample t test revealed significantly reduced activation in
right ventral striatum (x ? 12, y ? 15, z ? ?6; t ? 2.43, p ? .05
FWE-corrected for ventral striatal VOI) during anticipation of
gain versus neutral outcomes (Figure 3). Outside of the ventral
striatum, alcoholics did not show any significantly different brain
activation compared with healthy control subjects.
In healthy control subjects, anticipation of potential monetary
loss (contrast: loss ? neutral cues) was accompanied by signifi-
cant activation of bilateral ventral striatum, left medial dorsal
thalamus, bilateral putamen, bilateral parahippocampal gyrus
(BA 28 and 34), right middle occipital gyrus (BA 19), right
claustrum, left posterior cingulate (BA 30), right superior tempo-
ral gyrus (BA 22), and right cuneus (BA 18). Alcoholics also
showed activations of bilateral ventral striatum, right middle
frontal gyrus (BA 8), and right inferior frontal gyrus (BA 46)
(Table 2 in Supplement 1).
Alcoholics displayed a trendwise reduction of activation
compared with control subjects in right ventral striatum (x ? 15,
y ? 12, z ? ?3; t ? 2.27, p ? .07 FWE-corrected for ventral
striatal VOI). Again, outside of the ventral striatum, alcoholics did
not show any significantly different brain activation compared
with healthy control subjects (Figure 3).
Correlation Analyses Between Impulsivity and Anticipation
We correlated the differences in activation during 1) anticipa-
tion of monetary gain versus neutral outcomes and 2) anticipa-
tion of monetary loss versus neutral outcomes with the total
score of the Barratt Impulsiveness Scale in all alcohol-dependent
patients and healthy control subjects. During gain anticipation,
there was a significant association between impulsiveness and
brain activation in right ventral striatum (x ? 15, y ? 9, z ? 3;
F ? 23.35, p ? .001 uncorrected) and left anterior cingulate
cortex (ACC) (x ? 0, y ? 33, z ? ?3; F ? 21.80, p ? .001
uncorrected). Post hoc group-specific SPM analyses revealed
significant negative correlations in right ventral striatum (x ? 12,
y ? 9, z ? 3; t ? 3.83, p ? .05 FWE-corrected for main effect) and
ACC (x ? 0, y ? 33, z ? ?3; t ? 3.53, p ? .05 FWE-corrected for
main effect) (Figure 4). To test if the correlation finding was
specific for alcoholics compared with control subjects, the inter-
action between BIS-10 total score and group was tested in SPM
and revealed a group difference in ventral striatum (x ? 9, y ?
15, z ? ?6; t ? 2.36, p ? .059 FWE-corrected for ventral striatum
During anticipation of loss, there was a significant negative
association between impulsivity and brain activation in left
Figure 3. No increase in ventral striatal activation during gain anticipation in alcohol-dependent patients compared with healthy control subjects. (A) Box
plots with parameter estimates for the BOLD response during anticipation of loss (l), neutral (n), and gain (g) in healthy control subjects (red) and
x ? 0. Bottom: Box plots of differences in parameter estimates for the BOLD response during anticipation of gain? neutral within the peak voxel of the VOI.
green); displayed at MNI coordinate y ? 15; right side ? right hemisphere; plus sagittal view displayed at MNI coordinate x ? 0. Bottom: Box plots of
BOLD, blood oxygenation level-dependent; FWE, familywise error; L, left; MID, monetary incentive delay; MNI, Montreal Neurological Institute; P, posterior;
R, right; VOI, volume of interest.
738 BIOL PSYCHIATRY 2009;66:734–742
A. Beck et al.
superior temporal gyrus (BA 41) (x ? ?45, y ? ?30, z ? 18;
F ? 20.27, p ? .001 uncorrected) and right lateral globus pallidus
(x ? 12, y ? 3, z ? 3; F ? 19.02, p ? .001 uncorrected) in all
subjects. Post hoc group-specific SPM analyses revealed signifi-
cant negative correlations in right lateral globus pallidus (x ? 12,
y ? 0, z ? 3; t ? 2.91, p ? .05 FWE-corrected for main effect).
