Neurobiological correlates of problem gambling in a quasi-realistic blackjack scenario as revealed by fMRI

Article · February 2010with268 Reads
DOI: 10.1016/j.pscychresns.2009.11.008 · Source: PubMed
Abstract
In the present study we obtained functional magnetic resonance imaging (fMRI) data in occasional gamblers (OG) and problem gamblers (PG) during a quasi-realistic blackjack game. We focused on neuronal correlates of risk assessment and reward processing. Participants had to decide whether to draw or not to draw a card in a high-risk or low-risk blackjack situation. We assumed PG would show differences in prefrontal and ventral striatal brain regions in comparison to OG during risk assessment and due to the winning or losing of money. Although both groups did not differ in behavioral data, blood oxygen level dependent (BOLD) signals in PG and OG significantly differed in thalamic, inferior frontal, and superior temporal regions. Whereas PG demonstrated a consistent signal increase during high-risk situations and a decrease in low-risk situations, OG presented the opposite pattern. During reward processing as derived from contrasting winning vs. losing situations, both PG and OG groups showed an enhancement of ventral striatal and posterior cingulate activity. Furthermore, PG demonstrated a distinct fronto-parietal activation pattern which has been discussed to reflect a cue-induced addiction memory network which was triggered by gambling-related cues.
Neurobiological correlates of problem gambling in a quasi-realistic blackjack scenario
as revealed by fMRI
Stephan F. Miedl
a,c,
, Thorsten Fehr
a,b
, Gerhard Meyer
c
, Manfred Herrmann
a,b
a
Department of Neuropsychology and Behavioral Neurobiology, Center for Cognitive Sciences (ZKW), University of Bremen, Germany
b
Center for Advanced Imaging CAI Bremen, University of Bremen, Germany
c
Institute of Psychology and Cognition Research, University of Bremen, Germany
abstractarticle info
Article history:
Received 2 October 2009
Received in revised form 18 November 2009
Accepted 19 November 2009
Keywords:
fMRI
Thalamus
Addiction
Gambling
In the present study we obtained functional magnetic resonance imaging (fMRI) data in occasional gamblers (OG)
and problem gamblers (PG) during a quasi-realistic blackjack game. We focused on neuronal correlates of risk
assessment and reward processing. Participantshad to decidewhether to draw or not to draw a card in a high-risk
or low-risk blackjack situation. We assumed PG would show differences in prefrontal and ventral striatal brain
regions in comparison to OG during risk assessment and due to the winning or losing of money. Although both
groups did not differ in behavioral data, blood oxygen level dependent (BOLD) signals in PG and OG signicantly
differed in thalamic, inferior frontal, and superior temporal regions. Whereas PG demonstrated a consistentsignal
increase during high-risk situations and a decrease in low-risk situations, OG presented the opposite pattern.
During reward processing as derived from contrasting winning vs. losing situations, both PG and OG groups
showed an enhancement of ventral striatal and posterior cingulate activity. Furthermore, PG demonstrated a
distinct fronto-parietal activation pattern which has been discussed to reect a cue-induced addiction memory
network which was triggered by gambling-related cues.
© 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
Pathological gambling is characterized by a craving for gambling,
loss of control, and continuing gambling despite associated adverse
consequences. It is classied as an impulse control disorder in the
Diagnostic and Statistical Manual of Mental Disorders IV (DSM IV) with
a lifetime prevalence of 0.51% (Petry et al., 2005). From a clinical point
of view, pathological gambling is related to addictive behavior (Potenza,
2006), and there is emerging evidence that the underlying pathology on
a neuronal level compares to cue-related behavior in drug addiction
(Franklin et al., 2007). Gambling takes place in a complex social and
context-specic environment, which cannot easily be transferred into
an experimental setting. Furthermore, the gambling situation consists of
a variety of cognitive (problem solving, risk assessment) and emotional
(reward processing) behaviors, which may be prevalent in problem
gamblers. In the present study we introduced an experimental design
with a quasi-realistic blackjack game scenario to enhance ecological
validity, and to allow for the analysis of different episodes of the game.
We were particularly interested in separating the risk assessment and
reward processing periods of the game because both have been shown
to be impaired in pathological gambling (Reuter et al., 2005; Goudriaan
et al., 2006).
Growing evidence suggests that risk assessment/decision making
might be affected in pathological gambli ng, especially when
gam blers have to choose between ri sky and safe opt ions (Bechara,
2005). So far, most experimental data have been derived from the
Iow a Gambling Task (IGT), as developed by Bechara et al. (1994),and
introduced as a tool to measure risk-anticipation.Patientswith
ventromedial f rontal lobe ( VMF) damage (Cavedini et al., 2002),
patients with disinhibition behavior (substance dependencies,
psychopathy and attention decit/hyperactivity disorder; Blair,
2001), and pathological gamblers (Goudriaan et al., 2005) showed
impaired per formance on the IGT (Bechara et al., 1997). Tanabe et al.
(2007) showed a reduced activation in the right prefron tal cortex
during decision making in substance-dependent gamblers, as
com pared with substance-dependent controls on the IGT. These
authors relat ed gambling-associated problems to impaired working
memory, st imulus reward evaluation, or cue reactivity. Furthermore,
Brand et al. (2005, 2006) suggested dorsolateral prefrontal and
orb itofrontal dysfunctions in pathological gamblers both regi ons
are dis cussed to be inv olved in decision making. In particular, th e
orb itofrontal corte x was reported to be sensitive to the amount of
conict inherent to decisions (Rogers et al., 1999).
In addition to impaired risk evaluation and risk taking, it has been
shown that reward processing and the activity of the mesolimbic
Psychiatry Research: Neuroimaging 181 (2010) 165173
Corresponding author.Department of Neuropsychologyand Behavioral Neurobiology,
Center for Cognitive Sciences (ZKW), University of Bremen Cognium Building,
Hochschulring 18, D-28359 Bremen, Germany. Tel.: +49 421 218 68744; fax: +49 421
218 68759.
E-mail address: miedl@uni-bremen.d e (S.F. Miedl).
0925-4927/$ see front matter © 2009 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.pscychresns.2009.11.008
Contents lists available at ScienceDirect
Psychiatry Research: Neuroimaging
journal homepage: www.elsevier.com/locate/psychresns
dopaminergic reward system (Self and Nestler, 1998; Volkow et al.,
2002) may be affected in pathological gambling. Additionally, there is
evidence for a reduction in the sensitivity of the reward system.
Reuter et al. (2005) compared pathological gamblers with healthy
controls in a simple card-guessing game and demonstrated a ventral
striatal and ventromedial prefrontal hypoactivation in pathological
gamblers, which was positively correlated with gambling severity.
D-
Amphetamine, a non-specic dopamine agonist, was shown to prime
gambling motivation in problem gamblers (Zack and Poulos, 2004).
This points to a dysregulation of specic dopamine-related neuronal
reward processes in problem gamblers. More recently, the same
group (Zack and Poulos, 2007) demonstrated an enhancement of
reward and priming effects of a gambling episode (playing slot
machines) in pathological gamblers as compared with non-gamblers
after D2 antagonist intake.
The majority of studies investigating neuronal correlates in
pathological gambling are based on experimental settings which
simplify the complex gambling environment. There are only a few
studies which have introduced a more ecologically valid experimental
gambling design to examine neuronal or neuroendocrinological
responses while playing. Hewig et al. (2009) measured blood oxygen
level dependent (BOLD) signal while playing blackjack in healthy
participants. During the evaluative phase of decision making under
risk conditions, the authors demonstrated that excessively risky and
cautious decisions were associated with increased dorsal anterior
cingulate cortex activity, similar to observed correlations of the error-
related negativit y (ERN) amplit ude with both risk-taking and
decision-making behavior (Gehring and Fencsik, 2001; Gehring and
Willoughby, 2002) in a previous electroencephalography (EEG) study
using the same design (Hewig et al., 2007). Evidence for the inuence
of gambling for real money (Ladouceur et al., 2003; Wulfert et al.,
2005) was provided by Meyer et al. (2004), who compared the
neuroendocrine responses of problem gamblers and healthy controls
in two different situations: a real blackjack casino situation, where
gamblers invested their own money, and an experimental control
condition, where participants were playing for points in a laboratory
environment. The results showed higher levels of norepinephrine and
dopamine in problem gamblers compared with healthy participants
in a real money casino environment, which became not signicant
in the control condition. Additional support for the crucial role of the
applied stimulus material in relation to a specic addiction or disorder
came from Volkow et al. (2003). The authors reported an orbitofrontal
hypoactivation in addicted persons when confronted with natural
reinforcers in contrast to an activation of the same area to substance-
related cues. A similar nding was reported by Crockford et al. (2005),
who showed cue-induced dorsolateral prefrontal activity in patho-
logical gamblers during viewing of gambling-related cues. Thus, we
decided to design the present experimental blackjack game scenario
to be as realistic as possible, while considering necessary experimen-
tal efforts for methodologically correct parameterization.
