Neurophysiological correlates
of biased cognitive processing
in addiction
Marianne Littel
Neurophysiological correlates of biased
cognitive processing in addiction
Marianne Littel
Cover photography by Valo-x-Stock http://valo-x-stock.deviantart.com/
Cover photo manipulation by Esra Zengin http://sallysalander.deviantart.com/
Layout by Ferdinand van Nispen, Citroenvlinder-DTP.nl, Bilthoven
Printed by GVO drukkers & vormgevers B.V. | Ponsen & Looijen
ISBN 978-90-6464-551-8
Copyright © 2012 M. Littel
All rights reserved. No part of this thesis may be reproduced or transmitted in any
form, by any means, electronic or mechanical, without the prior written permission
of the author, or where appropriate, of the publisher of the articles.
The research presented in this dissertation was supported by the Netherlands
Organisation for Scientific Research (NWO, VIDI grant 016.08.322).
Neurophysiological correlates of biased
cognitive processing in addiction
Neurofysiologische correlaten van verstoorde
cognitieve verwerking in verslaving
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de
Erasmus Universiteit Rotterdam
op gezag van de
rector magnificus
Prof.dr. H.G. Schmidt
en volgens besluit van het College van Promoties.
De openbare verdediging zal plaatsvinden op
vrijdag 1 juni 2012 om 9:30 uur
door
Marianneke Littel
geboren te Willemstad
Promotiecommissie:
Promotor:
Prof.dr. I.H.A. Franken
Overige leden:
Prof.dr. J.W. van Strien
Prof.dr. R. Zwaan
Prof.dr. W. van den Brink
voor mijn ouders
Contents
Neurophysiological correlates of biased cognitive processing in addiction
Chapter 1
Electrophysiological indices of biased cognitive
processing of substance-related cues: a meta-analysis
Chapter 3
Psychometric properties of the brief Questionnaire on
Smoking Urges (QSU-Brief) in a Dutch smoker population
Chapter 2
Chapter 4
Chapter 5
Chapter 6
Chapter 7
Chapter 8
Chapter 9
Overall outline and hypotheses
The effects of prolonged abstinence on the processing
of smoking cues: an ERP study among smokers,
ex-smokers and never-smokers
Changes in the electroencephalographic spectrum in
response to smoking cues in smokers and ex-smokers
Implicit and explicit selective attention to smoking
cues in smokers indexed by brain potentials
Electrophysiological correlates of associative learning
in smokers: a higher-order conditioning experiment
Intentional modulation of the Late Positive Potential in
response to smoking cues by cognitive strategies in smokers
Reduced cognitive processing of alcohol cues in alcoholdependent patients seeking treatment: an ERP study
9
41
47
59
83
99
123
147
173
Chapter 10 Error-processing and response inhibition in excessive
computer game players: an ERP study
189
References
235
Chapter 11 Summary and discussion
211
Samenvatting (summary in Dutch)
265
Curriculum Vitae
297
Dankwoord (acknowledgements in Dutch)
291
Publications299
Chapter
1
Electrophysiological indices of biased
cognitive processing of substancerelated cues: a meta-analysis
Littel, M., Euser, A.S., Munafo, M.R., & Franken, I.H.A. (submitted
for publication). Electrophysiological indices of biased cognitive
processing of substance-related cues: a meta-analysis.
Chapter 1
Abstract
Several studies indicate that individuals with substance use disorders (SUD)
exhibit biases in the cognitive processing of substance-related stimuli. These
biases facilitate the detection and selection of substance cues and have been
argued to play a causal or perpetuating role in the reactivity to substance cues.
Indeed, mounting evidence suggests that these cognitive processing biases
are important in addictive behaviors. Two electrophysiological indices of
cognitive processing, the P300 and Late Positive Potential (LPP) components
of the event-related potential (ERP), are associated with the deployment of
attentional resources to motivationally relevant stimuli. In the present metaanalysis P300 and LPP amplitudes are used to investigate whether SUD persons
show enhanced cognitive processing of substance cues relative to neutral
cues as opposed to healthy control participants. Seventeen studies yielding
30 independent subject samples and six studies yielding ten independent
samples were selected for meta-analysis assessing cue-elicited P300 and LPP
responses, respectively. Results indicated the P300 and LPP amplitude effect
sizes were significantly larger in addicted participants than in controls. This
result can be explained by substance users’ motivated attention, that is, the
allocation of attention and memory resources to stimuli relevant to substance
users’ motivational states. Additional stratified moderator analyses revealed
that both P300 and LPP amplitudes were not moderated by electrode site (Fz vs.
Pz), type of substance used (stimulants vs. depressants), substance use status
(abstinent vs. non-abstinent), age, gender and task requirements (active vs.
passive paradigms). These results indicate that enhanced electrophysiological
processing of substance cues appears characteristic of SUD in general, might be
independent of recency of substance use, and presumably occurs irrespective
of task demands.
10
Meta-analysis of biased cognitive processing in SUD
Introduction
Cognitive processing biases in addiction
Addiction is a chronic, hard-to-treat condition characterized by cravings and
frequently occurring relapse. Over the years, addiction has been associated
with enhanced reactivity in response to substance-related stimuli such as
the sight or smell of drugs and drug paraphernalia. This drug cue reactivity is
comprised of a physiological component (e.g., skin conductance), a psychological
component (e.g., self-reported urges: for a review, see Carter & Tiffany, 1999),
and a cognitive component, i.e., individuals with substance use disorders (SUD)
exhibit biases in the cognitive processing of substance-related stimuli. These
biases, including biases in attention and memory, facilitate detection and selection
of substance cues and have been argued to play a causal or perpetuating role in
the physiological and psychological reactivity to substance cues. For example,
it is hypothesized by Franken (2003) that SUD individuals automatically detect
and orient attention toward substance-related stimuli. This increased attention
for substance cues diminishes attention left for alternative cues, enhances
substance-related cognitions, and causes subjective craving. The biological basis
of these processing biases can be explained by the incentive-sensitization theory
of addiction (Robinson & Berridge, 1993). This theory posits that repeated drug
administration causes a sensitization of dopaminergic neurotransmission in
the brain. Because of this sensitization drugs and drug-related stimuli acquire
incentive motivational properties, which can alter the way they are perceived
and processed. More specifically, substance-related stimuli will be perceived as
particularly salient and reinforcing, and attention will be preferentially allocated
to these stimuli (Franken, 2003; Robinson & Berridge, 1993; Ryan, 2002).
The existence of cognitive processing biases in addiction has been repeatedly
confirmed in studies using various behavioral tasks (see for reviews Field & Cox,
2008; Field, Munafò, & Franken, 2009). For example, utilizing modified Stroop
tasks, it has been demonstrated that SUD individuals are slower to color-name
substance-related words than neutral words relative to control participants (see
W. M. Cox, Fadardi, & Pothos, 2006). In addition, using visual probe tasks, it has
been shown that substance users exhibit faster reaction times to probes replacing
substance-related cues than to probes replacing neutral cues (e.g., Chanon, Sours,
& Boettiger, 2010; Franken, Kroon, & Hendriks, 2000; Mogg & Bradley, 2002).
Results from these studies as well as from studies utilizing other attentional
11
Chapter 1
paradigms (e.g., eye movement studies, dual task procedures, flicker induced
change blindness paradigms: see Field & Cox, 2008; Field et al., 2009), indicate
that substance users display enhanced attentional processing of substancerelated stimuli. In addition to substance-related attentional bias, substancerelated memory bias has also been demonstrated. For example, Franken, Rosso
and van Honk (2003) showed that after a picture color matching task, alcohol
patients recalled more alcohol pictures, but not neutral pictures, than light
drinkers. Using a more implicit measure of memory, i.e., a word-stem completion
task, McCusker and Gettings (1997) found that after a modified Stroop task, wordstems primed more gambling-related words in pathological gamblers than in
controls. Additional evidence comes from neuroimaging studies, in which it has
been shown that presenting substance-related stimuli to SUD individuals not
only increases activations in brain circuits involved in motivation and reward,
such as the amygdala and the ventral striatum, but also in brain circuits that are
normally involved in learning, memory and attention, such as the hippocampus,
the prefrontal cortex and the anterior cingulate cortex (David et al., 2005; Hyman,
2005; Luijten et al., 2010; Robbins & Everitt, 1996).
Mounting evidence suggests that cognitive processing biases are important in the
development and maintenance of addiction. Attentional bias has been associated
with poor treatment outcome (Carpenter, Schreiber, Church, & McDowell, 2006),
relapse following treatment (W. M. Cox, Hogan, Kristian, & Race, 2002; Garland,
Franken, & Howard, 2011; Marissen et al., 2006; Waters et al., 2003) and substance
consumption behavior (W. M. Cox, Pothos, & Hosier, 2007; Fadardi & Cox, 2009; Field
& Eastwood, 2005; Waters & Feyerabend, 2000), and both memory and attentional
bias have been associated with self-reported craving (Field et al., 2009; Franken et
al., 2003). All these studies indicate that attentional bias is an important concept
in addiction; at least it seems an important predictor of craving and substance use
behavior. However, direct evidence for a causal role of attentional bias in substance
use is lacking (e.g., Hogarth, Dickinson, Janowski, Nikitina, & Duka, 2008; Hogarth,
Dickinson, & Duka, 2009).
To summarize, addiction is characterized by cognitive processing biases. SUD
individuals are shown to selectively attend to substance-related stimuli and
memorize these at the cost of other stimuli. Associations between these biases and
craving, substance use and relapse highlight the importance of further investigation
in cognitive processing biases in addiction. In this meta-analysis we will focus on
12
Meta-analysis of biased cognitive processing in SUD
a relatively new method to asses cognitive processing of substance cues, i.e., the
measurement of Event-Related Potentials (ERPs) using electroencephalography
(EEG) techniques. ERP methodology provides a potentially more direct assessment
of attentional processing than conventional behavioral (reaction time/ accuracy)
data relying on indirect motor-responses.
Event-related potentials as index for cognitive processing bias
ERPs are manifestations of brain activities that occur in preparation for, or in
response to discrete events (Fabiani, Gratton, & Coles, 2000). They consist of several
peaks and troughs that tend to co vary in response to experimental manipulations.
Positive and negative deflections that have been associated with specific
information-processing operations are called components (Coles & Rugg, 1995).
Components are often labeled after their polarity (e.g., positive) and relative latency
(e.g., 300 ms) and vary in amplitude which presumably depicts the extent to which a
processing operation is engaged (Kok, 1990). In the present meta-analysis we focus
on two late positive ERP components that have been consistently associated with
attentional processes and have been studied most frequently in drug cue-reactivity
paradigms, namely the P3 or P300 and the Late Positive Potential (LPP).
The P300 component refers to a large positive deflection of the ERP, arising about
300-800 ms after stimulus presentation, which is typically maximal at medial
central and parietal electrode sites (e.g., Polich & Kok, 1995b; Pontifex, Hillman, &
Polich, 2009). This component has been extensively studied in oddball paradigms,
in which participants have to respond to non-frequent target stimuli presented
among a series of frequent non-target stimuli. In these paradigms the P300
amplitude tends to be more enhanced in response to targets as compared to nontargets (e.g., Donchin, 1981; Johnson, 1986), i.e., in response to stimuli that are
task-relevant and demand attention. Furthermore, the P300 amplitude in response
to targets is typically decreased when attention is directed elsewhere, for example
when it is occupied by a distracter task (Duncan-Johnson & Donchin, 1977;
Hillyard, Hink, Schwent, & Picton, 1973). Therefore it is generally believed that
the P300 reflects the mental processes underlying the deployment of attentional
resources to task-relevant stimuli (Donchin, 1981; Johnson, 1986; Polich & Kok,
1995b; Polich & Criado, 2006; Pontifex et al., 2009). Furthermore, the P300 has
been associated with the evaluation of task-relevant stimuli, and the subsequent
memory mechanisms engaged for these stimuli.
13
Chapter 1
Stimuli that are necessary for survival of the individual, i.e., stimuli that signal
threat or danger as well as stimuli that signal the availability of sex and food or
reward, tend to automatically attract attention (e.g., Ohman, Flykt, & Esteves,
2001). This automatic attention has been termed ‘motivated attention’, since it is
believed to originate from the individual’s drive state or motivation to survive and
appears dependent on motivational factors (e.g., approach or avoidance tendencies;
Lang, Bradley, & Cuthbert, 1997). Indeed, research has repeatedly shown that the
P300 component of the ERP is enhanced in response to pleasant and unpleasant
stimuli relative to neutral stimuli with most arousing pictures eliciting the
largest P300 amplitudes (for a review, see Hajcak, MacNamara, & Olvet, 2010). In
addition, when motivationally relevant stimuli are also task-relevant, i.e., when
they are targets in an oddball task, the difference between the P300 evoked by
motivationally relevant as compared to neutral stimuli has been observed to be
even larger (Ferrari, Codispoti, Cardinale, & Bradley, 2008; Schupp et al., 2007),
suggesting an interaction of directed and motivated attention.
More recent work in this field has focused on a P300-related component of the
ERP referred to as the Late Positive Potential (LPP). Whereas the P300 component
appears to be transient, the LPP is typically enhanced for several seconds after
the presentation of motivationally relevant stimuli (Cuthbert, Schupp, Bradley,
Birbaumer, & Lang, 2000; Hajcak & Olvet, 2008). It has been argued that the P300
and the LPP are functionally similar in that they reflect the same underlying
mental processes (Kok, 1997). From this point of view, the LPP reflects additional
attentive processing or a continuation of attentive processing of motivationally
relevant stimuli. However, research has indicated that the LPP has both a longer
duration and a different topography, with activity shifting from parietal electrode
sites to more fronto-central sites (Foti & Hajcak, 2008; Foti, Hajcak, & Dien, 2009;
Hajcak, Dunning, & Foti, 2007), indicating that the LPP and the P300 cannot be
considered entirely identical components (Foti et al., 2009). Although research on
this topic is limited, it indeed appears that the LPP reflects additional involvement
of memory encoding and storage (Koenig & Mecklinger, 2008) and that it is more
sensitive to top-down cognitive modulation (Hajcak et al., 2010). Because of the
functional overlap between the P300 and LPP, the two components have been
quantified in different windows across studies. For example, amplitudes in the
400-600 ms time window have been referred to as P300 in some studies (e.g.,
Schupp et al., 2007) and as LPP in other studies (e.g., Moser, Hajcak, Bukay, &
Simons, 2006), hindering the differentiation between effects on the P300 versus
14
Meta-analysis of biased cognitive processing in SUD
the LPP. However, recent results from an exploratory temporospatial principal
components analysis conducted by Foti et al. (2009) have indicated that the spatial
topography of the activity in the 300-600 ms time window, often referred to as
(part of) the LPP, is consistent with the P300. Furthermore, it was demonstrated
that activity shifted from posterior to anterior recording sites from about 1000 ms
after stimulus onset, implying that the two components are ideally separated
somewhere between 600 and 1000 ms after stimulus presentation. In this article
we chose to use the term P300 for positive ERP deflections between 300 and 800 ms,
even if they have been termed LPP in the reviewed studies, whereas we chose to
use LPP for the sustained positivity from 800 ms and onwards.
In addition to the P300 (300-800 ms), the LPP (> 800 ms) has also been found
to be enhanced in response to pleasant and unpleasant stimuli, as compared to
neutral stimuli (Cuthbert et al., 2000; Dunning & Hajcak, 2009; Foti & Hajcak,
2008; Hajcak et al., 2007; Hajcak & Olvet, 2008; Hajcak, Dunning, & Foti, 2009;
Thiruchselvam, Blechert, Sheppes, Rydstrom, & Gross, 2011). Thus, not only do
motivationally relevant stimuli capture attention and evaluation as reflected in
P300 ERP potentials, they also recruit more elaborate and sustained attention and
subsequent memory processes that are reflected in sustained late positive ERP
potentials.
Enhancement of late ERP components indicate motivated
attention for substance cues in SUD populations
Preferential processing of pleasant and unpleasant stimuli and stimuli signaling
emotion can be considered evolutionary relevant to all individuals, and indeed
motivated attention for these stimuli, as reflected by enhanced P300 and
LPP amplitudes, has been replicated across different samples. At some points
however, individuals can differ in their motivations, and consequently in what is
motivationally relevant. These individual differences in stimulus preference have
been found to be related to variation in the late ERP components. For example,
enlarged P300 amplitudes have been found for hungry compared to satiated
individuals in response to food cues (Nijs, Franken, & Muris, 2008; Nijs, Muris,
Euser, & Franken, 2010; Stockburger, Schmalzle, Flaisch, Bublatzky, & Schupp,
2009). From these results one would expect that SUD individuals, who are highly
motivated to seek out and consume drugs, would differ from healthy controls with
regard to electrophysiological processing of substance-related stimuli.
15
Chapter 1
As early as the 80’s, electrophysiological processing of an alcohol cue was
investigated in a small sample of alcoholics (Genkina & Shostakovich, 1983;
Shostakovich, 1987). It was found that alcoholics displayed enlarged P300
amplitudes in response to brief presentations of the word ‘vodka’, but not to
similarly presented neutral words, and that this effect was absent in control
participants. Warren and McDonough (1999) continued this line of research and
investigated ERP cue-reactivity in smokers. Instead of words they used pictures
which were repeated several times and presented for 150 ms. Results indicated
that the P300 was enlarged in response to smoking-related pictures as compared
to neutral pictures. Although this discrepancy was found in both smokers and nonsmokers, it was significantly larger for smokers at frontal and central electrode
sites. Comparable results were obtained in a replication study by the same
researchers two years later (McDonough & Warren, 2001). Because the results
were consistent with earlier studies showing that the P300 varies as a function
of affective value and motivational significance of stimuli, the authors concluded
that the effects were caused by the enhanced value or motivational significance of
smoking-related stimuli for smokers, with enhanced attentional allocation as the
underlying mechanism. Thus, the enhanced processing of smoking cues observed
in smokers, as reflected by enhanced P300 amplitudes, was assumed to reflect
the allocation of attentional resources to stimuli relevant to smokers’ smokingaddicted, motivational states.
In the years that followed, several researchers have sought to extent these
findings. With regard to smoking addiction, Littel and Franken (2007) confirmed
the finding that the P300 amplitude in response to smoking pictures compared
to neutral pictures is enhanced in smokers. In contrast to the results of Warren
and McDonough (1999) and McDonough and Warren (2001), they found the effect
to be absent in non-smokers, providing stronger evidence for the conclusion that
enhanced attentional processing of smoking-related stimuli is uniquely driven by
smokers’ smoking-addicted, motivational states. Furthermore, it was shown in
this study that the effects are uniquely attributable to the smoking-related content
of the pictures. Differential effects of physical properties of the visual stimuli
were ruled out, since all smoking and neutral pictures were closely matched for
all but the smoking-related features of content (i.e., the exact same backgrounds,
persons and number of persons in both types of stimuli). More recently, Versace et
al. (2011; 2011), demonstrated increased P300 amplitudes in response to smoking
cues relative to neutral cues in larger groups of smokers (n = 116 and n = 180)
16
Meta-analysis of biased cognitive processing in SUD
with varying dependence levels, with the most dependent smokers showing the
largest amplitude differences. Similar results were obtained by Littel and Franken
(2011), although no differences were observed between moderate and light
smokers in the latter study.
ERP cue-reactivity research has not been confined to alcoholics and smokers.
Up until now, enlarged P300 amplitudes (300-800 ms) elicited by substancerelated and neutral pictures and words have also been observed in cannabis users
(Wölfling, Flor, & Grüsser, 2008), cocaine users (Dunning et al., 2011; Sokhadze
et al., 2008), and heroin users (Franken, Stam, Hendriks, & van den Brink, 2003;
Lubman, Allen, Peters, & Deakin, 2008) as compared to healthy controls. In some
studies distinct analyses have been conducted for early and late P300 components
(Herrmann et al., 2000; Littel & Franken, 2007). In these studies both early and late
P300 amplitudes appeared to be enhanced in response to substance-related cues
as compared to neutral cues. Enhanced LPP amplitudes (> 800 ms) in response to
substance-related stimuli have been found in heroin users (Franken, Stam et al.,
2003) and cocaine users (Dunning et al., 2011; Franken et al., 2008; Van de Laar,
Licht, Franken, & Hendriks, 2004) relative to controls.
Of all ERP cue-reactivity studies in SUD individuals, two studies failed to find
significant effects. In a study by Hansenne et al. (2003), no ERP differences were
found between alcoholics and controls in response to alcohol-related words
relative to neutral words, whereas Jang et al. (2007) did not observe significant
differences between smokers’ and non-smokers’ ERPs in response to smokingrelated pictures. Both studies, however, included only 10 participants per group,
indicating that the absence of effects could have been caused by a lack of power.
Furthermore, in the first study alcohol and neutral words were not matched for
length, frequency or emotional valance and arousal, whereas the latter study
included smokers that smoked only two cigarettes per day, implying that the
smoker sample might not have been nicotine dependent.
All of the above studies utilized passive paradigms in which participants were
instructed to simply view presented words and pictures. There were no additional
task demands and stimuli were presented with equal probabilities. An advantage
of this design is that effects can neither be caused by high or low probability of
certain stimuli compared to others nor by specific task instructions that dictate
which stimuli are relevant. However, this passive paradigm also has a disadvantage.
17
Chapter 1
Although results from studies employing this paradigm have all been explained
in terms of substance users’ motivated attention toward substance-related cues,
one cannot be completely sure whether the effects are caused by the motivational
significance of substance stimuli or merely (or additionally) by participants’
intentional viewing strategies (Lubman, Allen, Peters, & Deakin, 2007). In other
words, it cannot be inferred whether the effects arise because of an implicit,
involuntary capture of attention or because of an explicit, voluntary choice to
focus attention on substance stimuli.
In order to solve this issue, several studies have been conducted in which attention
was manipulated (‘active’ studies). In an oddball study by Lubman et al. (2007)
heroin-dependent patients and controls were presented with frequent neutral and
heroin-related non-target stimuli, among which infrequent neutral and heroinrelated targets were presented. Stimuli were displayed for 500 ms and participants
had to respond to targets with a button press. ERPs were recorded in response
to neutral and heroin-related non-targets and all oddballs combined (neutral
and heroin-related targets). Results showed that heroin-dependent patients
responded with increased P300 amplitude to heroin non-targets as compared
to controls. These results are in line with the results from passive viewing
paradigms, but provide additional evidence that enhanced processing of substance
stimuli displayed by SUD individuals remains present when attention is directed
elsewhere. Therefore, substance-related stimuli are shown to capture the implicit,
involuntary attention of substance users. These results have been replicated in
cocaine users (Sokhadze et al., 2008) and smokers (Littel & Franken, 2010). In
these studies non-target substance stimuli were presented at shorter presentation
durations (respectively 200 ms and 333 ms) and without perceivable stimulus
intervals, making it even less probable that explicit, voluntary attention played a
role in the enhanced processing of non-target, task-irrelevant substance-related
stimuli. An additional comparison was made between responses to substancerelated target stimuli and neutral target stimuli in these studies. Results showed
that SUD individuals additionally displayed enhanced processing of substancerelated targets as compared to neutral targets, indicating that SUD individuals also
show enhanced voluntary attention toward substance cues. Similar results were
obtained in an oddball study by Namkoong et al. (2004) conducted in a sample of
alcoholics. In sum, motivated attention in addiction seems to be comprised of both
implicit and explicit attention for substance cues, the latter operating in addition
to selective attention directed by task-demands (i.e., responding to predefined
18
Meta-analysis of biased cognitive processing in SUD
targets). This is in agreement with the earlier mentioned assumption that there
exists an interaction between directed and motivated attention (Ferrari et al.,
2008; Schupp et al., 2007).
In addition to oddball paradigms, other ‘active’ paradigms have been employed
in two studies (Fehr, Wiedenmann, & Herrmann, 2006; Fehr, Wiedenmann, &
Herrmann, 2007). In the first study smokers and non-smokers performed a
smoking-related Stroop task in which they had to respond to the colors of the
words and to ignore word content. In the other study smokers and non-smokers
performed a smoking-related picture color matching task, in which they had to
respond to the colors of the pictures and ignore the picture content. Results of
both studies revealed that smokers, as compared to non-smokers, showed greater
late ERP positivity in response to primary smoking cues (e.g., cigarettes, people
smoking; Fehr et al., 2007) and secondary smoking cues (e.g., bus stops, kiosks;
Fehr et al., 2006; Fehr et al., 2007). These findings are in line with those obtained in
the substance-related oddball studies in that they provide evidence for enhanced
processing of substance cues while attention is directed elsewhere.
Despite the numerous indications of enhanced cognitive processing of substancerelated stimuli by SUD individuals on the electrophysiological level several issues
need to be resolved. First of all, the overall effect size needs to be confirmed by
a meta-analysis. As noted before, not all studies have found significant P300/
LPP amplitude differences between substance users and controls in response to
substance and neutral cues (Hansenne et al., 2003; Jang et al., 2007). Moreover, not
all studies reporting significant effects on the LPP component, have demonstrated
significant effects on the P300 component (Franken et al., 2008; Van de Laar et
al., 2004). These aberrant findings additionally indicate that the results of late
ERP enhancements to substance cues in substance users await a meta-analytic
approach. Secondly, P300 and LPP discrepancies are typically posteriorly
distributed in traditional oddball tasks and ERP studies of emotion (Hajcak
et al., 2010). Although this has also been found in some substance use studies
(Herrmann et al., 2000; Versace et al., 2011), several substance-specific ERP
studies have found frontally or fronto-centrally distributed differences between
substance users and controls in response to substance-related cues relative to
neutral cues (e.g., Littel & Franken, 2010; Sokhadze et al., 2008). Remarkably,
some studies only observed fronto-centrally distributed effects (Fehr et al., 2006;
Fehr et al., 2007; Littel & Franken, 2007; McDonough & Warren, 2001; Van de
19
Chapter 1
Laar et al., 2004; Warren & McDonough, 1999). Clearly, a meta-analytic approach
would elucidate this topographic issue. And finally, results from the ERP studies
in addiction have often been lumped together, despite differences in sample and
study characteristics, such as differences in type of substance dependence and
utilized paradigms. These different characteristics could have moderated the
observed effects. A meta-analytic approach would additionally clarify the role of
potential moderating variables.
Putative moderators: sample and study characteristics
With regard to sample characteristics, ERP studies have been conducted in
substance-users dependent on different types of substances. Some of these studies
were confined to individuals dependent on stimulants, i.e., nicotine and cocaine,
whereas others investigated in substance-related processing of individuals
dependent on depressants, i.e., heroin, alcohol, and cannabis. ERP differences
between substance users and controls might differ between different SUD samples.
In addition to type of substance use, the status of substance use varies between
studies. Whereas some studies examined current substance users, other studies
used participants that were abstinent at the time of testing, varying from at least
one week (e.g., Franken et al., 2008) to at least one month (Van de Laar et al., 2004).
Recency of substance use may relate to aberrant neural activity to substance cues
(Wilson, Sayette, & Fiez, 2004). For example, Littel and Franken (2007) found the
P300 to be significantly reduced in ex-smokers that were abstinent for at least
six months. In addition, Dunning et al. (2011) observed that both abstinent and
current cocaine users displayed enhanced P300/ LPP amplitudes in response to
cocaine cues, but that this effect was no longer discernible for current users during
a later LPP window. Furthermore, there is considerably variability between studies
with regard to participants’ ages, with averages ranging from 17 (Nickerson et al.,
2011) to 40 years (e.g., Sokhadze et al., 2008), as well as with regard to the relative
proportion of participating males and females, with percentages ranging from 15
(e.g., Littel & Franken, 2007) to 100 percent male (e.g., Franken, Stam et al., 2003).
Although the influence of age on cue-reactivity is unknown, gender appears to
have some influence, with females being more sensitive to substance cues than
males (e.g., Field & Duka, 2004).
With regard to study characteristics, different paradigms have been employed,
namely passive viewing, oddball and color matching paradigms. Differences
between these paradigms might influence the ERP discrepancies between
20
Meta-analysis of biased cognitive processing in SUD
substance users and controls, for example because the active paradigms require
either additional motor responses (button presses to targets) or additionally
call upon working memory (silently counting targets). Research indicates that
the P300 amplitude is attenuated in more complex paradigms (García-Larrea &
Cézanne-Bert, 1998). In addition, button-pressing specifically has been shown to
result in smaller P300 amplitudes and different P300 topographies than silentcounting (Salisbury, Rutherford, Shenton, & McCarley, 2001), which in turn has
been suggested to result in more prefrontal neural activity (Muller et al., 2003).
Rationale of the present meta-analytic investigation
The present meta-analysis was conducted in order to summarize and quantitatively
integrate the increasing amount of knowledge that is gained from empirical studies
addressing the relationship between late ERP amplitudes and substance-related
processing bias in SUD individuals. Two separate meta-analyses were conducted;
one for the P300 (300-800 ms) component of the ERP and one for the LPP (>
800 ms) component of the ERP. Both meta-analyses involved studies assessing
ERP reactivity to substance-related stimuli relative to neutral stimuli in SUD
participants compared with healthy control subjects. Analyses were restricted to
studies using pictures to elicit ERP cue-reactivity. Studies using words or sounds
were excluded, thereby maximizing the homogeneity between ERP responses
measured in included studies. To the authors’ knowledge, this is the first study that
is conducted to aggregate ERP amplitude findings in substance-related processing
bias research.
The primary objective of both analyses was to compute the overall effect size of
late ERP amplitude differences between stimulus conditions and groups. Special
emphasis was placed on the location of the ERP differences and for this purpose
effects on most frequently reported sites, i.e., frontal (Fz) and parietal (Pz) sites,
were compared. Furthermore, additional stratified moderator analyses by the
specific sample and study characteristics, as discussed above, were conducted in
order to assess potential moderating effects of those variables on ERP amplitude
outcomes and to provide an empirical basis for future ERP studies in SUD
individuals.
21
Chapter 1
Method
Literature search strategy
In the present meta-analytic investigation studies were initially selected for
inclusion based on an extensive literature search in various databases (Pubmed
and Scopus from 1980 to August 2011) using the key words: (P3 or LPP or SPW or
event-related potentials or ERP), cross-referenced with the key words (addiction or
drug use or drug abuse or drug dependence or substance use or substance abuse or
substance dependence or alcohol or ethanol or heroin or cocaine or opiate or cannabis
or marijuana or nicotine or smokers or tobacco) and (cue or stimuli or reactivity or
cue-reactivity or cue-exposure or conditioned responses or craving or attentional
bias or processing bias). Furthermore, reference lists of selected articles were
consulted to ensure that other relevant work was not overlooked.
Titles and abstracts of all findings were inspected and studies were selected for
further investigation if they examined P300 and/or LPP amplitudes as dependent
measure in cue-reactivity research. Selected papers were consulted and were
retained for the present meta-analytic investigation if they met the following
pre-defined inclusion criteria: (1) the procedure of the study had to involve a cuereactivity paradigm in which pictorial stimuli (cues) were used to elicit a P300 or
LPP response; (2) pictorial stimuli presented in the cue-reactivity paradigm had
to contain both substance-related cues and neutral-control cues; (3) the study had
to employ a typical ERP method (i.e., mean amplitudes or a maximum baseline-topeak method); (4) the study sample had to be comprised of SUD participants (i.e.,
alcoholics, smokers, and cannabis-, heroin-, and cocaine-dependent patients); and
(5) the study constituted independent subject samples.
In total, 29 studies were selected for inspection. Using the criteria outlined
above, three studies were excluded because they used words or sounds instead
of pictures to elicit ERP cue-reactivity (Fehr et al., 2006; Hansenne et al., 2003;
Heinze, Wölfling, & Grüsser, 2007), three studies were excluded because they did
not include a sample of SUD individuals (i.e., alcohol sensitive individuals, heavy
social drinkers; Bartholow, Henry, & Lust, 2007; Bartholow, Lust, & Tragesser,
2010; Herrmann, Weijers, Wiesbeck, Böning, & Fallgatter, 2001), and one study
was excluded because of the use of a dependent subject sample (Versace, Minnix
et al., 2011). For the remaining studies, it was checked whether specific P300/
LPP responses (means and standard deviations of the P300 amplitude/LPP in µV)
22
Meta-analysis of biased cognitive processing in SUD
as a function of cue (substance-related cue versus neutral cue) were reported. If
not, corresponding authors were contacted and asked to provide raw data or to
calculate the group means and standard deviations of the P300 amplitudes. Five
studies were excluded because of insufficient information permitting effect size
calculations (Genkina & Shostakovich, 1983; Lubman et al., 2008; McDonough
& Warren, 2001; Shostakovich, 1987; Warren & McDonough, 1999). Eventually,
17 studies were selected for the meta-analysis of the P300 cue-reactivity response
that yielded 30 independent samples (i.e., 13 control samples and 17 substancedependent samples). Furthermore, 6 studies that yielded 10 independent subject
samples (i.e., 4 control samples and 6 substance-dependent samples) were
selected for meta-analysis assessing the LPP cue-reactivity response. A brief
summary of each of the included studies, along with the sample characteristics
and methodological features, is presented in Table 1.