Interaction analyses revealed no significant group difference
during loss anticipation.
To further specify our findings, the BIS-10 subscores were
subjected to similar multiple regression analyses. The subscales
revealed similar correlations with ventral striatum and ACC
BOLD response during gain anticipation and a significant group-
by-covariate interaction in the ventral striatum for cognitive and
nonplanning subscales (Supplement 1).
The influence of possible confounds (motion, depression,
anxiety, socioeconomic status, intelligence, years of education,
and antisocial personality disorder) was controlled by additional
analyses (Supplement 1).
Neural Activity During Feedback of Monetary Gain and Loss
Healthy control subjects revealed significantly stronger acti-
vations than alcohol-dependent patients during loss-avoidance
feedback (contrast: successful vs. nonsuccessful loss-avoidance
trials) in left ventral striatum, left precuneus (BA 31), right
caudate tail, right claustrum, left middle temporal gyrus (BA 39),
right superior temporal gyrus (BA 22 and 42), left middle frontal
gyrus (BA 10), left insula, and left putamen, as well as right
transverse temporal gyrus (BA 41) (for within-group activation
see Table 3 in Supplement 1). There were no significant group
differences or within-group activations during gain feedback.
There were no significant correlations between any impul-
siveness scale and brain activation during feedback of gain or
loss-avoidance either in control subjects or alcohol-dependent
The current findings suggest that reduced ventral striatal
activation during reward expectation correlates with impulsivity
in alcohol-dependent patients.
Alcoholics showed a significantly reduced activation of the
ventral striatum during gain anticipation, as well as a trendwise
decrease in ventral striatal activation during loss anticipation.
These findings support the hypotheses that the ventral striatum is
involved in processing of reward-related cues and that the
reward system of alcoholics is less sensitive toward monetary
incentives (12). These findings are consistent with the notion that
the reward system in alcoholics is malfunctioning and may be
biased toward processing of alcohol-associated stimuli (61).
Grüsser et al. (62) showed that increased activation of the
striatum during presentation of alcohol pictures was positively
correlated with the prospective relapse risk, indicating that drugs
of abuse hijack and reorganize the priorities of reward circuitry,
so that they induce more appetitive behavior than cues for
conventional rewards (11).
To investigate links between impulsiveness and brain activity,
we correlated the activation during reward processing with the
Barratt Impulsiveness Scale. Activation of the ventral striatum
and anterior cingulate cortex during reward anticipation was
inversely correlated with impulsivity only in alcoholics but not in
control subjects. These findings are consistent with those of
Scheres et al. (34) and Ströhle et al. (35), who also found a
negative correlation between ventral striatal activation during
reward anticipation and impulsiveness in ADHD subjects. Impul-
sive subjects may have difficulty maintaining reward expectation,
even if they are responsive to reward outcomes, which might
contribute to increased delay discounting (29). Moreover, re-
duced neuronal responsiveness to anticipated reward may pro-
voke increased reward-seeking behavior as a means of compen-
For anticipation of loss, group differences in ventral striatum
were less marked and not associated with impulsivity. Both
groups showed a negative correlation between heightened im-
pulsivity and globus pallidum and superior temporal gyrus
activation. The superior temporal gyrus is implicated in speech
processing, as well as in complex cognitive tasks (like mental
rotation tasks) (65). The globus pallidus, a region of ventral
striatum connected with thalamus and prefrontal cortex, has
been associated with behavioral control (66,67). The present
Neurological Institute; P, posterior; R, right; VOI, volume of interest.