The pres ent functional magnetic r esonance imaging (fMRI) study
aimed at analyzing two major parts of the blackjack game, the
periods of risk assessment and reward processing, using an
ecologically valid quasi-realistic gambling scenario with compara-
tively high wagers in bot h pr oblem gamblers a nd oc casional
gam blers. Based on the above-mentioned studies, we hypothesized
that in problem gamblers both risk assessment and reward proces-
singmightbemodulatedbythegambling-relatednatureofthe
applied task: 1) For the period of risk assessment, we expect a signal
increase in inferior frontal/orbitofrontal and thalam ic brain region s,
and particularly in problem gamblers durin g high-risk situations. 2)
During reward processing, we expect a signal increase in the nuc leus
accumbens in both groups after win conditions. The qu asi -realis tic
task, including gambling for real money, is supposed to counteract or
override a generally observed hypoactivation in p athological g am-
blers associated with risk assessment and reward processing which is
rep orted in experimental setups with a lower gambling-authenticity
(Reuter et al., 2005; Tanabe et al., 2007).
2. Methods
2.1. Study participants
The study group consisted of 12 healthy male OG (range 25
49 years) and 12 male PG (range 2957 years). All participants were
right-handed according to a modied version of the Edinburgh
Handedness Questionnaire (Oldeld, 1971). Both groups did not differ
in age ( F[1,22] =2.97, P =0.1), smoking behavior (z = 1.7, P = 0.1),
and frequency of blackjack gambling (z = 0.6, P= 0.6; see also
Table 1). We decided to investigate only male participants, as the
prevalence of pathological gambling in men is reported to be two
times higher than in women (Grant and Potenza, 2004). Participants
were recruited through advertisements and were familiarized with
the gambling environment in the laboratory. Frequency of overall
gambling behavior (z = 2.7, P b 0.01), as well as the percentage of
income spent on gambling activities, was signicantly higher in PG
than in OG (F[1,22] =22.14, P b 0.01; see Table 1). The preferred forms
of gambling of PG were slot machines, roulette or internet-poker.
Prior to enrollment in the study, all participants underwent a
structured psychiatric interview.
OG and PG did not report a history of psychiatric or neurological
illness or regular drug use and were not under current medication. In
the PG group, ve participants were presented with a diagnosis of
problem gambling (3 or 4 criteria, Toce-Gerstein et al., 2003) and
seven participants had a diagnosis of pathological gambling ( 5
criteria) according to DSM IV. Furthermore, all individuals were
assessed with the German gambling questionnaire Kurzfragebogen
zum Glücksspielverhalten (KFG; Petry, 1996; derived from 20 items
as developed by Gamblers Anonymous). Instrumental (Cronbach's
alpha= 0.79) and retest (r =0.80) reliability of the scale are reported
to highly fulll the psychometric properties of a screening instrument
(Petry, 1996). This questionnaire contains 20 items (4-point Likert-
scale: 0 to 3 points) addressing lifetime gambling behavior. The
threshold for pathological gambling is set at 16 points. All PG scored
between 18 and 45 points, whereas OG scored between 0 and 12
points. In addition, all participants were evaluated with a German
version of the South Oaks Gambling Screen (SOGS; Lesieur and Blume,
1987). Participants who scored
5 points were classied as probable
pathological gamblers. All PG scored 6 on the SOGS, and OG
obtained 2. Both groups signicantly differed with respect to DSM IV
(F[1,22]= 48.58, P b 0.01), SOGS (F[1,22]=78.88, Pb 0.01), and KFG
scores (F[1,22]= 82.68, P b 0.01); see also Table 1. The study protocol
was designed according to the Code of Ethics of the World Medical
Association (Declaration of Helsinki, 1984) and was approved by the
local ethics committee. All participants were informed about the
procedure and gave written informed consent to participate.
Table 1
Demographic and clinical data of PG and OG (mean ±standard deviation).
PG (n =12) OG (n = 12)
Age 39.5±9.3 33.4± 8.0 F[1,22]= 2.97, P=0.10
Number of smokers
per group
10 6 z= 1.7, P = 0.09
DSM IV 4.9± 1.9 0.7 ± 0.9 F[1,22]= 48.58, Pb 0.01
SOGS 10.7±3.8 0.7 ± 0.7 F[1,22]= 78.88, Pb 0.01
KFG 28.2±7.9 5.3 ± 3.7 F[1,22]= 82.68, Pb 0.01
Percent of income
spent on gambling
57.1±34.7 7.3 ± 11.7 F[1,22]= 22.14, Pb 0.01
Blackjack frequency b 1 time/month b 1 time/month z= 0.6, P = 0.57
Frequency of overall
gambling behavior
N 3 times/week 3 times/month z= 2.7, P b 0.01
166 S.F. Miedl et al. / Psychiatry Research: Neuroimaging 181 (2010) 165173
2.2. Experimental design
The experimental blackjack task consisted of 206 trials (50 low-
risk, 50 high-risk, 50 ll-, and 56 validity-trials). The low-risk trials
reect situations in which the player started with 12 or 13 points
against the dealer's 7, 8, 9, or 10 points. Participants were informed
that they played against the computer. High-risk trials consisted of the
player with 15 or 16 points and the dealer with 7, 8, 9, or 10 points.
The probability of losing while drawing a card [P(lose|hit)] over all
low-risk trials was 0.34, and 0.56 over all high-risk trials. The trials
were designed in a way that according to the blackjack basis
strategy (Baldwin et al., 1956) in all high-risk and low-risk
situations a hit was more advantageous for the player than a stand
[P(lose|stand] =0.77]. Fill-trials were composed of cards with
pictures and numbers with no relation to the blackjack game, which
served as low-level baseline condition in further analyses not
reported here. Furthermore, we included 56 validity-trials, consisting
of aces (1 or 11 points), and starting-situations with 14, 17, 18, 19, 20
or 21 points for the player. These validity-trials should guarantee a
quasi-realistic blackjack scenario. Both ll- and validity-trials were
modeled separately and excluded from further analysis. The bet was
xed at 5 in low-risk and high-risk trials, and at 1 in validity-trials.
All trial elements were presented against a black background. A
trial started with a jeton representing a xed bet (1or5; frame 1,
see Fig. 1a) for 500 ms, followed by a white xation point for 1500 ms
(frame 2). Thereafter, three cards were presented for a maximum of
6000 ms; on the upper part of the screen, there was one card for the
dealer, and on the lower part of the screen, there were two. Within
this period the player had to decide whether he wanted to take
another card (hit; right mouse: left button click; index nger) or to
stand (no further card required; right mouse: right button click;
middle nger, frame 4). Thereafter, the dealer took cards according to
the ofcial blackjack rules (the dealer must hit until his total was 17 or
higher). Dependent on the player's response (hit or stand), the dealer
started to take another card 300 ms after the player decided to stand
(stand response), and 2000 ms after the player's hit (hit response).
The end of the round was presented for 3000 ms (frame 5), followed
by a 2000-ms information screen displaying the running total of the
player (frame 6) and a 2000-ms inter-trial xation point (frame 7).
Before the fMRI scanning session all participants had to perform
10 min of practice trials outside the scanner. At the beginning of the
game, each player started with a balance of 30. All study participants
were informed that they might lose their starting balance and that
they would receive the entire balance in cash at the end of the
experiment. Wins and losses followed a pre-determined course
independent from the player's decisions (see Fig. 1b). Trials were
presented in a pseudo-randomized non-stationary probabilistic
sequence (Friston, 2000). Participants lost 50% of the high-risk trials
and 50 percent of the low-risk trials, and always
nished the game
with a total amount of 52. In contrast to ofcial blackjack rules the
player was allowed to hit or to stand only one time per round.
2.3. fMRI data acquisition
While participants performed the tasks, functional MRI data were
collected on a 3 T Siemens Allegra scanner (Siemens, Erlangen, Germany)
using a gradient echoplanar imaging (EPI) sequence covering 44 axial
(ACPC), interleaved slices (3 mm thick) encompassing the entire
cerebrum and cerebellum (TR/TE= 2500/30 ms; FOV 192 mm). About
1000 volumes were obtained during each run. A T1-weighted structural
3D image of the brain was obtained using the MPRAGE sequence: 160
contiguous slices, TR= 2.3 s, TE= 4.38 ms, TI =900 ms, FA =8°, FOV
296× 296 mm, in-plane resolution 1× 1 mm, slice thickness 1 mm.