Moderator variables for stratified analysis
Additional stratified analyses were carried out in order to investigate the
influence of potential moderating variables on the cue-elicited P300 and LPP
amplitudes. This enabled us to delineate possible explanations for the observed
heterogeneity in results. The moderator variables were based on their theoretical
and methodological significance to cue-reactivity research, and can be classified
according to sample- and study characteristics. Table 2 summarizes the moderator
variables and their subcategory characterizations.
Most subgroup definitions were based on unequivocal classifications (e.g., Group,
and the study characteristics “Electrode Site” and “Task Requirements”). The
moderator variable “type of substance dependence” was defined according to
the action of the substance used (stimulants versus depressants). The variable
“substance use status of the SUD patients” was classified by current substance
users versus patients that were abstinent for at least one week prior to testing.
Classifications were based on a review of the description of sample selection of the
Method section of each study. Participants’ age and gender were also identified for
each sample and included as continuous moderator variables.
23
24
300-400 ms
400-700 ms
300-400 ms
500-750 ms
Franken, Stam,
Hendriks, & van
den Brink (2003)
Littel& Franken
(2007)
Littel & Franken
(2010)
Jang, Lee, Yang,
& Lee (2007)
350-600 ms
250-350 ms
300-400 ms
400-700 ms
250-750 ms
Franken et al.
(2008)
Franken et al.
(2004)
LPP
timeframes
27
21
-
-
10
-
700-1000 ms
19
1000-6000 ms
23
15
700-1000 ms
21
1000-2000 ms
750-1500 ms
-
33.5
14
27
21
23.3
21.6
30.6
29.3
-
10
37.2
26.7
45.6
16
19
29
33.3
14.3
100
100
76.2
100
46.7
96.4
Nicotine
Nicotine
Nicotine
Heroin
Cocaine
Cocaine
Nicotine
Cocaine
Abstinent
Abstinent
Current
Current
Stimulant
Stimulant
Stimulant
Current
Current
Current
Passive
picture
viewing
Color
matching
task
Passive
picture
viewing
Passive
picture
viewing
Passive
picture
viewing
Passive
picture
viewing
Passive
picture
viewing
Oddball
task
Substance
Task
use status
Depressant Abstinent
Stimulant
Stimulant
Stimulant
Stimulant
Mean
n substance n
% males
Substance
age
Substance
users (SUD) controls
SUD
Action
SUD
400-1000 ms 1000-2000 ms 28
Fehr,
400-500 ms
Wiedenmann, &
500-600 ms
Herrmann (2007)
Dunning et al.
(2011)
Author(s) and
P300
year of publication timeframes
Active
Passive
Passive
Passive
Passive
Passive
Active
Passive
Task
require
ments
Table 1. Summary of the studies included in the meta-analysis of P300 and LPP amplitude in relation to ERP cue-reactivity: sample and study
characteristics
Chapter 1
Wölfling, Flor, &
Grüsser (2008)
Versace et al.
(2011)
Sokhadze et al.
(2008)
Van de Laar,
Licht, Franken, &
Hendriks (2004)
Nickerson et al.
(2011)
Lubman, Allen,
Peters, & Deakin
(2007)
Namkoong et al.
(2004)
LPP
timeframes
450-750 ms
400-700 ms
-
-
-
15
180
25
700-1000 ms
26
1000-2000 ms
300-590 ms
300-400 ms
400-700 ms
13
20
13
-
-
-
30
300-600 ms
250-500 ms
280-600 ms
-
15
-
20
9
-
17
12
31
-
29
45.1
35.4
41.3
16.9
37.5
28.5
21.9
21.7
46.7
65
100
64
76.9
87.5
85.7
16.7
39.3
Cannabis
Nicotine
Cocaine
Cocaine
Cannabis
Alcohol
Heroin
Nicotine
Nicotine
Current
Abstinent
Current
Depressant Current
Stimulant
Stimulant
Stimulant
Depressant Abstinent
Depressant Abstinent
Current
Current
Oddball
task
Passive
picture
viewing
Oddball
task
Passive
picture
viewing
Passive
picture
viewing
Passive
picture
viewing
Oddball
task
Passive
picture
viewing
Passive
picture
viewing
Substance
Task
use status
Depressant Abstinent
Stimulant
Stimulant
Mean
n substance n
% males
Substance
age
Substance
users (SUD) controls
SUD
Action
SUD
600-1000 ms 1000-2000 ms 28
Littel & Franken
300-800 ms
(Littel & Franken)
Littel & Franken
(2011)
Author(s) and
P300
year of publication timeframes
Passive
Passive
Passive
Active
Passive
Active
Active
Passive
Passive
Task
require
ments
Table 1. Summary of the studies included in the meta-analysis of P300 and LPP amplitude in relation to ERP cue-reactivity: sample and study
characteristics
Meta-analysis of biased cognitive processing in SUD
25
Chapter 1
Table 2. Moderator variables with their descriptions and definitions
Moderator variables
Categories
Sample characteristics
Group
Controls
Substance Action
Stimulant
Substance use status
Age
SUD participants
Depressant
Current
Abstinent
Continuous variable
Gender
Continuous variable
Electrode Site
Pz
Task requirements
Passive
Study characteristics
Fz
Active
Definition
Healthy control participants
All addict samples (i.e., alcoholics,
smokers, cannabis users,
cocaine-dependent persons,
heroin-dependent persons)
All substances that produce
enhanced activity of the central
and peripheral nervous systems
(i.e., cocaine, nicotine)
All substances that produce central
nervous system depression (i.e.,
alcohol, cannabis, heroin)
Current substance users
Abstinent participants (10-30 days)
% males included in the study
All studies that measured the
P300 or LPP at electrode site Pz
All studies that measured the
P300 or LPP at electrode site Fz
Participants were passively exposed
to substance- and neutral cues
Participants were actively
exposed to substance- and
neutral cues (e.g., were required
to count/to press a button)
Data extraction and statistical method
Meta-analytical techniques were used to evaluate results from the ERP studies
that compared responses of participants with substance use disorders (alcoholics,
cigarette smokers, cannabis users and cocaine or heroin addicts) and controls
to substance-related versus neutral stimuli. For each study, the effect size was
calculated as the standardized difference between means for a within-subjects
comparison (i.e., the difference between responding to the substance-relevant cue
and responding to the substance-neutral control cue). Therefore the effect size
index (ES) for this meta-analysis is the standardized mean difference (SMD) for
the comparison of two dependent means. Effect sizes that showed an increase in
26
Meta-analysis of biased cognitive processing in SUD
P300 or LPP were assigned a positive value, while those that showed a decrease
were assigned a negative value.
For the primary meta-analysis, each independent study sample was represented
by one effect size. In other words, SUD and control samples were treated as
independent. However, some studies included multiple conditions or assessments
to elicit the P300/LPP from the same subjects (e.g., several time windows for
analyzing the P300/LPP amplitudes, multiple electrodes). If a study had two or
more conditions completed by the same participants, effects were averaged across
conditions so that each study only provided one effect size, except for stratified
moderator analysis: each condition (e.g., electrode site) was then considered as an
independent study with separate effect sizes calculated (Arends, Voko, & Stijnen,
2003; Van Houwelingen, Arends, & Stijnen, 2002).
Heterogeneity between and within studies was anticipated as a result of variation
in sample- and study characteristics. Therefore, the meta-analysis was performed
using random-effects models (DerSimonian & Laird, 1986). Random-effects
models are more conservative as compared to fixed effects models but give similar
results to a fixed-effect model when heterogeneity is low. To assess between-study
heterogeneity, we calculated (1) the Q statistic and (2) the I2 statistic. The I2 statistic
quantifies the proportion of between-study heterogeneity (conventionally, 0 % = no,
25% = low, 50% = moderate, and 75% = high).
The influence of the potential moderating variables in the additional stratified
moderator analyses on the P300/LPP effect size estimates was explored via a
priori defined subgroup-analyses. Subgroup analyses were conducted according
to the mixed effects model, in which studies within subgroups are pooled with the
random effects model, while tests for significant differences between subgroups
are performed with the fixed effects model. For comparison of moderator
variables, the Q-test on heterogeneity between groups was used, which provided
a Qb (Q-between) statistic. For continuous moderator variables, we used metaregression analysis to test whether there was a significant association between
the continuous variable and the effect size, as indicated with a Z-value and an
associated p-value.
Small study bias, such as publication bias against null results, was assessed using
graphical (i.e., funnel plot) and statistical (i.e., Egger’s test) methods. In addition,
27
Chapter 1
because these methods are not sensitive when there is little variation in sample size
across studies, we assessed the evidence for an excess of statistical significance
in included studies (Ioannidis, 2011). Briefly, this calculates the number of
statistically significant findings that would be expected given the achieved power
of each individual study, assuming the pooled effect size from the meta-analysis
is a reasonable estimate of the true underlying population effect. This is then
compared against the observed number of statistically significant findings using
a binomial test.
Potential moderator variables, sensitivity and small study bias analyses were
analyzed separately for both the P300 and LPP. All analyses were performed with
Comprehensive Meta Analysis 2.0 for Windows (Biostat, Englewood, USA) and the
overall effects are presented as the SMD with 95% confidence intervals (CI).
Results
P300
Overall effect size
Summary data from both meta-analyses are shown in Table 3. Individual effect
sizes (SMD), 95% confidence interval (CI), and p-value for each independent
subject sample included in the primary meta-analysis are presented in Figure 1.
Analysis of overall effects across all participants (controls and SUD participants
together across electrodes; k = 30), indicated that participants showed increased
P300 amplitudes when exposed to substance-related versus neutral stimuli
(SMD = 0.45, 95% CI = 0.30, 0.59, p < 0.001). High heterogeneity of effect size was
evident (I2 = 79.7%, Q = 142.99, p < 0.001).
In order to investigate whether the P300 effect was larger in substance-dependent
samples than in controls, a subgroup analysis was performed with Group (controls
versus SUD participants) as moderator variable. Stratification by Group indicated
that the P300 effect size was significantly larger in SUD participants (k = 17;
SMD = 0.61; 95% CI = 0.42, 0.79; p < 0.001), as compared to controls (k = 13; SMD = 0.23;
95% CI = 0.01, 0.45; p = 0.04), suggesting that SUD participants demonstrated
increased P300 amplitudes when exposed to substance-related versus neutral
stimuli. The difference between groups was statistically significant (p = 0.01).
28
Meta-analysis of biased cognitive processing in SUD
Healthy control samples were therefore omitted from subsequent analyses. As
expected, there was evidence of significant between-study heterogeneity for
P300 effect sizes in the SUD groups (across all substances; Qb = 81.17, p < .001,
I2 = 80.3%). Furthermore, the funnel plot (see Figure 2) and Egger’s test (t(15) = 2.26,
p = 0.04) demonstrated evidence of potential publication bias. However, the number
of statistically significant findings observed among these studies was consistent
with what would be expected given the average power of these studies (O: 13 vs.
E: 12.6). When we corrected the pooled effect size estimate using Duval and
Tweedie’s (2000) trim-and-fill method, based on the imputation of two hypothetical
studies, the overall effect size was slightly reduced (SMD = 0.52; 95% CI = 0.32,
0.71), but remained statistically significant (p < 0.001).
Stratified moderator analyses
Sample characteristics. A summary of the categorical moderator analysis results
for the sample- and study characteristics from all of the P300 cue-reactivity
studies is presented in Table 4. A series of analyses stratifying the main metaanalysis by different sample characteristics (type of substance used, substance
use status of the participant) did not reveal significant moderating effects on P300
amplitude. Additional continuous moderator analyses did not indicate significant
associations between P300 effect size and age or gender.
Study characteristics. Stratification by Electrode site (Pz versus Fz) and Task
requirements (passively versus actively exposed) revealed that neither Electrode
site nor task requirements had significant moderation effects on P300 amplitude.
29
Chapter 1
6XEJURXSZLWKLQVWXG\
6WXG\QDPH
6WDWLVWLFVIRUHDFKVWXG\
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
&RQWURO
'XQQLQJHWDO
)HKUHWDO
)UDQNHQHWDO
)UDQNHQHWDO
-DQJHWDO
/LWWHO)UDQNHQ
/LWWHO)UDQNHQ
/LWWHO)UDQNHQD
/XEPDQHWDO
1DPNRRQJHWDO
6RNKDG]HHWDO
9DQGH/DDUHWDO
:ROIOLQJHWDO
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
3DWLHQW
'XQQLQJHWDO
)HKUHWDO
)UDQNHQHWDO
)UDQNHQHWDO
)UDQNHQHWDO
-DQJHWDO
/LWWHO)UDQNHQ
/LWWHO)UDQNHQ
/LWWHO)UDQNHQD
/LWWHO)UDQNHQE
/XEPDQHWDO
1DPNRRQJHWDO
1LFNHUVRQHWDO
6RNKDG]HHWDO
9DQGH/DDUHWDO
9HUVDFHHWDO
:ROIOLQJHWDO
6WGGLIILQPHDQVDQG&,
6WGGLII
LQPHDQV
/RZHU
OLPLW
8SSHU
OLPLW
S9DOXH
Figure 1. Forest plots of studies in the meta-analysis examining the P300 effect (k = 17).
The forest plot shows the standardized effect sizes (SES) of each primary study on the
P300 amplitude in the meta-analysis. The horizontal lines represent 95% CI for the SES
in each individual study. The size of the dots represents the weights given to studies. The
black diamond shows the pooled standardized effect sizes (PSES) of all studies within the
subgroups (controls, k = 13 vs. all SUD participants, k = 17) using meta-analysis tools. The
white diamond shows the PSES of all studies together.
)XQQHO3ORWRI6WDQGDUG(UURUE\6WGGLIILQPHDQV
6WDQGDUG(UURU
6WGGLIILQPHDQV
Figure 2. Funnel plot of estimated effect sizes (Hedges’ g) against standard error of the
studies assessing the P300 effect.
30
Meta-analysis of biased cognitive processing in SUD
Table 3. Summary of meta-analytic results
Analysis
k
g (SE)
LPP amplitude effect (overall)
Controls
SUD participants
30
13
17
10
4
6
0.45 (0.07)
0.23 (0.11)
0.61 (0.09)
P300 amplitude effect (overall)
Controls
SUD participants
0.37 (0.07)
0.21 (0.11)
0.46 (0.07)
p
95% CI
0.30 – 0.59
0.01 - 0.45
0.42 – 0.79
Qb (1) = 5.94, p = .015
0.24 – 0.50
-0.01 – 0.42
0.33 – 0.60
Qb (1) = 3.99, p = .046
<.001
.04
<.001
<.001
.056
<.001
Note. k = number of samples; g = mean effect size; SE = standard error; CI = confidence
interval
Table 4. Summary of P300 meta-analytic results for each categorical moderator variable
(SUD participants only)
Subgroup analysis
Sample characteristics
Substance Action
Stimulant
Depressant
Substance use status
Current
Abstinent
Study characteristics
Electrode Site
Pz
Fz
Task Requirements
Passive
Active
k
g
95% CI
p
11
6
0.50
0.81
<.001
<.001
10
7
0.58
0.64
0.28 – 0.72
0.53 – 1.09
Qb (1) = 2.81, p = .093
0.31 – 0.85
0.44 – 0.84
Qb (1) = 0.12, p = .728
<.001
<.001
16
17
0.51
0.68
<.001
<.001
11
6
0.57
0.69
0.32 – 0.70
0.47 – 0.88
Qb (1) = 1.34, p = .248
0.36 – 0.78
0.28 – 1.10
Qb (1) = 0.27, p = .606
Note. k = number of samples; g = mean effect size; CI = confidence interval.
<.001
.001
LPP
Overall effect size
Individual effect sizes (SMD), 95% confidence interval (CI), and p-value for each
independent subject sample included in the meta-analysis are presented in Figure 3.
Analysis of overall effects across all participants (controls and SUD participants
31
Chapter 1
together across electrodes; k = 10) indicated that participants showed a significant
increased LPP when exposed to substance-related versus neutral stimuli. This was
evidenced by a positive effect size (SMD = 0.37, 95% CI = 0.24, 0.50, p < 0.001).
There was no significant heterogeneity of overall effect size across groups
(Qb = 12.95, p = 0.17; I2 = 30.5%).
By examining whether the LPP amplitude effect was larger for substance users
than healthy controls, a subgroup analysis was performed with group (controls
versus SUD participants) as moderator variable. Stratification by Group indicated
that the LPP effect size was larger in SUD participants (k = 6; SMD = 0.46;
95% CI = 0.33, 0.60; p < 0.001) as compared to controls (k = 4; SMD = 0.21;
95% CI = -0.01, 0.42; p = 0.06), suggesting that SUD participants demonstrated
increased LPP amplitudes when exposed to substance-related versus neutral
stimuli. The difference between groups was statistically significant (p = 0.05).
There were too few LPP studies to conduct meaningful small study bias/ excess of
significance analyses.
Stratified moderator analyses
Since the number of studies that examined the LPP amplitude effect included in the
present meta-analysis was very small, we only performed a stratified subgroup
analysis with Electrode site as moderator variable. Subgroup analysis revealed
that the LPP effect size across all SUD participants did not differ between Pz (k = 6,
SMD = 0.44, 95% CI = 0.31, 0.56, p < .001) and Fz (k = 6, SMD = 0.49, 95% CI = 0.30,
0.69, p < 0.001), Qb (1) = 0.24, p = 0.622.
6XEJURXSZLWKLQVWXG\
6WXG\QDPH
6WDWLVWLFVIRUHDFKVWXG\
6WGGLII
LQPHDQV
/RZHU
OLPLW
8SSHU
OLPLW
S9DOXH
&RQWURO
'XQQLQJHWDO
&RQWURO
)UDQNHQHWDO
&RQWURO
)UDQNHQHWDO
&RQWURO
9DQGH/DDUHWDO
3DWLHQW
'XQQLQJHWDO
3DWLHQW
)UDQNHQHWDO
3DWLHQW
)UDQNHQHWDO
3DWLHQW
)UDQNHQHWDO
3DWLHQW
/LWWHO)UDQNHQE
3DWLHQW
9DQGH/DDUHWDO
Figure 3. Forest plots of studies in the meta-analysis examining the LPP effect
32
Meta-analysis of biased cognitive processing in SUD
Discussion
Discussion of results
Results from the first meta-analysis indicate that substance users in general
display larger P300 amplitudes in response to substance-related pictorial stimuli
than in response to neutral pictorial stimuli, with an effect size of SMD = 0.61.
When corrected for publication bias, the estimated overall effect size was slightly
reduced, though remained statistically significant. This medium sized P300
discrepancy between stimuli in substance users was significantly larger than
the small discrepancy observed in healthy controls (SMD = 0.61 vs. SMD = 0.22,
respectively). Results from the second meta-analysis indicate that SUD persons
display larger LPP amplitudes in response to substance-related stimuli than
neutral stimuli, with an effect size of SMD = 0.46. This effect size was significantly
larger than that of controls (SMD = 0.46 vs. SMD = 0.21, respectively). Therefore,
results confirm that in both the P300 (300-800 ms) and LPP (> 800 ms) timeframe
of the ERP, substance users display enhanced electrophysiological processing of
substance stimuli as compared to neutral stimuli and control participants.
The present results expand results from a meta-analytic investigation by Carter and
Tiffany (1999) on physiological and psychological drug cue-reactivity. In this study
it was observed that substance users showed increased heart rate, increased sweat
gland activity, decreased skin temperature and increased craving in response to
substance cues relative to neutral cues. However, no control group was examined
in that study, and, hence, present meta-analysis is the first to show enhanced drug
cue-reactivity in substance users as compared to controls. Because P300 and
LPP amplitudes are associated with the recruitment of attentional resources, the
increased late ERP reactivity observed in the present study can be interpreted
in terms of enhanced cognitive processing of substance cues or ‘processing bias’.
Therefore results can also be considered consistent with results from behavioral
studies demonstrating attentional bias for substance-related material.
Stratified moderator analyses (first meta-analysis) revealed that none of the
predefined sample- and study characteristics explained the variability in effect
sizes across studies. The P300 amplitude was not moderated by type of substance
used (stimulants vs. depressants), substance use status (non-abstinent vs. abstinent
for 10-30 days), age, gender and task requirements (active vs. passive paradigms).
These results indicate that enhanced P300 reactivity to substance cues is
33
Chapter 1
independent of type of substance use and, hence, appears characteristic of SUD in
general. Results further indicate that enhanced P300 reactivity to substance cues
occurs irrespective of task demands. It is both present when attention is focused
on the cues and when attention is directed elsewhere. This could indicate that the
P300 effects arise because of an implicit, involuntary capture of attention instead
of or in addition to an explicit, voluntary choice to focus attention on substance
stimuli (Lubman et al., 2007). Furthermore, enhanced electrophysiological
processing appears to be independent of recency of substance use. Note that all
abstinent participants of the included studies received treatment. Therefore the
present findings do not generalize to individuals who have successfully quit using
substances and/or do not receive treatment (anymore).
Because of the topographic divergence across studies addressing
electrophysiological processing of substance cues, a second aim of the present
meta-analysis was to compute whether the effect was most pronounced at
frontal or parietal electrode sites. Stratified moderator analyses revealed that
both P300 and LPP amplitudes were not moderated by electrode site (Fz vs. Pz).
Thus, in contrast to studies of emotion, in which it is consistently observed that
emotional (pleasant and unpleasant) pictures elicit increased P300 amplitudes
over occipito-parietal or centro-parietal electrode sites (Hajcak et al., 2010;
Schupp, Junghöfer, Weike, & Hamm, 2003b; Schupp, Junghöfer, Weike, & Hamm,
2003a; Schupp et al., 2004; Schupp, Junghöfer, Weike, & Hamm, 2004), substanceelicited late ERP amplitudes can also be observed over frontal electrode sites.
Therefore we propose that future ERP drug-reactivity research should examine
frontal activity in addition to (centro)parietal activity. It has been suggested that
neural generators of the P300 invoke cortical regions that play a role in the cycle
of drug addiction. Substance-related stimuli become especially salient and receive
more attention than other cues, whereas at the same time the ability to control
behavior is decreased (Goldstein & Volkow, 2002). Recent meta-analyses on fMRI
studies investigating substance stimulus processing and craving (Chase, Eickhoff,
Laird, & Hogarth, 2011; Kuhn & Gallinat, 2011) indeed indicate that both regions
of attention (parietal cortex, prefrontal cortex) and cognitive control (anterior
cingulate cortex, prefrontal cortex) become activated in response to substance
cues, as well as regions implicated in appetitive processing, drug-expectation or
intention, and drug-seeking (orbito-frontal cortex, amygdala, ventral striatum).
Because amplitudes on specific EEG sites do not allow for making statements
on localization, future studies utilizing simultaneous fMRI/EEG techniques
34
Meta-analysis of biased cognitive processing in SUD
could provide more insight in specific neural generators of drug-cue elicited
electrophysiological responses.
To summarize, present meta-analysis demonstrates enhanced electrophysiological
processing of substance-related stimuli in substance users relative to controls
as reflected by enlarged P300 and LPP amplitudes. This enhanced processing
resembles processing of stimuli motivationally relevant to all individuals (aversive,
appetitive and emotional stimuli) and stimuli relevant to motivations that differ
between individuals (e.g., disorder-specific stimuli). Therefore substance users’
enhanced processing of substance cues can be explained by substance users’
motivated attention, that is, the allocation of attention and memory resources to
stimuli relevant to substance users’ motivational states. Furthermore, although
interpretive caution is warranted, the present study shows that ERP cue-reactivity
is comparable for conditions where attention is focused on the substance cues and
conditions were attention is directed to other tasks and stimuli. Therefore, the
present study provides evidence for the idea that motivated attention might be
both explicit and implicit in nature. In addition, the individual studies indicate
that P300 amplitudes to substance-related stimuli are correlated with rated
valence of substance cues (Herrmann et al., 2000; Littel & Franken, 2007; Warren
& McDonough, 1999) and rated arousal of substance cues (Franken, 2003).
Furthermore, increased P300 amplitudes to substance-related stimuli have been
demonstrated to correlate with self-reported craving in the majority of studies
(Franken, Stam et al., 2003; Littel & Franken, 2007; Lubman et al., 2008; Namkoong
et al., 2004). In a recent meta-analysis by Field et al. (2009) it was revealed that
there was a robust correlation between ERP cue-reactivity and self-reported
craving (r = 0.36), providing additional evidence for enhanced electrophysiological
processing as being indicative of enhanced motivation. Results fit well within
theoretical accounts of addiction, predicting that as a consequence of sensitization
of neurotransmission in the striatum due to repeated substance use, substance
cues acquire incentive motivational properties and are flagged as especially
salient (Franken, 2003; Robinson & Berridge, 1993; Ryan, 2002).
Directions for future research
A very important next step in this area of research is to examine the association
between processing bias and substance consumption or relapse. Littel and
Franken (2007) showed that ex-smokers with a mean quit duration of 1.4 years
display significantly reduced P300 amplitudes in response to substance cues
35
Chapter 1
relative to current smokers. Moreover, there were no P300 amplitude differences
between ex-smokers and non-smokers anymore. Importantly, results could
not be explained by nicotine dependence differences between groups, since exsmokers and smokers had similar levels of (retrospectively assessed) nicotine
dependence. However, it cannot be inferred from this study whether the observed
reduced processing bias is a cause or a consequence of current smoking status.
Does ERP cue-reactivity return to ‘normal’ after quitting or is reduced ERP cuereactivity predictive of successful quitting (and is enhanced ERP cue-reactivity
predictive of smoking)? Two ERP cue-reactivity studies have been conducted
providing indications for the latter (Bartholow et al., 2007; Bartholow et al.,
2010). In these studies P300 amplitudes in response to alcohol cues relative to
neutral cues were investigated in individuals low in alcohol sensitivity (LS) and
individuals high in alcohol sensitivity (HS). Low sensitivity to the acute effects
of alcohol is considered a genetically mediated risk factor for the development of
alcohol use disorders. Results demonstrated that LS individuals displayed more
enhanced P300 amplitudes in response to alcohol cues, but not neutral or other
motivationally relevant stimuli, than HS individuals. Most important, this pattern
remained significant even when differences in recent alcohol use, family history
or alcohol dependence levels were controlled for. Furthermore, Bartholow et al.
(2007) observed that alcohol cue-elicited P300 amplitude predicted drinking
prospectively. These findings indicate that enhanced P300 cue-reactivity may
precede the onset of substance use disorders, may be predictive of future substance
use, and therefore might serve as a marker or endophenotype for substance use and
dependence. It is necessary that future studies further investigate this possibility.
As suggested by Bartholow et al. (2010), one possible approach is to use P300 cuereactivity measured among abstinent substance users as a predictor of relapse.
Concerning another clinical relevant aspect, it has recently been discovered that
the late components of the ERP are sensitive to cognitive regulation strategies.
Several studies have demonstrated that the amplitude of the P300/LPP is reduced
when participants are instructed to decrease emotional responding to negative
and positive pictures using self-generated cognitive reappraisal strategies,
for example by imagining that the depicted situation gets worse or viewing the
pictures from an uninvolved, detached perspective (Hajcak, Moser, & Simons,
2006; Krompinger, Moser, & Simons, 2008; Moser et al., 2006). These P300/LPP
reductions were correlated with self-reported changes in emotional intensity.
A recent study among smokers indicates that it might be possible for substance
36
Meta-analysis of biased cognitive processing in SUD
users to similarly regulate substance-related processing (Littel & Franken,
2011). Results from this study showed that when smokers actively imagined how
pleasant it would be to smoke (pleasant condition), their P300/LPP (600-1000 ms)
in response to smoking cues increased, but when smokers actively focused on an
alternative stimulus (distraction condition) or thought of a rational, uninvolved
interpretation of the situation (rational condition), smoking-related late LPP
(1000-2000 ms) amplitude decreased to the processing level of neutral stimuli.
The distraction strategy even showed a tendency to decrease the LPP beyond the
processing level of passively viewed smoking pictures. Although the results of that
study are inconclusive in some ways, they indicate that another important step in
the field might be to further investigate whether enhanced drug cue-reactivity or
cognitive processing of substance cues can be deliberately controlled or reduced.
Furthermore, there are some issues that need to be resolved. Although there are
sufficient studies to assume that the enhanced electrophysiological processing
of substance cues appears to reflect motivated attention for these stimuli, it is
unclear what the exact association is with behavioral measures of attentional
bias and which aspects of attention are tapped by the P300 in the passive viewing
paradigms. There exist no studies that computed correlations between ERP
measures of processing bias and behavioral measures of attentional bias. Two
studies combined ERP measurements with behavioral measures of substancerelated attentional bias (Fehr et al., 2006; Fehr et al., 2007), utilizing a nicotine
Stroop task and a picture color matching task. Whereas clear ERP discrepancies
were observed between smokers and controls, smokers were not slower than
controls to color-name smoking-related words and pictures. The authors argue
that the absence of behavioral effects could be explained by the simplicity of the
tasks or the performance enhancing effects of nicotine, since participants were
not deprived. On the other hand, it might be that electrophysiological measures are
more sensitive and/or tap different or broader aspects of cognitive functioning.
Indeed, results from the meta-analysis by Field et al. (2009) show that ERP
measures show a stronger correlation with craving than behavioral measures,
suggesting that ERP measures are most sensitive in capturing motivated attention
for substance cues. However, future studies combining tasks need to shed light on
these issues.
Secondly, there have been some inconsistencies with regard to the involvement
of the early attention-related components of the ERP (< 300 ms) in the
37
Chapter 1
electrophysiological processing of substance-related cues. In studies of emotion,
several early components have been observed to differentiate between
motivationally relevant and control stimuli, including the P1, the N1, the P2, and
the Early Posterior Negativity (EPN; Carretie, Hinojosa, Martin-Loeches, Mercado,
& Tapia, 2004; Olofsson, Nordin, Sequeira, & Polich, 2008; Schupp, Junghöfer, Weike,
& Hamm, 2003a). These components tend to be maximal at occipital sites and have
been associated with early visual selection of or attention for stimulus features. In
ERP drug cue-reactivity studies, either no effects were found for early components
(N1; Franken, Stam et al., 2003; Namkoong et al., 2004), or these effects were not
moderated by group (N1; Dunning et al., 2011; Littel & Franken, 2010; Versace,
Minnix et al., 2011). In one study by Littel and Franken (2010), larger P1 amplitudes
were observed for positive and negative stimuli compared to smoking stimuli in
both smokers and non-smokers. Larger N1 amplitudes were observed for negative
and smoking stimuli, but also in both groups. Larger EPN amplitudes have been
observed for both substance-related cues and emotional cues as compared to
neutral cues (Dunning et al., 2011; Versace, Minnix et al., 2011), but again this
effect was present in both SUD and control participants (Dunning et al., 2011)
or was not compared to a control group (Versace, Minnix et al., 2011). Versace
et al. (2011) additionally demonstrated enhanced P1 amplitudes for smoking
stimuli but not for emotional or neutral stimuli but it is unclear whether this effect
is absent in non-smokers. It can be inferred from these studies that substancerelated stimuli might capture attention very early in the perceptual stream, but
that this is probably caused by other factors than substance use status, such as
arousal, luminosity or lower level perceptual features. Early components tend to
be very sensitive to extra-emotional features (e.g., M. M. Bradley, Hamby, Low,
& Lang, 2007; De Cesarei & Codispoti, 2006). Future studies should address the
role of early attentional processes in drug cue-reactivity while carefully matching
stimulus content and visual features.
Study limitations
A number of important issues warrant consideration when interpreting the results
of the present study. First of all, all non-significant moderator analyses must be
treated with caution. Although the first meta-analysis included sufficient studies
to compute overall effect sizes, the number of studies included in the subgroups
was relatively small. This might have resulted in insufficient power for certain
patterns to become significant. In addition, we were not able to compare groups
dependent on specific types of substances. Electrophysiological processing of
38
Meta-analysis of biased cognitive processing in SUD
substance-related stimuli appears to be understudied in cannabis and alcohol
dependence. Secondly, it might be that some relevant studies failed to find
significant P300 amplitude differences and therefore remained unpublished. In an
effort to address this problem, we calculated the Fail-Safe N statistics and used
the Trim-and-Fill method to model the potential impact of this. Results yielded
some evidence for a publication bias. Nevertheless, after we corrected for this bias
the P300 cue-reactivity effect size remained robust, although reduced slightly in
magnitude. This indicates that the observed effect size may be somewhat larger
than the true effect size and, hence, indicates that reported effect sizes might be
overestimated (Ioannidis, 2011). However, it should be noted that the number of
included studies was small, limiting our ability to detect modest publication bias
and an excess of statistically significant findings.
