Dysfunction of reward processing correlates with alcohol
craving in detoxified alcoholics
Jana Wrase,aFlorian Schlagenhauf,aThorsten Kienast,aTorsten Wüstenberg,b
Felix Bermpohl,aThorsten Kahnt,aAnne Beck,aAndreas Ströhle,a
Georg Juckel,cBrian Knutson,dand Andreas Heinza,⁎
aDepartment of Psychiatry and Psychotherapy, Charité, Universitätsmedizin Berlin, Charité Campus Mitte (CCM), Berlin, Germany
bDepartment of Medical Psychology, Georg-August-University, Göttingen, Germany
cDepartment of Psychiatry, Ruhr-University Bochum, Germany
dDepartment of Psychology, Stanford University, CA 94305, USA
Received 25 July 2006; revised 7 November 2006; accepted 15 November 2006
Available online 8 February 2007
Objective: Alcohol dependence may be associated with dysfunction of
mesolimbic circuitry, such that anticipation of nonalcoholic reward
fails to activate the ventral striatum, while alcohol-associated cues
continue to activate this region. This may lead alcoholics to crave the
pharmacological effects of alcohol to a greater extent than other
conventional rewards. The present study investigated neural mechan-
isms underlying these phenomena.
Methods: 16 detoxified male alcoholics and 16 age-matched healthy
volunteers participated in two fMRI paradigms. In the first paradigm,
alcohol-associated and affectively neutral pictures were presented,
whereas in the second paradigm, a monetary incentive delay task
(MID) was performed, in which brain activation during anticipation of
monetary gain and loss was examined. For both paradigms, we assessed
the association of alcohol craving with neural activation to incentive cues.
Results: Detoxified alcoholics showed reduced activation of the ventral
striatum during anticipation of monetary gain relative to healthy
controls, despite similar performance. However, alcoholics showed
increased ventral striatal activation in response to alcohol-associated
cues. Reduced activation in the ventral striatum during expectation of
monetary reward, and increased activation during presentation of
alcohol cues were correlated with alcohol craving in alcoholics, but not
Conclusions: These results suggest that mesolimbic activation in
alcoholics is biased towards processing of alcohol cues. This might
explain why alcoholics find it particularly difficult to focus on
© 2006 Elsevier Inc. All rights reserved.
Keywords: Ventral striatum; Reward system; fMRI; Alcoholism; Craving
Alcohol addiction is based on well-learned behavior of alcohol
consumption, which eventually becomes habitual (Everitt and Wolf,
2002; Robbins and Everitt, 2002). Stimuli associated with alcohol can
serve as conditioned cues that promote alcohol consumption (O’Brien
et al., 1998; Stewart et al., 1984). This process may be mediated by
recruitment of circuitry innervated by midbrain dopamine neurons,
since presentation of alcohol-associated stimuli elicits increased
activation of the striatum, anterior cingulate cortex (ACC), amygdala,
and medial prefrontal cortex (mPFC) in alcoholics, relative to healthy
2001; Tapert et al., 2004; Wrase et al., 2002).
However, associative learning alone is not sufficient to explain
the compulsive behavior of alcoholics. Chronic consumption of
alcohol and other drugs of abuse may additionally derange
motivation, leading to pathological “wanting” of the abused
substance (craving) (Berridge and Robinson, 2003), and excessive
attribution of incentive salience to drug-related cues. Thus, drug-
related cues may selectively and effectivelytriggerrelapse (Robbins
and Everitt, 1999; Robinson and Berridge, 2003). Changes wrought
in neural circuits by years of chronic abuse may not reverse, even
after extended abstinence, making addiction a chronic condition
(O’Brien, 2005). These changes could affect the function of
mesolimbic circuits in a number of ways. For instance, alcohol
abuse may simply blunt responsiveness to conventional reward-
indicating cues. On the other hand, alcohol abuse may “hijack”
mesolimbic circuitry—increasing the salience of drug-associated
cues at the expense of conventional reward-indicating cues (Kalivas
and Volkow, 2005; Nesse and Berridge, 1997).
fMRI-studies indicate that conventional reward-indicating cues
increase ventral striatal (including the nucleus accumbens) oxyge-
nation (hereafter, “activation”) in healthy volunteers (Aharon et al.,
NeuroImage 35 (2007) 787–794
⁎Corresponding author. Department of Psychiatry, Charité, Universitäts-
medizin Berlin (Charité Campus Mitte), Schumannstr. 20/21, 10117 Berlin,
Germany. Fax: +49 30 450517921.
E-mail address: Andreas.Heinz@charite.de (A. Heinz).
Available online on ScienceDirect (www.sciencedirect.com).
1053-8119/$ - see front matter © 2006 Elsevier Inc. All rights reserved.
2001; Breiter et al., 2001; Knutson et al., 2001b; Stark et al., 2005).
In detoxified alcoholics, alcohol cues can also activate the ventral
striatum (Braus et al., 2001; Myrick et al., 2004; Wrase et al., 2002).
However, it is not clear how detoxified alcoholics respond to
conventional reward-indicating cues, and whether this reflects their
neural response to alcohol cues. If alcoholics had a decreased
response to conventional reward-indicating cues but an increased
response to alcohol-associated cues, such an effect would be
consistent with the notion that drugs of abuse can “hijack” and
reorganize the priority of reward processing (Nesse and Berridge,
1997). To examine this possibility, we examined whether (1) cues
that predicted potential monetary gain would elicit less activation of
the ventral striatum in detoxified alcoholics than in healthy control
subjects, whether (2) alcohol cues would elicit more ventral striatal
types of activations would correlate with degree of alcohol craving.