2.4. fMRI data analysis
Image analysis was performed using SPM2 (http://www.l. ion.u cl.a c.
uk/spm/). For each session and participant, images were realigned to the
rs t image in the time series to correct for head motion . These realigned
images were spatially normalized into a standard stereotactic space
Fig. 1. (a) Trial description of and task elements in a quasi-realistic blackjack scenario, (b) pre-determined course of the game.
167S.F. Miedl et al. / Psychiatry Research: Neuroimaging 181 (2010) 165173
(Montreal Neurological Institute template) using a 12-parameter afne
model. These spatially normalized images were smoothed to minimize
noise and residual differences in gyral anatomy with a Gaussian lter set at
8 mm. Prior to statistical analysis a high pass lter (500 s) was applied to
remove global effects. Pre-processed data sets we re analyzed using a
second-level random effect model that accounts for both scan-to-scan and
participant-to-participant variability (Holmes, 1998).
The analysis of imaging data was restricted to low-risk and high-risk
trials. Furthermore, we focused on two different time periods in the
trials: 1) risk assessment, and 2) win or lose situations. The rst time
period covered the presentation of cards until the player's decision
(Fig. 1a, frames 3 to 4), the second period comprised the time between
the dealer's hit and the presentation of the information of winning or
losing and the runningtotal (Fig. 1a, frame 5 to 6). If the player's decision
to hit resulted in a higher totalthan21 (i.e., he lost the game), the second
period covering win or lose perception was restricted to frames 4 to 5.
Therefore, several trial elements and periods were modeled exclusively
by the standard hemodynamic response function, and included as
separate predictors (high-risk, low-risk, ll, validity, response, win, lose,
equal, running total) in the design matrix. Risk assessment and reward
processing were modeled as epochs; risk assessment with a duration
from presentation of the cards until the response was given (see Fig. 1,
frame 3 to 4); reward processing in a bust was modeled with a duration
of 2 s (see Fig. 1, frames 4 to 5), and in all other cases with a duration of
3s (see Fig. 1, frames 5 to 6). To compare PG and OG, second-level
analyses were performed by calculating t-statistics including rst level
contrast images for pre-determined condition effects at each voxel for
each participant and session for the following contrasts: high-risk vs.
low-risk, low-risk vs. high-risk, high-risk-hit vs. high-risk-stand, high-
risk-stand vs. high-risk-hit, win vs. lose, and lose vs. win. In a rst
analysis, we determined voxels showing a main effect of high-risk vs.
low-risk, low-risk vs. high-risk, high-risk-hit vs. high-risk-stand, and
high-risk-stand vs. high-risk-hit (Pb 0.001 uncorrected, cluster thresh-
old (k)=5). We further calculated an interaction analysis to test for
voxels showing larger contrasts in PG than in OG and vice versa. To
restrict the search volume to active regions showing a signicant main
effect of high-risk vs. low-risk, low-risk vs. high-risk, high-risk-hit vs.
high-risk-stand, or high-risk-stand vs. high-risk-hit, respectively, the
interaction analyses were masked inclusively by the corresponding
contrast of the rst group entered in the analysis (Pb 0.001 uncorrected;
k= 5). Thus, for an interaction analysis, PG vs. OG, the main effect of PG
was entered as an inclusive mask. We calculated the same analyses as
described above for the win vs. lose and lose vs. win conditions (FWE,
Pb 0.05; k=20), and we compared groups (PG vs. OG; OG vs. PG)
applying interaction analyses (Pb 0.001, k=20). Furthermore, conjunc-
tion analysis [conjunction null (Nichols et al., 2005)] was calculated
including win vs. lose conditions for PG and OG (P
b 0.05, FWE-corrected,
k= 20). Estimates of percentage signal change in the high-risk, low-risk,
win, and lose conditions were extracted from the signicant clusters of
brain regions for each participant using MarsBaR (Brett et al., 2002).
For the identication of activated anatomical structures, we
transformed MNI (Montreal Neurological Institute) template based
coordinates into Talairach coordinates (Talairach and Tournoux,
1988) using a matlab tool (mni2tal.m, http://imaging.mrc-cbu.cam.
ac.uk/imaging/MniTalairach) and determined the anatomical regions
using the Talairach Daemon Client software (http://ric.uthscsa.edu/
projects/talairachdaemon.html).
3. Results
3.1. Behavioral data
Response times (RT) and risk assessment (hit vs. stand) in PG and
OG did not differ signicantly. A repeated measures ANOVA for RTs
with the factors group (PG vs. OG)×risk (high-risk vs. low-risk)
showed no signicant main effect of group (F[1,22]= 0.9; P =0.3),
and group× risk interaction (F[1,22]=0.02; P= 0.9). Both groups
showed signicantly longer RTs in high-risk compared with low-risk
conditions (main effect of the factor risk; 1882± 629 ms vs. 1461 ±
452 ms; F[1,22]= 42.9, Pb 0.001).
A repeated measures ANOVA for RTs with the factors group (PG vs.
OG)×high-risk decision (high-risk-hit vs. high-risk-stand) revealed
neither a signicant main effect of group (F[1,22]=0.76; P=0.5) nor a
group×high-risk decision interaction (F[1,22]= 1.01; P=0.33). Both
groups showed signicantly faster RTs in high-risk-hit compared with
high-risk-stand conditions (main effect of the factor high-risk decision;
1868±600 ms vs. 2109±747 ms; F[1,22]=42.9, Pb 0.001). In the low-
risk condition only 10 out of 24 participants showed stand trials, and we
therefore did not compare the respective RTs.
In addition, PG and OG did not differ in risk assessment. A repeated
measures ANOVA for risk assessment with the factors group (PG vs.
OG)× decision behavior (percent high-risk-hit vs. percent low-risk-
hit) revealed no signicant main effect of group (F[1,22]=0.43;
P= 0.5) and no group ×decision behavior interaction (F[1,22] =0.02;
P= 0.9). Both groups showed a signicantly lower percentage of high-
risk compared with low-risk-hit trials (main effect of the factor
decision behavior; 63.4 ± 23.8% vs. 96.5 ±8.8%; F[1,22]= 52.8,
Pb
0.001; see also Table 2).
Furthermore, PG and OG did not differ in the number of bust trials
(in case participants draw another card and get over 21 points; see
also Table 2) in low-ri sk (F[1,22]= 3.4; P = 0.1 and hig h-risk
situations (F[1,22] =1.2; P =0.2).
A repeated measures ANOVA for the number of bust trials with the
factors group (PG vs. OG)× bust (high-risk bust vs. low-risk bust)
revealed neither a signicant main effect of group (F[1,22]=2.04;
P= 0.2) nor a group×bust interaction (F[1,22]=0.51; P = 0.5). Both
groups showed a signicantly lower number of bust trials in high-risk
compared with low-risk trials (main effect of the factor bust; 8.1± 3.4
vs. 12.5 ±0.9; F[1,22] =46.2, P b 0.001; see also Table 2).
3.2. Functional imaging data
Neural correlates of risk assessment, as revealed by contrasting
high-risk vs. low-risk conditions, resulted in signicant activation
patterns in PG (Pb 0.001, uncorrected) including bilateral frontal,
temporal, right parietal, and bilateral parahippocampal and right
thalamic regions (see Supplementary material Table 1S). The reverse
contrast revealed activation patterns in OG (P b 0.001 uncorrected; see
Supplementary material Table 1S) in bilateral frontal, right superior
temporal, and bilateral thalamic, insular, and parahippocampal
regions. This dissociation was conrmed by an interaction analysis
showing larger contrasts in OG compared with PG for low-risk vs.
high-risk conditions (right superior temporal gyrus, and left thala-
mus), and a signicant percent signal increase in PG compared with
OG in right superior temporal gyrus, right inferior frontal gyrus, and
right thalamus (see Table 3 and Fig. 2). Analyses contrasting high-
risk-hit vs. high-risk-stand and high-risk-stand vs. high-risk-hit trials
within PG and OG revealed no signicant activation patterns.
Table 2
Behavioral data of PG and OG (mean ± standard deviation).