A. Beck et al.
BIOL PSYCHIATRY 2009;66:734–742 739
findings suggest that reduced activation in these regions dur-
ing loss anticipation might also contribute to impulsivity. Loss
anticipation may be associated with serotonergic functioning,
whereas dopamine may be more prominent in gain anticipation
(68). Direct association of brain activity with dysfunction of
serotonin or dopamine neurotransmitters remain to be explored
in future positron emission tomography (PET) studies (69,70). No
associations were observed between feedback activation and
impulsivity, which indicates a specific relationship of neural
activation during anticipation and impulsiveness. In contrast,
Bjork et al. (71) reported a positive association between impul-
sivity and reward feedback, indicating an increased sensitivity of
impulsive subjects for reward delivery. During reward feedback,
a similar increased response was found in ADHD patients (35).
Differences during reward anticipation could be explained by
sample characteristics (younger sample of patients with polyva-
lent substance abuse/dependence, e.g., cocaine and Axis I
disorders), task design, and psychometric instruments (impulsiv-
ity facet of the Neuroticism Extraversion Openness-Five Factor
Inventory [NEO-FFI], which is related to neuroticism).
The present findings also revealed a significant negative
correlation between impulsivity and depression in alcoholics.
This is in line with findings that harm avoidance predicts levels of
depression (72), given that it has been suggested that harm
avoidance is correlated with impulsivity (19,73). However, on the
neural level, we did not observe a significant correlation between
anxiety and depression and the activation in the ventral striatum,
suggesting that alterations in the neuronal correlates of reward
expectation in alcoholics may be more strongly related to
Our results confirm a series of findings demonstrating in-
creased impulsivity in alcoholics compared with healthy control
subjects (22,23,74). In rats, Belin et al. (75) observed that high
impulsivity predicts the development of addiction-like behavior,
including persistent or compulsive drug taking in the face of
aversive outcomes. Beyond behaviors correlated with impulsiv-
ity, our study revealed direct neuronal correlates of impulsivity.
Impulsive behavior is also influenced by social factors. For
instance, primate studies revealed that social isolation in early
childhood leads to reduced serotonin turnover, which increases
anxious behavior, aggressive impulsive behavior, and enhanced
alcohol consumption (20). In alcoholics, heightened impulsivity
could therefore promote both the genesis and maintenance of
alcohol dependence and may result from both genetic and social
influences on monoaminergic neurotransmitter systems (76).
A potential limitation of our study might be the differences in
education, socioeconomic status, and IQ. However, these differ-
ences did not interfere with task performance, as behavioral data
clearly showed. Both groups also differed in the number of
cigarettes smoked per day, which was controlled for in all
Overall, our findings suggest that reduced ventral striatal
activation during reward anticipation and heightened impulsivity
seem to be important features for alcohol dependence. Future
studies will have to determine whether this dysfunction repre-
sents a preexisting condition or can be reversed over the course of
treatment (either pharmacological or psychotherapeutic). The
latter would emphasize the role of specific impulsivity-related
psychotherapeutic interventions within the therapy of addiction,
like attention focusing or stop techniques.
This study was supported by the German Research Founda-
tion (Deutsche Forschungsgemeinschaft; HE 2597/4-3; 7-3; Exc
257) and by the Bernstein Centre for Computational Neuro-
science Berlin (Bundesministerium für Bildung und Forschung
We thank Michael Koslowski for scanning and Michael Rapp
for statistical assistance.
Ms. Beck, Dr. Schlagenhauf, Mr. Wüstenberg, Dr. Hein, Dr.
Kienast, Mr. Kahnt, Dr. Schmack, Ms Hägele, Dr. Knutson, and
Dr. Wrase reported no biomedical financial interests or potential
conflicts of interest. Professor Heinz has received research fund-
ing from the German Research Foundation and the Bernstein
Center for Computational Neuroscience Berlin (German Federal
Ministry of Education and Research), Eli Lilly & Company,
Janssen-Cilag, and Bristol-Myers Squibb. Professor Heinz has
received Speaker Honoraria from Janssen-Cilag, Johnson &
Johnson, Lilly, Pfizer, and Servier.