Subjects and instruments
16 male alcoholics (mean age: 42.4±7.5, range: 25–57 years)
and 16 male age-matched healthy control subjects (mean age:
exceeding 4 mm from one scan to the next. The local ethics
committee approved the study, and written informed consent was
obtained from all participants after the procedures had been fully
explained. All patients were diagnosed as alcohol dependent
according to ICD-10 and DSM-IV criteria and had no other
psychiatric axis I disorders, no past history ofdependency or current
abuse of other drugs as verified by random urine drug testing and
SCID-interview (First et al., 2001). Patients had abstained from
alcohol in an in-patient detoxification treatment program for
11.5±7.5 days (range 5–37 days) as verified by random ad-
ministration of alcohol breath test. Besides medical care they had
received psycho-education in group therapy. All subjects were
free of benzodiazepine or chlormethiazole medication for at least
1 week. The severity of alcoholism was assessed with the Alcohol
Dependence Scale (Skinner and Horn, 1984). The severity of
alcohol craving during the last seven drinking days was measured
with the Obsessive Compulsive Drinking Scale (OCDS) (Anton,
2000) (Table 1). Healthy control subjects had no psychiatric axis I
or II disorder (SCID-interview) (First et al., 1997, 2001) and no
history of psychiatric disorders (including alcohol dependence) in
first-degree relatives, while 6 of the alcohol dependent subjects had
a positive family history for alcohol dependence.
The Hamilton Depression Rating Scale (Hamilton, 1960) was
used to quantify depressive symptoms. Before and after the fMRI-
paradigm mood was rated on a visual analog scale. All participants
were right-handed as confirmed by the Edinburgh Handedness
Inventory (Oldfield, 1971) (Table 1). We verified that smokers
were not in a withdrawal state during acquisition of scans.
Monetary incentive delay (MID) task
We used a “monetary incentive delay” (MID) task as described
previously (Knutson et al., 2001a). In brief, subjects were
examined with fMRI during trials in which they anticipated
potential monetary gain, loss, or no consequences. The partici-
pants’ monetary gain depended on their performance in a simple
reaction time task at the end of each trial, which required pressing a
button during the brief presentation of a visual target. Task details
are given in Fig. 1. Subjects were informed that they would receive
the monetary reward they earned after the scanning session, and the
money was shown in cash to them before entering the scanner.
During structural scans, participants completed a short practice
version of the task (in order to minimize later learning effects) for
which they did not receive monetary payment. Subsequently
functional scans were collected. After scanning, subjects retro-
spectively rated their own exertion in response to each of the seven
cues on a visual analog scale (Table 1).
We used affectively unpleasant, pleasant and neutral pictures
from the International Affective Picture System (Center for the
Study of Emotion and Attention [CSEA-NIMH], 1999) and
standardized alcohol-related pictures (Wrase et al., 2002). Neutral
stimuli were all inanimate and matched for complexity and colour
with the alcohol pictures, which have previously been shown to
elicit significant activation in mesolimbic and other regions in
alcoholics (Wrase et al., 2002). Each category included 18 pictures,
resulting in 72 total trials. fMRI results for the affective pictures
will be reported elsewhere. Participants were instructed to
passively view the stimuli, since even simple rating tasks can
influence amygdala activation (Taylor et al., 2003). To control for a
decrease in attention, participants had to confirm every viewed
picture with a button press with the right thumb. Pictorial stimuli
were presented for 2 s using an event-related design and were
arranged in an individually randomised order for each subject.
Intertrial interval (ITI) was randomly jittered between 2 and 4
acquisition times, which means that the duration between trial t and
t+1 was randomly varied between 4.6 and 9.2 s in order to sample
the hemodynamic response with different data points. A fixation
cross was presented during each ITI.
Functional magnetic resonance imaging
fMRI was performed on a 1.5 T scanner (Magnetom VISION
Siemens®) equipped with a standard circularly polarized head coil
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 %
Mean reaction time in ms
Total gain in euros
Hamilton Depression Scale
Stanford Sleepiness Scale
Edinburgh Handedness Inventory
Number of cigarettes per day
Years of education
Severity of alcohol dependence
Duration of alcohol dependence (years)
No. of detoxifications
788 J. Wrase et al. / NeuroImage 35 (2007) 787–794
(CP-Headcoil) using gradient echo–echo planar imaging. For
both fMRI-paradigms the following parameters were used: GE-
EPI, TE=40 ms, flip angle=90 degrees, matrix=64×64, voxel
size=4×4×3.3 mm. For the MID-paradigm 18 slices were
collected every 1.8 sec approximately parallel to the bicommis-
sural plane (ac–pc-plane), covering the inferior part of the
frontal lobe (superior border above the caudate nucleus), the
whole temporal lobe and large parts of the occipital region.
fMRI volume acquisitions were time-locked to the offset of each
cue and thus were acquired during anticipatory delay periods.