PG (n= 12) OG (n=12)
Response time High-risk 1992 ± 576 1772±684
Low-risk 1560 ± 417 1361±481
Response time per decision
in high-risk trials
High-risk-hit 2010±526 1726± 657
High-risk-stand 2168±720 2050 ± 801
Percentage of hit trials High-risk 61.2±25.7 65.7±22.7
Low-risk 93.8±11.1 97.2±5.8
Number of bust trials High-risk 7.3 ± 3.6 8.9± 3.2
Low-risk 12.2±1.1 12.8±0.6
168 S.F. Miedl et al. / Psychiatry Research: Neuroimaging 181 (2010) 165173
Contrasting win vs. lose conditions (P b 0.05; FWE-corrected)
produced widespread bilateral parietal, occipital, frontal, and subcorti-
cal activation patterns in both PG and OG (see Supplementary material,
Table 2S). Conjunction analysis (Pb 0.05, FWE-corrected; see Nichols
and Hayasaka, 2003) demonstrated that both groups showed common
activation patterns in brain regions related to reward processing (Ncl.
accumbens, bilateral frontal regions, left parietal regions [precuneus],
left occipital regions, bilateral cerebellum, left thalamus, and right
posterior cingulate gyrus; see Fig. 3 upper part and Table 4).
Comparing activation patterns in win vs. lose situation s between
groups [(winN lose) PG vs. OG; threshold Pb 0.001, uncorrected, Fig. 4
and Table 4] resulted in right superior frontal, left inferior parietal, and
left superior parietal signal increases; whereas the opposite contrast
(OGN PG) for the win vs. lose condition became insignicant. Further-
more, lose situations compared with win situations did not show
suprathreshold activations.
4. Discussion
Using fMRI, we investigated neuronal correlates in PG and OG
during a quasi-realistic blackjack scenario with respect to both risk
assessment and reward processing. The present data suggest that risk
assessment was associated with comparable arousal-related neuronal
networks in PG and OG. During high-risk situations, fronto-thalamic
brain activity was enhanced in PG, whereas OG showed a signicant
signal increase in low-risk conditions. Winning contrasted with losing
Table 3
Interaction analysis of brain regions activated during risk assessment Talairach co ordinates, anatomical regions and t-scores of the between group comparisons in the high-risk vs.
low-risk task contrasts for PG vs. OG (A) and in the low-risk vs. high-risk task contrasts for OG vs. PG (B) (all P b 0.001, uncorrected, k = 5).
Regions H A B
(high-risk
PG
N low-risk
PG
)N (high-risk
OG
N low-risk
OG
) (low-risk
OG
N high-risk
OG
)N (low-risk
PG
N high-risk
PG
)
t cl-size xy z t cl-size xy z
Inferior frontal gyrus R 3.92 18 46 42 5
Superior temporal gyrus R 4.28 50 67 25 1 5.25 80 67 21 8
R 4.34 50 53 13 6
R 4.75 80 59 31 7
R 3.97 8 44 27 7
Thalamus L 3.75 11 2 25 9
R 3.93 8 10 33 3
H, hemisphere; R, right; L, left.
Fig. 2. Activation pattern and percent signal change 1 SEM) derived from the high-risk vs. low-risk contrast in PG vs. OG (Pb 0.001 uncorrected) in the pulvinar nucleus of the
thalamus (Tha), inferior frontal gyrus (IFG), and superior temporal gyrus (STG). MNI-to-Talairach-transformed coordinates are given in Table 3.
169S.F. Miedl et al. / Psychiatry Research: Neuroimaging 181 (2010) 165173
money activated brain regions associated with reward processing in
both PG and OG, whereas an interaction analysis between groups
resulted in signicantly larger contrasts in fronto-parietal regions in
PG as compared to OG.
4.1. Risk assessment
Behavioral data derived from the risk assessment period of the
game did not show any differences in the percentage of hit trials
between groups, but signicantly slower RT for high-risk situations in
both groups. These data indicate that PG and OG did not differ on a
behavioral level and that PG did not run a higher risk during the
present blackjack game. The surplus in RT in high-risk situations
might be associated with a higher amount of response conict (Yang
et al., 2007). Although behavioral data were comparable in both
experimental groups, imaging data analysis during risk assessment
showed group-related differences. Analogously, Fehr et al. (2006)
showed a consistent colorword Stroop effect reected in both
behavioral and ERP data in a group of smokers and non-smokers.
However, smoking-cue-related interference processing in smokers
was only reected in ERP topographies, but not in the respective
behavioral data. This was explained by potential sequential (time-
consuming) and parallel (time-saving) mental processing steps in the
different mental domains.
To test the rst hypothesis, a region of interest analysis based on
percent signal change values was performed and indicated a dissocia-
tion between OG and PG in superior temporal gyrus, right inferior
frontal/orbitofrontal gyrus, and right medial pulvinar brain regions in
PG. OG presented increased activity during low-risk situations, whereas
PG showed a signal increase in high-risk trials. These results conrmed
the hypothesis of a signal increase in inferior frontal/orbitofrontal and
thalamic brain regions, especially in problem gamblers during high-risk
assessment. We suppose that this nding might be related to the highly
authentic gambling task of the present study.
The thalamus had been shown to be involved in addictive behavior
as reported by a variety of studies (Breiter et al., 1997; George et al.,
2001; Due et al., 2002; Potenza et al., 2003; Franklin et al., 2007; Wang
et al., 2007). The medial pulvinar nucleus of the thala mus is
reciprocally interconnected with the cingulate gyrus and other limbic
structures (Morgane et al., 2005) and is considered to play a crucial
role in learning and memory (Mitchell et al., 2008), emotional
experience and expression, drive (Sewards and Sewards, 2003;
Nummenmaa et al., 2008), and motivation (Schmahmann, 2003). In
addition, Leh et al. (20 08) showed that human pulvinar was
interconnected with subcortical structures (superior colliculus,
thalamus, and caudate nucleus) as well as with cortical regions
(primary and secondary visual areas, inferior temporal brain regions,
posterior parietal association areas [area 7], frontal eye eld, and
prefrontal areas). These data demonstrate the important role of the
pulvinar in human visual information processing and visuospatial
attention. However, the direction of the neuronal response is still the
subject of ongoing discussion. Tomasi et al. (2007) reported a
hypoactivation of the medio-dorsal thalamus and the lateral genic-
ulate body of the thalamus in cocaine abusers during a visuospatial
attention task, whereas George et al. (2001) showed alcohol cue-
induced dorsolateral prefrontal and anterior thalamus signal en-
hancement in alcoholics compared with healthy controls. Wang et al.
(2007) demonstrated that right thalamus, orbitofrontal, and dorso-
lateral prefrontal cortex activation was related to abstinence-induced
craving in smokers. These ndings might re
ect a dissociation within
thalamic activity with respect to an addiction-related nature of the
task. In c ommon decision maki ng paradigms (not referring to
addiction-related stimulus material) there is evidence for a thalamic
down-regulation in PG quite similar to the effects reported in cocaine
users (Goldstein et al., 2007; Tomasi et al., 2007), whereas the
Fig. 3. Activation pattern and percent signal change 1 SEM) derived from the win vs.
lose contrast in the conjunction {null} analysis for OG and PG (P b 0.05 FWE-corrected)
in the nucleus accumbens. MNI-to-Talairach-transformed coordinates are given in
Table 4.
Table 4
Brain regions activated during reward processing Talairach coordinates, anatomical
regions and t-scores of (A) a conjunction {null} analysis including win vs. lose contrasts
of PG and OG (Pb.05, FWE-corrected; k = 20), and (B) between group comparison (PG
vs. OG) in the win N lose contrast (P b 0.001, uncorrected, k = 20).
Regions H A B
Conjunction {null}
(win
PG
N lose
PG
) and
(win
OG
N lose
OG
)
(win
PG
N lose
PG
)N
(win
OG
N lose
OG
)
t cl-
size
xyztcl-
size
xyz
Subcallosal
gyrus
L 6.41 52 14 5 14
R 8.91 187 16 5 14
Superior
frontal gyrus
L 7.12 23 28 163
R 4.67 23 24 15 56
Inferior
parietal
lobule
L 4.68 86 32 55 44
Superior
parietal
lobule
L 4.03 86 28 50 55
Precuneus L 7.24 42 18 59 56
Lingual gyrus L 7.48 66 22 95 5
L 7.13 87 10 80 3
Cuneus L 7.58 87 8 79 8
L 7.94 66 18 99 0
L 7.13 31 12 60 9
Nucleus
accumbens
L 7.61 52 89 7
R 8.58 187 10 9 9
Cingulate
gyrus
R 7.80 72 2 24 29
Thalamus L 7.20 65 10 15 10
L 6.73 65 16 17 16
Cerebellum L 6.59 76
4 73 20
R 7.51 76 2 77 23
R 7.29 76 10 77 18
R 6.85 31 2 74 8
H, hemisphere; R, right; L, left.
170 S.F. Miedl et al. / Psychiatry Research: Neuroimaging 181 (2010) 165173
processing of addiction-related stimuli enhances thalamic activity in
PG, indicating cue-induced craving engagement.