Conclusions
The present meta-analysis demonstrates that enhanced drug cue-evoked
P300 (300-800 ms) and LPP (> 800 ms) amplitudes are characteristic of SUD in
general. This enhanced electrophysiological processing of substance-related
cues as compared to neutral cues can be interpreted in terms of substance users’
motivated attention. This interpretation is underlined by the strong association
between P300 reactivity and subjective craving (Field et al., 2009). Furthermore,
results indicate that the P300 amplitude effect is present under both implicit and
explicit cue presentation. Finally, the P300 amplitude effect has been observed
to have both a parietal and frontal distribution. This indicates that there exist
differences in brain topography between the processing of substance-related
and emotional stimuli, which might be related to cognitive control processes or
craving (Goldstein & Volkow, 2002). Because there are indications that enhanced
enlarged P300 amplitudes to substance cues may precede the onset of SUD and
may be predictive of future substance use (Bartholow et al., 2007; Bartholow et al.,
2010), future studies need to investigate whether P300 cue-reactivity can serve
as a marker or endophenotype for substance use and dependence. One suggestion
is to study P300 cue-reactivity in relation to substance consumption and relapse.
Furthermore, future studies should investigate ways to reduce or control drug
cue-reactivity and enhanced cognitive processing of substance-related cues.
39
Chapter
2
Overall outline and hypotheses
Chapter 2
As can be concluded from the meta-analysis described in Chapter 1, substance
dependence is characterized by enhanced cognitive processing of substancerelated stimuli relative to neutral stimuli as reflected by enlarged P300 and
LPP amplitudes. Enhancement of these late ERP components can be explained
by substance users’ motivated attention, that is, the allocation of attention and
memory resources to stimuli relevant to substance users’ motivational states.
Enhanced cognitive processing of substance-related stimuli has been associated
with both craving and substance use behaviors (consumption, relapse), and can
therefore be regarded as an important concept in addiction. Most studies have
investigated biased cognitive processing by means of behavioral measures such
as the addiction Stroop task and visual probe tasks. However, these measures
rely on reaction time data which can be very noisy because of multiple underlying
cognitive en motor processes. In a recent study it was shown that the majority
of behavioral measures assessing biased cognitive processing in addiction have
poor internal reliability (Ataya et al., 2011). This might explain the relatively
small correlation between these measures and craving (r = .18; Field et al., 2009).
In studies using more direct measures of cognitive processing, including ERP
measurements, the relation between processing bias and craving was shown to
be significantly larger (r = .36; Field et al., 2009). Because of the importance of
cognitive processing biases in addiction, and because of the advantages of ERP
methodology over traditional behavioral measures of processing biases, the main
goal of the present dissertation was to increase insight in electrophysiological
correlates of biased cognitive processing in addiction.
First, the validation of the Dutch translation of the brief Questionnaire on Smoking
Urges (QSU-brief) is described. Despite the frequent use of this questionnaire to
assess subjective cigarette craving in smoking-related processing bias research
(see also Chapters 4-8), it is still unknown whether it has acceptable psychometric
properties similar to its original English version. Therefore, Chapter 3 examines
the factor structure, internal consistency, and validity of the translated version of
the QSU-Brief in a Dutch smokers population (N= 208) utilizing a cross-sectional
design with random subject selection.
Influential addiction models state that substance-related stimuli acquire incentive
motivational and attention-grabbing properties because of an irreversible
sensitization of dopaminergic neurotransmission in the reward system of the brain
caused by repeated substance use. However, the issue whether processing bias is
42
Overall outline and hypotheses
a more or less permanent feature of nicotine addiction remains to be resolved.
Chapter 4 addresses the question whether smoking-associated processing bias is
still present in ex-smokers after prolonged abstinence. For this purpose, smokers,
ex-smokers and never-smokers were compared on their electrophysiological
brain responses (P300 and LPP) to passively viewed smoking-related and neutral
pictures. In addition, self-report measures of nicotine craving and pleasantness
ratings of smoking stimuli were obtained and compared between groups.
Whereas ERPs can provide us with information about discrete cortical activations
in response to specific events, EEG spectrum measures can enhance insight into
broader changes in brain activation (i.e., mental states) during prolonged stimulus
exposure, such as changes in alertness, vigilance, and arousal. In Chapter 5
EEG spectrum changes in smokers during a 30-s exposure to a neutral (pen)
and a smoking-related cue (lit cigarette) were examined. Furthermore, it was
investigated whether EEG frequency domain changes in response to smoking
cues are still present in ex-smokers after prolonged abstinence. Additionally, selfreport measures of cigarette craving and nicotine dependence were obtained and
compared between groups.
Although research has consistently shown that substance-dependent persons
show increased electrophysiological processing of substance-related stimuli, it
remains unclear whether these differences arise because of an implicit attention
capture or an explicit, voluntary choice to elaborate on the content of the stimuli.
Furthermore it is unclear whether substance users’ enhanced ERP response is
uniquely triggered by substance-related cues, i.e., whether there is a selective bias
for substance cues, or whether substance users show this bias to motivational
relevant stimuli in general, including positively or negatively valenced pictures
with certain arousing properties. Chapter 6 aims to enhance insight in the nature
(implicit vs. explicit) and specificity (substance vs. emotion) of substance-related
processing biases by manipulating attention for smoking and other motivationally
relevant cues in smokers and non-smokers using a rapid stream visual oddball task.
In addition to incentive sensitization of the brain through repeated substance
use, enhanced processing of substance cues is thought to arise through classical
conditioning. Classical conditioning theory predicts that with repeated drug use,
drug-related stimuli or contexts (conditioned stimuli, CS) become associated with
drug intake (unconditioned stimulus, UCS). Consequently, these stimuli acquire
43
Chapter 2
motivational significance and evoke conditioned drug responses or cue reactivity,
including enhanced electrophysiological processing of substance-related stimuli
(processing bias) as discussed in the present dissertation. Once the learning
process has taken place and the CS are able to elicit the conditioned drug responses,
the CS can be paired with new neutral stimuli or contexts, which will also acquire
associative strength and elicit conditioned drug responses. This process is called
second-order conditioning (higher-order conditioning; CS-CS learning) and can
lead to unlimited sequences of associations that presumably contribute to drugseeking in real world environments. Chapter 7 addresses the electrophysiological
correlates of higher-order learning processes associated with smoking addiction.
Smokers and non-smokers were presented with two neutral geometrical figures
that were repeatedly paired with either smoking-related or neutral pictures.
ERPs were recorded in response to the two conditioned geometrical figures and
compared between groups, conditions and first and last half of the experiment.
After the experiment, valence, arousal and craving ratings of the geometrical
figures were obtained and compared between groups.
Because cognitive processing biases have been associated with craving and relapse,
investigating in ways to regulate the processing of substance cues would be of
major clinical relevance. In Chapter 8 it is examined whether electrophysiological
indices of cognitive processing (early and late LPP amplitudes) can be intentionally
modulated by cognitive reappraisal strategies. The effects of three strategies
(a rational strategy, a distraction strategy, and a pleasant strategy) on cognitive
processing of smoking-related stimuli are assessed by comparing LPP amplitudes
in response to reappraised smoking stimuli to LPP amplitudes evoked by passively
viewed smoking-related and neutral stimuli. Additionally, it is investigated
whether enhanced attentive processing as reflected by enhanced LPP amplitudes
as well as the cognitive modulation of these amplitudes differs between nicotinedependent regular smokers and non-dependent light smokers.
As might be inferred from the meta-analysis in Chapter 1, electrophysiological
processing of substance cues is not unequivocally demonstrated in all substancedependent populations. In alcohol-dependent patients, for example, only one
study observed enhanced P300 amplitudes to alcohol cues as compared to neutral
cues (Namkoong et al., 2004), whereas two studies failed to find an effect on the
P300 (Hansenne et al., 2003; Herrmann et al., 2000). The main goal of Chapter 9
is to replicate the study described in Chapter 6 in a sample of alcohol-dependent
44
Overall outline and hypotheses
patients recruited at several addiction out-patient clinics. More specifically, this
chapter addresses whether alcohol-dependent patients show enhanced cognitive
processing bias relative to age-, education-, and gender matched controls as
reflected by enhanced P300 amplitudes and aims to enhance insight in both the
nature (implicit vs. explicit) and the specificity (substance vs. emotion) of this
alcohol-associated bias.
Not only has it been suggested that in addiction substance-related stimuli become
especially salient and receive more attention than other cues; it is has also been
proposed that at the same time the ability to control behavior is decreased
(Goldstein & Volkow, 2002). According to this view, enhanced motivation and
related cognitive processing biases as well as decreased cognitive control
contribute to substance use behaviors. Research has indeed confirmed that
substance-dependent persons show deficits on different indices of cognitive
control, including response inhibition and error-processing. These cognitive
control deficits might contribute to difficulties to resist the consumption of a
substance and the continuation of the behavior despite negative consequences.
We recently demonstrated reduced cognitive control in smokers using a Go/NoGo
task combined with ERP measurements (Luijten, Littel, & Franken, 2011). The
NoGo-N2 component of the ERP elicited by correctly inhibiting motor responses
was observed to be reduced in smokers relative to non-smokers.
Recently, scientific interest has grown for a putative new psychological disorder:
excessive computer gaming. Research on this topic is still in its infancy and
underlying neurobiological mechanisms have not yet been identified. Identifying
underlying mechanisms of excessive gaming might be useful for the identification
of those at-risk, a better understanding of the behavior, and the development of
interventions. It has been demonstrated in a recent study that excessive gamers
but not casual gamers show enhanced processing of game-related stimuli relative
to neutral stimuli as reflected by enhanced P3 amplitudes (Thalemann, Wölfling,
& Grüsser, 2007), which indicates that there might be similarities between
processing bias in excessive gaming and substance dependence. In chapter 10
mechanisms of cognitive control in excessive gaming are examined. Both errorprocessing and response inhibition were studied in excessive gamers and controls
using a Go/NoGo paradigm combined with ERP recordings. ERPs were compared
between errors and correct responses as well as between Go stimuli and correctly
inhibited NoGo stimuli in both groups.
45
Chapter
3
Psychometric properties of the brief
Questionnaire on Smoking Urges
(QSU-Brief) in a Dutch smoker population
Littel, M., Franken, I.H.A., & Muris, P. (2011). Psychometric properties
of the brief Questionnaire on Smoking Urges (QSU-Brief) in a Dutch
smoker population. Netherlands Journal of Psychology, 66, 44-49.
Chapter 3
Abstract
We investigated the reliability, validity, and factor structure of the 10-item
Questionnaire on Smoking Urges (QSU-Brief) in a Dutch smokers sample
(N = 208). The questionnaire displayed good internal consistency (Cronbach’s
alphas > 0.83), and scores were strongly correlated with three other rating
scales for measuring craving, urge, and desire for cigarettes, and moderately
linked to questionnaires that tap related constructs, such as cigarette
dependence. As in previous research, a two factor structure was revealed.
The first factor was best described by ‘the relief from nicotine withdrawal
or negative affect with an urgent and overwhelming desire to smoke’, and
appeared to be associated with negative affect, but not with positive affect.
The second factor reflected ‘the desire and intention to smoke’, and was neither
associated with positive nor negative affect. The factor structure, however,
slightly deviates from the original, English version of the QSU-Brief, which
might be explained by language differences. Overall, the Dutch translation
of the QSU-Brief offers a reliable, valid, and multidimensional assessment of
cigarette craving and appears suitable for use in a general population of young,
Dutch adults.
48
Psychometric properties QSU-brief
Introduction
Craving, which can be defined as “the desire to experience the effect(s) of
a previously experienced psychoactive substance” (UNDCP/WHO, 1992), is
considered an important concept in smoking addiction. Craving is often viewed
as the most difficult and aggravating withdrawal symptom during abstinence and
quitting (Orleans, Rimer, Cristinzio, Keintz, & Fleisher, 1991; Shiffman & Jarvik,
1976; West, Hajek, & Belcher, 1989). Moreover, several studies have shown that
craving hampers successful smoking cessation and that it correlates with relapse
after periods of abstinence (Allen, Bade, Hatsukami, & Center, 2008; Doherty,
Kinnunen, Militello, & Garvey, 1995; Ferguson, Shiffman, & Gwaltney, 2006; Killen,
Fortmann, Newman, & Varady, 1991; Killen & Fortmann, 1997; Niaura et al.,
1988; Orleans et al., 1991; Shiffman & Jarvik, 1976; Shiffman et al., 1997; Swan,
Ward, & Jack, 1996). Furthermore, the effects of positive outcome expectations
of smoking on relapse appear to be completely mediated by craving (Dijkstra &
Borland, 2003). Consequently, reliable assessment of craving is necessary in order
to predict relapse, improve cessation treatment, and understand the nature of
craving in general.
In the past decades, the majority of studies used single- or two-item questionnaires
to measure craving (see for an overview L. S. Cox, Tiffany, & Christen, 2001). With
such a restricted number of questions, the assessment of craving is rather onesided and the psychometric properties cannot be determined. In order to reliably
measure the multi-dimensional aspects of craving, Tiffany and Drobes (1991)
designed the 32-item Questionnaire on Smoking Urges (QSU). This self-report
instrument intends to capture several different aspects of craving, ranging from
positive expectations about the effects of smoking to more general, overwhelming
urges to smoke. Factor analysis indicated that the QSU consists of two clearly
distinguishable underlying factors, which can be described as ‘the desire and
intention to smoke with an anticipation of pleasure from smoking’ and ‘the relief
from nicotine withdrawal or negative affect with an urgent and overwhelming
desire to smoke’ (Tiffany & Drobes, 1991).
Nonetheless, because of its length, the QSU turned out to be less suitable in clinical
and laboratory settings where a fast assessment of the concept of craving is
important. With this in mind, Cox et al. (2001) developed the QSU-Brief, which is an
abbreviated version of the QSU consisting of only ten items that can be completed
49
Chapter 3
in about two minutes. The shortened scale has good reliability (α = 0.78 - 0.97) and
a two-factor structure that is generally well in keeping with that obtained for the
original QSU. However, two items (items 2 and 5) appeared to be ambivalent by
loading on both factors and consequently were, in spite of their inclusion in factor 2
of the original QSU, not assigned to any subscale of the QSU-Brief. In the QSU-Brief,
factor 1 includes the items 1, 3, 6, 7, and 10, while factor 2 includes the items 4, 8,
and 9 (L. S. Cox et al., 2001). These favorable psychometric properties have been
confirmed in further research (Cappelleri et al., 2007; Cepeda-Benito & Reig-Ferrer,
2004) and since then the 10-item QSU-Brief has been used in a wide variety of studies
(e.g., Attwood, O’Sullivan, Leonards, Mackintosh, & Munafò, 2008; B. P. Bradley,
Field, Healy, & Mogg, 2008; LaRowe, Saladin, Carpenter, & Upadhyaya, 2007).
So far, no cigarette craving questionnaire has been validated for the Dutch
population, and consequently studies have primarily relied on a non-validated
translation of the QSU-Brief (Littel & Franken, 2007; Littel, Franken, & Van Strien,
2009). Although this research has shown that the Dutch QSU-Brief appears to be
sensitive to tasks that are believed to enhance cigarette craving, such as viewing
smoking-related pictures, it is still unknown whether the Dutch translation of the
QSU-Brief has acceptable psychometric properties similar to its original English
version. Therefore, the main goal of the present study was to examine the factor
structure, internal consistency, and validity of a translated version of the QSUBrief in a Dutch smokers population utilizing a cross-sectional design with random
subject selection.
Since the two original QSU-Brief factors make reference to positive and negative
affect, an additional goal of the current study was to examine the correlations
between the QSU-Brief, and especially its subscales, and constructs reflecting
positive and negative mood. Research has shown that lower levels of positive
affect and higher levels of negative affect predict nicotine dependence (D. E.
McChargue, Cohen, & Cook, 2004b), that college smokers with elevated symptoms
of depression are more dependent on cigarettes than non-depressed peers (D. E.
McChargue, Cohen, & Cook, 2004a), and that relapse is preceded by increasing or
intense negative affect (Shiffman, 2005). In line with these results, negative mood
appears to be positively associated with craving (L. S. Cox et al., 2001). Positive
mood, however, is negatively related to craving in abstinent smokers who are
enrolled in a cessation program, but positively linked to craving in active smokers
in a laboratory setting (L. S. Cox et al., 2001). In the present study, it is predicted
50
Psychometric properties QSU-brief
that QSU-Brief factor 1 ‘the desire and intention to smoke with an anticipation
of pleasure from smoking’ is especially related to positive affect, whereas factor
2 ‘the relief from nicotine withdrawal or negative affect with an urgent and
overwhelming desire to smoke’ is mainly associated with high negative affect and
anhedonia. Obviously, such findings would provide further support for the validity
of the subscales of the QSU-Brief.
Method
Participants
The sample consisted of 208 smokers (58.7% female) with a mean age of 24.4 years
(SD = 7.9). Smokers’ mean score on the Fagerström Test for Nicotine Dependence
(FTND; Vink, Willemsen, Beem, & Boomsma, 2005) was 3.7 (SD = 2.3). 32.2 % of
the participants smoked on average between one and ten cigarettes per day, 51.4 %
smoked between 11 and 20 cigarettes per day, 13.9 % smoked between 21 and 30
cigarettes per day, and 2.4% smoked more than 31 cigarettes per day.
A subset (N = 184)1 of the participants was asked about smoking duration and quit
attempts. On average, this group had smoked for 8.0 years (SD = 7.9). 63.9% of
them made one or more quit attempts, with a mean total duration of 11.6 months
(SD = 16.1). Participants were recruited by advertisements on internet forums
and -communities and flyers distributed at the Erasmus University Rotterdam
(the Netherlands). They were not allowed to smoke during the completion of the
questionnaire.
Measures
Participants filled out an online questionnaire containing several questionnaires
and some additional demographics and rating scales. First of all, participants filled
out the Dutch translation of the QSU-Brief (L. S. Cox et al., 2001), which is a selfreport questionnaire measuring urges and cravings to smoke. As mentioned in the
introduction, the QSU-Brief consists of 10 items. Five items (items 1, 3, 6, 7, and
10) represent ‘the desire and intention to smoke with an anticipation of pleasure
from smoking’ and three (items 4, 8, and 9) reflect ‘the relief from nicotine
withdrawal or negative affect with an urgent and overwhelming desire to smoke’.
1 Since we combined data from several datasets that slightly differed in content, the number of participants is
different for some of the variables
51
Chapter 3
All items, including items 2 and 5, contribute to the total craving score. As in the
original version of the QSU (Tiffany & Drobes, 1991), a Likert-type scale ranging
from 1 (strongly disagree) to 7 (strongly agree) was used for the responses to
each question. The Dutch translation of the QSU-brief is based upon translations
by several persons, including native English speakers, that were discussed until
eventually consensus was reached.
In addition, all participants filled out the Dutch version of the Fagerström Test for
Nicotine Dependence (FTND; Vink, Willemsen, Beem, & Boomsma, 2005), which is
a 6-item measure assessing smoking habit and dependence. This questionnaire has
acceptable reliability (Cronbach’s alpha’s ranging from 0.66 - 0.71) and correlates
significantly with number of cigarettes smoked per day (correlations ranging from
0.50 - 0.60, all p’s > 0.05). The items of the FTND were scored according to the
scoring system described in Heatherton et al. (1991).
Further, a subset of the participants (N = 84) completed two mood questionnaires,
i.e., the Positive Affect Negative Affect Scales (PANAS; Watson, Clark, & Tellegen,
1988) and the Snaith-Hamilton Pleasure Scale (SHAPS; Snaith et al., 1995). The
PANAS consists of 20 items that either measure positive affect (PA; 10 items) or
negative affect (NA; 10 items). Each item refers to a mood state (e.g., proud, scared),
and participants rate the extent to which each mood state describes how they
feel at the moment of testing on a scale ranging from 1 (not at all or very slightly)
to 5 (extremely). High PA is thought to reflect high energy, concentration, and
pleasurable mood states, whereas low PA is characterized by sadness and lethargy.
Negative affect (NA), on the other hand, refers to distress and unpleasurable mood
states, with low NA reflecting a state of calmness and serenity (Watson et al.,
1988). The Dutch version of the PANAS, which was used in the present study, has
comparable satisfactory psychometric properties as the original questionnaire.
Cronbach’s alpha’s of the questionnaire ranges from 0.86 – 0.89, and moreover,
the questionnaire reliably differentiates between depressed patients and patients
with an anxiety disorder (Boon & Peeters, 1999).
The SHAPS is a 14-item self-report instrument measuring hedonic tone, i.e., the (in)
ability to experience pleasure (Snaith et al., 1995). The Dutch version, employed
in the current study, has excellent psychometric properties. The questionnaire
discriminates between clinical patients and non-clinical individuals. Cronbach’s
alpha’s are between 0.91 - 0.95 (Franken, Rassin, & Muris, 2007).
52
Psychometric properties QSU-brief
Finally, a subset of the participants reported smoking duration (i.e., for how
many years they had been smoking; N = 181), the degree of cigarette craving on
a scale ranging from 0 to 100 (N = 84), and urge and desire for a cigarette on a
Visual Analogue Scale (urge-VAS and desire-VAS; N = 84). These scales appear
to be reliable and responsive to cue exposure (e.g., Niaura et al., 1998; Sayette &
Hufford, 1994), they can be administered quickly and repeatedly, and they are
used regularly in smoking research (e.g., Mogg & Bradley, 2002; Shadel, Niaura, &
Abrams, 2001; Thewissen, van den Hout, Havermans, & Jansen, 2005; Thewissen,
Snijders, Havermans, van den Hout, & Jansen, 2006).
Procedure
To determine the factor structure of the Dutch translation of the QSU-Brief, an
exploratory principal components analysis with Promax rotation was conducted on
the 10 items. Reliability analysis was conducted to determine internal consistency
of the QSU-Brief and its subscales. In order to assess the validity of the QSUBrief, Spearman correlations were calculated between the QSU-Brief and other
questionnaires/rating scales. We selected Spearman correlation because data of the
QSU-Brief and the other craving rating scales displayed a non-normal distribution.
Results
Exploratory factor analysis
A principal components analysis was conducted on the 10-item QSU-Brief.
A Promax rotation was employed as the two subscales ‘desire and intention to
smoke’ and ‘anticipation of relieve from negative affect with an urgent desire to
smoke’ are considered to be non-orthogonal factors (Cappelleri et al., 2007; L. S.
Cox et al., 2001). Investigation of the scree plot pointed in the direction of a twofactor solution. These two factors had eigenvalues greater than one, i.e., 4.10 and
2.43, and respectively accounted for 41.02 % and 24.27 % of the variance. Eight
of the items had a loading of > 0.40 on one of the two factors. The other two items
(items 1 and 6) displayed substantial loadings (> 0.40) on both factors.
In general, the factor analysis revealed that most of the items loaded most
substantially on their hypothesized factor (see Table 1), that is, factor 1 consisted
of items 2, 4, 5, 8, and 9 and corresponds with factor 2 of the factor structure
described by Cox et al. (2001), i.e., ‘the relief from nicotine withdrawal or negative
53
Chapter 3
affect with an urgent and overwhelming desire to smoke’. Factor 2 contained the
items 1, 3, 6, 7, and 10 and was entirely in accordance with factor 1 of the original
QSU-Brief, i.e., ‘the desire and intention to smoke with an anticipation of pleasure
from smoking’. Items 2 and 5 were not included in their factor solution, but because
of high factor loadings and face validity, these items were nevertheless assigned to
factor 1 in the present study. Items 1 and 6 initially loaded on both factors, but
were assigned to factor 2 in order to avoid too much deviation from the original
factor structure.
Reliability
Cronbach’s alpha was 0.83 for the total score of the QSU-Brief, which indicates that
the Dutch translation of this questionnaire has adequate internal consistency. The
internal consistency of the separate factors was also good: factor 1 (items 2, 4, 5, 8,
and 9: α = 0.84) and factor 2 (items 1, 3, 6, 7, and 10: α = 0.84).
Table 1. Factor loadings for items of the Dutch QSU-Brief as obtained with a principal
components analysis
Original item (item number)
Dutch translation
All I want right now is a cigarette (5)
Het enige wat ik nu wil is een sigaret
Nothing would be better than smoking a cigarette right now (2)
Niets zou beter zijn dan nu een sigaret te roken
Smoking would make me less depressed (9)
Als ik nu mocht roken zou ik me minder depressief voelen
I would do almost anything for a cigarette now (8)
Ik zou er bijna alles voor over hebben om nu te mogen roken
I could control things better right now if I could smoke (4)
Ik zou alles beter onder controle hebben als ik nu mocht roken
I have an urge for a cigarette (6)
Ik ervaar een sterke drang om een sigaret te roken
If it were possible, I probably would smoke now (3)
Als het mogelijk was, zou ik waarschijnlijk nu een sigaret opsteken
A cigarette would taste good now (7)
Een sigaret zou me nu wel smaken
I am going to smoke as soon as possible (10)
Zodra dit mogelijk is, ga ik roken
I have a desire for a cigarette right now (1)
Ik verlang op dit moment naar een sigaret
1
0.87
Factor
2
0.85
0.81
0.70
0.65
0.64
0.53
0.87
0.82
0.42
0.80
0.79
Note. QSU-Brief = Brief Questionnaire on Smoking Urges; factor 1 = ‘the relief from nicotine
withdrawal or negative affect with an urgent and overwhelming desire to smoke’; factor 2
= ‘desire and intention to smoke’. Only loadings > 0.40 are shown.
54
Psychometric properties QSU-brief
Validity
Spearman’s correlation coefficients between the QSU-Brief total score and other
questionnaires/rating scales are reported in Table 2. As expected, scores on the
QSU-Brief were highly correlated with scores on the craving rating scale, ρ = 0.80,
p < 0.01, the desire-VAS, ρ = 0.77, p < 0.01, and the urge-VAS, ρ = 0.76, p < 0.01. In
addition, moderate, positive correlations were found between the QSU-Brief and the
FTND, ρ = 0.14, p < 0.05, and number of cigarettes smoked per day, ρ = 0.25, p < 0.01.
The subscale representing anticipation of relieve from negative affect with an
urgent desire to smoke (factor 1) was significantly correlated with negative affect
(PANAS-NA), ρ = 0.25, p < 0.01, whereas this appeared not true for the subscale
representing a desire and intention to smoke (factor 2), ρ = 0.16, ns. Neither of the
factors were significantly correlated with positive affect (PANAS-PA), ρ = - 0.02, ns
and ρ = - 0.01, ns. However, both factors were significantly correlated with the
SHAPS, ρ = 0.23, p < 0.01 (factor 1) and ρ = 0.22, p < 0.01 (factor 2). To recapitulate,
craving, especially the subscale ‘anticipation of relieve from negative affect with
an urgent desire to smoke’, is related to negative affect, but not necessarily to
positive affect.
Table 2. Spearman’s correlations between scores on the QSU-Brief and rating scales/
questionnaires tapping similar and related constructs
QSU-Brief
0-100 Craving rating scale
Desire-VAS
Urge-VAS
FTND
Cigarettes/day
QSU- 0-100 craving
Brief rating scale
0.80**
DesireVAS
0.76**
0.86**
UrgeVAS
0.77**
0.71**
0.73**
FTND
Cigarettes
per day
0.14* 0.25**
0.33** 0.42**
0.19
0.29**
0.26** 0.31**
0.75**
Note. QSU-Brief = Brief Questionnaire on Smoking Urges; VAS = Visual Analogue Scale;
FTND = Fagerström Test of Nicotine Dependence.
* p < 0.05
** p < 0.01
55
Chapter 3
Discussion
The present study investigated the factor structure, reliability, and validity of
the Dutch translation of the QSU-Brief. It can be concluded that the questionnaire
seems to be a reliable and valid measure of cigarette craving. The Dutch QSUBrief displayed good internal consistency, and scores on this scale were strongly
correlated with three other rating scales for measuring craving, urge, and
desire for cigarettes, and moderately linked to questionnaires that tap related
constructs, i.e., cigarette dependence and number of cigarettes smoked per day.
An exploratory factor analysis revealed a two factor structure, which is largely in
agreement with exploratory and confirmatory factor analyses of the English and
Spanish versions of the original QSU (Cepeda-Benito, Henry, Gleaves, & Fernandez,
2004; Davies, Willner, & Morgan, 2000), and the English and Spanish versions of
the 10-item QSU-Brief (Cappelleri et al., 2007; Cepeda-Benito & Reig-Ferrer, 2004).
The first factor, which corresponds with the second factor of the English QSU-Brief,
consisted of the items 2, 4, 5, 8, and 9, whilst items 1, 3, 6, 7, and 10 comprised factor 2.
In the present study, items 2 and 5 loaded convincingly on factor 1, whereas
they loaded ambivalently on two factors in previous studies. This discrepancy
might be explained by language differences. Items 2 and 5, i.e., ‘nothing would be
better than smoking a cigarette right now’ and ‘all I want right now is a cigarette’
convey quite extreme utterances, especially when they are literally translated
into Dutch. Because items as ‘smoking would make me less depressed’ (9) and
‘I could control things better right now if I could smoke’ (4) are also quite extreme
and rarely used in Dutch, it is not surprising that these items load on one and the
same factor. The abovementioned items constitute the subscale ‘the relief from
nicotine withdrawal or negative affect with an urgent and overwhelming desire to
smoke’, and the inclusion of items 2 and 5 makes this designation even more valid.
Furthermore, and in accordance with the findings by Cox et al. (2001), this subscale
was significantly correlated with negative affect and anhedonia, thereby yielding
further evidence for its relation to negative mood and withdrawal symptoms.
The second factor corresponds with the first factor of the original QSU-Brief,
although, in the present study, items 1 and 6 loaded on two factors. Item 6, however,
loaded considerably higher on factor 2 than on factor 1, and was thus assigned to
factor 2. In order to avoid too much deviation from the original factor structure,
item 1 was also assigned to this factor. An explanation for these items loading
on both factors might be again the Dutch language. Items 1 and 6 contain the
56
Psychometric properties QSU-brief
terms ‘desire’ and ‘urge’. Although Dutch people may use phrases such as ‘I have a
strong desire or urge for a cigarette, it is far more common to employ less strong
expressions, e.g., ‘I would like/fancy a cigarette’. Nevertheless, items 1 and 6 are
less extreme than the items assigned to factor one. This justifies the decision to
assign them to factor 2, which can be described as ‘desire and intention to smoke’.
We are careful with the addition of ‘anticipation of pleasure from smoking’ to the
name of this factor, because the subscale is neither significantly correlated with
positive nor negative affect and the individual items make no explicit reference
to any pleasure or reward that one can get from smoking. In previous research
(L. S. Cox et al., 2001), an ambiguous relation was found between positive affect
and craving in that it was negatively related to craving in abstinent, treatmentseeking smokers, but positively related to craving in active smokers in a laboratory
setting. If any relationship between ‘desire and intention to smoke’ and mood
exists, we expect this to be the case with depressive symptoms, since this subscale
correlated significantly with anhedonia.
In the present study only young adults with moderate to low smoking dependence
were questioned. Although their FTND score does not seem to deviate from
the average FTND score, which appears to be relatively low in the Netherlands
(<3; Fagerström & Furberg, 2008), future research needs to confirm the factor
structure and psychometric properties in older and more dependent smokers in
order to be able to generalize the results to the population.
Overall, the Dutch translation of the QSU-Brief offers a reliable, valid and multidimensional assessment of craving for cigarettes in a general population of young
adults and is suitable for being used in laboratory settings. However, it would be
useful if future research confirms the present factor structure in a Dutch sample,
since it slightly deviates from the original, English version.
57
Chapter
4
The effects of prolonged abstinence
on the processing of smoking cues:
an ERP study among smokers, exsmokers and never-smokers
Littel, M., & Franken, I.H.A. (2007). The effects of prolonged
abstinence on the processing of smoking cues: an ERP
study among smokers, ex-smokers and never-smokers.
Journal of Psychopharmacology, 21(8), 873-882.
Chapter 4
Abstract
Processing bias is an important feature of substance abuse. The issue whether
processing bias is a more or less permanent feature of nicotine addiction
remains to be resolved. The present study addresses the role of smoking
status on smoking-related processing bias. We employed Event-Related Brain
Potentials (ERPs) as measure of processing bias to investigate this issue.
Further, self-report measures of nicotine craving and pleasantness ratings
of smoking stimuli were obtained. Three groups, smokers, ex-smokers and
never-smokers, were compared on their electrophysiological brain response
to smoking-related and neutral pictures. The present study shows that both
the P300 and SPW amplitudes in response to smoking-related pictures are
significantly more enhanced for smokers than for ex-smokers and neversmokers at frontal and central sites, whereas the magnitude of the P300 and
SPW amplitudes in response to neutral pictures does not differ between the
three groups. Accordingly, it can be concluded that smokers show more bias
for smoking-related pictures than ex-smokers and smokers. Because there is
no significant difference between the P300 and SPW amplitudes of ex-smokers
and never-smokers, it can also be concluded that ex-smokers display the same
(low) level of processing bias as never-smokers. In addition, nicotine-craving
ratings and pleasantness ratings of smoking stimuli were higher in smokers
compared to ex-smokers. It can be concluded that the smoking-related craving,
pleasantness rating, and processing bias decreases after a period of prolonged
abstinence.