Six fMRI volumes were acquired per trial, resulting in 450 total
For the alcohol-cue paradigm, 24 slices covering the whole
head parallel to ac–pc plane were collected every 2.3 s. Five fMRI
volumes were acquired per trial, resulting in 380 total volumes. For
anatomical reference a 3D MPRAGE (Magnetization Prepared
Rapid Gradient Echo, TR=9.7 ms; TE=4 ms; flip angle 12
degrees; matrix=256×256, voxel size 1×1×1 mm) image data set
was acquired. Head movement was minimized using a vacuum
fMRI data analysis
Three alcoholic subjects only participated in one paradigm
(MID), resulting in a total of 16 subjects for MID paradigm and a
total of 13 subjects in the alcohol cue paradigm. Functional MRI
data were analyzed using SPM2 (http://www.fil.ion.ucl.ac.uk/spm).
The first three volumes of each functional time series were
discarded in order to avoid non steady-state effects caused by T1
saturation. Sinc interpolation was used to realign all volumes to the
middle volume to correct for between-scan movements and to
remove signals correlated with head motion. All six movement
parameters (translation; x, y, z and rotation; pitch, roll, yaw) were
included in the statistical model. The structural 3D data set was co-
registered with the first T2* image. The coregistered structural
image was spatially normalized to the standard template provided
by the Montreal Neurological Institute (MNI-Template) using an
automated spatial transformation (12-parameter affine transforma-
tion followed by non-linear iterations using 7×8×7 basis
functions), resulting in an isometric voxel size of 4×4×4 mm3.
The normalized images were smoothed with a Gaussian kernel
designed to optimize the signal-to-noise ratio in small subcortical
structures of interest (e.g., the nucleus accumbens) (full width at
half maximum=4 mm).
Functional MRI data were then analyzed in the context of the
general linear model (GLM) approach. Changes in the BOLD
response can be assessed by linear combinations of the estimated
GLM parameters (beta values) and are contained in the individual
contrast images (equivalent to percent signal change or effect size).
For the MID task, this analysis was performed by modelling the
different conditions (“anticipation of gain”, “anticipation of loss”
and “anticipation of no outcome”) as explanatory variables
Fig. 1. Task structure for a representative trial. In each trial, volunteers saw one of seven shapes (“cue”; 250 ms), which indicated that they would, in a few
moments, be able to respond and either win or avoid losing different amounts of money (3.00 euros, 0.60 euros, or 0.10 euros) or that they should respond for no
monetary outcome. The different cues are shown at the bottom of the figure. Cues signaling potential gain were denoted by circles, potential loss was denoted by
squares,andno monetaryoutcomewas denotedby triangles;the possible amountof moneythat subjects wereableto win wasindicatedby onehorizontalline for
0.10 euro, two horizontal lines for 0.60 euro and three horizontal lines for 3.00 euro. Similarly, loss cues signaled the possibility of losing the same amounts of
money.After the cue,volunteerswaitedavariableinterval(delay: 2250,2500or 2750ms) andthenrespondedtoa whitetarget squarethat appearedfor avariable
length of time (target) by pressing a button. To succeed in a given trial, volunteers had to press the button while the target was visible. Immediately after target
presentation, feedback appeared (“feedback”; 1.650 ms), notifying volunteers that they had won or lost money and indicating their cumulative total at that point.
Through an adaptive algorithm for the time of the target presentation, subjects succeeded on average in 67% of the trials. The inter-trial interval was 4000 ms.
Trial types were randomly ordered within each session.
789J. Wrase et al. / NeuroImage 35 (2007) 787–794
convolved with the gamma variate function described by Cohen
(1997) and similar to the method of Knutson et al. (2001b) and
Breiter et al. (2001). For the alcohol paradigm we used the
hemodynamic response function (HRF) provided by SPM2 with
time deviation, that contrasts the BOLD response during alcohol-
related vs. neutral pictures.
To detect group differences between alcohol dependent
patients and healthy controls, individual contrast images (i.e. the
BOLD response differences) of all subjects in each group were
included in a second level random effects analysis, comparing
within-group activation with a one-sample t test and between-
group differences with a two-sample t test. For this exploratory
analysis we used a significance level of p<0.001 uncorrected and
a cluster threshold of k>1. To test the hypothesis of activation in
the ventral striatum elicited by both paradigms, SPM’s small
volume correction (S.V.C.) was performed using binary masks
from the publication-based probabilistic MNI-atlas (Fox and
Lancaster, 2002) at the threshold of 0.5 probability (please refer to
html, access date 19.06.2006) and a significance level of p<0.05
FWE-corrected for the volume of interest (VOI, left and right
Correlation analysis between ventral striatal activation and
In the confirmatory analysis, we tested the hypothesis that the
severity of alcohol craving (1) is negatively correlated with ventral
striatal activation during anticipation of monetary reward versus
nonreward but (2) is positively correlated with ventral striatal
activation during viewing of alcohol versus neutral pictures. These
contrasts within the ventral striatal VOI were defined by a
probabilistic map (see above) and were correlated with the craving
score of the obsessive compulsive drinking scale (OCDS (Anton,
2000)) using SPM2 simple regression analysis with a significance
level of p<0.05 FWE-corrected for the VOI.
We also explored correlations between alcohol craving and
functional activations outside of the ventral striatum as indicated
in the group contrast (alcoholics vs. controls) for both paradigms.