Furthermore, the inferior frontal cortex (IFG) was found to be
differentially activated in PG and OG in the high-risk vs. low-risk
contrast. Potenza et al. (2003) related decreased activation in the right
orbitofrontal cortex, basal ganglia, and thalamus in pathological
gamblers compared with controls, while viewing videotaped scenar-
ios with gambling content, to impulse regulation. Crockford et al.
(2005), however, showed increased right dorsolateral prefrontal and
right parahippocampal activity in pathological gamblers when
compared with controls during gambling-related video material
without an involvement of any actual gambling. The conicting data
were discussed with respect to differences in stimulus material and
cue-induced craving (Wilson et al., 2004). These data might indicate
that both IFG as well as the medial pulvinar activation in high-risk
situations might reect a cue-induced signal increase in PG. High-risk
situations, characterized by physiological arousal, euphoria, distrac-
tion, and perceived control (Legg England and Gotestam, 1991), might
serve as an addiction-cue in PG, while the low-risk situation signies a
safe hit in OG.
The higher right superior temporal gyrus (STG) activation in high-
risk vs. low-risk task conditions in PG might be related to feature
based serial exploratory search (Ellison et al., 2004) or intuitive
judgments (Ilg et al., 2007). Similar to brain activity of the IGT
reported in the anticipation phase (Lin et al., 2008), in the present task
environment, the right STG might have been involved in updating the
card constellation, especially in PG during high-risk assessment.
To conclude, neuronal correlates of problem gambling in the
present quasi-realistic blackjack scenario might be characterized by
an expectancy shift from looking for secure low-risk conditions in
OG to seeking for thrilling high-risk situations in PG.
4.2. Win or lose situations and reward processing
With regard to our second hypothesis the nucleus accumbens
showed a signal increase in win situations and a signal decrease in lose
situations in both PG and OG corroborating the sensitivity of the
dopaminergic reward system for reward processing in gambling (Blum
et al., 1996). The present ndings, showing a consistent ventral striatal
signal increase across groups, corroborate the crucial role of the nucleus
accumbens in human reward processing (Knutson et al., 2001). In
contrast to Reuter et al. (2005), we did not nd a differential activation
of the accumbens in PG and controls, which might support our
hypothesis that applying a quasi-realistic experimental paradigm will
adjust striatal signal decrease in PG during reward processing.
Additionally, this nding might also be related to differences in the
experimental task design. In the study of Reuter et al. (2005),
participants of a simple card-guessing game realized the moment of
winning/losing the game immediately, whereas in the present quasi-
realistic blackjack scenario, participants had rst to calculate the sum of
the card values before perceiving the win or loss information. Therefore,
the relative high stimulus-complexity in the present experiment might
have modulated reward processing (Fehr et al., 2007b).
Comparing PG with OG in the win vs. lose condition resulted in left
inferior parietal, left superior parietal, and right premotor activations.
The inferior parietal sulcus (IPS) has been reported to be activated
during number tasks (Dehaene et al., 2003; Fehr et al., 2007a, 2008
),
while experiencing gains and losses (Lin et al., 2008), and is expected
to play a crucial role in the amodal representation of quantity
(Dehaene et al., 2004). Smolka et al. (2006) described left inferior
parietal and right premotor cortex activation in a smoking-cue
paradigm and demonstrated that brain activity in these regions
correlated positively with the degree of nicotine dependence.
Therefore, left parietal and right premotor activation, frequently
associated with visuospatial attention (McCarthy et al., 1997; Kirino
et al., 2000), motor preparation (Toni et al., 2002), and imagery
(Rizzolatti and Craighero, 2004), might reect neuronal correlates to
the preparation of further gambling activity. Corroborating this line of
argumentation, Fehr and colleagues (2006, 2007b) discussed early
and late fronto-parietal ERP-differences between smokers and non-
smokers during different nicotine-cue interference tasks as being
related to addiction memory-related cue driven activation of neurally
established perceptionaction cycle networks.
4.3. Limitations of the present study
There are several limitations of the present study. First, the OG group
did not represent normal control participants naïve to the experi-
mental condition as described in other studies related to the eld of
investigation. However, in order to focus on the differences between
Fig. 4. Activation patterns and percent signal change 1 SEM) resulting from the win vs. lose contrast in PG vs. OG (Pb 0.001 uncorrected) in the inferior parietal lobe/superior
parietal lobe (IPL/SPL), and superior frontal gyrus (SFG). MNI-to-Talairach-transformed coordinates are given in Table 4.
171S.F. Miedl et al. / Psychiatry Research: Neuroimaging 181 (2010) 165173
normal and problematic gambling, we triedto test individuals who were
equally familiar with blackjack gambling without demonstrating
pathological gambling behavior. Consequently, PG and OG were rather
similar with respect to blackjack gambling experience but signicantly
different with respect to overall gambling frequency. Second, ecological
validity in the present experiment was limited as participants knew that
they were playing against the computer in comparison to a casino
situation playing against a person, a fact that also might have inuenced
neuronal activation patterns. In addition, only seven out of 12 PG
fullled all DSM IV criteria for pathological gambling. Hence, it could be
that the PG group in the present study suffered from a milder form of
pathological gambling, although all participants in the PG group fullled
the SOGS and KFG criteria for pathological gambling. Another
shortcoming of the present study is that we only discriminated between
smokers and non-smokers. It would have been more appropriate to
assess the amount of cigarettes per day to test for differences regarding
smoking behavior between groups more precisely. Furthermore, the
grouprelated age difference was close to signicant (P=0.10) with a
trend to higher age in the PG group, where compensatory activity due to
declining neuronal structures (Park and Reuter-Lorenz, 2008)mayhave
slightly biased our fMRI results. Another limitation of the study is that
we could not clearly discriminate between risk assessment and
preparatory stages of decision making. Different levels of risk implicated
different decision behavior in both groups: Whereas hits and stands did
not differ in high-risk situations, low-risk situations were almost
exclusively characterized by hits. Nevertheless, as PG and OG did not
differ in hit rate and response times (difference of high-risk-hit vs. high-
risk-stand), and fMRI analyses did not demonstrate signicant effects in
both PG and OG in high-risk-hit vs. high-risk-stand and high-risk-stand
vs. high-risk-hit contrasts, we assume that the reported results reect
group-specic differences in assessing different levels of perceived risk
and not differences in decision making.
4.4. Conclusion
The present data suggest that problem gambling within a quasi-
realistic and ecologically valid experimental setting is reected in
differential brain activation patterns during risk assessment and
during reward processing. High-risk situations enhanced arousal-
related brain networks in PG, whereas low-risk situations activated
similar brain areas in OG. High-risk situations as well as winning
money might serve as central addiction-cues for PG, which in turn
activate addictive gambling behavior. In contrast to recent ndings,
the present data did not point to an alteration of reward-related
activity in the accumbens in PG.
Acknowledgements
The present study was supported by a research grant (11/174/05)
of the Bremen University Research Commission to Gerhard Meyer and
Manfred Herrmann and by the BMBF Neuroimaging Program
(01GO0202) from the Center for Advanced Imaging (CAI) to Manfred
Herrmann). We would like to thank Dorit Kliemann and Juliana
Wiechert for assistance during data acquisition.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.pscychresns.2009.11.008.
References
Baldwin, R.R., Cantey, W.E., Maisel, H., McDermott, J.P., 1956. The optimum strategy in
blackjack. Journal of the American Statistical Association 51 (275), 429439.
Bechara, A., 2005. Decision making, impulse control and loss of willpower to resist
drugs: a neurocognitive perspective. Nature Neuroscience 8 (11), 14581463.
Bechara, A., Damasio, A.R., Damasio, H., Anderson, S.W., 1994. Insensitivity to future
consequencesfollowing damage to human prefrontal cortex. Cognition 50 (13), 715.
Bechara, A., Damasio, H., Tranel, D., Damasio, A.R., 1997. Deciding advantageously
before knowing the advantageous strategy. Science 275 (5304), 12931295.
Blair, R.J., 2001. Neurocognitive models of aggression, the antisocial personality
disorders, and psychopathy. Journal of Neurology, Neurosurgery and Psychiatry 71
(6), 727731.
Blum, K., Sheridan, P.J., Wood, R.C., Braverman, E.R., Chen, T.J., Cull, J.G., Comings, D.E.,
1996. The D2 dopamine receptor gene as a determinant of reward deciency
syndrome. Journal of the Royal Society of Medicine 89 (7), 396400.
Brand, M., Kalbe, E., Labudda, K., Fujiwara, E., Kessler, J., Markowitsch, H.J., 2005.
Decision-making impairments in patients with pathological gambling. Psychiatry
Research 133 (1), 9199.