60
Effects of prolonged abstinence on processing bias
Introduction
Substance use disorders are associated with processing biases for drug-related
stimuli (for reviews see Field, Mogg, & Bradley, 2006; Franken, 2003). These
processing biases are thought to emerge because of the motivational and
attention-grabbing properties of drug cues (Robinson & Berridge, 1993). For drugdependent persons these stimuli are extremely attractive, become the focus of
attention, and are able to elicit approach behaviors such as drug seeking and drug
consumption (Robinson & Berridge, 1993). The hyperattentive state of drug users
that is associated with drugs and drug-related stimuli is called attentional bias.
The incentive-sensitization theory of Robinson and Berridge (1993) provides an
explanation for this bias in drug abuse patients. This theory predicts that repeated
administration of drugs causes a sensitization of dopamine neurotransmission in
the striatum, which in turn causes drugs and drug-associated stimuli to acquire
incentive motivational properties. This ‘incentive salience’ or relevance of stimuli
for reinforcement makes the drug-associated stimuli extremely ‘wanted’ and
therefore a greater proportion of attentional resources is allocated to them.
Further, because the neurobiological substrates of this wanting system are
irreversibly sensitized, it is implicitly hypothesized that this enhanced processing
does not decrease after abstinence. A related theoretical account of addiction
is Franken’s model of attentional bias (2003), in which it is speculated that the
presence of attentional bias may increase drug-related cognitions, enhance the
signaling of drug cues and diminish the attentional resources left for alternative
cues, all of which in turn may strengthen the enhanced processing of drugs and
drug-related stimuli. Furthermore, Franken’s model predicts that craving is
reciprocally associated with the attentional processing of drug-related stimuli.
That is, the presence of craving results in enhanced processing of drug-related
stimuli and vice versa.
The clinical importance of this enhanced processing has been demonstrated in
several studies that found a relation between relapse rates and performance on
the emotional Stroop task, a measure of attentional bias. This relation has been
demonstrated in smokers (Waters et al., 2003), alcoholics (W. M. Cox et al., 2002),
cocaine (Carpenter et al., 2006), and heroin dependent subjects (Marissen et al.,
2006). In a study using a visual probe task, this relation between relapse and
processing bias was not observed (Waters, Shiffman, Bradley, & Mogg, 2003),
suggesting that only the Stroop task has predictive value.
61
Chapter 4
Substance-related processing bias has been demonstrated in heroin (Franken,
Kroon, Wiers, & Jansen, 2000; Lubman, Peters, Mogg, Bradley, & Deakin, 2000)
and cocaine abusers (Hester, Dixon, & Garavan, 2006) and heavy alcohol drinkers
(Duka & Townshend, 2004; Field, Mogg, Zetteler, & Bradley, 2004). In addition, also
smokers exhibit this processing bias (Ehrman et al., 2002; T. M. Gross, Jarvik, &
Rosenblatt, 1993; Waters & Sayette, 2006). Processing bias for smoking-related
stimuli is present in both light to moderate smokers (Waters & Feyerabend,
2000) and heavy smokers (Waters, Shiffman, Bradley et al., 2003), and in both
smokers who abstained from smoking for a couple of hours and smokers who
recently smoked (Rusted, Caulfield, King, & Goode, 2000). Lifetime consumption
of nicotine and extent of smoking dependence appear unrelated to this bias
(Waters & Feyerabend, 2000; Waters, Shiffman, Bradley et al., 2003), but number
of unsuccessful quitting attempts as well as attitudes against smoking appear to
be respectively positively and negatively correlated (B. P. Bradley, Mogg, Wright,
& Field, 2003; Johnsen, Thayer, Laberg, & Asbjornsen, 1997). However, some
studies did not find a positive relationship between processing bias and indices
of smoking behavior such as the number of cigarettes smoked per day (Hogarth,
Mogg, Bradley, Duka, & Dickinson, 2003; Hogarth, Dickinson, & Duka, 2005; Mogg,
Field, & Bradley, 2005), implicating that this relationship is far from clear and that
more research on this topic is needed.
A relatively new approach to assess the processing of drug-related stimuli is the
measurement of event-related potentials (ERP) using electroencephalography
(EEG) techniques. The ERP consists of several time-locked components, all of
which reflect one or more information-processing operations. The amplitude of the
components presumably depicts the extent to which an information-processing
operation is engaged (for reviews, see Coles, Gratton, & Fabiani, 1990; Gehring,
Gratton, Coles, & Donchin, 1992). ERP has several advantages above reaction time
measurements. ERP methodology provides a potentially more direct assessment
of processing bias than conventional reaction time data since brain activity can be
measured directly without relying on motor-responses. Further, it is possible to
derive some indications of neural generators. For example, it is known that early
visual ERP components are associated with activity in the extra striate cortex
(e.g., Schupp, Junghöfer, Weike, & Hamm, 2003b), although of course, neuroimaging
methodology such as fMRI is more suitable for this goal. In addition, ERP
methodology is suitable to study the temporal dynamics of the processing. That is,
early components of the ERP are thought to reflect the more automatic, stimulus-
62
Effects of prolonged abstinence on processing bias
driven cortical processing of a stimulus, whereas later components most likely
reflect more voluntary, top-down controlled processing (Carretie et al., 2004).
ERP research addressing the processing of drug-related stimuli show that the
later ERP components, such as the P300 and the Slow Positive Wave (SPW), are
enhanced in drug use populations, in contrast to earlier components and drug
naïve populations (e.g., Franken, Stam et al., 2003). This is in line with behavioral
data showing that processing biases are only found when stimuli are presented
above the threshold of awareness, i.e., do not operate in preconscious processes
(B. P. Bradley, Field, Mogg, & De Houwer, 2004; Franken, Kroon, Wiers et al., 2000;
Mogg & Bradley, 2002).
Although there is still some debate on the exact meaning of these late components,
it is widely believed that they reflect attentive processing as well as the activation
of motivational and arousal systems in the brain (Cuthbert et al., 2000; Lang
et al., 1997; Schupp et al., 2000). In recent ERP studies of addiction, it has been
found that these late ERP components are adequate indices of the processing of
drug-related stimuli (Franken, Stam et al., 2003; Van de Laar et al., 2004). More
specifically, enhanced P300 and SPW amplitudes resulting from the processing of
drug-related stimuli have been found in alcohol, cocaine, and heroin dependent
patients (Franken, Stam et al., 2003; Herrmann et al., 2000; Herrmann et al., 2001;
Namkoong et al., 2004; Van de Laar et al., 2004).
As for smokers, Warren and McDonough (1999) were the first to study processing
biases using ERPs. In accordance with the results of aforementioned studies
(e.g., Namkoong et al., 2004), they found significant ERP discrepancies between
smoking-related and neutral pictures at the P412, a component similar to the P300.
This discrepancy was significantly larger for smokers than for never-smokers at
Fz and Cz electrodes, indexing enhanced processing of smoking-related stimuli.
Never-smokers also showed a difference in P412 amplitude between smokingrelated and neutral cues in this study, but the location of this difference was,
in contrast to smokers, more posterior, being most pronounced at central and
parietal-temporal sites. Additional analyses revealed that the effects for neversmokers were smaller than for smokers and therefore Warren and McDonough
(1999) assume that their P300-like component indeed reflected the allocation
of attentional resources toward information relevant to the smokers’ tobaccoaddicted, incentive-motivational states. The effects found in never-smokers could
63
Chapter 4
have been due to task demands, arising from the realization that the study dealt
with cigarette smoking. In contrast to cocaine and heroin-dependent patients
(Franken, Stam et al., 2003; Van de Laar et al., 2004), Warren and McDonough
did not observe significant differences between smokers and nonsmokers on
the SPW component. Recently, Fehr et al. (2006) demonstrated an attentional
bias in smokers for smoking-related words compared to neutral words and nonsmokers. This bias was associated with frontal relative positivity in the P300 time
frame. Although these results are difficult to compare with those of Warren and
McDonough (1999) because of different methodology (Word Stroop vs. passive
picture viewing), both studies showed similar ERP activation patterns.
There are indications that processing biases are associated with subjective drug
craving (for reviews, see Field et al., 2006; Franken, 2003). Recent research (Field
et al., 2004) suggests that this relationship is bidirectional in nature: drug craving
results in enhanced processing of drug-cues, but processing biases may result
in enhanced craving. ERP measures of processing bias, i.e., enhanced P300 and
SPW amplitudes have been found to correlate with drug craving, confirming this
relationship. Namkoong et al. (2004) report subjective craving to be increased
after drug-related picture presentation, and, moreover, this increase correlates
significantly with P300 amplitude. Approximately the same is true for heroin
abusers, who show a significant correlation between self-reported craving and SPW
amplitude (Franken, Stam et al., 2003). Furthermore, in a study in which cocaine
abusers are classified as ‘low cravers’ or ‘high cravers’, the latter show a more
pronounced SPW in response to cocaine cues relative to neutral cues (Franken,
Hulstijn, Stam, Hendriks, & van den Brink, 2004). However, it must be noted that not
all ERP studies of addiction find correlations between processing bias and craving
(Van de Laar et al., 2004). The relation between processing bias and craving has
also been demonstrated in smokers (Mogg, Bradley, Field, & De Houwer, 2003;
Waters, Shiffman, Bradley et al., 2003). Nevertheless, processing bias as measured
by ERPs failed to correlate with urge to smoke (Warren & McDonough, 1999).
Clearly, more research is needed in order to resolve these issues.
Studies addressing the time-course of attentional biases in ex-smokers are scarce. A
recent study using a dot-probe measure of attentional bias reveals that ex-smokers
have an intermediate bias for smoking-related stimuli, falling in between smokers
and nonsmokers (Ehrman et al., 2002). A second study using the modified Stroop
paradigm reveals that there is actually no significant difference in attentional
64
Effects of prolonged abstinence on processing bias
bias between never-smokers and ex-smokers (Munafò, Mogg, Roberts, Bradley, &
Murphy, 2003), indicating that processing biases do not appear to be a permanent
feature of nicotine addiction. These results, in particular the latter, are in contrast
with one specific notion of the incentive-sensitization model (Robinson & Berridge,
1993), predicting that the neuro-adaptations are more or less permanently present,
suggesting that drug-related cues retain their incentive-motivational properties
after cessation of drug use. And therefore, processing biases will persist after
cessation of smoking. Apparently, the issue whether attentional bias is a more or
less permanent feature of nicotine addiction remains to be resolved. The present
study addresses the question whether a smoking-associated processing bias is
still present in ex-smokers after prolonged abstinence.
In order to investigate the permanency of smoking-related processing bias in
ex-smokers, we conducted an ERP study in which we compared ex-smokers’
later ERP components in response to smoking-related and neutral pictures with
those of smokers and never-smokers. Following the results of the aforementioned
studies, one of the main hypotheses of the present study is that ex-smokers
have less processing bias for smoking-related cues than smokers and that this
bias approximates to never-smokers’ bias, i.e., smoking-related cues are less
motivational relevant for ex-smokers than for smokers and equally insignificant
for ex-smokers as for never-smokers. Therefore it is hypothesized that ex-smokers,
in response to smoking-related pictures, show less enhanced P300 and SPW
amplitudes than smokers, whereas the P300 and SPW amplitudes of ex-smokers
and never-smokers have approximately the same magnitude.
Since there is some evidence that attentional bias is associated with craving levels
(Field et al., 2006; Franken, 2003), we also assessed smokers’ and ex-smokers’
subjective craving scores. It is expected that in the present study ex-smokers
will report less subjective craving than smokers. Furthermore, the present study
investigated the differences between smokers, ex-smokers and never-smokers
in arousal and valence judgments of the smoking-related and neutral pictures.
Previous studies show that smokers evaluate smoking-related pictures more
positively than neutral stimuli (Geier, Mucha, & Pauli, 2000; Hogarth & Duka,
2006; Mogg et al., 2003), whereas never-smokers evaluate them more negatively
than neutral stimuli (Mogg et al., 2003). It is unknown how ex-smokers will judge
the stimuli and if their scores will differ from those of smokers or never-smokers.
Finally, both subjective craving and arousal and valence judgments are correlated
65
Chapter 4
with ERP amplitude. Because positive correlations are found in prior studies with
drug-dependent individuals (Franken, Stam et al., 2003; Franken et al., 2004;
Namkoong et al., 2004), they are predicted to be positively associated.
In addition, because Warren and McDonough (1999) not only found significant
differences between smokers and never-smokers at the P300 component, but
also at the N268 component (similar to the N300), differences between smokers,
ex-smokers and never-smokers at this latter component were exploratively
investigated in the present study.
Method
Subjects
Twenty-two smokers, 21 ex-smokers and 24 never-smokers were initially
recruited by an advertisement placed at the psychology department of the
Erasmus University Rotterdam (The Netherlands). All participants were screened
by telephone for study eligibility (smoker status, and cigarettes/day). Smokers
were eligible if they smoked ten cigarettes or more per day. Ex-smokers were
participants who quit smoking at least six months ago and did not smoke a single
cigarette within that period. Never-smokers were included if they had not smoked
more than three cigarettes in their lifetime.
Seven participants (1 smoker; 3 ex-smokers; 3 never-smokers) were excluded from
the analyses because of excessive artifacts in the EEG-signal (>50% of the epochs),
resulting in a final group of 21 smokers (mean age 21.6 years, SD = 2.5 years),
18 ex-smokers (mean age 23.1 years, SD = 4.1 years) and 21 never-smokers
(mean age 19.6 years, SD = 1.2 years). The age difference between the groups was
significant, F(2,59) = 8.1, p < 0.01. Never-smokers were younger than smokers
and ex-smokers. However, no correlations were found between age and the ERP
measures (N300, P300 and SPW), indicating that age is not a confounding variable2.
Smokers and ex-smokers did not differ in smoking duration (smokers = 4.8 years,
SD = 2.8 years; ex-smokers = 5.3 years, SD = 3.0 years; t(37) = 0.56, p = 0.58) nor in
nicotine dependence (smokers’ Fagerström Test of Nicotine Dependence (FTND)
score = 3.6, SD = 2.2; ex-smokers’ FTND score = 2.7, SD = 2.4; t(37)= 1.30, p = 0.20).
2 In addition, age was added as covariate in all analyses. No significant main or interaction effect of age was found.
Therefore, we report the analyses without age as covariate.
66
Effects of prolonged abstinence on processing bias
Smokers smoked 10 to 30 cigarettes a day (13.6%: approximately 10 cigarettes,
81.8%: 11-10 cigarettes, 4.5%: 21-30 cigarettes). Ex-smokers smoked also 10 to
30 cigarettes a day (42.9%: approximately 10 cigarettes, 38.1%: 11-20 cigarettes,
19.0%: 21-30 cigarettes). The mean quit duration of ex-smokers was 1.4 years
(SD = 1.8). The groups consisted predominantly of undergraduate psychology
students, who received course credit or a small financial compensation for
participation. The study was approved by the institutional ethical board.
Experimental stimuli
Stimuli consisted of 16 different smoking-related pictures and 16 nonsmokingrelated, neutral pictures. The smoking-related stimuli consisted of ten digital
photographs of persons holding, lighting up, or smoking a cigarette and six
photographs of attributes related to smoking activity, such as packs of cigarettes
and a burning cigarette in an ashtray. These scenes with smoking-related cues
represented situations are known to produce smoking cue-reactivity in smokers
(Niaura, Abrams, Pedraza, Monti, & Rohsenow, 1992) and are associated with
smoking relapse in ex-smokers (Baer & Lichtenstein, 1988). The neutral stimuli
consisted of photographs identical to the smoking-related photographs (i.e., the
same persons, same location, same pose), only without visible smoking activity, and
displaying neutral objects (e.g., a spoon) that were unrelated to smoking behavior.
This way, it was controlled for contrast, brightness and other possible confounding
factors. All pictures were presented full-screen on a 15” color monitor located at
approximately eye level about 1 m in front of the participants.
Procedure
Smokers were asked to abstain from smoking for at least one hour before the
experiment. This short period of smoking deprivation served to reduce possible
acute nicotine effects on ERP amplitude, which are found in several studies
(Houlihan, Pritchard, & Robinson, 1996; Pritchard, Sokhadze, & Houlihan,
2004). Participants were tested alone in a light and sound-attenuated room.
After obtaining informed consent, participants completed a questionnaire about
demographics and smoking history. After completion, participants were seated
in the EEG chair and electrodes were attached. Instructions were to sit relaxed
and still, and to carefully attend to all pictures without employing distracting
thoughts. Then the task was started.
67
Chapter 4
Each of the 16 smoking-related and 16 neutral stimuli was repeated four times
resulting in a total of 128 stimulus presentations. Stimulus presentations from
the two categories were varied in a quasi-random fashion to prevent order and
‘oddball’ effects. There were no successions of more than four stimuli from the
same category. Stimuli were presented for 2000 ms, with an inter-stimulus interval
randomly varying from 1800 to 2200 ms (with an average of 2000 ms).
After the picture viewing, electrodes were removed and smokers and ex-smokers
filled-out the brief Questionnaire on Smoking Urges (QSU-brief; L. S. Cox et al.,
2001). In addition, all participants rated the pictures on their arousal and valence
properties. After having completed the experiment, subjects received their course
credit or financial compensation.
Self-report measures
Demographic and smoking history data were self-reported (age, smoking duration,
and period(s) of abstinence). Subjective craving was measured by the 10-item
QSU-brief (Cox et al., 2001). This 10-item questionnaire was adapted from the
Questionnaire on Smoking Urges (QSU; Tiffany & Drobes, 1991) and consists of two
subscales: “desire and intention to smoke”, and “reduction of negative affect and
withdrawal symptoms”. These subscales have adequate psychometric properties
(L. S. Cox et al., 2001).
Strength of smoking habit was assessed by means of the Dutch version of
the Fagerström Test for Nicotine Dependence (FTND; Vink et al., 2005). This
questionnaire has good reliability and correlates significantly with number of
cigarettes smoked per day. The FTND consists of six items, which are scored according
to the scoring system described in Heatherton et al. (1991). Ex-smokers filled-out
the FTND retrospectively. Retrospectively assessed FTND scores have adequate
psychometric properties (Hudmon, Pomerleau, Brigham, Javitz, & Swan, 2005).
Valence and arousal properties of the pictures were assessed by 10 cm Visual
Analog Scales (VAS). The valance scale ranged from very pleasant (0 cm) to very
unpleasant (10 cm); the arousal scale ranged from doesn’t arouse me (0 cm) to
arouses me much (10 cm). For this task, the pictures were printed in color ink on
white paper.
68
Effects of prolonged abstinence on processing bias
Physiological measures
EEG was measured with a digital BioSemi amplifier using Ag/AgCl electrodes at
34 scalp sites according to the International 10/20 system (32 standard channels
including left and right mastoid locations). The vertical electro-oculogram (VEOG)
was recorded with two Ag/AgCl electrodes located above and underneath the left
eye. The horizontal electro-oculogram (HEOG) was recorded with two Ag/AgCl
electrodes located at the outer canthus of each eye. All signals were digitized on a
PC with a sample rate of 256 Hz and 24-bit A/D conversion. Off-line, EEG and EOG
were filtered using a 0,1-30 Hz (24 dB/Oct roll off) band-pass filter. Four scalp
electrodes (Fz, Cz, Pz, Oz) were used in the present study.
Data reduction and analysis
EEG and EOG recordings were segmented in 950 ms epochs, including 100 ms prestimulus baseline. Gratton and Coles algorithm (Gratton, Coles, & Donchin, 1983)
was used for correction of vertical and horizontal eye movements, and eye blinks.
After ocular correction all segments with an EEG activity above -/+ 100 μV were
excluded from further analysis. A 100 ms pre-stimulus baseline correction was
applied, and epochs were averaged across trials. Overall grand averages were
obtained for each picture category in the three groups. The resulting ERP-waves
were visually inspected and appeared to correspond well with ERP-waves usually
reported in response to visual stimuli (see Figure 1 for a representation of the
separate waves at electrode Fz). Three ERP components were investigated: a
negative waveform captured by a 220-300 ms time window (most similar to the
traditional N300), a positive waveform from 300 to 400 ms (traditional P300)
and a slow positive wave captured by a 500-750 ms time window (SPW). Mean
maximum amplitudes were computed per group and stimulus category for the
aforementioned time windows. Since no clear peaks were observed in the 500-750
ms time range, area measurement (mean activity) was applied here.
69
Chapter 4
—9 )]
6PRNHUVQHXWUDO
6PRNHUVVPRNLQJ
1RQVPRNHUVQHXWUDO
1RQVPRNHUVVPRNLQJ
([VPRNHUVQHXWUDO
([VPRNHUVVPRNLQJ
1
3
63:
PV
Figure 1. Average event-related potentials for all groups and conditions at electrode Fz
Statistical analysis
For each time interval, ERP effects were assessed by performing repeatedmeasurement analyses of variance (ANOVA) on the four midline electrode sites
(Oz, Pz, Cz, and Fz). Group (smokers, ex-smokers, and never-smokers) served as
the between-subjects factor, and cue type (neutral versus smoking-related) and
midline site (Oz, Pz, Cz and Fz) served as within-subjects factors. This resulted
in a 4 (midline site) x 2 (cue) x 3 (group) repeated measures ANOVA. To assess
relationships between cue-evoked ERP amplitudes, craving levels, and valance/
arousal assessments, Spearman correlation coefficients were calculated between
significant ERP amplitudes, self-reported craving levels, and valence/ arousal
judgments. Arousal and valance ratings of the pictures were tested using two
2 (cue type) x 3 (group) repeated-measurement ANOVA’s. To examine exact
differences between groups and cues, pairwise post-hoc follow-up analyses with
Bonferroni correction were applied to all ANOVAs. Greenhouse-Geisser correction
was applied to all ANOVAs when necessary. An alpha-level of 0.05 was used for all
statistical tests.
70
Effects of prolonged abstinence on processing bias
Results
Because our interest concerned mainly group differences, only Group and Group
interaction effects are reported. In order to reduce the number of ERP results, and
in line with the hypotheses of the study, we report only significant (or bordersignificant) Group or Group-interaction effects. The averaged ERP waveforms
on the neutral and smoking-related stimuli for smokers, ex-smokers, and neversmokers are displayed in figures 2-4. In tables 1-3, the mean and standard
deviations of the ERP components are displayed.
—9 )]
PV
Figure 2. Average event-related potentials at the frontal (Fz) site for smokers in response
to neutral pictures (grey) and smoking-related pictures (black)
71
Chapter 4
—9 )]
PV
Figure 3. Average event-related potentials at the frontal (Fz) site for ex-smokers in
response to neutral pictures (grey) and smoking-related pictures (black)
—9 )]
PV
Figure 4. Average event-related potentials at the frontal (Fz) site for non-smokers in
response to neutral pictures (grey) and smoking-related pictures (black)
72
Effects of prolonged abstinence on processing bias
Table 1. Mean amplitudes (in μV) of N300, P300 and slow positive wave (SPW) on smokingrelated and neutral cues at midline sites (Fz, Cz, Pz, and Oz) for ex-smokers.
Component
N300
P300
SPW
Site
Fz
Cz
Pz
Oz
Fz
Cz
Pz
Oz
Fz
Cz
Pz
Oz
Msmoking-related (SD)
- 9.39 (6.00)
- 6.38 (7.32)
2.09 (7.39)
7.08 (2.59)
- 0.28 (6.21)
4.33 (7.35)
12.02 (7.27)
11.42 (3.42)
0.63 (6.11)
5.29 (6.47)
8.73 (5.05)
6.31 (3.10)
Mneutral (SD)
- 9.99 (6.39)
- 6.72 (7.03)
1.86 (7.55)
7.56 (2.84)
- 2.45 (5.48)
1.25 (6.13)
9.20 (5.96)
10.40 (3.37)
- 1.53 (4.75)
2.90 (4.60)
6.84 (3.54)
5.88 (3.21)
Table 2. Mean amplitudes (in μV) of N300, P300 and slow positive wave (SPW) on smokingrelated and neutral cues at midline sites (Fz, Cz, Pz, and Oz) for smokers.
Component
N300
P300
SPW
Site
Fz
Cz
Pz
Oz
Fz
Cz
Msmoking-related (SD)
- 9.39 (6.00)
- 9.99 (6.39)
7.08 (2.59)
- 0.28 (6.21)
4.33 (7.35)
7.56 (2.84)
- 2.45 (5.48)
1.25 (6.13)
- 6.38 (7.32)
2.09 (7.39)
Pz
12.02 (7.27)
Cz
5.29 (6.47)
Oz
Fz
Pz
Oz
Mneutral (SD)
- 6.72 (7.03)
1.86 (7.55)
9.20 (5.96)
11.42 (3.42)
10.40 (3.37)
8.73 (5.05)
6.31 (3.10)
6.84 (3.54)
5.88 (3.21)
0.63 (6.11)
- 1.53 (4.75)
2.90 (4.60)
73
Chapter 4
Table 3. Mean amplitudes (in μV) of N300, P300 and slow positive wave (SPW) on smokingrelated and neutral cues at midline sites (Fz, Cz, Pz, and Oz) for never-smokers.
Component
N300
P300
SPW
Site
Fz
Cz
Pz
Oz
Fz
Cz
Pz
Oz
Fz
Cz
Pz
Oz
Msmoking-related (SD)
- 6.18 (5.19)
- 3.53 (4.97)
0.76 (4.70)
4.78 (4.80)
0.52 (4.64)
3.17 (4.80)
8.88 (6.37)
9.49 (5.90)
0.97 (3.58)
4.74 (3.92)
6.19 (4.51)
4.05 (4.33)
Mneutral (SD)
- 7.18 (4.45)
- 4.78 (4.33)
0.33 (4.75)
4.95 (5.42)
- 0.94 (4.02)
2.40 (4.30)
8.05 (6.42)
9.31 (5.84)
0.49 (3.82)
3.64 (3.50)
6.05 (4.37)
4.65 (3.61)
N300
For the N300 peak, no significant main effect of Group (G), Group (G) x Cue (C)
interaction nor G x C x Site (S) interaction effects could be observed (F < 0.55, ns)
on any of the midline sites.
P300
For the P300 peak, a G x C interaction effect was found, F(3,171) = 3.83, p < 0.05.
Post-hoc comparisons revealed no significant differences between smokers, exsmokers and never-smokers on neutral cues. On the smoking-related cues P300
amplitude was significantly larger for smokers than for never-smokers (p < 0.005)
or ex-smokers (p < 0.05), whilst no significant differences were found between
ex-smokers and never-smokers (p = 1). Besides a significant G x C interaction,
a G x C x S interaction effect was found, F(3,171) = 3.46, p < 0.01. Post-hoc analyses
showed that at none of the single electrodes (Fz, Cz, Pz, and Oz) groups differed
in their P300 response to nonsmoking-related, neutral pictures. However, in
response to smoking-related pictures several differences were found. At Fz, P300
amplitude was more enhanced for smokers than for never-smokers (p < 0.05) and
ex-smokers (p < 0.05). At this site never-smokers’ P300 amplitude did not differ
from ex-smokers’ P300 amplitude (p = 1). Almost the same differences were found
at Cz: Smokers showed a more enhanced P300 amplitude than never-smokers
(p < 0.005), ex-smokers showed a less enhanced P300 amplitude than smokers
(p < 0.05), but ex-smokers’ P300 amplitude did not differ from the never-smokers’
74
Effects of prolonged abstinence on processing bias
amplitude (p = 1). At Pz, P300 amplitude differed between smokers and neversmokers (p < 0.005). No effects were found at Oz.
SPW
No significant G x C interaction was found for the SPW, F(3,171) = 2.39, p = 0.10.
However, a G x C x S interaction effect was revealed, F(3,171) = 2.39, p < 0.05.
Follow-up comparisons showed that the three groups did not differ in SPW
response to neutral pictures on any of the sites. These comparisons also revealed
that in response to smoking-related pictures smokers’ SPW amplitude at Fz was
significantly more enhanced than never-smokers’ SPW amplitude (p < 0.05) and
ex-smokers’ SPW amplitude (p < 0.05). However, ex-smokers’ and never-smokers’
SPW amplitudes did not differ at Fz (p = 1). At Cz, smokers displayed a significantly
larger SPW amplitude than never-smokers (p < 0.05). The difference between
smokers and ex-smokers nearly reached significance (p = 0.061), whereas no
difference between ex-smokers and never-smokers was found (p = 1). At Pz and Oz
electrodes no significant SPW amplitude differences were found between groups.
Self-reported craving and ERP waves
The scores on the post-exposure QSU-brief were significantly higher for smokers
(M = 39.4, SD = 9.0) than for ex-smokers (M = 19.11, SD = 9.6), t(37)= 6.83, p < 0.01.
This difference is mainly the result of a difference between the groups in scores on
the first subscale ‘desire and intention to smoke’, t(37) = 10.03, p < 0.01. Smokers
did not differ from ex-smokers on the second subscale ‘reduction of negative affect
and withdrawal symptoms’, t(37)= 1.84, ns.
No significant correlations were observed between SPW amplitude differences
(response to smoking-related cues minus response to neutral cues) on the four
midline sites on the one hand and self-reported craving on the other. However, at the
P300 peak, Fz amplitude difference correlated significantly with the first subscale
of the QSU-brief, “desire and intention to smoke”, r = 0.32, p < 0.05. Therefore, the
greater the desire and intention to smoke, the larger the Fz amplitude in response
to smoking-related pictures relative to the Fz amplitude in response to neutral
pictures.
75
Chapter 4
Valence and arousal properties of smoking-related pictures and
ERP waves
Concerning arousal and valence ratings, we were interested in Cue (C) x Group
(G) effects. Therefore, we only report these interaction effects. See table 4 for the
mean ratings. On the arousal ratings we found a significant C x G interaction effect,
F(2,57) = 7.47, p < 0.05. Post hoc tests showed that there were no group effects for
the neutral cues (all p’s ns). However, a group effect for the smoking-related cues
was found. The arousal score of smoking-related pictures was significantly greater
for smokers than for never-smokers (p < 0.05), but did not differ between smokers
and ex-smokers (p = 0.90) and ex-smokers and never-smokers (p = 0.240), which
implicates that the arousal ex-smokers experienced falls in between the arousal
smokers and never-smokers experience in response to smoking-related pictures.
Within-group differences in self-reported arousal were observed between neutral
and smoking-stimuli among smokers (p < 0.001) and ex-smokers (p < 0.001), but
not in never-smokers (p = 0.08). Both smokers and ex-smokers reported more
arousal on smoking pictures than on neutral pictures.
Table 4. Mean self-reported arousal and valence ratings (SD) of the three samples.
Arousal
Valence
Neutral
Smoking
Neutral
Smoking
Never-smokers
2.7 (1.4)
3.5 (2.5)
4.9 (1.2)
6.3 (1.9)
Smokers
2.3 (1.7)
5.6 (2.0)
4.8 (0.5)
4.2 (1.2)
Ex-smokers
2.8 (1.8)
4.8 (2.2)
4.6 (0.7)
6.3 (1.4)
Note. Arousal scores range from 1 (does not arouse me) - 10 (arouses me very much).
Valence scores range from 1 (very pleasant) - 10 (very unpleasant).
Concerning valence ratings, we also found a significant C x G interaction effect
F(2,57) = 18.17, p < 0.001. Post-hoc tests showed that there were no group effects for
the neutral cues (all p’s ns). However, the valence score of smoking-related pictures
was greater for smokers than never-smokers (p < 0.001), greater for smokers than
ex-smokers (p < 0.001) and did not differ between never-smokers and ex-smokers
(p = 1). This implicates that ex-smokers and never-smokers found smoking-related
pictures less pleasurable than smokers and that ex-smokers found these pictures
as unpleasant as never-smokers. Within-group differences in self-reported valence
were observed between neutral and smoking-stimuli were found and ex-smokers
(p < 0.05), and never-smokers (p < 0.05). Never-smokers and ex-smokers evaluated
smoking-related pictures as less pleasurable than neutral pictures. Self-reported
valence differences were also observed between neutral and smoking-stimuli in
76
Effects of prolonged abstinence on processing bias
smokers (p < 0.05). In contrast to ex-smokers and never-smokers, smokers rated
smoking-related pictures significantly more pleasurable than neutral pictures.
Correlations between SPW amplitude differences on midline sites and both arousal
and valence difference scores (evaluation of smoking-related pictures minus
neutral pictures) were not significant. In addition, no significant correlations were
found between P300 amplitude difference and valence difference score. For this
component, however, a significant correlation emerged between Fz amplitude
difference and valence difference score (r = -0.29, p < 0.05) and between Cz
amplitude difference and valence difference score (r = -0.26, p < 0.05), suggesting
that the greater the frontal and central amplitude difference in response to
smoking-related pictures and neutral pictures, the lower the valence score given
to the smoking-related pictures relative to neutral pictures, that is, the more
pleasurable the smoking-related pictures are found.