In the exploratory analysis, we further assessed correlations
between the activation contrasts for the MID task and the alcohol
paradigm described above and potentially confounding variables
(age, severity of alcohol dependence (Skinner and Horn, 1984),
duration of alcohol dependence,number of detoxifications, duration
of abstinence before scanning, number of cigarettes smoked per
day, and severity of depression (Hamilton, 1960) using SPM2
simple regression analysis and a significance level of p<0.001,
Alcoholics reported stronger alcohol craving than healthy
controls (t=10.65, p<0.001). No significant group difference was
observed in the Hamilton Depression Rating Scale (t=1.53,
p=0.142) (Hamilton, 1960). Performing the MID task, there was
also no significant difference between alcoholics and controls in
mean hit rate (t=1.20, p=0.240), the mean reaction time (t=0.01,
p=0.996), the amount of money that was gained (t=1.89,
p=0.068), self-reported alertness during the task assessed with
the Stanford Sleepiness Scale (Hoddes et al., 1973) (t=−0.36,
p=0.720), or self-reported effort to achieve monetary gains
(t=0.91, p=0.371) or to prevent losses as assessed with visual
analog scales (VAS) (t=0.24, p=0.810) (Table 1). For the mood
ratings: there were no main effects of diagnosis (F(1,42)=0.013,
p=0.910) and time (F(1,42)=0.229, p=0.634) nor an interaction
diagnosis×time (F(1,42)=1.33, p=0.245). This shows that the
stimulus materials did not elicit significant changes in mood
Anticipation of gain
During anticipation of monetary gain versus no outcome,
healthy control subjects showed significant bilateral activation of
the ventral striatum including the nucleus accumbens (Table 2). In
detoxified alcoholics, reward anticipation did not produce
significant activations in the ventral striatum. Instead, alcoholics
showed activations in the thalamus and in the lateral orbital frontal
cortex (Table 2). Activation in the ventral striatum was only visible
in the alcoholic group at a more liberal threshold on the right side
(p<0.01; t=3.23; (x y z)=(12 15 −4)).
When controls were directly compared with detoxified alco-
holics, they showed significantly stronger activations in the left
ventral striatum, in the right posterior putamen and in the right head
of the caudate nucleus during gain versus nongain anticipation
(Table 2, Fig. 2). This difference was not due to alterations in the
time course of the BOLD response (see Supplementary Fig. S1 for
the ventral striatum (x y z)=(−16 12 −4)). Compared with healthy
controls, alcoholics showed no stronger activations for this contrast.
In alcoholics but not controls, the severity of alcohol craving
correlated significantly and inversely only with the activation in the
bilateral ventral striatum during the anticipation of monetary gain
versus no outcome (left: t=3.61, p=0.014; (x y z)=(−8 15 −4);
the less the ventral striatum activated in response to the presentation
of monetary reward cues (Fig. 3).
and not to the significant activations in the posterior putamen or
caudate nucleus (see controls vs. alcoholics, Table 2), we also
correlated these activations with alcohol craving. Both correlations
were not significant (Supplementary Fig. S2). Furthermore, the
correlations between alcohol craving and both the activations in the
putamen and caudate were significantly smaller than the correlation
Brain activation elicited by the anticipation of potential gain compared with
the anticipation of no outcome (CON=controls, ALC=alcohol-dependent
Region Talairach coordinates
ControlsVentral striatum left
4.62Alcoholics Inferior frontal gyrus,
Caudate nucleus head
790J. Wrase et al. / NeuroImage 35 (2007) 787–794
between craving and the activation in the left ventral striatum
(z=2.79, p=0.0026 and z=3.25, p=0.00058, respectively).
Anticipation of loss
When healthy control subjects anticipated potential monetary
loss, they also showed significant activation of the bilateral ventral
striatum, including the nucleus accumbens. Exploratory analysis
additionally revealed activation in the left lateral globus pallidus
(Table 3). In detoxified alcoholics, no significant activation was
found in the ventral striatum during anticipation of potential loss.
Instead, alcoholics activated the left lateral orbital frontal cortex,
and left thalamus (Table 3).
When controls were directly compared with detoxified alco-
holics, they showed a significantly stronger activation in the left
ventral striatum during loss anticipation (Fig. 2). Compared with
healthy controls, alcoholics displayed a stronger activation in the
left middle frontal gyrus (Table 3).
Comparison of gain versus loss anticipation in alcoholics and
Even though we observed significant activation in the ventral
striatum during both gain and loss anticipation in healthy control
subjects, the increase was much stronger during gain compared to
loss anticipation (left: t=5.06, p=0.002; (x y z)=(−12 12 −4)).
However, there was no significant difference between anticipated
gain and loss in alcoholics.
Alcohol cues compared to neutral control cues
Pictures of alcohol-associated versus neutral cues elicited
significant activation in alcoholics in the right ventral striatum,
but not in healthy control subjects (Table 4). In the exploratory
analysis alcoholics showed also activation in the thalamus, middle
occipital gyrus, posterior cingulate and middle and superior
temporal gyrus (Table 4). In the direct comparison between
alcohol dependent patients and healthy controls, alcoholics
exposed a stronger activation in the thalamus, precuneus and
middle temporal and occipital gyrus. The ventral striatum was only
observable as a trend (t=3.20, p=0.102; (x y z)=(−20 4 0);
t=3.14, p=0.083; (x y z)=(16 0 0)).