Brand, M., Labudda, K., Markowitsch, H.J., 2006. Neuropsychological correlates of decision-
making in ambiguous and risky situations. Neural Networks 19 (8), 12661276.
Breiter, H.C., Gollub, R.L., Weisskoff, R.M., Kennedy, D.N., Makris, N., Berke, J.D.,
Goodman, J.M., Kantor, H.L., Gastfriend, D.R., Riorden, J.P., Mathew, R.T., Rosen, B.R.,
Hyman, S.E., 1997. Acute effects of cocaine on human brain activity and emotion.
Neuron 19 (3), 591611.
Brett, M., Johnsrude, I.S., Owen, A.M., 2002. The problem of functional localization in the
human brain. Nature Reviews. Neuroscience 3 (3), 243249.
Cavedini, P., Riboldi, G., Keller, R., D'Annucci, A.,Bellodi, L., 2002. Frontal lobe dysfunction in
pathological gambling patients. Biological Psychiatry 51 (4), 334341.
Crockford, D.N., Goodyear, B., Edwards, J., Quickfall, J., el-Guebaly, N., 2005. Cue-induced
brain activity in pathological gamblers. Biological Psychiatry 58 (10), 787795.
Dehaene, S., Piazza, M., Pinel, P., Cohen, L., 2003. Three parietal circuits for number
processing. Cognitive Neuropsychology 20, 487506.
Dehaene, S., Molko, N., Cohen, L., Wilson, A.J., 2004. Arithmetic and the brain. Current
Opinion in Neurobiology 14 (2), 218224.
Due, D.L., Huettel, S.A., Hall, W.G., Rubin, D.C., 2002. Activation in mesolimbic and
visuospatial neural circuits elicited by smoking cues: evidence from functional
magnetic resonance imaging. American Journal of Psychiatry 159 (6), 954960.
Ellison, A., Schindler, I., Pattison, L.L., Milner, A.D., 2004. An exploration of the role of the
superior temporal gyrus in visual search and spatial perception using TMS. Brain
127 (10), 23072315.
Fehr, T., Wiedenmann, P., Herrmann, M., 2006. Nicotine Stroop and addiction memory
an ERP study. International Journal of Psychophysiology 62 (2), 224232.
Fehr, T., Code, C., Herrmann, M., 2007a. Common brain regions underlying different
arithmetic operations as revealed by conjunct fMRI-BOLD activation. Brain Research
1172, 93102.
Fehr, T., Wiedenmann, P., Herrmann, M., 2007b. Differences in ERP topographies during
color matching of smoking-related and neutral pictures in smokers and non-smokers.
International Journal of Psychophysiology 65 (3), 284293.
Fehr, T., Code, C., Herrmann, M., 2008. Auditory task presentation reveals predominantly
right hemispheric fMRI activation patterns during mental calculation. Neuroscience
Letters 431 (1), 3944.
Franklin, T.R., Wang, Z., Wang, J., Sciortino, N., Harper, D., Li, Y., Ehrman, R., Kampman, K.,
O'Brien, C.P., Detre, J.A., Childress, A.R., 2007. Limbic activation to cigarette smoking
cues independent of nicotine withdrawal: a perfusion fMRI study. Neuropsycho-
pharmacology 32 (11), 23012309.
Friston, K.J., 2000. Experimental design and statistical issues. In: Mazziotta, J.C., Toga, A.W.
(Eds.), Brain Mapping: The Disorders. Academic Press, San Diego, CA, pp. 33
58.
Gehring, W.J., Fencsik, D.E., 2001. Functions of the medial frontal cortex in the
processing of conict and errors. Journal of Neuroscience 21 (23), 94309437.
Gehring, W.J., Willoughby, A.R., 2002. The medial frontal cortex and the rapid
processing of monetary gains and losses. Science 295 (5563), 22792282.
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-specic cues. Archives of General Psychiatry 58 (4),
345352.
Goldstein, R.Z., Alia-Klein, N., Tomasi, D., Zhang, L., Cottone, L.A., Maloney, T., Telang, F.,
Caparelli, E.C., Chang, L., Ernst, T., Samaras, D., Squires, N.K., Volkow, N.D., 2007. Is
decreased prefrontal cortical sensitivity to monetary reward associated with impaired
motivation and self-control in cocaine addiction? American Journal of Psychiatry 164
(1), 4351.
Goudriaan, A.E., Oosterlaan, J., de Beurs, E., van den Brink, W., 2005. Decision making in
pathological gambling: a comparison between pathological gamblers, alcohol
dependents, persons with Tourette syndrome, and normal controls. Brain Research.
Cognitive Brain Research 23 (1), 137151.
Goudriaan, A.E., Oosterlaan, J., de Beurs, E., van den Brink, W., 2006. Psychophysiolog-
ical determinants and concomitants of decient decision making in pathological
gamblers. Drug and Alcohol Dependence 84 (3), 231239.
Grant, J.E., Potenza, M.N., 2004. Pathological Gambling: A Clinical Guide To Treatment.
Amer Psychiatric Pub Inc., Washington, DC.
Hewig, J., Trippe, R., Hecht, H., Coles, M.G., Holroyd, C.B., Miltner, W.H., 2007. Decision-
making in blackjack: an electrophysiological analysis. Cerebral Cortex 17 (4), 865877.
Hewig, J., Straube, T., Trippe, R.H., Kretschmer, N., Hecht, H., Coles, M.G.H., Miltner, W.H.R.,
2009. Decision-making under risk: an fMRI study. Journal of Cognitive Neuroscience
21 (8), 16421652.
Holmes, A.P., Friston K.J., 1998. Generalisability, random effects, and population
inference. Neuroimage 7, 754.
Ilg, R., Vogeley, K., Goschke, T., Bolte, A., Shah, J.N., Poppel, E., Fink, G.R., 2007. Neural
processes underlying intuitive coherence judgments as revealed by fMRI on a
semantic judgment task. NeuroImage 38 (1), 228238.
Kirino, E., Belger, A., Goldman-Rakic, P., McCarthy, G., 2000. Prefrontal activation
evoked by infrequent target and novel stimuli in a visual target detection task: an
172 S.F. Miedl et al. / Psychiatry Research: Neuroimaging 181 (2010) 165173
event-related functional magnetic resonance imaging study. Journal of Neurosci-
ence 20 (17), 66126618.
Knutson, B., Adams, C.M., Fong, G.W., Hommer, D., 2001. Anticipation of increasing
monetary reward selectively recruits nucleus accumbens. Journal of Neuroscience
21 (16), RC159.
Ladouceur, R., Sévigny, S., Blaszczynski, A., O'Connor, K., Lavoie, M.E., 2003. Video
lottery: winning expectancies and arousal. Addiction 98 (6), 733738.
Legg England, S., Gotestam, K.G., 1991. The nature and treatment of excessive gambling.
Acta Psychiatrica Scandinavica 84 (2), 113120.
Leh, S.E., Chakravarty, M.M., Ptito, A., 2008. The connectivity of the human pulvinar: a
diffusion tensor imaging tractography study. International Journal of Biomedical
Imaging 2008, 789539.
Lesieur, H.R., Blume, S.B., 1987. The South Oaks Gambling Screen (SOGS): a new
instrument for the identi cation of pathological gamblers. American Journal of
Psychiatry 144 (9), 11841188.
Lin, C.H., Chiu, Y.C., Cheng, C.M., Hsieh, J.C., 2008. Brain maps of Iowa gambling task.
BMC Neuroscience 9, 72.
McCarthy, G., Luby, M., Gore, J., Goldman-Rakic, P., 1997. Infrequent events transiently
activate human prefrontal and parietal cortex as measured by functional MRI.
Journal of Neurophysiology 77 (3), 16301634.
Meyer, G., Schwertfeger, J., Exton, M.S., Janssen, O.E., Knapp, W., Stadler, M.A.,
Schedlowski, M., Kruger, T.H., 2004. Neuroendocrine response to casino gambling
in problem gamblers. Psychoneuroendocrinology 29 (10), 12721280.
Mitchell, A.S., Browning, P.G., Wilson, C.R., Baxter, M.G., Gaffan, D., 2008. Dissociable
roles for cortical and subcortical structures in memory retrieval and acquisition.
Journal of Neuroscience 28 (34), 83878396.
Morgane, P.J., Galler, J.R., Mokler, D.J., 2005. A review of systems and networks of the
limbic forebrain/limbic midbrain. Progress in Neurobiology 75 (2), 143160.
Nichols, T., Hayasaka, S., 2003. Controlling the familywise error rate in functional
neuroimaging: a comparative review. Statistical Methods in Medical Research 12
(5), 419446.