Supplementary material cited in this article is available
1. World Health Organization (2008): World Health Statistics. Geneva:
World Health Organization.
2. Adinoff B (2004): Neurobiologic processes in drug reward and addic-
3. DiChiara G (1997): Alcohol and dopamine. Alcohol Health Res World
4. Wrase J, Gruesser SM, Klein S, Diener C, Hermann D, Flor H, et al. (2002):
tion in alcoholics. Eur Psychiatry 17:287–291.
5. Braus DF, Wrase J, Grusser S, Hermann D, Ruf M, Flor H, et al. (2001):
Alcohol-associated stimuli activate the ventral striatum in abstinent
Differential brain activity in alcoholics and social drinkers to alcohol
cues: Relationship to craving. Neuropsychopharmacology 29:393–402.
7. Aharon I, Etcoff N, Ariely D, Chabris CF, O’Connor E, Breiter HC (2001):
Beautiful faces have variable reward value: fMRI and behavioral evi-
dence. Neuron 32:537–551.
8. Stark R, Schienle A, Girod C, Walter B, Kirsch P, Blecker C, et al. (2005):
responses of the brain. Biol Psychol 70:19–29.
9. Breiter HC, Aharon I, Kahneman D, Dale A, Shizgal P (2001): Functional
gains and losses. Neuron 30:619–639.
10. Knutson B, Adams CM, Fong GW, Hommer D (2001): Anticipation of
increasing monetary reward selectively recruits nucleus accumbens.
11. Nesse RM, Berridge KC (1997): Psychoactive drug use in evolutionary
perspective. Science 278:63–66.
et al. (2007): Dysfunction of reward processing correlates with alcohol
craving in detoxified alcoholics. Neuroimage 35:787–794.
13. Kalivas PW, Volkow ND (2005): The neural basis of addiction: A pathol-
ogy of motivation and choice. Am J Psychiatry 162:1403–1413.
M, et al. (2007): Immediate reward bias in humans: Fronto-parietal net-
works and a role for the catechol-O-methyltransferase 158(Val/Val) ge-
notype. J Neurosci 27:14383–14391.
15. Garavan H, Hester R (2007): The role of cognitive control in cocaine
dependence. Neuropsychol Rev 17:337–345.
associated with impaired motivation and self-control in cocaine addic-
17. Salo R, Ursu S, Buonocore MH, Leamon MH, Carter C (2009): Impaired
study. Biol Psychiatry 65:706–709.
18. Bechara A (2005): Decision making, impulse control and loss of will-
power to resist drugs: A neurocognitive perspective. Nat Neurosci
740 BIOL PSYCHIATRY 2009;66:734–742
A. Beck et al.
19. Cloninger CR (1987): Neurogenetic adaptive-mechanisms in alcohol-
ism. Science 236:410–416.
20. Heinz A, Mann K, Weinberger DR, Goldman D (2001): Serotonergic dys-
function, negative mood states, and response to alcohol. Alcohol Clin
al. (2008): The role of behavioral impulsivity in the development of
alcohol dependence: A 4-year follow-up study. Alcohol Clin Exp Res 32:
22. Swann AC, Dougherty DM, Pazzaglia PJ, Pham M, Moeller FG (2004):
23. Virkkunen M, Kallio E, Rawlings R, Tokola R, Poland RE, Guidotti A, et al.
(1994): Personality profiles and state aggressiveness in Finnish alco-
holic, violent offenders, fire setters, and healthy-volunteers. Arch Gen
24. Dom G, D’Haene P, Hulstijn W, Sabbe B (2006): Impulsivity in abstinent
a discounting task. Addiction 101:50–59.
25. Lejoyeux M, Feuche N, Loi S, Solomon J, Ades J (1998): Impulse-control
disorders in alcoholics are related to sensation seeking and not to im-
pulsivity. Psychiatry Res 81:149–155.