The severity of alcohol craving correlated significantly and
positively only with the activation elicited by alcohol versus
neutral pictures in the right ventral striatum (t=4.50, p=0.016;
(x y z)=(20 4 −4)) in alcohol dependent subjects. Higher craving
was associated with a stronger activation. This association was not
observed in healthy volunteers.
Fig. 2. Brain activation in the ventral striatum elicited by visual stimuli that indicate potential gain versus no outcome (left) or potential loss versus no outcome
Fig. 3. Negative correlation between the severity of alcohol craving measured with the Obsessive–Compulsive Drinking Scale and brain activation in the ventral
striatum elicited by the contrast anticipation of monetary gain versus no outcome.
791 J. Wrase et al. / NeuroImage 35 (2007) 787–794
and not to the significant activations in the group comparison in the
thalamus, middle occipital gyrus or precuneus (see controls vs.
alcoholics, Table 4), we also correlated these activations with
alcohol craving but found no significant associations (Supplemen-
tary Fig. S3). Furthermore, these correlations were significantly
ventral striatum (z=2.81, p=0.0025, z=2.61, p=0.0045, z=3.56,
There were no significant associations between potentially
confounding variables such as the duration of abstinence before
scanning, severity of alcohol dependence, number of detoxifica-
tions, duration of dependence, number of cigarettes smoked per
day, severity of clinical depression, age, and the activation elicited
by alcohol vs. neutral pictures or anticipation of gain vs. no
outcome in the ventral striatum.
To the best of our knowledge, these findings demonstrate for
the first time that abstinent alcoholics show reduced activation of
the ventral striatum including the nucleus accumbens during
anticipation of conventional (i.e. monetary) rewards, and that this
reduction is correlated with increased alcohol craving. However,
alcoholic patients displayed increased activation in the ventral
striatum when confronted with alcohol cues, which was associated
with increased alcohol craving. These findings are not consistent
with a global deficit of brain activation in alcoholics (e.g., due to
hypoperfusion), or a selective blunting account of ventral striatal
activation, but instead with the notion that alcoholism selectively
diverts motivational resources away from conventional rewards
towards drug rewards (Garavan et al., 2000; Nesse and Berridge,
1997). Furthermore, explorative analysis showed that alcoholics
also invoked the orbitofrontal cortex to process anticipation of
monetary gain and loss, while controls used more subcortical
systems to do the same.
In healthy control subjects, activation of the ventral striatum
was elicited by both monetary reward and loss cues, which might
be interpreted as consistent with a “salience” interpretation of
ventral striatal activation (Zink et al., 2004). In healthy controls,
direct comparison of gain and loss anticipation revealed signifi-
cantly greater ventral striatal activation during gain anticipation,
which is consistent with earlier reports, and a predominant reward
anticipation account of ventral striatal activation (Knutson et al.,
2001a). Ikemoto and Panksepp (1999) have suggested that in the
nucleus accumbens, stressful situations can also induce phasic
dopamine release if an unpleasant outcome can be avoided by a
motor response. A similar situation is found in this study, where
subjects were able to avoid monetary loss with a fast motor
response. Such an activation of the ventral striatum was not found
in detoxified alcoholics. The lack of differential ventral striatal
activation during anticipation of gain versus loss in alcoholics may
be due to the low levels of activation shown during anticipation of
all monetary incentives by this group. Failure to activate brain
circuitry associated with anticipation of new, reward-indicating
stimuli may thus contribute to craving for long established, well-
known effects of alcohol.
While drug and drug-associated stimuli can elicit dopamine
release in the ventral striatum and thus reinforce drug intake during
the acquisition of dependent behavior, chronic drug intake is
associated with neuroadaptation of glutamatergic neurotransmis-
sion in the ventral striatum and limbic cortex (McFarland et al.,
2003; Vorel et al., 2001). After alcohol withdrawal, cue exposure
may stimulate an attentional response towards cues regularly
associated with drug intake via glutamatergic neurotransmission in
the prefrontal cortex and ventral striatum (Kalivas and Volkow,
2005) and NMDA receptor antagonist may therefore play an
important role in the treatment of alcoholism (Krystal et al., 2003).
Low presynaptic dopamine production and release and down-
regulated D2 receptors in the ventral striatum (Heinz et al., 2004,
2005; Martinez et al., 2005) may interfere with the ability of
detoxified alcoholics to adequately respond to new, reward-
associated stimuli. Therefore, alcoholics may suffer from a deficit
crave alcohol-associated effects during exposure to alcohol cues.
However,thisdoesnotmean thatactivationintheventralstriatum is
studies have shown that decreased dopamine release in the striatum
results in reduced baseline activation in the orbitofrontal cortex and
the cingulate gyrus but enhanced activation in these same regions in
response to drug-related cues, which was correlated with drug
Brain activation elicited by the anticipation of potential loss compared with
the anticipation of no outcome (CON=controls, ALC=alcohol-dependent
SampleRegion Talairach coordinates
Controls Ventral striatumleft
15 −7 4.46
11 −7 4.12
4 0 5.39
19 −4 4.69
Lateral globus pallidus left
Inferior frontal gyrus,
CON>ALC Ventral striatum
ALC>CON Middle frontal gyrus,
12 −4 3.40
51 1 4.46
Brain activation elicited by alcohol pictures>neutral pictures (CON=
controls, ALC=alcohol-dependent patients)
Region Talairach coordinates
Middle occipital gyrus,
Posterior cingulate, BA 29 right
Middle temporal gyrus,
Superior temporal gyrus,
Middle occipital gyrus,
Precuneus, BA 31
Middle temporal gyrus,
−63 −31 −5 4.64
right63 −38 13 4.47
792 J. Wrase et al. / NeuroImage 35 (2007) 787–794
this study, but in the ventral striatum. While fMRI cannot address
cue-elicited BOLD response, future multimodal studies that
combine functional brain activation and neuroreceptor imaging
may help to elucidate the neurotransmitter correlates of altered
reward system activation in alcoholics.