Discussion
The present study investigated processing bias of smoking-related stimuli in
smokers, ex-smokers and never-smokers employing ERP measurements. Several
hypotheses were formulated concerning group differences in smoking-related
and neutral cue-evoked ERP waves. The main hypothesis was that smokingrelated pictures have greater motivational salience and therefore smokers would
display enhanced processing of these pictures compared to ex-smokers and neversmokers. Since enhancement of amplitudes of later ERP components is believed
to reflect increased processing, it was hypothesized, more specifically, that
smokers would show more enhanced amplitudes of the later ERP components in
response to smoking-related pictures than ex-smokers and never-smokers. This
hypothesis is confirmed by the results of the present study. Both the P300 and
SPW amplitudes in response to smoking-related pictures are significantly more
enhanced for smokers than for ex-smokers and never-smokers at frontal and
central sites, whereas the magnitude of the P300 and SPW amplitudes in response
to neutral pictures does not differ between the three groups. Accordingly, it can
be concluded that smokers show more bias for smoking-related pictures than exsmokers and smokers. Because there is no significant difference between the P300
and SPW amplitudes of ex-smokers and never-smokers, it can also be concluded
that ex-smokers display the same amount of processing bias as never-smokers.
77
Chapter 4
From previous studies using electrophysiological measures of emotional
information processing it has become apparent that motivational relevant
stimuli, such as emotional pictures attract attention (Cuthbert et al., 2000; Lang
et al., 1997; Lang, Bradley, & Cuthbert, 1998; Schupp, Junghöfer, Weike, & Hamm,
2003b; Vuilleumier, 2005). It has been suggested that increased slow waves of
the ERP reflect an increased allocation of attentional resources to motivational
relevant (emotional) stimuli, also described as “motivated attention” (Lang et al.,
1997; Schupp et al., 2004). In this context, the present findings on the processing
of smoking stimuli are in line with studies using more specific attentional bias
measures of smoking-related processing, such as the smoking Stroop task.
Our findings are fully in line with the results of Munafo et al. (2003), who found
a significant difference in attentional bias between smokers and ex-smokers but
no difference between ex-smokers and never-smokers. Although the findings are
generally in line with an incentive sensitization view of addiction (Robinson &
Berridge, 1993), the present findings also contradict the prediction of Robinson
and Berridge that the neural adaptations, which result in an enhanced processing
of drug cues in former drug users, are a permanent feature of addiction. The
current findings suggest that in ex-smokers, at least to some extent, extinction
of the cortical reactivity towards smoking cues has taken place. The P300 and
SPW differences between smokers and ex-smokers and the absence of these
differences between ex-smokers and never-smokers in the present study show
that the property of smoking-related cues to enhance cortical processing is not a
permanent feature and, at least partly, reversible.
The P300 discrepancies between smokers and never-smokers found in the present
study are to a large extent in accordance with the results reported by Warren
and McDonough (1999). The smokers in their study showed larger positive P412
(comparable to the P300 component in our study) differences at Cz and Fz to
the two types of stimuli (smoking-related minus neutral stimuli) compared to
never-smokers. The P300 differences between smokers and never-smokers in the
present study are comparable, in that they are also found at Fz and Cz. In contrast
to our study, Warren and McDonough (1999) did not find any differences between
smokers and never-smokers on the SPW component. The finding that there are SPW
differences between smokers, never-smokers and ex-smokers in the present study
is in line with results from previous studies among heroin abusers (Franken, Stam
et al., 2003) and cocaine abusers (Franken et al., 2004; Van de Laar et al., 2004).
78
Effects of prolonged abstinence on processing bias
A possible explanation for the presence of SPW differences in the present study but
their absence in the study of Warren and McDonough (1999), is the utilization of
different stimulus material. Present material appears to be more attractive than
the material used by Warren and McDonough. The smokers in their study did not
evaluate the smoking-related stimuli significantly more pleasurable than neutral
stimuli, whilst the smokers in the present study do. Besides, the stimuli presented
in the present research were shown for a longer time compared to Warren and
McDonough (2000 versus 150 ms), allowing more elaborative processing of the
stimuli, which results in a larger SPW component.
It should be noted, however, that the distinction between the P300 and the SPW
component is rather arbitrary. It is possible that these two components in fact
represent only one component, i.e., one information processing operation. The
positive wave from 400 to 750 ms, labeled SPW in the present study, could be part
of or an extension of the P300.
In the present study no significant N300 differences were found between
smokers, ex-smokers and never-smokers. This is in contrast with Warren and
McDonough (1999) who found a difference between smokers and never-smokers
on the N268 component. In smokers, the amplitude of this N300-like component
was significantly more enhanced in response to neutral cues than in response to
smoking-related cues. The authors suggested that this component probably depicts
the neutral pictures’ lack of fit to the smokers’ functional or subjective tobaccoaddicted states. A possible explanation for this inconsistency on the N300 between
the present study and that of Warren and McDonough (1999) is that the smokers
in the present study had more difficulty detecting differences between smokingrelated and neutral pictures because they were practically identical except for the
presence or absence of smoking activity. However, future investigation of the N300
component in addiction is necessary.
Another objective of the present study was to examine the relationship between
processing bias and drug craving. After picture viewing smokers report significantly
more craving than ex-smokers, which is congruent with smokers exhibiting greater
amplitude differences than ex-smokers. Furthermore, the craving subscale ‘desire
and intention to smoke’ appears to be significantly and positively correlated
with frontal P300 amplitude. The robustness of this finding is confirmed by the
correlation between the frontal P300 amplitude and the valence judgment. These
79
Chapter 4
findings implicate that the more pleasant the smoking-related pictures are found
and the more desire and intention to smoke they induce, the greater the frontal
P300 amplitude difference between smoking-related and neutral pictures, i.e., the
more enhanced the attentive processing of smoking-related cues.
The correlation with only one aspect of craving (i.e., ‘desire and intention to smoke’
but not ‘reduction of negative affect’) is consistent with a positive-incentive view of
addiction (Robinson & Berridge, 1993; Stewart, de Wit, & Eikelboom, 1984), which
predicts that not the negative withdrawal symptoms, but mainly the positiveincentive features of drug stimuli elicit craving.
It should be noted that only the frontal P300 amplitude correlates with selfreported craving, and that this relation is moderate. The absence of a stronger
correlation, observed in other studies addressing the relation between drug
craving and processing bias (Franken, Stam et al., 2003; Franken et al., 2004;
Namkoong et al., 2004), may be attributable to the fact that smoking cravings are
less explicit than cocaine, heroin, and alcohol craving.
A limitation of the present study is that we did not measure the perceived
availability to smoke cigarettes. Although all groups had technically the same
opportunity to smoke after the experiment, it might be that the perceived
availability of ex-smokers was reduced because of their higher motivation to
keep abstinent. It is known that perceived availability is associated with craving
levels (Wertz & Sayette, 2001) and other drug-related responses (Hogarth &
Duka, 2006; Wilson et al., 2004). The influence of perceived availability on ERP
measures of processing has not been addressed before. Another limitation is that
our smokers and ex-smokers samples mainly consisted of relatively light-smokers.
Only a small minority of the (ex-) smokers smoked more than 20 cigarettes a day.
It awaits further study whether the present findings can be generalized to heavy
(ex-)smokers. Another point that should be addressed in future research is how
long after smoking cessation the processing biases to smoking cues persist. In the
present study we found no evidence for a processing bias in ex-smokers who were
abstinent for at least 6 months. Inclusion of a recently abstinent group would yield
more insight in the time course of the extinction of processing biases.
The main conclusion of the present study is that smokers and ex-smokers process
smoking-related pictures differently, whereas ex-smokers and nonsmoker appear
80
Effects of prolonged abstinence on processing bias
to process smoking-related pictures more or less in the same way. The slow
components of the ERP are more enhanced for smokers than for ex-smokers in
response to smoking-related pictures, whereas there are no significant ERP
differences between ex-smokers and never-smokers. This indicates that exsmokers show less processing bias for smoking-related cues than smokers and
above all that this bias diminishes to the bias level of never-smokers. Therefore,
it appears that processing bias is not a permanent feature of nicotine addiction.
Furthermore, we found a relation between amplitudes of the P300 component and
self-reported craving.
81
Chapter
5
Changes in the electroencephalographic
spectrum in response to smoking
cues in smokers and ex-smokers
Littel, M., Franken, I.H.A., & Van Strien, J.W. (2009). Changes in the
electroencephalographic spectrum in response to smoking cues
in smokers and ex-smokers. Neuropsychobiology, 59(1), 43-50.
Chapter 5
Abstract
The aims of the present study were to investigate the changes in the
electroencephalographic (EEG) spectrum in smokers during exposure to a
neutral and a smoking-related cue to determine whether these EEG changes
are still present in ex-smokers after prolonged abstinence and to examine the
relationship between the power in each spectral bandwidth and subjective
craving. EEG frequencies in response to a smoking-related and a neutral
cue were examined in 23 smokers and 21 ex-smokers, who quit smoking for
1.4 years on average. Additionally, self-report measures of cigarette craving
and nicotine dependence were obtained. The spectral power of each bandwidth
was computed, log-transformed, and analyzed using a within-subject design.
Differences between EEG activity under neutral and smoking conditions were
correlated with differences between pre- and post-experimental subjective
craving. Increases in reward craving (desire and intention to smoke) were
associated with reduced theta activity, whereas increases in withdrawal
craving (reduction of negative affect and withdrawal symptoms) were
correlated with increases in both delta and higher alpha power. Furthermore,
in smokers, but not in ex-smokers, a significant beta power increase was
observed between the neutral condition and the smoking condition. Since the
beta band is associated with arousal, attention, and alertness, it is suggested
that the beta increase in response to the smoking cue might reflect an enhanced
allocation of resources to smoking-related stimuli, i.e., a processing bias, which
is an important feature of substance abuse. Since ex-smokers do not respond
to the smoking cue with beta activity enhancement, we preliminarily conclude
that smoking cues do not arouse ex-smokers or capture their attention as much
as they do in smokers.
84
EEG spectrum changes
Introduction
Cigarette smoking has been associated with increases in alertness and mood
changes (e.g., Adan, Prat, & Sanchez-Turet, 2004; Church, 1989; Conrin, 1980;
Jarvik, 1991; Knott, Harr, Ilivitsky, & Mahoney, 1998; Knott, 2001; Knott et al.,
2005). These findings are supported by measures using electroencephalography
(EEG) techniques, which reveal that nicotine administration leads to strong
increases in electrophysiological activity, i.e., scalp-recorded activity shifts from
low (delta, theta, alpha-1) to high (alpha-2, beta) frequencies, indicating a state of
arousal. The reverse is true for nicotine abstinence: deprivation causes increases
in theta power, and leads to reductions in both alpha and beta frequency (for an
overview, see Domino, 2003; Teneggi et al., 2004). Decreases in alpha frequency
have been associated with slow reaction time (Surwillo, 1961; Surwillo, 1963),
diminished arousal and decreased vigilance (Knott & Venables, 1977; Ulett &
Itil, 1969). Increases in theta power are correlated with drowsiness (Matousek &
Petersen, 1983; Ulett & Itil, 1969) and the transition from wakefulness to sleep
(Kooi, Tucker, & Marshall, 1978). These changes in EEG spectrum in response
to smoking abstinence persist until at least 1 month after quitting (Gilbert et
al., 1999). As far as we know, there are no studies investigating EEG activity in
prolonged (>1 month) smoking abstinence, and therefore it is unknown how EEG
power and frequency will develop after a 1-year period of abstinence.
Besides shifts from high to low frequencies, smoking deprivation has been
associated with shifts regarding the balance of alpha activity between the left
versus the right frontal hemisphere (Zinser, Fiore, Davidson, & Baker, 1999). When
people are presented with appetitive stimuli, and are motivated to approach these,
they tend to display relatively greater left than right activation; conversely, people
presented with aversive stimuli or people under withdrawal-associated conditions
display relatively greater right than left frontal activation (Coan & Allen, 2004;
Davidson, Ekman, Saron, Senulis, & Friesen, 1990; Fox, 1991; Harmon-Jones, 2004;
Pizzagalli, Sherwood, Henriques, & Davidson, 2005). If smokers who are deprived
for 24 h are exposed to cigarette cues or anticipate smoking, they show left frontal
asymmetry, i.e., greater left than right frontal hemispheric activation, suggesting
an enhanced approach motivation as a result of the deprivation (Zinser et al., 1999).
Most EEG spectrum research in smokers is focused on the spontaneous
electrophysiological changes during nicotine intake and/or abstinence. Only
85
Chapter 5
few studies have addressed the question whether EEG power and frequency are
affected by (imaginary) exposure to cigarette cues. Knott et al. (Knott et al., 2008)
found smoking urge scripts, depicting scenes with persons experiencing a desire
to smoke, to increase both beta and theta activity, whereas they found no-urge
scripts, depicting the same scenes without smoking desires, to have no effects
on the EEG frequency domains. Comparable results were observed in cocaine
users: during cocaine-craving-related guided imagery, both theta and beta power
increased. During active cocaine paraphernalia handling and video viewing only
beta activity was enhanced. Changes in delta power were more dependent on task:
during imagery, there was an increase in delta, whereas a drop was observed in the
paraphernalia and video task (Reid et al., 2003). In contrast, Knott et al. (Knott et
al., 2008) found delta activity to be decreased in response to urge-related imagery,
but only in males. Alpha activity remained unaffected by urge and no-urge scripts
and drug-related stimuli in both smokers and cocaine users (Knott et al., 2008;
Reid et al., 2003).
It has been suggested (Knott et al., 2008; Reid et al., 2003) that the EEG changes
in the frequency domain as a result of urge-related imagery reflect drug craving.
However, either no correlation analyses between the power in each frequency band
and self-reported craving were performed (Knott et al., 2008) or they led to nonsignificant results (Reid et al., 2003). It might also be possible that the EEG changes
reflect other processes such as a general enhancement of arousal, or an enhanced
allocation of cognitive resources to smoking-related cues, i.e., a processing bias,
which is a concept associated with craving (Field et al., 2006; Franken, 2003; Mogg
et al., 2003; Waters, Shiffman, Bradley et al., 2003). At present, it is not clear whether
self-reported cigarette craving is related to changes in EEG frequency and power.
Changes in beta, theta and delta might be neurophysiological indices of selfreported craving, since urge scripts are able to provoke changes only in these
frequency domains. Encouraged by the structure of the Questionnaire on Smoking
Urges (QSU-brief; L. S. Cox et al., 2001), which subdivides craving into ‘desire
and intention to smoke’ (reward craving) and ‘reduction of negative affect and
withdrawal symptoms’ (withdrawal craving), based on respectively appetitiveincentive models of craving (e.g., Baker, Morse, & Sherman, 1986; Robinson &
Berridge, 2003; Stewart et al., 1984; Wise, 1988) and associative-withdrawal
models of craving (e.g., Poulos, Hinson, & Siegel, 1981; Siegel, 1983; Wikler, 1980),
Knott et al. (2008) make several assumptions. First, they assume delta reductions
86
EEG spectrum changes
and beta increments, which are smoking-like EEG changes, to be correlated
with reward craving. The second subscale, or withdrawal craving, is expected
to correlate with theta increases, a withdrawal-like EEG change. However, these
assumptions need to be confirmed.
Within the present study, we examined the EEG spectrum changes in smokers
during exposure to a neutral and a smoking-related cue. In line with the results
of Knott et al. (Knott et al., 2008), we expect smokers’ beta and theta activity to
increase in response to the smoking-related cue, but not in response to the neutral
cue. Because of conflicting findings concerning the direction of the changes in the
delta frequency (Knott et al., 2008; Reid et al., 2003), we only hypothesize a change
in delta in response to the smoking cue. Furthermore, the present study addressed
the question whether EEG frequency domain changes in response to smoking cues
are still present in ex-smokers after prolonged abstinence (> 1 year). Recently,
research has shown that ex-smokers, at least to some extent, exhibit an extinction
of the cortical reactivity towards smoking cues (Littel & Franken, 2007). In the
present study, we expected to find differences between smokers and ex-smokers
in line with these results, i.e., that ex-smokers will not respond with changes in
delta, beta and theta activity to the smoking-related cues, or at least to a lesser
degree than smokers.
Moreover, we expected increases in self-reported craving, as measured with the
QSU-brief, to be correlated with increases in beta and theta EEG activity and de
creases in delta EEG activity. In accordance with the assumptions of Knott et
al. (2008), we expected reward craving (‘desire and intention to smoke’) to be
associated with delta reductions and beta increases, and withdrawal craving
(‘reduction of negative affect and withdrawal symptoms’) with theta increases.
Finally, because of their greater approach motivation in smoking-related contexts,
we expect more left frontal alpha asymmetry (left > right) in smokers than in
ex-smokers during the presentation of the smoking cue. Since ex-smokers tend
to evaluate smoking cues as less pleasurable than smokers do and also tend to
evaluate smoking cues as less pleasurable than neutral cues (Littel & Franken,
2007), they will probably show more alpha activity in the right frontal hemisphere
than smokers.
87
Chapter 5
Method
Subjects
Twenty-two smokers and 21 ex-smokers (partly the same as reported in Littel &
Franken, 2007) participated in this study, which was approved by the institutional
ethical board. Smokers (23.3% males, mean age 21.5 years, SD = 2.4) smoked at
minimum 10 cigarettes a day. Ex-smokers (9.3% males, mean age 23.3 years,
SD = 3.9) quit smoking at least 6 months ago and did not smoke a single cigarette
within that period. Smokers and ex-smokers did not differ significantly in age
(t(41) = 1.9, p = 0.07), smoking duration (smokers = 4.8 years, SD = 2.7 years; exsmokers = 5.3 years, SD = 3.1 years; t(41) = 1.6, p = 0.56) or nicotine dependence
(smokers’ Fagerström Test of Nicotine Dependence (FTND) score = 3.5, SD = 2.3;
ex-smokers’ FTND score = 2.7, SD = 2.5; t(41) = 1.0, p = 0.31). Furthermore, sex
ratio was equal in both groups (χ2(1) = 3.4, p = 0.07). The mean quit duration of
ex-smokers was 1.4 years (SD = 1.7). Because of the marginally significant sex
ratio, sex was added as covariate in all analyses. No significant main or interaction
effect of sex was found. Therefore, we report the analyses without sex as covariate.
The groups consisted predominantly of undergraduate psychology students, who
received course credit or a small financial compensation for participation.
Procedure
Smokers were asked to abstain from smoking for at least 1 h before the
experiment. This short period of smoking deprivation served to reduce the acute
effects of nicotine on electro-cortical arousal and accordingly to decrease the
differences between smokers and ex-smokers. First of all, participants completed
a questionnaire about demographics and smoking history: the FTND (Heatherton
et al., 1991). After completion, participants were seated in an EEG chair in a
sound- and light-attenuated room; electrodes were attached, and a pictorial task
was presented (reported elsewhere, see Littel & Franken, 2007). Instructions
were to sit relaxed and still and to carefully attend to the cues without employing
distracting thoughts. After the task was completed, the subjects filled out the
QSU-brief (L. S. Cox et al., 2001). In the present experiment, two cues were
presented. The first cue consisted of a pen on a small dish (control/neutral
condition), the second consisted of a lit cigarette on the same dish (smokingrelated condition). Each cue was located at approximately eye level about 1 m in
front of the participants and was presented for 30 s. After the cue presentation,
electrodes were removed and subjects filled out the QSU-brief for the second time.
88
EEG spectrum changes
After having completed the experiment, subjects received their course credit or
financial compensation.
Self-report measures
Demographic and smoking history data were self-reported (age, period(s) of
abstinence, and smoking duration). Craving was assessed twice by means of the
QSU-brief (L. S. Cox et al., 2001), before and after cue presentation. This 10-item
questionnaire is adapted from the QSU (Tiffany & Drobes, 1991) and consists of
two subscales: ‘desire and intention to smoke’, and ‘reduction of negative affect and
withdrawal symptoms’. These subscales have adequate psychometric properties
(L. S. Cox et al., 2001).
Strength of smoking habit was measured with the Dutch version of the FTND
(Heatherton et al., 1991; Vink et al., 2005). This questionnaire consists of 6 items,
which are scored according to the scoring system described by Heatherton et al.
(1991). The FTND has good reliability and correlates significantly with number of
cigarettes smoked per day. Ex-smokers answered the questions retrospectively.
Retrospectively assessed FTND scores also have adequate psychometric properties
(Hudmon et al., 2005).
Physiological measures
EEG was recorded with a digital BioSemi Active-Two system, using active Ag/
AgCl electrodes at 34 scalp sites according to the International 10/10 system
(32 standard channels mounted in an elastic cap and two mastoid locations, M1
and M2, which were used for off-line re-referencing; ACNS, 2006). The vertical
electro-oculogram (VEOG) was recorded with two active Ag/AgCl electrodes
located above and underneath the left eye. The horizontal electro-oculogram
(HEOG) was recorded with two Ag/AgCl electrodes located at the outer canthus
of each eye. An additional active electrode (CMS, common mode sense) and a
passive electrode (DRL, driven right leg) were used to comprise a feedback loop for
amplifier reference. All signals were digitized with a sample rate of 1,024 Hz and
24-bit A/D conversion with a low-pass filter of 134 Hz. Offline, the EEG signals were
referenced to the mathematically linked mastoids, and EEG and EOG were phaseshift-free filtered using a 0.1- to 40-Hz (24 dB/octave roll off) band-pass filter.
89
Chapter 5
Data reduction and analysis
EEG and EOG recordings were segmented in two 30-second epochs and corrected for
vertical and horizontal eye movements and eye blinks using the Gratton and Coles
algorithm (Gratton et al., 1983). After the EOG correction, we excluded segments
containing artifacts. The absolute difference between two values in a segment was
not allowed to exceed 200 μV. To analyze band power changes during exposure to
the smoking, each EEG epoch was divided in 30 1-second segments. For smokers
under the neutral condition, the mean number of artifact-free segments was 28.41.
For ex-smokers, this was 29.29. For smokers under the smoking condition, the
mean number of artifact-free segments was 29.18. For ex-smokers under the same
condition, this was 29.10. Each segment was Fast Fourier transformed using a
Hanning window of 10%. For each condition, the Fast Fourier transforms (FFTs) were
averaged and delta (0.75–3.75 Hz), theta (3.75–7.75 Hz), alpha-1 (7.75–10.75 Hz),
alpha-2 (10.75–13.75 Hz), alpha-total (7.75–13.75 Hz) and beta (13.75–29.75 Hz)
band power was measured. For each of these 6 frequency bands and each of the
32 electrodes, log-transformed [log (x)] power (μV2) was calculated.
Statistical analyses
The electrodes were divided into 4 clusters: left frontal (LF; Fp1, AF3, F7, F3, FC5),
right frontal (RF; Fp2, AF4, F8, F4, FC6), left posterior (LP; CP5, CP1, P3, PO3, O1),
and right posterior (RP; CP6, CP2, P4, PO4, O2; Dien & Santuzzi, 2005). For each
of the 6 frequency bands, a 4 (cluster) x 2 (condition: pen, cigarette) x 2 (group:
smokers, ex-smokers) repeated-measures analysis of variance (RM ANOVA) was
performed. Greenhouse-Geisser correction was applied when applicable.
To assess relationships between EEG activity changes and increase in craving as
a consequence of exposure to the cue, Spearman’s ρ coefficients were calculated
between the difference between the mean cluster activity in the smoking and
neutral condition (mean increase or decrease in cluster activity between the two
conditions) in each of the bands and the difference between the post- and preexposure QSU-brief scores. We selected Spearman correlation because of nonnormal distributed data of the QSU-brief. In order to measure left frontal alpha
asymmetry, two RM ANOVAs were performed. The first was a 2 (frontal clusters)
x 2 (condition) x 2 (group) RM ANOVA; the second was almost identical except for
the frontal clusters. In this analysis, frontal clusters were replaced by the specific
frontal electrodes F3 and F4, which are similar to the locations used by Zinser et
al. (1999). An alpha level of 0.05 was used for all statistical tests.
90
EEG spectrum changes
Results
Because our interest mainly concerned group differences, only group and groupinteraction effects are reported. Furthermore, and in order to reduce the number
of results, we report only significant (or borderline significant) effects.
EEG frequency bands
Beta exhibited a significant cluster x condition x group interaction effect,
F(3,123) = 4.37, p < 0.05. For smokers, post-hoc tests with Bonferroni adjustments
revealed a significant beta power increase between the neutral and smoking
condition at the left posterior cluster (p < 0.05). Additionally, this beta increase
was nearly significant at the left frontal and right posterior clusters (both p
values = 0.08). In contrast, for ex-smokers, no significant beta increase was
observed. However, at the right posterior cluster, a decrease in beta power almost
reached significance (p = 0.09). These results are confirmed by a follow-up t-test,
which showed a significant difference between smokers and ex-smokers for the
difference between the neutral and the smoking condition at both the left posterior
cluster, t(41) = –2.14, p < 0.05, and the right posterior cluster, t(41) = –2.50,
p < 0.05. To summarize, when exposed to smoking-related cues, smokers’ beta
increased, yet ex-smokers’ beta did not change (or even displayed a non-significant
tendency to decrease; figure 1). These group differences are found particularly at
posterior sites.
For the theta band, a cluster x condition x group effect was also found,
F(3,123) = 6.48, p < 0.01. However, post-hoc analyses revealed no significant
differences between groups or conditions. In addition, we computed group
differences on cue-induced change scores at each cluster using t tests. Again, this
did not result in significant effects. The significant interaction was probably caused
by differences between clusters dependent on condition and group. These effects
are beyond our scope of interest. No interactions were found between clusters,
conditions, and groups in the other EEG bands.
91
Chapter 5
ORJEHWDSRZHUGLIIHUHQFH
6PRNHUV
([VPRNHUV
/)
5)
/3
53
Figure 1. Differences between log beta power in response to the smoking cue and the
neutral cue for smokers and ex-smokers at left frontal (LF), right frontal (RP), left posterior
(LP), and right posterior (RP) clusters (including error bars).
Alpha asymmetry
When comparing activity at all left hemisphere electrodes to activity at all right
hemisphere electrodes, no significant differences are found between groups or
conditions for alpha-1, alpha-2, and alpha-total (all p values > 0.42). In addition,
when comparing activity at F3 (left) and F4 (right) electrodes, no significant alpha
differences are found either (all p values > 0.143).
Self-reported craving
A time (pre-QSU vs. post-QSU) x group ANOVA revealed significant differences
between groups on scores at pre-QSU, F(1,41) = 57.27, p < 0.001, and post-QSU,
F(1,41) = 59.68, p < 0.001. The difference between the mean post-exposure QSUbrief total score and the mean pre-exposure QSU-brief total score was significantly
larger for smokers (M = 6.0, SD = 6.7) than for ex-smokers (M = 1.8, SD = 4.2),
t(41) = 2.36, p = 0.05. This difference between smokers and ex-smokers was
mainly the result of smokers’ larger difference between pre- and post-experiment
scores on the first QSU subscale ‘desire and intention to smoke’, t(41) = 2.36,
p < 0.05. Smokers did not differ from ex-smokers on differences between the preand posttest on the second subscale ‘reduction of negative affect and withdrawal
symptoms’, t(41) = 0.85, p = 0.40. When analyzing groups separately, ex-smokers
report no increase in craving at all , t(20) = 1.98, p = 0.06, whereas smokers report
a strong increase, t(21) = 4.23, p < 0.001.
92
EEG spectrum changes
EEG and craving among smokers
Because ex-smokers did not show any increase in craving in response to the
smoking-related cue, we only examined the data of the smokers for the EEG craving
correlation analysis. Only correlations between QSU subscales, i.e., ‘desire and
intention to smoke’, and ‘reduction of negative affect and withdrawal symptoms’,
and mean increase or decrease in activity per frequency band will be reported
below. Increases in scores on the first subscale of the QSU (desire and intention)
are significantly correlated with a left posterior decrease in theta activity,
ρ = –0.51, p < 0.05. Furthermore, increases in scores on the second subscale of the
QSU (negative affect and withdrawal) are significantly correlated with left and
right frontal increases in delta activity, ρ = 0.45, p < 0.05 and ρ = 0.53, p < 0.05,
respectively, and a left posterior increase in higher alpha (alpha-2) activity,
ρ = 0.57, p < 0.01.
Conclusions
The present study investigated EEG spectrum changes in smokers and ex-smokers
in response to a neutral (pen) and a smoking-related cue (lit cigarette). In line with
our hypotheses, and in line with the results of Knott et al. (Knott et al., 2008),
a significant increase in beta power between the two conditions was observed in
the smokers. Ex-smokers did not show such an increase in beta power. If anything,
their beta tended to decrease (borderline significance) as a consequence of the
smoking cue. In general, beta power increments are thought to reflect increases in
cortical arousal (Neidermeyer, 1999) and have been associated with perception,
cognition, the orienting response, attentive-like behavior and increased activation
in an attentional alertness/vigilance network (Egner & Gruzelier, 2001; Egner
& Gruzelier, 2004; Haenschel, Baldeweg, Croft, Whittington, & Gruzelier, 2000;
Singer, 1993; Wrobel, 2000). In studies addressing addiction, beta power increases
have been associated with exposure to a cocaine cue, cocaine craving-related
guided imagery (Reid et al., 2003), exposure to smoking urge scripts (Knott et al.,
2008), and both actual nicotine (Domino, 2003; Teneggi et al., 2004) and cocaine
administration (Herning, Jones, Hooker, Mendelson, & Blackwell, 1985; Herning,
Glover, Koeppl, Phillips, & London, 1994; Reid, Flammino, Howard, Nilsen, &
Prichep, 2006).
93
Chapter 5
Although we cannot exclude the possibility that our beta power enhancement
merely reflects an arousal increment, the beta power enhancement may reflect
a processing bias for drug-related stimuli. Processing bias refers to the enhanced
processing of drug-related stimuli, which is a consequence of their strong
acquisition of incentive motivational properties (Robinson & Berridge, 2003).
During the course of drug use, drug-associated stimuli become extremely salient
and a greater proportion of attentional resources is allocated to them than to
other (rewarding) stimuli (Franken, 2003; Robinson & Berridge, 2003). Studies
investigating smoking cue reactivity by means of Stroop tasks, visual probe tasks
and the later components of event-related potentials (ERP: P300; late positive
potential) have shown a processing bias in smokers (e.g., Ehrman et al., 2002;
T. M. Gross et al., 1993; Warren & McDonough, 1999; Waters & Sayette, 2006), but
not in ex-smokers (Littel & Franken, 2007; Munafò et al., 2003). Because of the
association of beta with arousal and attention, the above-mentioned processing
bias might be in accordance with a beta increase in the processing of drug-related
cues. Egner and Gruzelier (2001; 2004) have demonstrated this relationship more
directly: beta training appears to be reliably correlated with the enhancement of
P300 ERP component amplitudes.
In the present study, the smokers’ increase in beta activity probably reflects their
enhanced processing of smoking-related cues, i.e., their processing bias. This idea
is strengthened by the topography of the beta increase, which is predominantly
posterior in nature and thus reflects arousal in temporal and occipital parts of
the brain. Activation of the visual cortex is in agreement with previous findings
that smokers, compared to non-smokers, maintain their gaze longer on smokingrelated cues than on neutral cues (Mogg et al., 2003), which is also considered an
indication of enhanced attention and processing bias. Since ex-smokers do not
show increases in beta activity, we can preliminarily conclude that this might
reflect an absence of processing bias in this population, which is in line with
results from previous studies (Littel & Franken, 2007; Munafò et al., 2003). The
smoking-related cue does not lead to the same amount of arousal and attention
in ex-smokers as it does in smokers. Oddly enough, arousal and attention tend to
decrease in ex-smokers. Although this result is not significant and should therefore
be interpreted cautiously, it might reflect some kind of avoidance mechanism, e.g.,
caused by disgust, or automated coping strategy, i.e., a cognitive distraction or
relaxation technique. Further research on this topic is necessary.
94
EEG spectrum changes
In contrast to the results of Knott et al. (2008), we found no significant changes in
delta and theta activity between conditions, nor did we find any differences between
the two groups. These differences between our results and those of Knott et al. may
have been caused by differences in the experimental manipulation. Knott et al.
obtained their recordings during imagery-elicited cravings, and not in response to
in vivo cues, where craving processes may be experienced differently. However, we
did find delta and theta frequency bands to be correlated with subjective craving.
Knott et al. (2008) suggested that reductions in delta and increments in beta, both
smoking-like EEG changes, and increases in theta, a withdrawal-like EEG change,
could be paralleled by neural substrates reflecting reward craving and neural
substrates reflecting withdrawal craving, respectively. As for theta and delta, just
about the opposite was observed. Increases in reward craving were significantly
correlated with reductions in left posterior theta activity. Furthermore, increases
in withdrawal craving were associated with increases in frontal delta activity.