26. Dougherty DM, Marsh-Richard DM, Hatzis ES, Nouvion SO, Mathias CW
27. Marczinski CA, Combs SW, Fillmore MT (2007): Increased sensitivity to
the disinhibiting effects of alcohol in binge drinkers. Psychol Addict
28. Bechara A, Dolan S, Hindes A (2002): Decision-making and addiction
(part II): Myopia for the future or hypersensitivity to reward? Neuropsy-
29. Hariri AR, Brown SM, Williamson DE, Flory JD, de Wit H, Manuck SB
(2006): Preference for immediate over delayed rewards is associated
with magnitude of ventral striatal activity. J Neurosci 26:13213–13217.
30. Moeller FG, Barratt ES, Dougherty DM, Schmitz JM, Swann AC (2001):
Psychiatric aspects of impulsivity. Am J Psychiatry 158:1783–1793.
31. Green L, Myerson J (2004): A discounting framework for choice with
delayed and probabilistic rewards. Psychol Bull 130:769–792.
32. Kuntsche E, Knibbe R, Gmel G, Engels R (2006): Who drinks and why? A
review of socio-demographic, personality, and contextual issues be-
33. Tobler PN, O’Doherty JP, Dolan RJ, Schultz W (2007): Reward value
coding distinct from risk attitude-related uncertainty coding in human
reward systems. J Neurophysiol 97:1621–1632.
hyporesponsiveness during reward anticipation in attention-deficit/
hyperactivity disorder. Biol Psychiatry 61:720–724.
35. Strohle A, Stoy M, Wrase J, Schwarzer S, Schlagenhauf F, Huss M, et al.
(2008): Reward anticipation and outcomes in adult males with atten-
tion-deficit/hyperactivity disorder. Neuroimage 39:966–972.
36. van Gaalen MM, van Koten R, Schoffelmeer ANM, Vanderschuren LJMJ
(2006): Critical involvement of dopaminergic neurotransmission in im-
pulsive decision making. Biol Psychiatry 60:66–73.
37. Cardinal RN, Pennicott DR, Sugathapala CL, Robbins TW, Everitt BJ
(2001): Impulsive choice induced in rats by lesions of the nucleus ac-
cumbens core. Science 292:2499–2501.
38. Cardinal RN, Howes NJ (2005): Effects of lesions of the nucleus accum-
rewards in rats. BMC Neurosci 6:37.
college students. Addict Behav 28:1167–1173.
40. Mitchell JM, Fields HL, D’Esposito M, Boettiger CA (2005): Impulsive
41. Cloninger CR, Bohman M, Sigvardson S (1981): Inheritance of alcohol-
abuse—Cross-fostering analysis of adopted men. Arch Gen Psychiatry
42. Higley JD (2001): Individual differences in alcohol-induced aggres-
sion—A nonhuman-primate model. Alcohol Res Health 25:12–19.
43. King JA, Tenney J, Rossi V, Colamussi L, Burdick S (2003): Neural sub-
strates underlying impulsivity. Roots of Mental Illness in Children 1008:
45. Hamilton M (1960): A rating scale for depression. J Neurol Neurosurg
Angstinventar. Weinheim, Germany: Beltz Verlag.
47. First MB, Spitzer RL, Gibbon M, Williams J (2001): Structured Clinical
Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition
48. Skinner HA, Horn JL (1984): Alcohol Dependence Scale: Users Guide. To-
ronto: Addiction Research Foundation.
49. Anton RF (2000): Obsessive-compulsive aspects of craving: Develop-
ment of the Obsessive Compulsive Drinking Scale. Addiction 95:S211–
50. First M, Spitzer R, Gibbon M, Williams J (1997): Structured Clinical Inter-
view for DSM-IV Personality Disorders (SCID-II). Washington, DC: Ameri-
can Psychiatric Press, Inc.