elicited increased activation in the thalamus, posterior cingulate
cortex, temporal cortex, and in areas associated with visual
processing such as the middle occipital gyrus. It has previously
been shown that emotionally salient stimuli, in contrast to neutral
et al., 1998). As in the present study, these brain areas were also
activated by alcohol cues in previous studies (Grusser et al., 2004;
Tapert et al., 2004; Braus et al., 2001). In alcoholics, increased brain
activation elicited by alcohol pictures was observed in the thalamus,
which conveys sensory input to the frontal cortex (George et al.,
dose of the dopamine D2 receptor antagonist amisulpride (Hermann
et al., 2006). The posterior cingulate has previously been implicated
in processing of alcohol- and drug-related stimuli (Garavan et al.,
2000; Kosten et al., 2006; Tapert et al., 2003) and is thought to
function as a rapid relay for incentive valuation and episodic
memories (Kosten et al., 2006). Altogether, these additional brain
areas may participate in incentive salience attribution to visual cues
associated with alcohol.
Exploratory analyses also revealed greater activation in healthy
controls than in alcoholics in posterior putamen and caudate head
during the anticipation of monetary gain. In alcoholics, reduced
activation of these brain areas was not correlated with the severity
of alcohol craving, while the failure to activate the ventral striatum
was associated with stronger alcohol urges. The caudate has been
implicated in linking reward to behavior (Knutson and Cooper,
2005) and could be associated with greater learning benefit from
reward in healthy controls. Interestingly, anticipation of both gain
and loss elicited stronger activation of the orbitofrontal cortex and
thalamus in alcoholics. These hyperactivations may compensate
for dysfunction of the ventral striatum. Altered connectivity
between the OFC and limbic brain areas has been described in
opiate addicts (Daglish et al., 2003). Further studies should assess
the connectivity between frontal and subcortical brain areas in
alcoholism and its possible contribution to alcohol craving.
A potential limitation of the study is a difference in cigarettes
smoked per day between alcoholics and controls. However, in this
sample, cigarettes smoked per day did not correlate significantly
with activation observed in the ventral striatum. Previous fMRI
for task performance, because the task-dependent BOLD response
we matched controls and alcohol-dependent patients for hit rate and
amount of money gained and for their effort to achieve monetary
gains and to avoid loss. Since money is a secondary reinforcer, it
would be interesting to assess whether alcoholics also show
alterations in brain activation elicited by primary reinforcers.
Taken together, these findings suggest that detoxified alcoholics
failed to activate the ventral striatum during the anticipation of
conventional monetary rewards, and this decreased activation was
associated with alcohol craving. On the other hand, alcoholics
displayed increased activation of the ventral striatum when
confronted with alcohol cues, and this increased activation was
associated with alcohol craving. Since alcoholics’ ventral striatal
recruitment appears to be biased towards processing of alcohol
cues, alcoholics may find it particularly difficult to focus on
conventional rewards, and thus may find it difficult to seek and
enjoy alcohol-free situations. Further studies will have to examine
the role of this reprioritization of ventral striatal recruitment on the
risk for relapse in detoxified alcoholics, and whether long-term
abstinence or therapy can facilitate a normalization of reward
This study was supported by the German Research Foundation
(Deutsche Forschungsgemeinschaft; HE 2597/4-2). We do not
have any commercial or financial involvements that might present
an appearance of a conflict of interest.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.neuroimage.2006.11.043.
Aharon, I., Etcoff, N., Ariely, D., Chabris, C.F., O’Connor, E., Breiter, H.C.,
2001. Beautiful faces have variable reward value: fMRI and behavioral
evidence. Neuron 32, 537–551.
Anton, R.F., 2000. Obsessive–compulsive aspects of craving: development
of the Obsessive Compulsive Drinking Scale. Addiction (Abingdon,
England) 95, S211–S217.
Berridge, K.C., Robinson, T.E., 2003. Parsing reward. Trends Neurosci. 26,
Braus, D.F., Wrase, J., Grusser, S., Hermann, D., Ruf, M., Flor, H., Mann,
K., Heinz, A., 2001. Alcohol-associated stimuli activate the ventral
striatum in abstinent alcoholics. J. Neural Trans. 108, 887–894.
Breiter, H.C., Aharon, I., Kahneman, D., Dale, A., Shizgal, P., 2001.
Functional imaging of neural responses to expectancy and experience of
monetary gains and losses. Neuron 30, 619–639.
Callicott, J.H., Egan, M.F., Mattay, V.S., Bertolino, A., Bone, A.D.,
Verchinksi, B., Weinberger, D.R., 2003. Abnormal fMRI response of the
dorsolateral prefrontal cortex in cognitively intact siblings of patients
with schizophrenia. Am. J. Psychiatry 160, 709–719.