This latter result is in line with results of Reid et al. (2006), who found that selfreported cocaine craving was correlated with delta power during the first 5 min
following cocaine self-administration. Nevertheless, correlations between selfreported craving and EEG activity were absent in most other studies (Liu, Vaupel,
Grant, & London, 1998; Reid et al., 2003). However, in contrast to the present study,
these studies concern cocaine addiction, cocaine paraphernalia and cocainerelated craving questionnaires. Differences in substances of abuse might cause
the inconsistencies in correlations between EEG activity and subjective craving.
There are no studies on nicotine addiction that have adequately addressed the
correlation between EEG oscillations and craving. In addition to reductions in theta
and increases in delta, we found increases in alpha-2 activity to be correlated with
craving. Specifically, increases in left posterior higher alpha were significantly
associated with increases in withdrawal craving. This is in contrast with results
from all other EEG frequency domain studies on exposure to drug cue and craving,
which reported either no relation (Knott et al., 2008; Reid et al., 2003; Reid et al.,
2006) or a borderline significant negative correlation (Liu et al., 1998) between
alpha power and craving. Finally, neither alpha-1 nor beta activity was associated
with self-reported craving.
Besides differences between smokers and ex-smokers in EEG spectrum changes
as a result of the presented cues, and a relation between these EEG changes
and craving, we hypothesized that smokers, compared to ex-smokers, would
display greater approach motivation to the smoking cue than to the neutral cue.
95
Chapter 5
We expected this to be reflected by a greater left frontal alpha asymmetry in
the smokers than in the ex-smokers. However, we did not find any alpha power
differences between groups or hemispheres. This is not in line with the results of
Zinser et al. (1999), who found EEG asymmetry to be increased as a consequence of
seeing a cigarette. Nevertheless, their study differed from ours in many ways. First
of all, smokers had been abstinent for 24 h, whereas smokers in our study had been
deprived for only 1 h. Second, smokers in the above-mentioned study were told
that they were allowed to smoke immediately after the experiment, whereas our
smokers did not receive any instructions, except for concentrating on the stimulus
at hand. Both deprivation and anticipation of drug use are thought to augment
drug motivation (e.g., T. M. Gross et al., 1993; Piasecki, Kenford, Smith, Fiore, &
Baker, 1997; Sayette & Hufford, 1994; Shiffman, Paty, Gnys, Kassel, & Hickcox,
1996; Wilson, Sayette, Delgado, & Fiez, 2005; Zinser, Baker, Sherman, & Cannon,
1992), and although we successfully manipulated craving levels, which also reflect
augmented drug motivation (Robinson & Berridge, 2003), the differences in
deprivation and anticipation might have had an influence on the absence of left
frontal alpha asymmetry in the present study.
A limitation of the current study is that the subjects performed a pictorial task first
(reported elsewhere, see Littel & Franken, 2007). This task, in which 16 different
smoking-related pictures and 16 neutral pictures were presented 4 times, may
have impacted the present measurements. Although the pen and the lit cigarette
were new, unexpected, multimodal (sight and smell) and more realistic than the
stimuli in the pictorial task, and although the presentation of these cues led to a
significant increase in subjective craving in smokers, the continuous presentation
of so many smoking cues may have caused some habituation, which may have
changed or reduced the effects. Research investigating the effects of repeated
drug exposure has shown that both physiological reactivity and self-reported
cue reactivity decrease over time (Marissen, Franken, Blanken, van den Brink, &
Hendriks, 2007), which makes it plausible that the EEG activity of smokers in the
current experiment could have been subject to habituation. However, both groups
were presented with the same cues and the same amount of cues, and in spite
of possible habituation, the significant group differences still remain. Another
limitation is that we did not counterbalance the order of the two cues across
participants. Counterbalancing was not possible because of the smoke and smell a
lit cigarette produces. During the presentation of the pen, the participants would
still have smelled the cigarette and the condition would not have been neutral
96
EEG spectrum changes
anymore. A third limitation is that we used self-report to validate ex-smoker status.
In the future, smoking status should be validated with a biochemical marker.
The main conclusion of the present study is that smokers, but not ex-smokers,
show an increase in beta activity in response to a smoking-related cue compared
to a neutral cue. Since activity in the beta frequency band has been associated with
heightened arousal, attention, alertness and enhancement of the P300 component
of the ERP, the increase in beta activity in the current study might reflect an
enhanced allocation of resources to smoking-related stimuli, i.e., a processing bias,
which was found in smokers in previous studies and is a very important feature of
substance abuse. Since ex-smokers do not respond to the smoking cue with beta
activity enhancement, we preliminarily conclude that smoking cues do not arouse
ex-smokers as much as they arouse smokers. Furthermore, we are the first to
establish the relationship between EEG spectrum changes in response to smokingrelated cues and self-reported cigarette craving by means of correlation analyses.
In smokers, reward craving (‘desire and intention to smoke’) is associated with
reduced theta activity. Withdrawal craving (‘reduction of negative affect and
withdrawal symptoms’), on the other hand, holds positive correlations with
increases in both delta and alpha-2 activity.
97
Chapter
6
Implicit and explicit selective
attention to smoking cues in smokers
indexed by brain potentials.
Littel, M. & Franken, I.H.A. (2011). Implicit and explicit selective
attention to smoking cues in smokers indexed by brain
potentials. Journal of Psychopharmacology, 25(4), 503-513.
Chapter 6
Abstract
Substance use disorders are characterized by cognitive processing biases,
such as automatically detecting and orienting attention towards drug-related
stimuli. However, it is unclear how, when and what kind of attention (i.e.,
implicit, explicit) interacts with the processing of these stimuli. In addition, it
is unclear whether smokers are hypersensitive to emotionally significant cues
in general or to smoking-related cues in particular. The present ERP study
aimed to enhance insight in drug-related processing biases by manipulating
attention for smoking and other motivationally relevant (emotional) cues in
smokers and non-smokers using a visual oddball task. Each of the stimulus
categories served as a target (explicit attention; counting) or as a non-target
(implicit attention; oddball) category. Compared to non-smokers, smokers’
P300 (350-600 ms) was enhanced to smoking pictures under both attentional
conditions. P300 amplitude did not differ between groups in response to
positive, negative, and neutral cues. It can be concluded from this study that
attention manipulation affects the P300 differently in smokers and nonsmokers. Smokers display a specific bias to smoking-related cues, and this bias
is present during both explicit and implicit attentional processing. Overall, it
can be concluded that both explicit and implicit attentional processes appear
to play an important role in drug-related processing bias.
100
Implicit and explicit attention in processing bias
Introduction
Drug use disorders are characterized by cognitive processing biases for drugrelated stimuli (for reviews, see Field et al., 2006; Field & Cox, 2008; Franken,
2003). It is hypothesized that drug users automatically detect and orient their
attention toward drug-related stimuli, which in turn diminishes attentional
resources left for alternative cues, enhances drug-related cognitions, and causes
subjective craving (Franken, 2003). These processes are thought to have mutual
excitatory relationships with each other. Consequently the drug user gets caught
in a vicious circle of increasing attention and craving. Both craving and attentional
bias have been associated with drug use and relapse (e.g., Killen & Fortmann, 1997;
Marissen et al., 2006).
The emergence of these processing biases can be explained by the incentivesensitization theory (Robinson & Berridge, 1993), which posits that repeated
administration of a drug causes a sensitization of dopaminergic neurotransmission
in the brain. Subsequently, both the drug itself and the drug-related stimuli acquire
incentive motivational properties. In other words, the sensitized dopaminergic
system causes the drug and drug-related stimuli to be perceived as particularly
salient, reinforcing, and ‘wanted’, which in turn leads to a greater allocation of
attentional resources to them. This hypothesis is confirmed in studies among
humans, which show less attention for drug-related stimuli in heroin users after a
single dose of the dopamine antagonist haloperidol (e.g., Franken, Hendriks, Stam,
& Van den Brink, 2004).
Research confirms that drug users exhibit an excessive attentional focusing on
drug-related cues. Utilizing attention tasks such as the emotional Stroop, dualtask procedures, the flicker-induced change blindness paradigm, and visual probe
and attentional cuing tasks, attentional bias has been demonstrated in various
drug use disorders, including smoking addiction (Ehrman et al., 2002; Field &
Cox, 2008; Mogg et al., 2003). For example, smokers are slower than non-smokers
to color name smoking-related words on the smoking Stroop task (Munafò et al.,
2003). Furthermore, smokers maintain their gaze on smoking stimuli longer than
on neutral stimuli (Mogg et al., 2003).
101
Chapter 6
Event-related potential studies of addiction and craving
A relatively new approach to assess the processing of drug-related stimuli, and
associated biases, is the measurement of Event-Related Potentials (ERP) using
electroencephalography (EEG) techniques. Two components of the ERP are of
particular interest in drug use research, i.e., the P300 and the related slow positive
wave. These components have been associated with attention allocation, intensity
of processing, the closure of perceptual events and activation of immediate memory
(Kok, 2001; Polich & Kok, 1995a). Furthermore, it is assumed that enhancement
of these late ERP components reflects motivational (emotional) engagement,
motivated attention, and the activation of arousal systems in the brain (Cuthbert
et al., 2000; Lang et al., 1997; Schupp et al., 2000).
ERP studies of visual processing in addiction show that these later ERP components
are more enhanced in drug users than in controls in response drug-related stimuli.
This result has been obtained in alcoholics (Herrmann et al., 2000; Herrmann
et al., 2001; Namkoong et al., 2004), heroin users (Franken, Stam et al., 2003;
Lubman et al., 2007; Lubman et al., 2008), cocaine users (Franken et al., 2008; Van
de Laar et al., 2004), cannabis users (Wölfling et al., 2008), and smokers (Littel
& Franken, 2007; McDonough & Warren, 2001; Warren & McDonough, 1999). In
all smoking cue-reactivity studies, a centro-frontally distributed enhancement of
P300 amplitude has been found in response to smoking cues relative to neutral
cues in smokers compared to non-smokers (Littel & Franken, 2007; McDonough &
Warren, 2001; Warren & McDonough, 1999). Littel and Franken (2007) found an
additional frontally distributed interaction effect on the slow positive wave (400750 ms), which is in accordance with results from studies among patients addicted
to other drugs (e.g., Franken et al., 2004; Van de Laar et al., 2004).
These ERP indices of processing biases are associated with subjective craving
(for reviews, see Field et al., 2006; Franken, 2003). Research repeatedly shows
that ERP waves, i.e., enhanced P300 and slow positive wave amplitudes, correlate
significantly with subjective drug craving (Franken, Stam et al., 2003; Franken et
al., 2004; Namkoong et al., 2004). A recent meta-analysis over all drugs of abuse
found an overall correlation of r = 0.37 between late positive waves (including
the P300 and slow positive wave) in passive viewing paradigms and self-reported
craving (Field et al., 2009). However, it must be noted that not all ERP studies of
addiction find correlations between processing bias and craving (Van de Laar
et al., 2004).
102
Implicit and explicit attention in processing bias
Focusing on smoking studies only, a correlation between ERP amplitudes and
craving for cigarettes is not unambiguously established. Warren and McDonough
(1999) failed to find such a correlation and Littel and Franken (2007) only found
a correlation between P300 amplitude at the Fz electrode and the first subscale
of the QSU-brief, ‘desire and intention to smoke’. In general, ERP measures of
processing bias are moderately associated with self-reported craving. This
association appears to be larger for illicit drugs compared to alcohol and tobacco
(Field et al., 2009).
To recapitulate, it has become clear from these studies that smokers and nonsmokers process smoking-related pictures differently. Because enhancement of
late ERP components is associated with the allocation of attentional resources to
motivational relevant stimuli (Cuthbert et al., 2000; Lang et al., 1997; Schupp et
al., 2000), and is moderately correlated with subjective craving (Field et al., 2009),
the enlarged P300 in the smoking studies is believed to be induced by the smokers’
allocation of attentional resources toward information relevant to their tobaccoaddicted, incentive-motivational states (Warren & McDonough, 1999). This would
be in accordance with the aforementioned theories of addiction (Franken, 2003;
Robinson & Berridge, 1993) and results from the majority of behavioral studies
employing paradigms like the Stroop and visual cuing tasks (Field & Cox, 2008).
Role of attention in ERP processing bias
However, all ERP smoking cue-reactivity studies used passive viewing paradigms
in which attention was not manipulated. Moreover, it is still unclear how, when and
what kind of attention (i.e., implicit, explicit) interacts with the electrophysiological
processing of drug-related stimuli in drug-dependent patients. As far as we know,
there have only been two studies that used ERP methodology outside a passive
viewing paradigm (Fehr et al., 2006; Fehr et al., 2007). Fehr and colleagues
presented smokers and non-smokers with a smoking-related Stroop task and a
smoking-related picture color matching task while measuring ERP. On both tasks,
smokers displayed a right frontal relative positivity in the P300 time frame that
appeared to be associated with cue interference, indicating a possible association
between P300 amplitude and attentional processing. Furthermore, Fehr et al.
(2006) showed a P100 modulation for verbal smoking-related stimuli, which might
indicate that smokers are affected by smoking-related stimuli during very early
stages of information processing. However, in addition to the smoking words and
pictures, Fehr et al. (2006; 2007) use ‘secondary smoking words and pictures’,
103
Chapter 6
such as bus stop, kiosk, for which it is unknown to what extent they affect cue
reactivity, task interference and/or craving in smokers. Moreover, non-smokers
also showed some interference effects -although at different electrode sites-, and
these effects were not exclusively elicited by smoking-related words and pictures.
To conclude, because the present focus and methodology fairly differs from the
focus and methodology used in the aforementioned smoking cue-reactivity
studies, it is difficult to make comparisons and draw conclusions regarding the
issue at hand, i.e., the exact role of attention in ERP processing bias.
Specificity of processing bias
Apart from this issue, it is also unclear whether drug users’ enhanced ERP response
is uniquely triggered by drug-related cues, i.e., whether there is a selective bias for
drug cues, or whether drug users are hyperresponsive to motivational relevant
stimuli in general, such as to positively or negatively valenced pictures with certain
arousing properties. For example, Stormark et al. (2000) found a greater Stroop
interference for negatively valenced words in alcohol-dependent patients compared
to healthy controls. In line with this, Bauer and Cox (1998) showed that differences
between alcoholics and controls in Stroop interference for alcohol-related words
disappeared when making use of affective control stimuli. Furthermore, cocaine
abusers with high craving levels displayed a more enlarged slow positive wave in
response to emotional valenced stimuli than low cocaine cravers (Franken et al.,
2004). In contrast, Lubman et al. (2008) demonstrated that heroin abusers only
displayed P300 processing biases for heroin-related cues. However, no differences
were found between P300 amplitudes in response to affective cues and neutral
cues, whilst the control group did show significant differences between these.
Instead of a hyperreactivity, these results would support a hyporeactivity to
emotional significant stimuli. Recently, Lubman et al. (Lubman et al., 2009)
replicated these findings. Using a variety of psychophysiological measures, they
convincingly showed that heroin users demonstrated reduced responsiveness
to natural reinforcers, i.e., pleasant stimuli. A plausible explanation for these
enhanced and decreased responses to emotional cues might be impaired affect
regulation, which is often linked to drug abuse (e.g., Thorberg & Lyvers, 2006).
Unfortunately, research on the processing of general emotional stimuli among
smokers is limited. It has been shown that nicotine administration (nicotine
patches) directly affects emotional processing in that amplitudes evoked by
emotionally negative pictures are enhanced compared to amplitudes evoked by
104
Implicit and explicit attention in processing bias
emotionally neutral and positive pictures (Gilbert et al., 2004). When employing a
difficult information processing task, nicotine decreases distraction by negative
and smoking-related stimuli and promotes attention to task-related stimuli
(Gilbert et al., 2007). Nevertheless, these results reflect the direct pharmacological
effects of nicotine intake, and cannot be generalized to cue-reactivity due to
smoking status.
The present study
Attention is thought to play a major role in smoking-related processing biases, but
it is not fully understood whether this role is implicit, explicit or both. In addition,
it is unclear whether smokers are hypersensitive to emotionally significant
cues in general or to smoking-related cues in particular. Therefore, the main
goal of the present study was to enhance insight in smoking-related processing
bias by manipulating attention (explicit or implicit) for smoking cues and other
motivationally relevant cues, i.e., positive and negative cues, in smokers and
non-smokers.
The relationship between attention and motivational significance was recently
studied by Schupp et al. (2007) utilizing a rapid and continuous stream of positive,
negative and neutral pictures, with each picture category serving as target and
non-target in separate series (oddball paradigm). Targets were explicitly attended
(silently counted); non-targets were assumed to be implicitly attended, since it is
widely believed that emotional stimuli are intrinsically significant and command
priority processing (Vuilleumier, 2005). It was demonstrated that explicit attention
and emotional significance effects operated additively on earlier processing stages
(early posterior negativity (EPN); 200-350 ms), but synergistically on later ERP
components (P300; 400-600 ms). In other words, the interaction of emotion and
attention appears to be merely present at later information processing stages.
The present study utilizes an adapted version of the abovementioned design of
Schupp et al. (2007). In order to investigate drug-related processing biases, we
added a third oddball category, i.e., a category of smoking-related pictures, and
a second group of participants, i.e., smokers. No passive viewing condition was
employed. Because in previous addiction research results have only been obtained
on later components of the ERP, and because we are mainly interested in the
abovementioned emotion-attention interaction, the focus of the present study will
be on the P300. Because Fehr et al. (2006) showed a P100 modulation for verbal
105
Chapter 6
smoking-related stimuli, the early ERP components (P100 and N100) will be
exploratively investigated.
Hypotheses
The main hypothesis of the current study is that smokers will display a processing
bias similar to the biases found in previous studies (Littel & Franken, 2007;
McDonough & Warren, 2001; Warren & McDonough, 1999). This bias will be
stronger for smoking cues than for general emotional cues and will be present
under both implicit and explicit attention conditions. The P300 will be larger for
smokers than non-smokers in response to smoking cues compared to positive,
negative and neutral cues. P300 amplitude will be larger for explicitly attended
stimuli. Yet, since it is hypothesized that attentional bias is at least partly implicit
in nature (e.g., Mogg et al., 2003), we also expect to find group differences and
differences between the stimulus types in the implicit attention condition.
Since there is evidence that attentional bias is associated with craving levels (Field
et al., 2006; Field et al., 2009; Franken, 2003), we assessed smokers’ subjective
craving scores before and after the task. It is hypothesized that craving levels will
increase between pre- and posttest and that this increase will be correlated with
P300 magnitude. Furthermore, the present study investigated the differences
between smokers and non-smokers on arousal and valence judgments of the
positive, negative and smoking-related pictures. Previous studies show that
smokers evaluate smoking-related pictures more positively than neutral stimuli
(Geier et al., 2000; Hogarth & Duka, 2006; Mogg et al., 2003), whereas non-smokers
evaluate them more negatively than neutral stimuli (Mogg et al., 2003). Because
positive, negative, and smoking pictures were matched on arousal levels, we expect
all pictures to be equally arousing for smokers. For non-smokers, we expect the
emotional stimuli to be more arousing. Correlations between arousal and valence,
CO level, nicotine dependence and P300 amplitude will be investigated in order to
receive more information on the factors that modulate the P300.
106
Implicit and explicit attention in processing bias
Method
Participants
Twenty-seven smokers and 27 non-smokers were recruited at the Erasmus
University Rotterdam (the Netherlands). They were all students and received
either course credit or financial compensation for participation. Smokers were
included if they smoked > 10 cigarettes a day. Smokers (mean age 23.3 years,
SD = 3.1 years) had a smoking duration of 7.1 years (SD = 3.0), smoked 15.1 cigarettes
a day on average (SD = 5.3), had a mean score of 3.8 (SD = 1.9) on the Fagerström
Test for Nicotine Dependence (FTND; Vink et al., 2005), and had a mean carbon
monoxide (CO) level of 12.5 parts per million (Ppm; SD = 7.5) at the time of testing.
Non-smokers (mean age 21.7, SD = 2.3) had smoked 2.6 (SD = 7.8) cigarettes in their
lifetime. They had a mean CO level of 1.0 Ppm (SD = 1.2) and differed significantly
from smokers on this last measure, t(52) = 7.85, p < 0.001. Smokers (33.3% male)
and non-smokers (29.6% male) did not differ on sex ratio, χ2(1) = 0.09, p = 1, and
the number of ambidextrous, right- and left-handed participants was equal in both
groups χ2(2) = 0.86, p = 0.65. The participants provided written informed consent
for the protocol approved by the institutional ethical board.
Experimental stimuli
Stimuli consisted of 150 neutral pictures, 22 positive pictures (animals),
22 negative pictures (garbage), and 22 smoking-related pictures (smoking-related
attributes, e.g. cigarettes and lighters, and people smoking). Oddball pictures were
selected from one category to prevent category effects. All of the neutral pictures,
all of the positive pictures, and six of negative pictures were selected from the
International Affective Picture System (IAPS; Lang, 1995). The other 16 negative
pictures were selected via internet search. The 22 smoking pictures were selected
from a database and were the same as those used in (Littel & Franken, 2007).
Previous studies indicate that smoking-related pictures are only moderately
arousing for smokers (e.g., Littel & Franken, 2007), so the positive and negative
pictures in this study could not be too arousing either. Instead of the erotic and
mutilation pictures that are usually adopted in studies of emotion, we chose to
present subjects with somewhat less arousing animal and garbage pictures. This
way we were able to match the arousal levels of the positive, negative and smoking
stimuli (4.5, 4.5 and 4.6, respectively), and thus to control for the effects of nonspecific arousal on ERP amplitude.
107
Chapter 6
To make sure that our positive pictures were more positively valenced than
our negative pictures, the most positively valenced positive (M = 7.3) and most
negatively valenced negative pictures (M = 3.2) were selected from the IAPS.
Participants rated all pictures on arousal and valence properties.
Procedure
Smokers were instructed to abstain from smoking for at least one hour in order
to avoid direct effects of nicotine on task performance and ERP signals. They
were told that this would be checked with a smoke analyzer. Subjects were tested
alone in a light and sound-attenuated room. After obtaining written informed
consent, subjects proceeded to a non-invasive CO Ppm estimate using the EC50
Micro III Smokerlyzer® (Bedfont Scientific, Medford, NJ, USA), a portable device
which measures breath carbon monoxide levels. In addition, subjects filled out
questionnaires about demographics, smoking history, cigarette craving (smokers)
and smoking dependence (smokers). After completion, participants were seated
in a comfortable chair and electrodes were attached. Then the task was explained
and instructions were given.
The experiment consisted of three separate stimulus conditions. In each condition,
subjects were asked to silently count the (1) animal, (2) garbage, or (3) smoking
pictures. The order of the three stimulus conditions was counterbalanced across
subjects. Within each condition, pictures from every category, including the neutral
category, were repeated three times, resulting in 66 animal, 66 garbage and 66
smoking pictures (132 not counted/ implicitly attended; 66 counted/ explicitly
attended) per condition. Pictures were presented for 333 ms in a continuous stream
without perceivable inter-stimulus intervals. This fast-stimulus presentation
procedure was adopted from Schupp et al. (2007) and served to enhance attention
for the stimuli by increasing perceptual demands and make the identification of
target stimuli more challenging. The pictures were presented in a perceptually
random order. However, there were no successions of two or more targets or nontargets (figure 1).
108
Implicit and explicit attention in processing bias
1HXWUDO
PV
1HXWUDO
PV
1HXWUDO
PV
6PRNLQJ
&
PV
1HXWUDO
PV
1HXWUDO
PV
3RVLWLYH
PV
1HXWUDO
Figure 1. Study design. Participants were presented with three blocks of frequent neutral
pictures and infrequent (oddball) smoking, animal and garbage pictures, all presented for
333 ms. In each block they had to count pictures from one of the three categories (C).
At irregular intervals, the stream of pictures was stopped and subjects were asked
to report the number of target pictures they had identified. They had to make a
choice between four on-screen options by pressing a corresponding button.
Participants immediately received feedback (correct or incorrect). All answers
were recorded. The tests and test intervals were the same for all participants.
After the picture viewing, electrodes were removed and smokers filled out the
craving questionnaire for the second time. Subsequently, all participants rated the
pictures on their valence and arousal properties. Both for stimulus presentation
and valence and arousal judgments e-prime® software (Psychology Software
Tools, Pittsburgh, PA, USA) was used.
Self-report measures
Smoking history and demographic data were self-reported (sex, age, smoking
duration, number of cigarettes a day). Handedness was measured with a 10-item
Dutch handedness questionnaire, i.e., the “Vragenlijst voor handvoorkeur” (Van
Strien, 1992), which has been shown to have excellent reliability.
Smoking dependence was measured by the Dutch version of the Fagerström Test
for Nicotine Dependence (FTND; Vink et al., 2005), which has good reliability and
holds a significant correlation with number of cigarettes smoked per day. The
109
Chapter 6
FTND is composed of six items, which are scored according to the scoring system
described in Heatherton et al. (1991).
Subjective craving was assessed by means of the QSU-brief (L. S. Cox et al.,
2001; Littel, Franken, & Muris, 2011). This questionnaire was adapted from the
Questionnaire on Smoking Urges (QSU; Tiffany & Drobes, 1991) and consists of
two subscales: ‘desire and intention to smoke’ (reward-craving) and ‘reduction of
negative affect and withdrawal craving’ (withdrawal-craving). The QSU-brief and
its subscales have adequate psychometric properties (Cox et al., 2001).
Arousal and valence properties of the positive, negative, and smoking-related
pictures were assessed by a computerized Self Assessment Manikin (SAM; M. M.
Bradley & Lang, 1994), which is a non-verbal pictorial assessment technique that
directly measures the pleasure and arousal associated with a person’s affective
reaction to stimuli. The arousal scale ranged from a relaxed, sleepy figure to an
excited, wide-eyed figure; the valence scale ranged from a frowning, unhappy
figure to a smiling, happy figure.
Electroencephalogram (EEG) recording and signal processing
The electroencephalogram (EEG) was recorded with a digital Active-Two system
(BioSemi, Amsterdam, the Netherlands), using active Ag/AgCl electrodes at
34 scalp sites according to the International 10/10 system (32 standard channels
mounted in an elastic cap and two mastoid locations, which were used for off-line
re-referencing; ACNS, 2006). The vertical electro-oculogram (VEOG) was recorded
with two active Ag/AgCl electrodes located above and underneath the left eye. The
horizontal electro-oculogram (HEOG) was recorded with two Ag/AgCl electrodes
located at the outer canthus of each eye. An additional active electrode (CMS –
common mode sense) and a passive electrode (DRL – driven right leg) were used
to comprise a feedback loop for amplifier reference. All signals were digitized with
a sampling rate of 512 Hz, a 24-bit A/D conversion, and a low pass filter of 134 Hz.
Offline, data was processed with BrainVision Analyzer 2 (Brain products GmbH,
Munich, Germany).
First of all, the EEG signals were referenced to the mathematically linked mastoids
and EEG and EOG were phase-shift-free filtered using a 0.1–30 Hz (24 dB/Octave roll
off) band-pass filter. EEG and EOG recordings were segmented in 800 ms epochs,
including 100 ms pre-stimulus baseline. For correction of vertical and horizontal
110
Implicit and explicit attention in processing bias
eye movements and eye blinks we applied automatic processing algorithms, i.e.,
Gratton and Coles algorithm (Gratton et al., 1983). After ocular correction, the ERPs
were baseline corrected. Artifact rejection criteria were minimum and maximum
baseline-to-peak −75 to +75 μV, and a maximum allowed voltage skip (gradient) of
50 μV. Epochs were averaged across trials. Number of artifact-free epochs did not
differ between groups and stimulus conditions (smoking-explicit: smokers M = 63;
non-smokers M = 64, positive-explicit: smokers M = 64; non-smokers M = 65,
negative-explicit: smokers M = 63; non-smokers M = 64, smoking-implicit: smokers
M = 127; non-smokers M = 129, positive-implicit: smokers M = 127; non-smokers
M = 129, negative-implicit: smokers M = 127; non-smokers M = 129, all p’s ns).
Overall grand averages were obtained for each attention condition and picture
category in the two groups, yielding six conditions per group (smoking-explicitly
attended; smoking-implicitly attended/ oddball; positive-explicit; positiveimplicit; negative-explicit; negative-implicit).
Analyses
Resulting ERP-waves were visually inspected and both a N100 (maximum negative
peak in the time window from 50 - 80 ms) and a P100 (maximum negative peak in
the time window from 110 - 150 ms) were identified. In contrast to (Schupp et al.,
2007), no clear EPN could be observed. Most important, in the 350 - 600 ms time
window a clear P300 was identified. For each component, mean activities (average
amplitude in the time-window) were computed per group, attention and stimulus
category.
Because of the short stimulus presentations and the absence of inter stimulus
intervals, P300 waveforms overlapped with waveforms of the following stimuli,
resulting in somewhat deviant amplitude values. Nevertheless, it is unlikely that
this has confounded the results. First of all, positive, negative, and smoking pictures
never appeared in succession, but were always followed by neutral pictures with
low arousal and moderate valence levels. Accordingly, neither the attention nor the
stimulus effect is likely to be contaminated by systematic differences in emotional
valence of the subsequent stimuli. Secondly, the P300 appeared with similar
polarity, topography and latency as in previous studies (see Schupp et al., 2007,
and figures 2 and 3).
111
Chapter 6
For the P300, ERP effects were assessed by performing repeated-measurement
analyses of variance (ANOVA) on crossed lateral and caudal sites, including
15 electrodes (F7, F3, Fz, F4, F8, T7, C3, Cz, C4, T8, P7, P3, Pz, P4, P8), resulting in
a 5 (laterality) x 3 (caudality) x 2 (attention) x 3 (stimulus) x 2 (group) repeated
measures ANOVA. Since N100 and P100 components are predominantly present
at posterior electrodes (PO3, O1, Oz, O2, and PO4), two 5 (electrode site) x 2
(attention) x 3 (stimulus) x 2 (group) repeated measures ANOVAs were conducted
for these components.
Arousal and valence ratings of the pictures and results of the counting task were
analyzed using three 3 (stimulus) x 2 (group) repeated-measurement ANOVA’s. To
examine exact differences for the significant group, stimuli, and attention condition
interactions, pairwise post-hoc follow-up analyses with Bonferroni correction
were applied to all ANOVAs. Greenhouse-Geisser correction was applied to all
ANOVAs (uncorrected df’s are reported).
To determine whether craving was significantly increased after picture viewing, a
paired t-test was performed (pre- versus posttest craving). To assess relationships
between cue-evoked ERP amplitudes, self-reported craving, CO level, nicotine
dependence level, and valence/arousal assessments, Spearman correlation
coefficients were calculated between significant ERP amplitudes, increases in
craving between pre- and post-measure, CO measures, FTND score, and valence/
arousal judgments. An alpha-level of 0.05 was used for all statistical tests.
112
Implicit and explicit attention in processing bias
6PRNLQJSLFWXUHV±([SOLFLWDWWHQWLRQ
VPRNHUVYVQRQVPRNHUV
)]
)]
&]
&]
3]
3]
6PRNLQJSLFWXUHV±,PSOLFLWDWWHQWLRQ
VPRNHUVYVQRQVPRNHUV
)]
)]
&]
&]
3]
3]
Figure 2. Average event-related potentials (ERPs) at Fz, Pz and Cz for smokers (grey) and
non-smokers (black) in response to explicitly and implicitly attended smoking stimuli.
6PRNLQJSLFWXUHVVPRNHUVYVQRQVPRNHUV
)]
)]
&]
&]
3]
—9
3]
1HJDWLYHSLFWXUHVVPRNHUVYVQRQVPRNHUV
)]
)]
&]
&]
3]
3]
3RVLWLYHSLFWXUHVVPRNHUVYVQRQVPRNHUV
)]
)]
&]
&]
3]
PV
SUHGHILQHGWLPH
ZLQGRZIRU
DPSOLWXGHDYHUDJLQJ
QRQVPRNHUV
VPRNHUV
3]
Figure 3. Average event-related potentials (ERPs) at Fz, Cz and Pz for smokers (grey) and
non-smokers (black) in response to smoking, negative and positive stimuli.
113
Chapter 6
Results
Behavioral and self-reported data
Counting task
On the counting task, no S x G interactions were found, F(2,104) = 0.14, p = 0.87,
indicating that smokers and non-smokers counted stimuli from the positive,
negative, and smoking-related stimulus conditions equally well.
CO level and nicotine dependence
Smokers had a mean CO level of 12.5 parts per million (Ppm; SD = 7.5) at the time
of testing. Non-smokers had a mean CO level of 1.0 Ppm (SD = 1.2) and differed
significantly from smokers, t(52) = 7.85, p < 0.001.
Craving
QSU-score increased significantly between the first measure (before the task;
M = 18.19, SD = 13.70) and the second measure (after the task; M = 40.44,
SD = 13.73), t(26) = 2.71, p < 0.05. This effect appeared to be driven by the increase
in scores on the first subscale, ‘desire and intention to smoke’, t(26) = 2.78, p < 0.05.