Nichols, T., Brett, M., Andersson, J., Wager, T., Poline, J.B., 2005. Valid conjunction
inference with the minimum statistic. NeuroImage 25 (3), 653660.
Nummenmaa, L., Hirvonen, J., Parkkola, R., Hietanen, J.K., 2008. Is emotional contagion
special? An fMRI study on neural systems for affective and cognitive empathy.
NeuroImage 43 (3), 571580.
Oldeld, R.C., 1971. The assessment and analysis of handedness: the Edinburgh
inventory. Neuropsychologia 9 (1), 97113.
Park, D.C., Reuter-Lorenz, P., 2008. The adaptive brain: aging and neurocognitive
scaffolding. Annual Review of Psychology 60 (1), 173196.
Petry, J., 1996. Psychotherapie der Glücksspielsucht. Beltz/Psychologie Verlags Union,
Weinheim.
Petry, N.M., Stinson, F.S., Grant, B.F., 2005. Comorbidity of DSM-IV pathological gambling
and other psychiatric disorders: results from the National Epidemiologic Survey on
Alcohol and Related Conditions. Journal of Clinical Psychiatry 66 (5), 564574.
Potenza, M.N., 2 006. Should addictive diso rders include non-substance-related
conditions? Addiction 101 (Suppl 1), 142151.
Potenza, M.N., Steinberg, M.A., Skudlarski, P., Fulbright, R.K., Lacadie, C.M., Wilber, M.K.,
Rounsaville, B.J., Gore, J.C., Wexler, B.E., 2003. Gambling urges in pathological
gambling: a functional magnetic resonance imaging study. Archives of General
Psychiatry 60 (8), 828836.
Reuter, J., Raedler, T., Rose, M., Hand, I., Glascher, J., Buchel, C., 2005. Pathological
gambling is linked to reduced activation of the mesolimbic reward system. Nature
Neuroscience 8 (2), 147148.
Rizzolatti, G., Craighero, L., 2004. The mirror-neuron system. Annual Review of
Neuroscience 27, 169192.
Rogers, R.D., Owen, A.M., Middleton, H.C., Williams, E.J., Pickard, J.D., Sahakian, B.J.,
Robbins, T.W., 1999. Choosing between small, likely rewards and large, unlikely
rewards activates inferior and orbital prefrontal cortex. Journal of Neuroscience 19
(20), 90299038.
Schmahmann, J.D., 2003. Vascular syndromes of the thalamus. Stroke 34 (9), 22642278.
Self, D.W., Nestler, E. J., 1998. Rel apse to d rug-seeking: neural and molec ular
mechanisms. Drug and Alcohol Dependence 51 (12), 4960.
Sewards, T.V., Sewards, M.A., 2003. Representations of motivational drives in mesial cortex,
medial thalamus, hypothalamus and midbrain. Brain Research Bulletin 61 (1), 2549.
Smolka, M.N., Buhler, M., Klein, S., Zimmermann, U., Mann, K., Heinz, A., Braus, D.F., 2006.
Severity of nicotine dependence modulates cue-induced brain activity in regions
involved in motor preparation and imagery. Psychopharmacology 184 (34), 577588.
Talairach, J., Tournoux, P., 1988. Co-planar stereotaxic atlas of the human brain. Thieme,
New York.
Tanabe, J., Thompson, L., Claus, E., Dalwani, M., Hutchison, K., Banich, M.T., 2007.
Prefrontal cortex activity is reduced in gambling and nongambling substance users
during decision-making. Human Brain Mapping 28 (12), 12761286.
Toce-Gerstein, M., Gerstein, D.R., Volberg, R.A., 2003. A hierarchy of gambling disorders
in the community. Addiction 98 (12), 16611672.
Tomasi, D., Goldstein, R.Z., Telang, F., Maloney, T., Alia-Klein, N., Caparelli, E.C., Volkow,
N.D., 2007. Thalamo-cortical dysfunction in cocaine abusers: implications in
attention and perception. Psychiatry Research 155 (3), 189201.
Toni,I., Thoenissen, D., Zilles, K., Niedeggen, M., 2002. Movement preparation and working
memory: a behavioural dissociation. Experimental Brain Research 142 (1), 158162.
Volkow, N.D., Fowler, J.S., Wang, G.J., Goldstein, R.Z., 2002. Role of dopamine, the frontal
cortex and memory circuits in drug addiction: insight from imaging studies.
Neurobiology of Learning and Memory 78 (3), 610624.
Volkow, N.D., Fowler, J.S., Wang, G.J., 2003. The addicted human brain: insights from
imaging studies. Journal of Clinical Investigation 111 (10), 14441451.
Wang, Z., Faith, M., Patterson, F., Tang, K., Kerrin, K., Wileyto, E.P., Detre, J.A., Lerman, C.,
2007. Neural substrates of abstinence-induced cigarette cravings in chronic
smokers. Journal of Neuroscience 27 (51), 1403514040.
Wilson, S.J., Sayette, M.A., Fiez, J.A., 2004. Prefrontal responses to drug cues: a
neurocognitive analysis. Nature Neuroscience 7 (3), 211214.
Wulfert, E., Roland, B.D., Hartley, J., Franco, C., Wang, N., 2005. Heart rate arousal and
excitement in gambling: winners versus losers. Psychology of Addictive Behaviors
19 (3), 311316.
Yang, J., Li, H., Zhang, Y., Qiu, J., Zhang, Q., 2007. The neural basis of risky decision-
making in a blackjack task. NeuroReport 18 (14), 15071510.
Zack, M., Poulos, C.X., 2004. Amphetamine primes motivation to gamble and gambling-
related semantic networks in problem gamblers. Neuropsychopharmacology 29
(1), 195207.
Zack, M., Poulos, C.X., 2007. A D2 antagonist enhances the rewarding and priming
effects of a gambling episode in pathological gamblers. Neuropsychopharmacology
32 (8), 16781686.
173S.F. Miedl et al. / Psychiatry Research: Neuroimaging 181 (2010) 165173
    • e, functional connectivity analyses revealed, in PrG, relative to HC, an increased VS connectivity in regions including occipital fusiform gyrus, posterior cingulate cortex, superior and middle temporal gyrus. Connectivity, between the VS seed and the occipital fusiform and the middle temporal gyrus, was also positively correlated with SOGS scores.Miedl et al. (2010)assessed, using fMRI, neural correlates of 12 PrG and 12 HC male individuals within a quasi-realistic blackjack scenario where participants had to choose whether or not to draw a card in highrisk and low-risk gaming situations. No behavioral differences were found between the groups. In fact, both PrG and HC showed a significantly lower
    [Show abstract] [Hide abstract] ABSTRACT: Decreased cognitive control over the urge to be involved in gambling activities is a core feature of Gambling Disorder (GD). Cognitive control can be differentiated into several cognitive sub-processes pivotal in GD clinical phenomenology, such as response inhibition, conflict monitoring, decision-making, and cognitive flexibility. This article aims to systematically review fMRI studies, which investigated the neural mechanisms underlying diminished cognitive control in GD. We conducted a comprehensive literature search and collected neuropsychological and neuroimaging data investigating cognitive control in GD. We included a total of 14 studies comprising 499 individuals. Our results indicate that impaired activity in prefrontal cortex may account for decreased cognitive control in GD, contributing to the progressive loss of control over gambling urges. Among prefrontal regions, orbital and ventromedial areas seem to be a possible nexus for sensory integration, value-based decision-making and emotional processing, thus contributing to both motivational and affective aspects of cognitive control. Finally, we discussed possible therapeutic approaches aimed at the restoration of cognitive control in GD, including pharmacological and brain stimulation treatments.
    Full-text · Article · Apr 2017
    • It follows that experienced poker players might be able to " let go " of unfavorable outcomes from previous actions, and consequently they might be better skilled at regulating themselves when facing monetary risky decisions (Laakasuo et al., 2014; Palomäki et al., 2013 Palomäki et al., , 2014). Besides, previous neuroimaging studies on decision-making in gamblers have been undertaken with individuals suffering from gambling disorders and recruited from addiction treatment centers (Choi et al., 2012; van Holst, Veltman, Büchel, van den Brink, & Goudriaan, 2012) or did not control for gamblers' preferred type of gambling (e.g., poker vs. slot-machine; Balodis et al., 2012; Brevers et al., 2015a; Chase & Clark, 2010; Miedl, Fehr, Meyer, & Herrmann, 2010; Peters, Miedl, & Büchel, 2013; Power, Goodyear, & Crockford, 2012; van Holst, Chase, & Clark, 2014). This could have biased gambler participants' approach towards monetary risk-taking (Lorains et al., 2014; Turner, 2014).