Center for the Study of Emotion and Attention [CSEA-NIMH], 1999. The
International Affective Picture System (Photographic Slides). Centers
for Research in Psychophysiology, University of Florida, Gainesville,
Cohen, M.S., 1997. Parametric analysis of fMRI data using linear systems
methods. NeuroImage 6, 93–103.
Daglish, M.R.C., Weinstein, A., Malizia, A.L., Wilson, S., Melichar, J.K.,
Lingford-Hughes, A., Myles, J.S., Grasby, P., Nutt, D.J., 2003.
Functional connectivity analysis of the neural circuits of opiate craving:
“more” rather than “different”? NeuroImage 20, 1964–1970.
Everitt, B.J., Wolf, M.E., 2002. Psychomotor stimulant addiction: a neural
systems perspective. J. Neurosci. 22, 3312–3320.
First, M.B., Spitzer, R.L., Gibbon, M., Williams, J., 1997. Structured
Clinical Interview for DSM-IV Personality Disorders, (SCID-II).
American Psychiatric Press, Inc., Washington, DC.
Interview for DSM-IV-TR Axis I Disorders, Research VersionPatient
Research. New York State Psychiatric Institute, New York.
Fox, P.T., Lancaster, J.L., 2002. Opinion: mapping context and content: the
brain map model. Nat. Rev., Neurosci. 3, 319–321.
793 J. Wrase et al. / NeuroImage 35 (2007) 787–794
Garavan, H., Pankiewicz, J., Bloom, A., Cho, J., Sperry, L., Ross, T.,
Salmeron, B., Risinger, R., Kelley, D., Stein, E., 2000. Cue-induced
cocaine craving: neuroanatomical specificity for drug users and drug
stimuli. Am. J. Psychiatry 157, 1789–1798.
George, M.S., Anton, R.F., Bloomer, C., Teneback, C., Drobes, D.J.,
Lorberbaum, J.P., Nahas, Z., Vincent, D.J., 2001. Activation of
prefrontal cortex and anterior thalamus in alcoholic subjects on exposure
to alcohol-specific cues. Arch. Gen. Psychiatry 58, 345–352.
Grusser, S.M., Wrase, J., Klein, S., Hermann, D., Smolka, M.N., Ruf, M.,
Weber-Fahr, W., Flor, H., Mann, K., Braus, D.F., Heinz, A., 2004. Cue-
induced activation of the striatum and medial prefrontal cortex is
associated with subsequent relapse in abstinent alcoholics. Psychophar-
macology 175, 296–302.
Hamilton, M., 1960. A rating scale for depression. J. Neurol., Neurosurg.
Psychiatry 23, 56–62.
Heinz, A., Siessmeier, T., Wrase, J., Hermann, D., Klein, S., Grusser-
Sinopoli, S.M., Flor, H., Braus, D.F., Buchholz, H.G., Grunder, G.,
Schreckenberger, M., Smolka, M.N., Rosch, F., Mann, K., Bartenstein,
P., 2004. Correlation between dopamine D-2 receptors in the ventral
striatum and central processing of alcohol cues and craving. Am. J.
Psychiatry 161, 1783–1789.
Heinz, A., Siessmeier, T., Wrase, J., Buchholz, H.G., Grunder, G.,
Kumakura, Y., Cumming, P., Schreckenberger, M., Smolka, M.N.,
Rosch, F., Mann, K., Bartenstein, P., 2005. Correlation of alcohol
craving with striatal dopamine synthesis capacity and D-2/3 receptor
availability: a combined [18F]DOPA and [18F]DMFP PET Study in
detoxified alcoholic patients. Am. J. Psychiatry 162, 1515–1520.
Hermann, D., Smolka, M.N., Wrase, J., Klein, S., Nikitopoulos, J., Georgi,
A., Braus, D.F., Flor, H., Mann, K., Heinz, A., 2006. Blockade of cue-
induced brain activation of abstinent alcoholics by a single administra-
tionof amisulprideas measuredwithfMRI.Alcohol.,Clin.Exp.Res.30,
Hoddes, E., Zarcone, V., Smythe, H., Phillips, R., Dement, W., 1973.
Quantification of sleepiness: a new approach. Psychophysiology 10,
Hommer, D.W., 1999. Functional imaging of craving. Alcohol Res. Health
Ikemoto,S.,Panksepp,J., 1999.The role of nucleusaccumbensdopamine in
motivated behavior: a unifying interpretation with special reference to
reward-seeking. Brain Res. Brain Res. Rev. 31, 6–41.
Kalivas, P.W., Volkow, N.D., 2005. The neural basis of addiction: a
pathology of motivation and choice. Am. J. Psychiatry 162, 1403–1413.
Knutson, B., Adams, C., Fong, G., Walker, J., Hommer, D., 2001a.
Parametric fMRI confirms selective recruitment of nucleus accumbens
during anticipation of monetary reward. NeuroImage 13, S430.
Knutson, B., Cooper, J.C., 2005. Functional magnetic resonance imaging of
reward prediction. Curr. Opin. Neurol. 18, 411–417.
Knutson, B., Adams, C.M., Fong, G.W., Hommer, D., 2001b. Anticipation
of increasing monetary reward selectively recruits nucleus accumbens.