There was no increase in scores on the second subscale, ‘the relief from nicotine
withdrawal or negative affect with an urgent and overwhelming desire to smoke’,
although there was a trend to significance, t(26) = 1.91, p = 0.07.
Arousal and valence
On both arousal and valence judgments significant S x G interactions were found,
respectively F(2,104) = 10.48, p < 0.001 and F(2,104) = 21.49, p < 0.001. Smokers
rated smoking pictures as significantly more arousing than non-smokers,
t(52) = 3.01, p < 0.01. They also found smoking pictures more positive than nonsmokers, t(52) = 6.42, p < 0.001. Groups did not differ on valence and arousal
judgments of the positive and negative pictures (all p’s > 0.25). As intended, there
was no difference within the smokers group between arousal of smoking, negative,
and positive cues (all p’s > 0.22). Non-smokers also found positive pictures as
arousing as negative pictures, t(26) = 1.19, p < 0.001. However, smoking pictures
were rated by non-smokers as less arousing than positive and negative cues,
respectively t(26) = 4.25, p < 0.001 and t(26) = 6.19, p < 0.001.
Smokers rated positive pictures more positively than smoking cues and negative
cues, respectively t(26) = 2.51, p < 0.05 and t(26) = 14.21, p < 0.001. Negative
114
Implicit and explicit attention in processing bias
pictures were rated more negatively than smoking pictures, t(26) = 8.92, p < 0.001.
The same pattern was observed in non-smokers: smoking pictures were more
positive than negative pictures, t(26) = 4.63, p < 0.001, but more negative than
positive pictures, t(26) = 12.78, p < 0.001, and positive cues were rated as more
positively than negative cues, t(26) = 15.13, p < 0.001. See table 1 for all mean
arousal and valence ratings.
Table 1. Mean self-reported arousal and valence ratings (SD) of the smokers and nonsmokers
Arousal
Valence
Positive
Negative
Smoking
Positive
Negative
Smoking
Electrophysiological data
Smokers
3.6 (1.5)
3.4 (1.3)
3.8 (2.2)
6.8 (0.9)
2.4 (1.1)
5.9 (1.4)
Non-smokers
3.6 (1.7)
4.0 (2.0)
2.3 (1.3)
6.8 (0.9)
2.6 (1.1)
3.6 (1.2)
P300
On the P300 wave, a significant Stimulus (S) x Group (G) interaction effect was
found, F(2,104) = 3.36, p < 0.05. Post-hoc comparisons revealed that smokers and
non-smokers did not differ on P300 amplitude in response to positive stimuli
(p = 0.15) and negative stimuli (p = 0.54). However, smokers’ P300 response to
smoking-related pictures was significantly larger than that of non-smokers
(p < 0.01). See figure 3 for P300 amplitudes in response to smoking, positive, and
negative cues.
Furthermore, a significant S x G x Attention (A) interaction was observed,
F(2,104) = 3.04, p = 0.05. Post-hoc analyses showed that the aforementioned
significant interaction between smokers and non-smokers of P300 amplitude to
smoking-related pictures was present in both the implicit and the explicit attention
condition (respectively p = 0.035 and 0.003). See figure 2 for P300 amplitudes
to implicitly and explicitly attended smoking cues. In neither of the attention
conditions, smokers and non-smokers differed in their P300 response to negative
and positive stimuli (all p’s ns).
Additionally, a significant Lateral (L) x S x G interaction effect was found. Smokers
displayed a significantly more enhanced P300 amplitude in response to smoking
115
Chapter 6
cues than non-smokers on all five lateral clusters (F7, T7, P7, p = 0.05; F3, C3, P3,
p = 0.004; Fz, Cz, Pz, p = 0.002, F4, C4, P4, p = 0.002; and F8, T8, P8, p = 0.036).
On neither of the lateral clusters, smokers and non-smokers differed in P300
amplitude elicited by positive and negative cues (all p’s ns).
In addition to the group effects, a significant main effect for Stimulus,
F(2,104) = 69.42, p < 0.001, a significant main effect for Attention, F(1,52) = 238.90,
p < 0.001, and a significant A x S interaction, F(2,104) = 55.66, p <0.001, was found.
P300 amplitude in response to negative pictures was smaller than P300 amplitude
in response to smoking and positive pictures (both p’s < 0.001). Furthermore, P300
in response to explicitly attended stimuli was more enhanced than in response to
implicitly attended stimuli (p < 0.001). In the explicit attention condition, P300
responses to all stimuli differed from each other (smoking > positive > negative; all
p’s < 0.01), whereas in the implicit attention condition responses to negative and
smoking-related pictures did not (positive > negative, smoking; ns).
Early components
In contrast to the P300, neither on the P100 peak, nor on the N100 peak, group
interaction effects were found, all p’s ns. On the P100, both a significant Stimulus
effect, F(2,104) = 5.37, p < 0.05, and a significant Attention effect, F(1,52) = 10.34,
p < 0.01, were found. Post-hoc comparisons revealed that positive stimuli elicited
larger P100 amplitudes than smoking stimuli (p < 0.05) and that there was a trend
for negative stimuli to elicit more positive P100 amplitudes than smoking stimuli
(p = 0.06). There was no difference between P100 in response to negative cues and
P100 in response to positive cues. The P100 amplitude in response to implicitly
attended stimuli appeared to be larger than in response to explicitly attended
stimuli (p < 0.01).
On the N100 peak, both a significant Stimulus effect, F(2,104) = 35.97, p < 0.001,
and a significant Attention effect were found, F(1,52) = 37.67, p < 0.001. Posthoc tests showed that both negative and smoking stimuli evoked larger N100
amplitudes than positive stimuli (both p’s < 0.01), but did not differ from each
other. Furthermore, explicitly attended stimuli elicited larger N100 amplitudes
than implicitly attended stimuli (p < 0.001). Furthermore, a significant A x S
interaction effect, F(2,104) = 19.72, p < 0.01 was observed. The enlargement of
the N100 evoked by negative stimuli did not differ between the implicit and
explicit attention condition, whereas the N100 in response to positive and
116
Implicit and explicit attention in processing bias
smoking pictures was more enlarged in the explicit than in the implicit condition
(both p’s < 0.001).
Correlations
Nicotine dependence and CO level
In smokers, CO level correlated significantly with P300 amplitude to explicitly
attended smoking cues on electrodes F4 (ρ = 0.40, p < 0.05) and F8 (ρ = 0.47, p <
0.05), indicating that more enhanced CO levels are related to more enhanced rightfrontal P300 amplitudes in response to explicitly attended smoking stimuli. No
correlations were found between FTND score and P300 in response to smoking,
positive or negative stimuli.
Craving
Increases on the QSU-brief were negatively correlated with the P300 to explicitly
attended smoking cues on electrodes Pz (ρ = -0.38, p < 0.05) and F8 (ρ = -0.39, p <
0.05), indicating that decreases in subjective craving are related to more enhanced
right-central/parietal P300 waves in response to explicitly attended smoking
stimuli. No other significant correlations between ERP amplitude and subjective
craving were found.
Arousal and valence
In the smokers group analyses revealed no correlations between arousal and
valence judgments of the smoking pictures and ERP amplitude in response to the
smoking pictures.
Discussion
The main goal of the present study was to examine smoking-related processing
bias by manipulating attention (i.e., explicit versus implicit conditions) for smoking
cues and other motivationally relevant cues (i.e., positive and negative cues) in
smokers and non-smokers. It was hypothesized that in both attention conditions
the P300 would be larger for smokers than non-smokers in response to smoking
cues compared to positive, negative and neutral cues.
This hypothesis is confirmed by the results of the present study. P300 amplitude
to smoking-related cues was more enhanced in smokers than in non-smokers,
117
Chapter 6
irrespective of attention condition. This implies that smokers display a processing
bias that is similar to biases observed in previous smoking studies (Littel &
Franken, 2007; McDonough & Warren, 2001; Warren & McDonough, 1999) and
other addiction studies (Franken, Stam et al., 2003; Franken et al., 2008; Herrmann
et al., 2000; Herrmann et al., 2001; Lubman et al., 2007; Lubman et al., 2008;
Namkoong et al., 2004; Van de Laar et al., 2004; Wölfling et al., 2008).
Moreover, this processing bias is present during both implicit and explicit attention.
The results show that when smoking-related stimuli are presented as oddballs in a
continuous stream of neutral stimuli, they automatically attract smokers’ attention
to a greater extent than non-smokers’ attention. So, even when the smokers
are instructed to pay attention to non-smoking cues, they automatically and
unintendedly pay attention to smoking-related stimuli. In addition, if instructions
are to explicitly pay attention to and count the smoking-related stimuli, smokers
do this in a more elaborate and/ or motivated way than non-smokers.
More important, smokers and non-smokers did not differ in P300 amplitude to
positive and negative stimuli in general, confirming the hypothesis that smokingrelated processing bias is very selective and specific and is not caused by some
sort of hyperresponsivity to motivationally relevant stimuli in general. This is
partly in line with a study among heroin users (Lubman et al., 2008), showing that
heroin users only exhibit ERP processing biases for heroin-related cues. However,
in contrast to this study and the Lubman et al. (2009) study, no hyporeactivity to
emotional stimuli was found either. Smokers appear to respond normal to general
motivationally relevant stimuli.
Early components of the ERP
In line with the majority of addiction ERP studies, no differences were found
between smokers and non-smokers on the P100 and N100 components of the ERP,
indicating that there are no differences between these groups with regard to early
oriented attention for implicitly or explicitly attended smoking-related stimuli.
This is in contrast with the results of Fehr et al. (Fehr et al., 2006). Smokers in
their study showed increased P100 and N100 amplitudes in response to Stroop
interference caused by smoking-related words. The discrepancy between the
studies might be explained by the fact that the present study did not comprise
any interference, i.e., there was no information to be consciously ignored during
neither attention conditions. Furthermore, our study did not comprise any
118
Implicit and explicit attention in processing bias
reading, because our experiment consisted of pictures instead of (primary and
secondary) words.
Behavioral measures
Although it has been demonstrated that ERP processing biases are related to
subjective craving (Field et al., 2009), in smokers no clear-cut relation has been
found yet. In the present study craving increased between pre- and post-task
measures, but this increase correlated negatively with P300 amplitude on two
right-central/parietal electrode sites. Although this is difficult to explain from a
theoretical point of view, it might be that attending to the smoking cues in this
paradigm is quite difficult and associated with increased cognitive efforts and
therefore reduced craving.
Perhaps craving for cigarettes cannot be compared to craving for other, illicit
drugs, for which the correlation with ERP amplitude is clearer and always positive
(Field et al., 2009). Period of abstinence is considerably shorter (1-2h compared
to >2 weeks; e.g., Lubman et al., 2008), cigarettes are evidently more readily
available than illicit drugs, and in smoking addiction both the pleasurable effects
and withdrawal symptoms are of less relevance.
Arousal and valence ratings of the smoking-related pictures did not correlate with
P300 amplitude to implicitly or explicitly attended stimuli. In contrast, CO level
correlated with P300 amplitude to explicitly attended smoking stimuli on several
frontal electrode sites. Higher CO levels are related to more enhanced right-frontal
P300 amplitudes in response to explicitly attended smoking stimuli.
To summarize, smokers display an increase in craving between the first measure
(before the task) and the second measure (after the task). Besides, they find
smoking pictures more positive and more arousing than non-smokers. In contrast
with several studies on illicit drugs (e.g., Franken, Stam et al., 2003; Lubman et al.,
2008), but in line with studies in smokers (e.g., Warren & McDonough, 1999), these
measures do not have an unequivocal relationship with ERP responses. However,
CO level and nicotine dependence level appear to have some relation to frontal
ERP responses, but only in the explicit attention condition, indicating that more
severe smokers might process smoking-related stimuli in a more elaborate and/ or
motivated way than lighter smokers, whereas it is possible that they do not differ
in their automatic and unintended attention to smoking-related stimuli.
119
Chapter 6
Limitations
In the present study the emotional stimuli had relatively moderate levels of
valence. This can be interpreted as both a strength and a weakness. Because we
were able to match arousal levels, the P300 differences we found could not be
ascribed to arousal differences. However, the negative pictures elicited amplitudes
substantially smaller than what is common in ERP studies of emotion. This might
have been caused by the fact that some of the pictures in the negative (garbage)
category were more difficult to recognize and categorize than pictures in the
positive (animal) and smoking categories. After all, garbage is a broadly based
concept that includes dirt, trash, rubbish bags, litter bins etc. It is possible that not
all garbage pictures in both the explicit and implicit attention condition actually
captured attention. However, garbage pictures elicited enhanced P100 and N100
amplitudes and were rated significantly more negative than positive and smoking
pictures, providing support for the suitability of the present research design.
Another explanation for the reduced amplitudes to negative pictures is that the
negative pictures were inanimate, whereas the positive pictures were animate.
However, smoking pictures were both animate and inanimate, but still yielded the
largest ERP effects.
The short stimulus presentations in combination with the absence of inter
stimulus intervals caused the P300 waveforms to overlap with the waveforms
of subsequent stimuli. This resulted in divergent P300 amplitudes and makes it
difficult to directly compare our study to other studies. Nevertheless, it is unlikely
that it has confounded the (group) effects. First of all, positive, negative, and
smoking pictures never appeared in succession, but were always followed by
neutral pictures with low arousal and average valence levels. Accordingly, neither
the attention nor the stimulus effect is likely to be contaminated by systematic
differences in emotional valence of the subsequent stimuli. Moreover, the P300
appeared with similar polarity, topography and latency as in previous studies (see
Schupp et al., 2007, and figures 2 and 3).
Of course it is possible that because of the fast stimulus presentations and the
absence of perceivable ISIs participants elaborated less on the pictures. However,
we had several reasons to present the stimuli this way. First of all, we wanted to
enhance attention for the stimuli by increasing perceptual demands and making
the counting more challenging. Secondly, we wanted the implicit stimuli to be as
implicit as possible, but still visible and not different from the explicit stimuli.
120
Implicit and explicit attention in processing bias
Finally, we didn’t want the task to last too long to prevent participants from
getting bored and drowsy. In addition, we wanted to adopt the procedure which
was published by Schupp et al. (2007), and which turned out to be an adequate
method to investigate implicit and explicit attention.
It should be noted that in the present study only the explicit attention condition,
and not the implicit condition, calls upon working memory capacity because of the
intermediate storage and rehearsal of counted numbers in short term memory. This
might have interacted with category-related picture processing in the counting but
not in the pure oddball task. However, if working memory capacity interacts with
picture processing, this would very likely be the case in both smokers and nonsmokers and does probably not account for the ERP differences we found between
the groups on both the implicit and explicit processing of smoking pictures.
Another point that should be noted in future research is that data on number of
cigarettes smoked before testing as well as time to the last cigarette were not
questioned, whereas these variables might co vary with cue reactivity and craving.
Conclusion
The current ERP study is the first to demonstrate that smokers display a processing
bias that cannot be attributed to hyperreactivity to motivationally relevant cues in
general or to hyporeactivity to emotional cues, but is specific to smoking-related
cues. Moreover, this is the first ERP study in which smokers’ attention for smoking
cues is manipulated and it can be concluded that processing bias is present in both
explicit and implicit attentional processing. Smokers display both an implicit and
explicit attentional bias to smoking cues in particular.
Concerning the societal impact, these results emphasize that enlarged P300
amplitudes in response to both implicitly and explicitly presented drug cues may
provide an indicator of important psychological mechanisms relevant to addiction.
Therefore, future research should also focus on the possibilities to change drugrelated implicit and explicit attentional biases.
121
Chapter
7
Electrophysiological correlates of
associative learning in smokers:
a higher-order conditioning experiment
Littel, M. & Franken, I.H.A. (2012). Electrophysiological
correlates of associative learning in smokers: A higherorder conditioning experiment. BMC Neuroscience, 13(8).
Chapter 7
Abstract
Classical conditioning has been suggested to play an important role in the
development, maintenance, and relapse of tobacco smoking. Several studies
have shown that initially neutral stimuli that are directly paired with smoking
are able to elicit conditioned responses. However, there have been few human
studies that demonstrate the contribution of higher-order conditioning to
smoking addiction, although it is assumed that higher-order conditioning
predominates learning in the outside world. In the present study a higherorder conditioning task was designed in which brain responses of smokers
and non-smokers were conditioned by pairing smoking-related and neutral
stimuli (CS1smoke and CS1neutral) with two geometrical figures (CS2smoke and
CS2neutral). ERPs were recorded to all CSs. Data showed that the geometrical
figure that was paired with smoking stimuli elicited significantly larger
P2 and P3 waves than the geometrical figure that was paired with neutral
stimuli. During the first half of the experiment this effect was only present in
smokers whereas non-smokers displayed no significant differences between
both stimuli, indicating that neutral cues paired with motivationally relevant
smoking-related stimuli gain more motivational significance even though they
were never paired directly with smoking. These conclusions are underscored
by self-reported evidence of enhanced second-order conditioning in smokers.
It can be concluded that smokers show associative learning for higher-order
smoking-related stimuli. The present study directly shows the contribution of
higher-order conditioning to smoking addiction and is the first to reveal its
electrophysiological correlates. Although results are preliminary, they may
help in understanding the etiology of smoking addiction and its persistence.
124
Higher-order conditioning of processing bias
Introduction
Classical conditioning has been suggested to play an important role in the
development, maintenance, and relapse of drug use (e.g., Davis & Gould, 2008;
Hyman, 2005; O’Brien, O’Brien, Mintz, & Brady, 1975; Poulos et al., 1981;
Torregrossa, Corlett, & Taylor, 2011). Classical conditioning theory predicts that
with repeated drug use, drug-related stimuli or contexts (conditioned stimuli,
CS) become associated with drug intake (unconditioned stimulus, UCS), and
consequently, in the course of time, these stimuli acquire motivational significance
and evoke conditioned drug responses or cue reactivity, such as subjective
craving, drug seeking behaviors, or changes in physiological measures (e.g.,
Carter & Tiffany, 1999; Drummond, 2000; Drummond, Litten, Lowman, & Hunt,
2000; Thewissen et al., 2005). Once the learning process has taken place and the
CS are able to elicit the conditioned drug responses, the CS can be paired with
new neutral stimuli or contexts, which will also acquire associative strength and
elicit conditioned drug responses or cue reactivity. This process is called secondorder conditioning (higher-order conditioning; CS-CS learning) and can lead to
unlimited sequences of associations that presumably contribute to drug-seeking
in real world environments (Everitt, Dickinson, & Robbins, 2001; Gewirtz & Davis,
2000; Schindler, Panlilio, & Goldberg, 2002).
Classical conditioning requires the storage of a neural representation of the
associations between conditioned incentive stimuli and the conditioned responses
they elicit. These associations are both reflexive and cognitively-mediated, and,
accordingly, multiple neural mechanisms of learning and memory seem to be
involved (White, 1996). Potential substrates are the amygdala, which is thought to
be implicated in emotional processing involving discrete cues; the hippocampus,
which is assumed to play a major role in contextual learning; the striatum, which
mediates procedural and habit learning; and cortical systems such as the anterior
cingulate cortex and the orbitofrontal cortex that have more regulatory and general
information processing functions (Robbins, Ersche, & Everitt, 2008). Under normal
circumstances, these neural substrates are involved in behaviors that are needed
for survival, such as obtaining food, sex, and other natural rewards. However, after
repeated drug use they are recruited or ‘hijacked’ by the drugs of abuse, producing
maladaptive behavioral and cellular changes that maintain addiction (Davis &
Gould, 2008; Hyman, 2005).
125
Chapter 7
A variety of animal studies provides evidence for the role of associative learning in
drug use. With regard to nicotine addiction, animals increase self-administration
of nicotine in the presence of stimuli that were previously paired with nicotine
administration and they display preferences for contexts that were previously
paired with nicotine administration (e.g., Grabus et al., 2005; Le Foll & Goldberg,
2006). In addition, several animal studies have demonstrated that environmental
cues can maintain and reinstate drug seeking behaviors and drug administration
(e.g., Caggiula et al., 2001; Le Foll & Goldberg, 2006). In addition to firstorder nicotine conditioning, second-order nicotine conditioning has also been
demonstrated in animals (e.g., Goldberg, Spealman, & Goldberg, 1981; Spealman
& Goldberg, 1982). For example, Goldberg et al. (1981) showed that monkeys
press a lever at high rates under a second-order schedule of reinforcement in
which lever pressing produces a visual stimulus that is occasionally predictive of
nicotine administration.
In human research, it has been shown that smokers show increased physiological
reactions (e.g., heart rate, skin conductance) and report higher levels of craving
following the presentation of smoking-related stimuli than following the
presentation of non-smoking stimuli (for reviews, see Bevins & Palmatier, 2004;
Carter & Tiffany, 1999; Chiamulara, 2005). However, it can only be assumed that
these responses reflect prior classical conditioning; only studies in which UCS-CS
associations are formed ad hoc, i.e., only studies in which initially neutral stimuli
are paired with smoking within an experimental paradigm can be decisive on
this issue. Lazev et al. (1999) were the first to directly support the conditioning
hypothesis in smokers. It was demonstrated in their study that after pairing
with smoking, visual, olfactory, and auditory stimuli increase pulse rate and selfreported cigarette craving. In several other studies it was shown that smokers
report higher levels of craving when exposed to a cue or context that has been
paired with the occurrence of smoking than when exposed to a cue or context
paired with the nonoccurrence of smoking (Dols, Willems, van den Hout, & Bittoun,
2000; Dols, Hout, Kindt, & Willems, 2002; Thewissen et al., 2005). Moreover,
it was shown that smokers show greater approach tendency towards smoking
stimuli that were presented in the presence of a cue predicting nicotine intake
(Thewissen, Havermans, Geschwind, van den Hout, & Jansen, 2007), that smokers
exhibit greater preparatory physiological responses, i.e., skin conductance and
facial electromyographic responses, greater salivary responses and enhanced EEG
beta power in the presence of abstract cues paired with smoking (Field & Duka,
126
Higher-order conditioning of processing bias
2001; Mucha, Pauli, & Angrilli, 1998; Winkler et al., 2010) and that they attend
selectively to discriminative cues that signal the availability of tobacco-smoke
reinforcement (e.g., Hogarth, Dickinson, & Duka, 2003; Hogarth & Duka, 2006; and
see Hogarth, Dickinson, & Duka, 2010 for an overview).
Second-order conditioning in smoking addiction has been less extensively studied
in humans. However, some studies have been conducted that can be considered
second-order conditioning studies. In these studies, neutral cues were paired
with the expectancy of winning or losing cigarettes (instead of really obtaining
cigarettes or real smoking). The neutral cues that were associated with the
expectancy of winning cigarettes elicited greater attentional bias, enhanced
drug-seeking behavior and consumption, and more pleasurable mood states than
cues that were associated with the expectancy of losing cigarettes (e.g., Hogarth,
Dickinson, Wright, Kouvaraki, & Duka, 2007; Hogarth et al., 2010). However, no
studies have been conducted in which neutral cues were paired with conditioned
smoking-related cues.
Since conditioned cue-reactivity appears to play such an important role in the
continuation of smoking behavior and relapse after periods of abstinence, and
since smoking has so many deleterious effects on health, investigating associative
learning in smoking addiction into greater detail is of major relevance. Little is
known about the neural correlates of classical conditioning in smoking addiction
and even less is known about the contribution of higher-order conditioning to
smoking addiction. According to Gewirtz and Davis (Gewirtz & Davis, 2000),
studying the latter is particularly important, since only little human learning
involves the direct pairing of stimuli or contexts with powerful, unconditioned
reinforcers. Higher-order conditioning presumably predominates learning in the
outside world.
One way to study the neural correlates of associative learning in smokers is
by measuring event-related potentials (ERP) using electroencephalography
(EEG) techniques. ERPs are electrophysiological brain responses to internal or
external stimuli, which consist of several time-locked components that convey a
certain amplitude magnitude. They are particularly suited to study differences
in cognitive processing, because the magnitude of the amplitude can provide us
with information about the extent of engagement, whereas the locations can teach
us more about the neurobiological generators. It is assumed that enhancement of
127
Chapter 7
the later components of the ERP, i.e., the P3 and the Late Positive Potential (LPP)
reflects enhanced motivated attention for the stimuli presented (Cuthbert et al.,
2000; Hajcak et al., 2010; Lang et al., 1997; Schupp et al., 2000).
In studies of cognitive processes and biases in addiction, these later components
of the ERP are of particular relevance. Several studies show that the P3 and the
LPP are larger in drug users than in controls in response to drug-related stimuli
compared to neutral stimuli (e.g., Franken, Stam et al., 2003; Franken et al., 2004;
Herrmann et al., 2000; Herrmann et al., 2001; Wölfling et al., 2008). In smokers,
a centro-frontally distributed enlargement of P3 and LPP amplitudes has been
found in response to smoking cues relative to matched neutral cues, whereas nonsmokers show no difference in P3 and LPP amplitudes to both stimulus categories
(Littel & Franken, 2007; Littel & Franken, 2010; McDonough & Warren, 2001;
Warren & McDonough, 1999). These ERP studies among smokers provide evidence
for the assumption that smoking-related cues capture attentional resources and
that smokers exhibit enhanced motivated attention towards smoking-related
stimuli. This assumption is underlined by the fact that in most studies of drug
addiction P3 and LPP amplitudes have been found to correlate with subjective
craving (for an overview, see Field et al., 2009).
Recently, Franken et al. (2011) used ERP methodology in order to study classical
conditioning of emotional stimuli. They found an increased P3 in response to
initially neutral stimuli (CS) that predicted the occurrence of emotional pictures
compared to CS that predicted the occurrence of neutral pictures. These results
demonstrate that conditioning processes, including higher-order conditioning
processes (emotional pictures have acquired motivational relevance during life),
can be measured with ERPs, and, moreover, that the P3 is a suitable index of
acquired motivational relevance.
The present study was conducted in order to examine higher-order learning
processes associated with smoking addiction and its electrophysiological
correlates. Based on the experimental paradigm used by Franken et al. (2011), a
second-order smoking conditioning task utilizing ERP methodology was designed.
Smoking-related and neutral stimuli (pictures; CS1smoke and CS1neutral) were
paired with two geometrical figures (CS2smoke and CS2neutral). Both smokers’ and
non-smokers’ ERPs in response to the CS1 and the preceding CS2 were recorded
throughout the experiment. The abovementioned measure, i.e., enhanced
128
Higher-order conditioning of processing bias
motivated attention for stimuli reflected by increased P3 amplitudes in response
to CS2, was used as an outcome measure for conditioning to have taken place.
The rationale is that if relatively neutral and meaningless figures become capable
of differentially eliciting increased P3 amplitudes (which are associated with
enhanced motivated attention and preference for stimuli) after pairing with
meaningful and motivationally relevant pictures, associative learning must have
taken place.
We predict that the geometrical figures will become associated with the
pictures and will elicit enhanced P3 amplitudes. In smokers, we expect this P3
enhancement to be more evident for CS2smoke than for CS2neutral, whereas we expect
to find no differences between stimuli in non-smokers. Furthermore, we predict
that smokers will rate the CS2smoke as more positively valenced, more arousing, and
eliciting more subjective craving than the CS2neutral. No rating differences (valence,
arousal) are expected in non-smokers.
Although the P3 is the best-described ERP index of attention, several earlier
ERP components have been associated with attention processing and shown to
co vary with intrinsic motivational properties of stimuli, including the P1, N1,
and P2 (Carretie et al., 2004; Delplanque, Lavoie, Hot, Silvert, & Sequeira, 2004;
Potts, 2004). These components were investigated in an exploratory manner and
differential enhancement of these components was also regarded as indication for
associative learning.
Methods
Participants
Thirty smokers (5 males, 25 females) and 31 non-smokers (5 males, 26 females)
participated in the present study. They were recruited from the Erasmus University
Rotterdam (the Netherlands) and received either financial compensation or
course credit for participation. Non-smokers (mean age 20.5 years, SD = 1.9) were
included if they had smoked fewer than 5 cigarettes in their lifetimes (mean = 1.1,
SD = 1.5). Smokers (mean age 21.9 years, SD = 3.0) were eligible if they smoked at
least 10 cigarettes per day on average (mean = 15.6, SD = 4.2). Smokers had a mean
score of 4.4 (SD = 1.9) on the Dutch version of the Fagerström Test for Nicotine
Dependence (FTND; Vink et al., 2005), which suggests that they had low to medium
129
Chapter 7
levels of dependence. Furthermore, they had a mean carbon monoxide (CO) level
of 12.6 parts per million (Ppm; SD = 7.8), which differed significantly from nonsmokers’ CO level (mean Ppm = 1.1, SD = 1.1), t(59) = 8.17, p < 0.001. The present
study was approved by the local ethics committee of the Institute of Psychology.
All participants provided informed consent.
Stimuli and experimental paradigm
Forty smoking-related pictures (people smoking or holding cigarettes and
smoking-related objects) and 40 neutral pictures selected from the International
Affective Picture System (IAPS; Lang, 1995) served as CS1. These pictures were
paired with two geometrical figures, i.e., a green pyramid and a red cube, which
served as CS2. Each trial started with a fixation cross which was presented for 1000
ms. Subsequently, one of the two geometrical figures was presented in the upper
half of the screen with a duration of 800 ms. After 400 ms of CS2 presentation, a
smoking-related (CS1smoke) or neutral (CS1neutral) picture was added in the center
of the screen. This CS1 remained visible for 400 ms. Inter trial interval was 500
ms. See figure 1 for a schematic representation of the experimental paradigm. For
each participant, the same CS2 was always paired with the same CS1 type (e.g., the
cube was always paired with the smoking pictures). Pairing combinations were
counterbalanced across subjects. In total there were 160 trials: 80 CS2-neutral
trials and 80 CS2-smoking trials. All CS2-CS1 pairs were presented in a random
order. Each CS1 was presented 4 times. After 40 CS2-CS1 pairs, participants
received 15 second breaks.
130
Higher-order conditioning of processing bias
Figure 1. Schematic representation of the experimental paradigm
Procedure
Smokers were instructed to abstain from smoking for at least one hour prior to the
experiment in order to avoid direct effects of nicotine on ERP signals. They were
told that this was absolutely necessary and that it would be checked with a breath
analyzer. After obtaining written informed consent, participants filled out several
questionnaires on demographics, smoking history, and smoking dependence
(smokers). After completion, participants were seated in a comfortable chair which
was positioned in a light and sound-attenuated room. After the attachment of the
electrodes, participants proceeded to a non-invasive CO Ppm estimate utilizing
the EC50 Micro III Smokerlyzer® (Bedfont Scientific, Medfort, NJ, USA), a portable
device that measures breath carbon monoxide levels. Participants were instructed
to sit still, to focus on the fixation cross in the center of the screen, to blink in
between stimulus presentations, and to carefully watch the stimuli without
employing distracting thoughts. Furthermore, they were explicitly told to search
for an association between the figures and the pictures. After picture viewing,
participants rated the CS2smoke and CS2neutral on valence and arousal properties
utilizing a 10 cm Visual Analogue Scale (VAS). In addition, smokers rated both CS2
131
Chapter 7
on their capacity to elicit cigarette craving. Finally, all participants were asked
to fill out what they thought the association was between the two geometrical
figures and the pictures.
Self-report measures
All participants reported sex and age. Additionally, smokers reported smoking
duration and number of cigarettes per day, whereas non-smokers reported number
of cigarettes smoked in their lifetimes. Smoking dependence was measured with
the Dutch version of the Fagerström Test of Nicotine Dependence (FTND; Vink et
al., 2005), which is based on the original version by Heatherton et al. (1991). The
questionnaire consists of six items and has good reliability (Vink et al., 2005). All
participants rated the conditioned geometrical figures on valence and arousal
properties by means of a 10 cm VAS. Smokers also rated the conditioned figures on
craving properties, again by means of a 10 cm VAS. Since it has been hypothesized
that contingency awareness is necessary for learned motivation in humans (for an
overview, see Hogarth & Duka, 2006), contingency awareness was tested by asking
participants to write down the associations between the geometrical figures and
the pictures.
Electroencephalogram (EEG) recording and signal processing
The electroencephalogram (EEG) was recorded using a BioSemi Active-Two
amplifier system (BioSemi, Amsterdam, the Netherlands) from 34 scalp sites
(International 10-10 system; ACNS, 2006; Jurcak, Tsuzuki, & Dan, 2007) using
active Ag/AgCl electrodes mounted in an elastic cap. Six additional electrodes
were attached. Two electrodes were attached to the left and right mastoids,
two to the outer canthi of both eyes (horizontal electro-oculogram; HEOG), and
two to the infra-orbital and supra-orbital regions of the eye (vertical electrooculogram; VEOG). Both an active electrode (CMS – common mode sense) and a
passive electrode (DRL – driven right leg) were used to comprise a feedback loop
for amplifier reference. Signals were recorded online with a low pass filter of 134
Hz and digitized with a 512 Hz, 24-bit A/D converter. Offline, the EEG signals were
referenced to the mathematically linked mastoids and EEG and EOG were filtered
with a band pass of 0.01-80 Hz (phase shift-free; 24 dB/octave slope). CS1 data
were segmented in epochs of 900 ms, including 100 ms pre-stimulus baseline,
whereas CS2 data were segmented in epochs of 600 ms, including 100 ms prestimulus baseline. Ocular correction (Gratton et al., 1983) was applied and epochs
containing an EEG signal exceeding ± 75 μV were excluded from the average.