    [Show abstract] [Hide abstract] ABSTRACT: Individuals have a tendency to be more risky in their choices after having experienced a monetary loss, than after a reward. Here, we examined whether prior outcomes influence differently the patterns of neural activity of individuals who are used to taking monetary risk, namely poker players. High-frequency poker players and non-gamblers were scanned while performing a controlled task that allowed measuring the effect of prior outcomes on subsequent decisions. Both non-gamblers and poker players took more risks after losing a gamble than after winning one. Neuroimaging data revealed that non-gamblers exhibited higher brain activation than poker players when pondering a decision after losing, as compared to after winning. The opposite was found in poker players. This differential pattern of activation was observed in brain regions involved in high-order motor processes (the dorsal premotor cortex). These results suggest that gambling habits introduce significant changes in action preparation during decision-making following wins and losses.
    Full-text · Article · Jan 2017
    • This suggests that, during the IGT, OFC activity might essentially underlie self-regulatory affective processes (Bechara, Damasio & Damasio 2000). By contrast, during the card games used in previous fMRI studies (Miedl et al. 2010; van Holst et al. 2012; Brevers et al. 2015), participants computed their choices based on explicit information (probabilities and values of the potential rewards/losses displayed on the screen) and (ultimately) independently of previous choice outcomes. During these more simple card games, the OFC might be primarily involved in cue reactivity (see also Goudriaan et al. 2010).
    [Show abstract] [Hide abstract] ABSTRACT: The aim of this study was to examine the impact of different neural systems on monetary decision making in frequent poker gamblers, who vary in their degree of problem gambling. Fifteen frequent poker players, ranging from non-problem to high-problem gambling, and 15 non-gambler controls were scanned using functional magnetic resonance imaging (fMRI) while performing the Iowa Gambling Task (IGT). During IGT deck selection, between-group fMRI analyses showed that frequent poker gamblers exhibited higher ventral-striatal but lower dorsolateral prefrontal and orbitofrontal activations as compared with controls. Moreover, using functional connectivity analyses, we observed higher ventral-striatal connectivity in poker players, and in regions involved in attentional/motor control (posterior cingulate), visual (occipital gyrus) and auditory (temporal gyrus) processing. In poker gamblers, scores of problem gambling severity were positively associated with ventral-striatal activations and with the connectivity between the ventral-striatum seed and the occipital fusiform gyrus and the middle temporal gyrus. Present results are consistent with findings from recent brain imaging studies showing that gambling disorder is associated with heightened motivational-reward processes during monetary decision making, which may hamper one's ability to moderate his level of monetary risk taking. © 2015 Society for the Study of Addiction.
    Full-text · Article · Mar 2016
    • Additionally, research has shown reduced VMPFC activation in PG undertaking a Probabilistic Reversal Task, where participants were given positive reinforcement for their correct responses (monetary gain) and punished for giving incorrect answers (monetary loss) (De Ruiter et al. 2009). In contrast, several studies have found increased activity in the mesocorticolimbic brain regions, such as experiments that vary the amount of risk involved (e.g. Miedl et al. 2010) or that use different probabilities of winning or losing varying amounts of money (e.g. Van Holst et al. 2012b).
    [Show abstract] [Hide abstract] ABSTRACT: In light of the upcoming eleventh edition of the International Classification of Diseases (ICD-11), the question arises as to the most appropriate classification of 'Pathological Gambling' ('PG'). Some academic opinion favors leaving PG in the 'Impulse Control Disorder' ('ICD') category, as in ICD-10, whereas others argue that new data especially from the neurobiological area favor allocating it to the category of 'Substance-related and Addictive Disorders' ('SADs'), following the decision in the fifth revision of the Diagnostic and Statistical Manual of Mental Disorders. The current review examines important findings in relation to PG, with the aim of enabling a well-informed decision to be made with respect to the classification of PG as a SAD or ICD in ICD-11. Particular attention is given to cognitive deficits and underlying neurobiological mechanisms that play a role in SADs and ICDs. These processes are impulsivity, compulsivity, reward/punishment processing and decision-making. In summary, the strongest arguments for subsuming PG under a larger SAD category relate to the existence of similar diagnostic characteristics; the high co-morbidity rates between the disorders; their common core features including reward-related aspects (positive reinforcement: behaviors are pleasurable at the beginning which is not the case for ICDs); the findings that the same brain structures are involved in PG and SADs, including the ventral striatum. Research on compulsivity suggests a relationship with PG and SAD, particularly in later stages of the disorders. Although research is limited for ICDs, current data do not support continuing to classify PG as an ICD.
    Article · Mar 2016
    • Étant donné le rôle attribué à ces régions dans les processus mnésiques, émotionnels et visuels, les auteurs ont interprété ces résultats comme un signe que les indices de jeu possèdent une saillance exacerbée chez les joueurs pathologiques. Enfin, deux études se sont intéressées à la phase d'évaluation du risque dans un jeu de blackjack [41, 56] . Les résultats ont mis en évidence une augmentation d'activité dans le striatum et le cortex orbito-frontal des joueurs pathologiques dans les essais à haut risque en comparaison aux essais à faible risque.
    [Show abstract] [Hide abstract] ABSTRACT: Although most people consider gambling as a recreational activity, some individuals lose control over their behavior and enter a spiral of compulsive gambling leading to dramatic consequences. In its most severe form, pathological gambling is considered a behavioral addiction sharing many similarities with substance addiction. A number of neurobiological hypotheses have been investigated in the past ten years, relying mostly on neuroimaging techniques. Similarly to substance addiction, a number of observations indicate a central role for dopamine in pathological gambling. However, the underlying mechanism seems partly different and is still poorly understood. Neuropsychological studies have shown decision-making and behavioral inhibition deficits in pathological gamblers, likely reflecting frontal lobe dysfunction. Finally, functional MRI studies have revealed abnormal reactivity within the brain reward system, including the striatum and ventro-medial prefrontal cortex. These regions are over-activated by gambling cues, and under-activated by monetary gains. However, the scarcity and heterogeneity of brain imaging studies currently hinder the development of a coherent neurobiological model of pathological gambling. Further replications of results and diversification of approaches will be needed in the coming years in order to strengthen our current model. © 2015 médecine/sciences – Inserm.
    Full-text · Article · Sep 2015
    • Ventromedial activation has also been found to be reduced when participants were given positive reinforcement for their correct responses (monetary gain) and punished for giving incorrect answers (monetary loss) on a probabilistic reversal task (e.g. de Ruiter et al., 2009). By contrast, fMRI studies that varied the amount of risk involved (e.g. Miedl et al., 2010 ) or used tasks with different probabilities of winning or losing varying amounts of money (e.g. van Holst et al., 2012b) have found increased activity in the mesocorticolimbic brain regions.
    [Show description] [Show abstract] [Hide description] [Hide abstract] DESCRIPTION: Neural correlates of reward processing and response inhibition in disordered gambling: The role of depressive symptomatologyABSTRACT: Introduction: Cognitive features that play an important role in the development and maintenance of disordered gambling (DG) are impulsivity and sensitivity to reward. However, behavioural and neuroimaging findings are less consistent as one would expect when looking at the diagnostic criteria of the disorder. Given that almost half of the DGs suffer from comorbid depressive symptomatology our aim was to assess its impact on reward processing and response inhibition. Methods: We presented two different tasks during 3 tesla functional magnetic resonance imaging to study the neurobiological correlates of 1) effort-related reward processing and 2) motor response inhibition (work in progress) in a large cohort of disordered gamblers (DGs) and healthy controls (HCs). Depressive symptoms were assessed using the Beck Depression Inventory (BDI). DGs and HCs were divided into subgroups (“high” and “low”) based on their BDI scores. Results: Neither effort-related monetary reward processing nor inhibition-related brain activation differed between the complete groups of HCs and DGs. Notably, for both tasks we found a significant Group × BDI interaction. During receipt of monetary reward, DGs with higher BDI scores compared to DGs with lower BDI scores showed greater brain activity in the right insula cortex and dorsal striatum. No differences were observed for HCs with higher versus lower BDI scores. During successful response inhibition, DGs with higher BDI scores compared to those with lower scores revealed significantly diminished activity in a parieto-frontal network including inferior and middle frontal gyrus, anterior cingulate cortex, supplementary motor area and postcentral gyrus among other stopping-relevant regions. No differences were observed for HCs with higher versus lower BDI scores. Conclusion: Comorbid depressive symptomatology in DGs has a significant impact on effort-related reward processing and inhibition-related brain activation. Our findings strengthen the need for subgroup comparisons in future investigations of disordered gambling as part of a personalized medicine approach.
    File · Presentation · Aug 2015
Show more