J. Neurosci. 21, 1–5 (art-RC159).
Kosten, T.R., Scanley, B.E., Tucker, K.A., Oliveto, A., Prince, C., Sinha, R.,
Potenza, M.N., Skudlarski, P., Wexler, B.E., 2006. Cue-induced brain
activity changes and relapse in cocaine-dependent patients. Neuropsy-
chopharmacology 31, 644–650.
Krystal, J.H., Petrakis, I.L., Krupitsky, E., Schutze, C., Trevisan, L.,
D’Souza, D.C., 2003. NMDA receptor antagonism and the ethanol
intoxication signal—from alcoholism risk to pharmacotherapy. Ann. N.
Y. Acad. Sci. 1003, 176–184.
Lang, P.J., Bradely, M.M., Fitzsimmons, J.R., Cuthbert, B.N., Scott, J.D.,
Moulder, B., Nangia, V., 1998. Emotional arousal and activation of the
visual cortex: an fMRI analysis. Psychophysiology 35, 199–210.
Martinez, D., Gil, R., Slifstein, M., Hwang, D., Huang, Y., Perez, A.,
Kegeles, L., Talbot, P., Evans, S., Krystal, J., Laruelle, M., Abi-
Dargham, A., 2005. Alcohol dependence is associated with blunted
dopamine transmission in the ventral striatum. Biol. Psychiatry 58,
McFarland, K., Lapish, C.C., Kalivas, P.W., 2003. Prefrontal glutamate
release into the core of the nucleus accumbens mediates cocaine-
induced reinstatement of drug-seeking behavior. J. Neurosci. 23,
Myrick, H., Anton, R.F., Li, X.B., Henderson, S., Drobes, D., Voronin, K.,
George, M.S., 2004. Differential brain activity in alcoholics and social
drinkers to alcohol cues: relationship to craving. Neuropsychopharma-
cology 29, 393–402.
Nesse, R.M., Berridge, K.C., 1997. Psychoactive drug use in evolutionary
perspective. Science 278, 63–66.
O’Brien, C.P., 2005. Anticraving medications for relapse prevention: a
possible new class of psychoactive medications. Am. J. Psychiatry 162,
O’Brien, C.P., Childress, A.R., Ehrman, R., Robbins, S.J., 1998.
Conditioning factors in drug abuse: can they explain compulsion?
J. Psychopharmacol. 12, 15–22.
Oldfield, R., 1971. The assessment and analysis of handedness: the
Edinburgh inventory. Neuropsychologia 9, 97–113.
Robbins,T.W.,Everitt,B.J.,1999.Drug addiction: badhabitsaddup.Nature
Robbins, T.W., Everitt, J., 2002. Limbic-striatal memory systems and drug
addiction. Neurobiol. Learn. Mem. 78, 625–636.
Robinson, T.E., Berridge, K.C., 2003. Addiction. Annu. Rev. Psychol. 54,
Schneider, F., Habel, U., Wagner, M., Franke, P., Salloum, J.B., Shah, N.J.,
Toni, I., Sulzbach, C., Honig, K., Maier, W., Gaebel, W., Zilles, K.,
2001. Subcortical correlates of craving in recently abstinent alcoholic
patients. Am. J. Psychiatry 158, 1075–1083.
Skinner, H.A., Horn, J.L., 1984. Alcohol Dependence Scale: Users Guide.
Addiction Research Foundation, Toronto.
Stark, R., Schienle, A., Girod, C., Walter, B., Kirsch, P., Blecker, C., Ott, U.,
Schafer, A., Sammer, G., Zimmermann, M., Vaitl, D., 2005. Erotic and
disgust-inducing pictures—differences in the hemodynamic responses
of the brain. Biol. Psychol. 70, 19–29.
Stewart, J., Dewit, H., Eikelboom, R., 1984. Role of unconditioned and
conditioned drug effects in the self-administration of opiates and
stimulants. Psychol. Rev. 91, 251–268.
Tapert, S., Cheung, E., Brown, G., Lawrence, R., Paulus, M., Schwinsburg,
A., Meloy, M., Brown, S., 2003. Neural response to alcohol stimuli in
adolescents with alcohol use disorder. Arch. Gen. Psychiatry 60,
Tapert, S., Brown, G., Baratta, M., Brown, S., 2004. fMRI BOLD response
to alcohol stimuli in alcohol dependent young women. Addict. Behav.
Taylor, S.F., Phan, K.L., Decker, L.R., Liberzon, I., 2003. Subjective rating
of emotionally salient stimuli modulates neural activity. NeuroImage 18,
Volkow, N.D., Fowler, J.S., Wang, G.J., Swanson, J.M., 2004. Dopamine in
drug abuse and addiction: results from imaging studies and treatment
implications. Mol. Psychiatry 9, 557–569.
Vorel, S.R., Liu, X., Hayes, R.J., Spector, J.A., Gardner, E.L., 2001. Relapse
to cocaine-seeking after hippocampal theta burst stimulation. Science
Wrase, J., Gruesser, S.M., Klein, S., Diener, C., Hermann, D., Flor, H.,
Mann, K., Braus, D.F., Heinz, A., 2002. Development of alcohol-
associated cues and cue-induced brain activation in alcoholics. Eur.
Psychiatry 17, 287–291.
2004. Human striatal responses to monetary reward depend on saliency.
Neuron 42, 509–517.
794J. Wrase et al. / NeuroImage 35 (2007) 787–794