51. Oldfield RC (1971): The assessment and analysis of handedness: The
Edinburgh Inventory. Neuropsychologia 9:97–113.
52. Hollingshead AA (1975): Four-Factor Index of Social Status. New Haven,
CT: Department of Sociology, Yale University.
54. Hoddes E, Zarcone V, Smythe H, Phillips R, Dement WC (1973): Quan-
tification of sleepiness: A new approach. Psychophysiology 10:431–
55. Jensen MP, Karoly P, Braver S (1986): The measurement of clinical pain
intensity—a comparison of 6 methods. Pain 27:117–126.
56. David SP, Munafo MR, Johansen-Berg H, Mackillop J, Sweet LH, Cohen
RA, et al. (2007): Effects of acute nicotine abstinence on cue-elicited
ventral striatum/nucleus accumbens activation in female cigarette
smokers: A functional magnetic resonance imaging study. Brain Imag-
57. Tanabe J, Crowley T, Hutchison K, Miller D, Johnson G, Du YP, et al.
(2008): Ventral striatal blood flow is altered by acute nicotine but not
withdrawal from nicotine. Neuropsychopharmacology 33:627–633.
58. Wilson SJ, Sayette MA, Delgado MR, Fiez JA (2008): Effect of smoking
opportunity on responses to monetary gain and loss in the caudate
Brain: An Approach to Medical Cerebral Imaging. Stuttgart, Germany:
Georg Thieme Verlag.
61. Gilman JM, Hommer DW (2008): Modulation of brain response to emo-
Cue-induced activation of the striatum and medial prefrontal cortex is
associated with subsequent relapse in abstinent alcoholics. Psycho-
63. Reuter J, Raedler T, Rose M, Hand I, Glascher J, Buchel C (2005): Patho-
logical gambling is linked to reduced activation of the mesolimbic
reward system. Nat Neurosci 8:147–148.
65. Mourao-Miranda J, Ecker C, Sato JR, Brammer M (2009): Dynamic
changes in the mental rotation network revealed by pattern recogni-
tion analysis of fMRI data. J Cogn Neurosci 21:890–904.
cortical circuits—parallel substrates for motor, oculomotor, prefrontal
and limbic functions. Prog Brain Res 85:119–146.
67. Alexander GE, DeLong MR, Strick PL (1986): Parallel organization of
functionally segregated circuits linking basal ganglia and cortex. Annu
68. Daw ND, Kakade S, Dayan P (2002): Opponent interactions between
serotonin and dopamine. Neural Netw 15:603–616.
69. Heinz A, Siessmeier T, Wrase J, Hermann D, Klein S, Grusser SM, et al.
(2004): Correlation between dopamine D(2) receptors in the ventral
striatum and central processing of alcohol cues and craving. Am J Psy-
70. Volkow ND, Wang GJ, Telang F, Fowler JS, Logan J, Childress AR, et al.
(2006): Cocaine cues and dopamine in dorsal striatum: Mechanism of
craving in cocaine addiction. J Neurosci 26:6583–6588.
A. Beck et al.
BIOL PSYCHIATRY 2009;66:734–742 741
71. Bjork JM, Smith AR, Hommer DW (2008): Striatal sensitivity to reward Download full-text
72. Trouillet R, Gana K (2008): Age differences in temperament, character
and depressive mood: A cross-sectional study. Clin Psychol Psychother
73. Heinz A, Dufeu P, Kuhn S, Dettling M, Graf K, Kurten I, et al. (1996):
Psychopathological and behavioral correlates of dopaminergic sen-
sitivity in alcohol-dependent patients. Arch Gen Psychiatry 53:1123–
psychological and psychopathological conceptualization. Nervenarzt
76. Verdejo-Garcia A, Lawrence AJ, Clark L (2008): Impulsivity as a vulnera-
risk research, problem gamblers and genetic association studies. Neu-
742 BIOL PSYCHIATRY 2009;66:734–742
A. Beck et al.