132
Higher-order conditioning of processing bias
After baseline correction, epochs were averaged across trials and overall grand
averages were obtained for the two CS1 conditions (CS1smoke and CS1neutral) and for
the first and the last 40 trials of the two CS2 conditions (CS2smoke and CS2neutral).
The resulting ERP waves were visually inspected and appeared to correspond well
with ERP waves usually reported in response to visual stimuli. Regarding the CS1,
a clear P3 was identified in the 300-800 ms time window (see Figure 2). Regarding
the CS2, a P3 component was observed in the 280-500 ms timeframe (see Figure 3).
In addition, P1, N2, and P2 components were identified in the 100-150 ms, the 150200 ms, and the 250-280 ms timeframe, respectively. For all components elicited
by the CS1 and CS2, mean activity (area measurement) was computed per group
and stimulus category. Brain Vision Analyzer (Brain Products, Germany) was used
for all offline EEG analyses.
Analyses
For each component ERP effects were assessed by performing repeatedmeasurement analyses of variance (ANOVAs). Based on current source density
(CSD) maps for differences in brain activity between CS1 conditions (figure 2)
10 electrodes of interest were selected, i.e., FC1, Fz, FC2, C3, Cz, C4, CP1, Pz, Cp2,
and Oz. For analyzing P2 and P3 components elicited by CS2 conditions three
midline electrode sites (Fz, Cz, Pz) were selected (figure 3). Analyses of P1 and N1
components were restricted to occipital electrode sites, i.e., PO3, O1, Oz, O2, PO4.
Group (smokers versus non-smokers) served as the between-subjects factor. CS1
stimulus type (neutral versus smoking-related), block (first block, second block),
and electrode site served as within-subjects factors in the ANOVA on CS1, and CS2
stimulus type (neutral versus smoking-related), block (first block, second block),
and electrode site served as within-subjects factors in the ANOVAs on CS2. Arousal,
valence, and craving ratings of the geometrical figures were tested using two 2
(stimulus) x 2 (group) repeated-measurement ANOVAs (arousal and valence) and
an independent t-test (craving). Because of missing data, two participants were
excluded from this analysis.
Greenhouse-Geisser correction was applied to all ANOVAs (uncorrected df’s are
reported). All significant effects and effects showing trends towards significance
were further analyzed using pairwise comparison post-hoc tests. An alpha-level of
0.05 was used for all statistical tests.
133
Chapter 7
ȝ9
ȝ9
6PRNHU
)]
&]
3PV
&66PRNH1HXWUDO
%ORFN
ȝ9
PV
3]
%ORFN
6PRNHU
ȝ9
PV
2]
1RQVPRNHU
ȝ9P
ȝ9P
ȝ9P
ȝ9
PV
PV
ȝ9
1RQVPRNHU
)]
&]
&6VPRNHEORFN
&6QHXWUDOEORFN
&6VPRNHEORFN
&6QHXWUDOEORFN
ȝ9
PV
3]
ȝ9
PV
PV
2]
PV
Figure 2. Smokers’ (upper) and non-smokers’ ERPs (lower) in response to CS1smoke and
CS1neutral in Block1 and Block 2. Current Source Density (CSD) maps represent differences in
activity between CS1smoke and CS1neutral in the 300-800 ms timeframe (P3)
134
Higher-order conditioning of processing bias
ȝ9
6PRNHU
)]
ȝ9
&]
%ORFN
%ORFN
6PRNHU
PV
3PV
&66PRNH1HXWUDO
PV
ȝ9
3]
1RQVPRNHU
ȝ9P
ȝ9P
ȝ9P
ȝ9
1RQVPRNHU
)]
ȝ9
PV
%ORFN
&]
%ORFN
ȝ9
3PV
&66PRNH1HXWUDO
PV
6PRNHU
1RQVPRNHU
3]
PV
ȝ9P
ȝ9P
ȝ9P
&6VPRNHEORFN
&6QHXWUDOEORFN
&6VPRNHEORFN
&6QHXWUDOEORFN
PV
Figure 3. Smokers’ (upper) and non-smokers’ ERPs (lower) in response to CS2smoke
and CS2neutral in Block 1 and Block 2. Current Source Density (CSD) maps represent
differences in activity between CS2smoke and CS2neutral in the 280-500 ms timeframe (P3) and
the 200-280 ms timeframe (P2)
135
Chapter 7
Results
First-order Conditioned Stimuli (pictures; CS1)
P3
See figures 2 and 4 for mean P3 amplitudes per group, CS1 and block. A significant
main effect for CS1 was observed F(1,59) = 94.17, p < 0.001. Smoking pictures
elicit larger P3 amplitudes than neutral pictures across participants. In addition, a
significant CS1 x Block interaction was found, F(1,59) = 12.02, p = 0.001. Post-hoc
tests revealed that the P3 amplitude in response to CS1smoke is larger during the
first 40 trials than during the last 40 trials (p = 0.009), whereas there is a trend
for the P3 in response to CS1neutral to be larger during the last 40 trials compared to
the first 40 trials (p = 0.064). However, in both blocks the CS1smoke elicit larger P3
amplitudes than the CS1neutral (both p’s < 0.001). Furthermore, a significant CS1 x
Group effect, F(1,59) = 6.57, p = 0.013, was found. Post-hoc comparisons revealed
that smokers respond with significantly larger P3 amplitudes to CS1smoke than to
CS1neutral (p = 0.001), whereas non-smokers show no amplitude difference between
the two CS1 (p = 0.239). In addition, a trend towards a significant CS1 x Electrode
x Group effect, F(3,177) = 2.05, p = 0.064, was observed. Post-hoc tests showed
larger P3 amplitudes for CS1smoke in smokers relative to controls at all electrodes
(all p’s < 0.01), except for Oz.
Figure 4. Mean P3 amplitudes for smokers and non-smokers per Block and CS1
Second-order Conditioned Stimuli (geometrical figures; CS2)
P3
See figures 3 and 5 for mean P3 amplitudes per group, CS2 and block. First of
all, the main effect for CS2 showed a trend towards significance, F(1,59) = 0.075,
with CS2smoke showing larger amplitudes than CS2neutral across all participants and
blocks indicating a conditioning effect for all participants in response to figures
associated with smoking pictures. Furthermore, a significant CS2 x Block x Group
136
Higher-order conditioning of processing bias
interaction effect was observed, F(1,59) = 4.45, p = 0.039. Post-hoc comparisons
revealed that during the first block, smokers’ P3 amplitude in response to the
CS2smoke is significantly larger than their P3 amplitude to the CS2neutral (p = 0.019),
whereas non-smokers show no difference between P3 amplitude in response to
the CS2smoke and CS2neutral during the first 40 trials (p = 0.502). Neither smokers,
nor non-smokers show differences in P3 amplitude in response to CS2smoke and
CS2neutral during the second block (smokers: p = 0.557; non-smokers: p = 0.142).
Furthermore, although not significant, smokers show trends to respond with an
enlarged P3 to CS2neutral than non-smokers during the second block (p = 0.095)
and an enhanced P3 to CS2neutral in the second block compared to the first block
(p = 0.078). Furthermore, a significant CS2 x Block x Electrode x Group interaction
was found, F(3,177) = 6.26, p = 0.003. Post-hoc comparisons showed that in the first
block, smokers display a more positive P3 in response to CS2smoke than to CS2neutral
at Fz (p = 0.028), Cz (p = 0.018) and Pz (p = 0.091), whereas non-smokers display
a larger P3 to CS2smoke than to CS2neutral during the second block (Fz: p = 0.078,
Cz: p = 0.046). In addition, during the second block smokers’ P3 in response to
CS2neutral becomes greater than non-smokers’ P3 in response to CS2neutral at Cz
(p = 0.046). Moreover, although the conditioning effect for CS2smoke does not change
from the first to the second block, smokers show enhancement of P3 amplitudes
for CS2neutral in second block as compared to the first block at Fz (p = 0.052) and
Cz (p = 0.064). Non-smokers, show the opposite pattern, i.e., no changes of the
conditioning effect for neutral cues between the first and the second block, but an
enhancement of P3 amplitudes for CS2smoke in the second block as compared to the
first block at Fz (p = 0.030).
Figure 5. Mean P3 amplitude for smokers and non-smokers per Block and CS2
Early components
No significant main or interaction effects were found for the P1. For the N1,
a significant main effect was found for Block, F(1,59) = 5.98, p = 0.017, indicating
137
Chapter 7
that N1 amplitudes are larger during block 1 than during block 2. No other main
or interaction effects were observed for this component. For the P2 a significant
main effect for CS2 was observed, F(1,59) = 4.28, p = 0.043. Figures associated
with smoking pictures elicit larger P2 amplitudes than figures associated with
neutral figures across participants and blocks. In addition, a significant CS2 x
Block x Electrode x Group interaction was found, F(2,118) = 8.54, p = 0.001. Posthoc comparisons revealed that in block 1 smokers show enlarged P2 amplitudes
in response to CS2smoke as compared to CS2neutral at Fz (p = 0.029) and Cz (p = 0.049)
electrodes. During this first block non-smokers also show enlarged P2 amplitudes
in response to CS2smoke as compared to CS2neutral but in contrast to the effect found
in smokers, this effect not present at Fz and Cz, but at Pz (p = 0.040). During the
second block no significant main or interaction effects were found. See figures 3
and 6 for mean P2 amplitudes per group, CS2 and block.
Behavioral measures
All participants were aware of the CS2-CS1 relation. Smokers’ craving ratings (selfreported craving elicited by CS) were significantly higher for the CS2smoke than for
the CS2neutral, t(28) = 5.98, p < 0.001. Furthermore, on both arousal and valence
judgments of the geometrical figures significant CS2 x Group interactions were
found, respectively F(1,57) = 28.79, p < 0.001 and F(1,57) = 18.80, p < 0.001. Posthoc tests showed that smokers rate the CS2smoke as significantly more arousing
than non-smokers (p < 0.001). They also find the CS2smoke more positive than
do non-smokers, (p < 0.001). Smokers and non-smokers do not differ in arousal
judgment of the CS2neutral (p = 0.191). However, non-smokers find the CS2neutral
significantly more pleasant than smokers (p < 0.001). See figure 7 for smokers’
and non-smokers’ mean valence, arousal, and craving scores.
Figure 6. Mean P2 amplitude for smokers and non-smokers per Block and CS2
138
Higher-order conditioning of processing bias
Figure 7. Smokers’ and non-smokers’ mean self-reported arousal and valence ratings and
smokers’ mean self-reported craving ratings of the CS2smoke and CS2neutral measured with a
10 cm Visual Analogue Scale (VAS). Error bars represent standard deviations
Discussion
The aim of the present study was to examine higher-order classical conditioning
processes associated with smoking addiction by employing a direct cortical
measure, i.e., brain activity as measured by ERPs. Brain responses of smokers and
non-smokers were conditioned by pairing smoking-related and neutral stimuli
(CS1smoke and CS1neutral) with two geometrical figures (CS2smoke and CS2neutral). All
subjects were consciously aware of the CS2-CS1 associations. ERPs were recorded
to both CS1 and preceding CS2.
With regard to the CS1, results from the present study replicate the previously
observed finding that smokers exhibit a processing bias for smoking-related
stimuli(Littel & Franken, 2007; McDonough & Warren, 2001; McDonough &
Warren, 2001). At frontal, central and parietal sites, P3 components of the ERP
were larger in response to smoking cues than in response to neutral cues for
smokers compared to non-smokers. This implies that our smoking stimuli were
suitable to serve as CS1 in the current study. Furthermore, it was observed that
smoking cues elicited larger P3 amplitudes than neutral cues across participants,
indicating that smoking cues in general captured more attention than neutral cues.
139
Chapter 7
With regard to the CS2, we expected that through second-order associative
learning the neutral figures would become associated with the pictures that
followed them and would therefore start to elicit equal conditioned responses,
i.e., motivated attention and preference as reflected by increased ERP amplitudes.
We predicted this ERP enlargement to be more evident for CS2smoke than for
CS2neutral in smokers, whereas we expected to find no differences between stimuli
in non-smokers. Although all clearly discernable ERP components (P1, N1, P2,
and P3) were analyzed, P3 components were of special interest since they have
been associated with motivated attention for the cues presented (Cuthbert et
al., 2000; Hajcak et al., 2010; Lang et al., 1997; Olofsson et al., 2008; Schupp et
al., 2000) and because previous studies show that they are enlarged in smokers
in response to smoking-related pictures that are suggested to have acquired
motivational significance through prior first-order classical conditioning (Littel &
Franken, 2007; McDonough & Warren, 2001; Warren & McDonough, 1999). Results
showed that P3 amplitudes were enlarged in response to CS2smoke as compared
to CS2neutral across participants. This implies an overall conditioning effect for
geometrical figures associated with smoking cues. This finding is in accordance
with electrophysiological responding to the CS1 (i.e., smoking cues capturing
more attentional resources than neutral cues in general). Comparable results
were obtained for the P2. P2 amplitudes were enlarged in response to CS2smoke as
compared to CS2neutral across participants. Although the P2 is reported less often
than the P3 in picture processing, there are indications that this earlier component
is also sensitive to automatic attention capture and could be modulated by stimulus
valence (Carretie et al., 2004; Delplanque et al., 2004; Potts, 2004).
Furthermore, in line with our primary hypotheses, data from the present study
showed that in smokers, the CS2smoke, i.e., the geometrical figure that was paired
with smoking stimuli, elicited significantly greater P3 amplitudes than the
CS2neutral during the first half of the experiment, whereas no differences between
CS2 conditions were found in non-smokers. Similarly, the CS2smoke was shown to
elicit larger P2 amplitudes than CS2neutral in smokers but not in non-smokers. These
results suggest that smokers, compared to non-smokers, show more enhanced
associative learning for smoking cues than for neutral cues even though these cues
were never paired directly with an UCS (i.e., smoking). Furthermore, it underscores
the idea that addiction affects basic learning and memory systems, and that their
neural substrates, normally involved in obtaining more conventional goals, are
recruited by the drugs of abuse (Carretie et al., 2004; Robbins et al., 2008).
140
Higher-order conditioning of processing bias
Besides the electrophysiological evidence, we also observed self-reported evidence
of enhanced second-order conditioning in smokers: smokers reported more cueelicited craving for the CS2smoke compared to the CS2neutral. Furthermore, smokers
found the CS2smoke more arousing and more pleasurable than non-smokers.
Besides, they rated the CS2neutral as less pleasurable than non-smokers. As can be
observed in figure 7, the valence differences between smokers and non-smokers
with regard to the CS2smoke were partly driven by the finding that non-smokers rate
the CS2smoke as less pleasurable or more aversive than smokers, since smokers find
the CS2smoke and the CS2neutral equally pleasurable. Yet, the group differences remain
and indicate that smokers and non-smokers engage in different learning patterns.
Direct first-order smoking conditioning has already been demonstrated in several
previous studies. In these studies it was shown that smokers report more craving
and show greater approach bias, attentional bias and physiological responses in
response to cues paired with the presence of smoking than in response to cues
paired with the absence of smoking (Dols et al., 2000; Dols et al., 2002; Lazev et al.,
1999; Thewissen et al., 2005; Thewissen et al., 2007). Although no explicit reference
was made to it, second-order conditioning has already been demonstrated in
several studies in which neutral cues were paired with the expectancy of winning
or losing cigarettes. The neutral cues that were associated with the expectancy
of winning cigarettes elicited greater attentional bias, enhanced drug-seeking
behavior and consumption, and more pleasurable mood states than cues that
were associated with the expectancy of losing cigarettes (e.g., Hogarth et al.,
2007; Hogarth et al., 2010). The present study is the first to demonstrate direct
second-order conditioning in smoking addiction by combining neutral stimuli
with conditioned smoking pictures.
During the course of the experiment, the group-specific conditioning effects of
the present study seem to reverse; during the last half of the experiment smokers’
P3 amplitudes in response to the CS2neutral increased relative to the first half of
the experiment, whereas non-smokers’ P3 amplitudes to the CS2smoke increased
relative to the first half experiment. Moreover, smokers’ P3 amplitude in response
to CS2neutral became significantly larger than that of non-smokers across all
electrodes, whereas non-smokers’ P3 amplitude to CS2smoke became significantly
larger than that of smokers at Pz. With regard to the P2, all significant conditioning
effects found during the first half of the experiment disappeared during the second
half (see figures 5 and 6 for a visual representation of these results).
141
Chapter 7
Although great interpretive caution is warranted, it appears that non-smokers
show a slowly progressing learning curve for CS2 smoke, i.e., stimuli that are more
attentively processed (see results on CS1). Smokers, on the other hand, show
a steeper learning curve for CS2 smoke during the first trials of the experiment,
suggesting an initially enhanced associative learning for smoking cues as
compared to non-smokers. However, this enhanced smoking-related associative
learning declines after a certain amount of time, and seems to be replaced by a
delayed conditioning for the neutral cues.
There exist several possible explanations for this finding. First of all, secondorder conditioning is intrinsically weaker than first-order conditioning and
appears typically to be transient (Gewirtz & Davis, 2000). It is argued that after
a small number of trials second-order learning reaches a maximum and starts
to decline with further training. Gewirtz and Davis (2000) posit that this is
caused by the development of conditioned inhibition; the CS2 becomes a signal
for the nonoccurrence of reinforcement and therefore inhibits the elicitation of
conditioned responses. In line with this is the hypothesis that drug use expectancy
is necessary for learned motivation in humans (for an overview, see Hogarth &
Duka, 2006). Although second-order conditioning develops faster than conditioned
inhibition, the latter is the strongest of the two phenomena. Therefore, a plausible
explanation for decrement of the conditioning effect for smoking cues during
the second part of the experiment might be that after some time smokers lose
their interest in the cues paired with smoking stimuli, since they predict no real
smoking and subsequent reinforcement and start focusing on the cues paired with
neutral stimuli instead. Evidence for this explanation is provided by the results
from the CS1 by Block interaction analyses, which showed that P3 amplitudes in
response to smoking pictures were larger during the first trials than during the
last trials, whereas there was a trend for the P3 in response to neutral pictures
to be larger during the last trials compared to the first trials. However, because
scores on the valence, arousal, and craving VAS, which were collected at the end of
the conditioning session, still show a clear subjective conditioning effect, it can be
argued whether smokers really lost their interest in the CS2smoke during the second
block due to conditioned inhibition or the absence of contingencies. A plausible
alternative explanation could be merely fatigue or boredom, since the experiment
conveys many repetitions of the same figures and pictures.
142
Higher-order conditioning of processing bias
Conclusions
The present findings suggest that smokers and non-smokers show associative
learning during a higher-order conditioning experiment in response to neutral
cues that are paired with smoking-related stimuli as measured with ERP indices
of attention. Furthermore, results indicate that this associative learning is more
pronounced in smokers than in non-smokers during the first half of the experiment.
These effects are not only found on the electrophysiological level, as reflected by
enlarged P2 and P3 components of the ERP, but are also self-reported. Therefore, the
results indicate that for smokers neutral cues that are paired with motivationally
relevant smoking-related stimuli gain more motivational significance, at least for
a short period of time, even though they were never paired directly with an UCS.
We acknowledge that we only investigated the effects of explicit conditioning
as we instructed participants to pay attention to the presented contingencies.
Therefore, the conclusions are limited by the fact that they only pertain to explicit
conditioning. Although several studies have shown that in addictive behaviors
implicit processes may have limited value in conditioning effects (Hogarth &
Duka, 2006), the present paradigm could be employed to further elucidate this
role on the neurophysiological level. Furthermore, it must be noted that there
was an overrepresentation of female participants. Because there were no gender
differences between groups, these could not have accounted for the observed
group differences. In addition, we assume that the functional meaning attributed
to P2 and P3 responses can be safely applied to electrophysiological responding to
CS2. We acknowledge that there exists no certainty regarding this issue. However,
some evidence for this assumption can be derived from the study by Franken
et al. (2011). In this study an increased P3 was observed in response to neutral
stimuli that predicted the occurrence of emotional pictures compared to neutral
stimuli that predicted the occurrence of neutral pictures, indicating that the P3 is
a suitable index of acquired motivational relevance.
This is the first study to directly show the contribution of higher-order
conditioning to smoking addiction. Furthermore, it is the first study to reveal the
electrophysiological correlates of higher-order conditioning in smoking. Replication
studies are warranted, ideally using a design in which actual smoking is paired
with certain neutral cues, which are in turn paired with other neutral cues.
143
Chapter 7
Although results are preliminary, they may help in understanding the etiology
of smoking addiction and its persistence. Craving and relapse might not be
triggered by concrete cues and contexts only, but also, or predominantly, by more
complex and divergent cues and contexts which do not necessarily have intrinsic
motivational value, but have motivational value that is acquired through the
processes of higher-order conditioning.
144
Chapter
8
Intentional modulation of the Late Positive
Potential in response to smoking cues
by cognitive strategies in smokers
Littel, M. & Franken, I.H.A. (2011). Intentional modulation of
smoking-related motivated attention by cognitive strategies
in smokers: An ERP study. PLoS ONE, 6(11): e18898.
Chapter 8
Abstract
Attentional bias is considered an important concept in addiction since it has
been found to correlate with subjective craving and is strongly associated with
relapse after periods of abstinence. Hence, investigating in ways to regulate
attention for drug cues would be of major clinical relevance. The present
study examined deliberate, cognitive modulation of motivated attention for
smoking cues in smokers. The effects of three different reappraisal strategies
on an electrophysiological measure of attentive processing were investigated.
Early and late LPP components in response to passively viewed neutral and
smoking pictures were compared with LPPs in response to smoking pictures
that were reappraised with three different reappraisal strategies. Results
show that when smokers actively imagine how pleasant it would be to smoke
(pleasant condition), their early LPP in response to smoking cues increases,
but when smokers actively focus on an alternative stimulus (distraction
condition) or think of a rational, uninvolved interpretation of the situation
(rational condition), smoking-related late LPP amplitude decreases to the
processing level of neutral stimuli. Present results are the first to indicate that
smoking cue-elicited LPP amplitudes can be modulated by cognitive strategies,
suggesting that attentive processing of smoking cues can be intentionally
regulated by smokers with various levels of dependence. Although cognitive
strategies can lead to enhanced processing of smoking cues, it is not completely
clear whether cognitive strategies are also successful in reducing smokingrelated motivated attention. Although findings do point in this direction, the
present study is best considered preliminary and a starting point for other
research on this topic. A focus on the distraction strategy is proposed, as there
are indications that this strategy is more successful than the rational strategy
in decreasing LPP amplitude.
148
Intentional modulation of processing bias
Introduction
Substance users display cognitive processing biases towards drug-related stimuli
(Field & Cox, 2008). For example, drug users are slower than healthy controls to
color name drug-related words on the modified Stroop task (Munafò et al., 2003)
and maintain their gaze on drug-related stimuli longer than on neutral stimuli
(Mogg et al., 2003). These biases in cognitive processing are thought to emerge
because of the motivational and attention-grabbing properties of drug-related
stimuli (Robinson & Berridge, 1993). According to the incentive-sensitization
theory, drug-related stimuli have acquired these properties through repeated
drug administration which causes a sensitization of dopamine neurotransmission
in the striatum. The ‘incentive salience’ or relevance of stimuli for reinforcement
makes the drug-associated stimuli extremely ‘wanted’ and therefore a greater
proportion of attentional resources is allocated to them. This drug specific
allocation of attentional resources, or attentional bias, is believed to diminish
attentional resources left for alternative cues, enhances drug-related cognitions,
and causes subjective craving (Franken, 2003).
Enhanced attentional processing of drug cues has indeed been found to be
associated to a certain extent with subjective craving in various drug-dependent
populations (Field et al., 2009). Attentional bias has also been associated with
relapse after periods of abstinence, indicating that persons having higher degrees
of attentional bias demonstrate higher relapse rates. This relation has been
found in smokers (Waters et al., 2003), alcoholics (W. M. Cox et al., 2002), cocaine
(Carpenter et al., 2006) and heroin dependent patients (Marissen et al., 2006).
In addition, some recent studies (Fadardi & Cox, 2009; Field & Eastwood, 2005)
demonstrated that higher levels of attentional bias for alcohol-related stimuli
increases the motivation to drink alcohol and that higher levels of attentional
bias for smoking stimuli increases the urge to smoke in males (Attwood et al.,
2008). Together, these studies show that attentional bias may play a causal role
in addictive behaviours and underline the clinical importance of the concept of
attentional bias in alcohol and drug addiction.
Recently, Kober et al. (2009) showed that the intensity of subjective craving can
be intentionally modulated by cognitive regulation strategies. Smokers and nonsmokers were presented with smoking-related and food-related stimuli that are
thought to elicit self-reported craving. Participants were instructed to think about
149
Chapter 8
the immediate consequence of consuming the pictured substance or the longterm consequences of repeatedly consuming the substance. Reports of craving
were reduced when smokers considered the long-term consequences associated
with smoking compared to when they considered the immediate consequence
of smoking. Furthermore, Kober et al. (2010) demonstrated that these craving
regulations were supported by the same prefrontal systems of the brain, i.e.,
the dorsolateral prefrontal cortex, the ventral lateral prefrontal cortex, and the
dorsolateral prefrontal cortex – striatal pathways, as the regulation of other
appetitive desires (craving for food) and emotion in general.
The regulation of emotion utilizing several different cognitive regulation strategies
has been extensively investigated. It has been shown that when participants
actively try to reinterpret the meaning of emotional stimuli, autonomic arousal,
facial expression, and reports of emotion can be modulated (Dillon & Labar, 2005;
J. J. Gross & Levenson, 1997; J. J. Gross, 1998; Hajcak & Nieuwenhuis, 2006; Jackson,
Malmstadt, Larson, & Davidson, 2000). Furthermore, using fMRI methodology,
cognitive regulation strategies have been linked to increased activation of
cognitive control regions of the brain, such as the prefrontal cortex, and decreased
activation of affective appraisal structures, such as the amygdala (McRae et al.,
2010; Ochsner & Gross, 2008). Another common method for studying the effects of
cognitive regulation on brain activity has been the use of Event-Related Potentials
(ERPs). The later components of the ERP, the P300 and the related Late Positive
Potential (LPP), have been found to be enhanced following the presentation of both
positive and negative compared to neutral pictures and words (Hajcak et al., 2009)
and are associated with directed attention toward task-relevant information and
facilitated perceptual processing of motivationally relevant stimuli. The P300
component appears to be transient, but the LPP can be enhanced for several seconds
after the presentation of emotional stimuli (Cuthbert et al., 2000). Because of its
sustained duration and its sensitivity to affective properties of pictorial stimuli,
the LPP is particularly suited to study the impact of cognitive regulation strategies
on emotional responding (Hajcak et al., 2009).
Several studies have demonstrated that the amplitude of the LPP is reduced when
participants are instructed to decrease emotional responses to negative and
positive pictures using self-generated cognitive reappraisal strategies, for example
by imagining that the depicted situation gets worse or viewing the pictures from an
uninvolved, detached perspective (Hajcak & Nieuwenhuis, 2006; Krompinger et al.,
2008; Moser et al., 2006). These LPP reductions were correlated with self-reported
150
Intentional modulation of processing bias
changes in emotional intensity. Furthermore, when presenting participants with
externally provided reappraisal frames (negative and neutral descriptions prior
to the presentation of negative pictures), the magnitude of the LPP is decreased to
pictures with neutral reappraisal frames (Foti & Hajcak, 2008) and enhanced to
pictures that were preceded by negative instructions (MacNamara, Foti, & Hajcak,
2009; MacNamara, Ochsner, & Hajcak, 2010). Together, these studies suggest that
cognitive regulation strategies, whether self-generated or externally provided, can
effectively modulate emotional responding. This applies to self-reported emotion
as well as motivated attention, i.e., the activation of attentional and motivational
systems of the brain (Schupp et al., 2000), as reflected by LPP magnitude.
ERP studies of substance dependence have demonstrated that the P300 and LPP
of substance users relative to healthy controls is enhanced in response to drugrelated stimuli compared to neutral stimuli. This result has been obtained in
alcoholics (Herrmann et al., 2000; Herrmann et al., 2001; Namkoong et al., 2004),
heroin users (Franken, Stam et al., 2003; Lubman et al., 2007; Lubman et al.,
2008), cocaine users (Dunning et al., 2011; Franken et al., 2008; Van de Laar et
al., 2004), cannabis users (Wölfling et al., 2008), and smokers (Littel & Franken,
2007; Littel & Franken, 2010; McDonough & Warren, 2001; Warren & McDonough,
1999). Similar to the view that enhanced P300 an LPP in response to emotional
stimuli reflects enhanced motivated attention to these stimuli, it is assumed that
the enhancement of the late ERP components in substance users reflects their
motivated and elaborate attention for drug-related stimuli. This is underlined
by the finding that P300 and LPP amplitudes are significantly correlated with
subjective craving (Field et al., 2009).
The aim of the present study was to examine whether it is possible to modulate
smokers’ attentional processing of smoking pictures by deliberate cognitive
regulation strategies. Because the enhancement of the late components of the ERP
in response to drug cues has been associated with enhanced processing of and
motivated attention for smoking cues and because these components, especially
the LPP, are sensitive to cognitive regulation strategies in healthy participants,
we used LPP amplitude as an outcome measure for successful or unsuccessful
regulation of motivated attention for smoking cues in smokers.
To test our hypotheses, we presented the participants with smoking-related
stimuli under three instructional conditions. The first instruction was to imagine
151
Chapter 8
how pleasant or delicious it would be to smoke the cigarettes depicted in the
pictures. The other instructions consisted of a distraction strategy, naming the
dominant color in the pictures, and a strategy in which the participants had to
view the pictures from an uninvolved perspective, making up a rational story
about the content of the pictures. ERPs in response to the reappraised pictures
were compared with ERPs in response to passively viewed, non-reappraised
smoking pictures and neutral pictures. It was hypothesized that the amplitude of
the LPP would be increased during the instruction to think of the pleasant aspects
of smoking cigarettes (pleasant condition) and decreased during instructions
to focus on other aspects of the smoking pictures or to rationally reinterpret
the smoking pictures (distraction and rational condition). Because we expected
that, depending on the regulation strategy used, the magnitude of the LPP would
change over time (Foti et al., 2009), we analyzed both the early LPP (600-1000 ms)
and the late LPP (1000-2000 ms). There were no specific hypotheses about which
regulation strategy would be most successful or how the different LPPs would
change over time.
A second aim of the present study was to test whether enhanced attentive processing
as reflected by enhanced LPP amplitudes as well as the cognitive modulation of
these amplitudes differs between regular smokers and light smokers that do not
smoke every day of the week, but at least two days per week for at least two years.
Previous studies show that light smokers experience less craving than regular
smokers when exposed to smoking cues (Kober et al., 2009; Sayette, Martin, Wertz,
Shiffman, & Perrott, 2001; Shiffman, Kassel, Paty, Gnys, & Zettler-Segal, 1994;
Shiffman, Paty, Kassel, Gnys, & Zettler-Segal, 1994; Shiffman, Paty, Gnys, Kassel,
& Elash, 1995), but are equally distracted by smoking cues when performing a
reaction time task (Sayette et al., 2001), and do not differ from regular smokers in
the extent to which they can regulate their craving levels in response to smoking
cues (Kober et al., 2009). It is unknown whether light smokers differ from regular
smokers in the electrophysiological processing of smoking cues associated with
increased attentional resources and whether they are capable of regulating this
attentive processing.
152
Intentional modulation of processing bias
Methods
Ethics Statement
The study was conducted in accordance with the Declaration of Helsinki and
all procedures were carried out with the adequate understanding and written
informed consent of the subjects. The study protocol was approved by the ethics
committee of the Institute of Psychology of Erasmus University.
Participants
Twenty-eight regular tobacco smokers (mean age = 21.7, SD = 2.57, 39.3 % male, 60.7 %
female) and 22 light smokers (mean age = 21.00, SD = 1.85, 45.5 % male, 54.5 %
female) participated in the present study. They were recruited at the Erasmus
University Rotterdam (the Netherlands) and received either course credit or
financial compensation. Smokers were included if they smoked > 10 cigarettes per
day; light smokers were included if they did not smoke every day of the week, but
at least 2 days per week for at least 2 years. Smokers smoked 99.25 cigarettes per
week on average (SD = 29.08), 14.18 cigarettes per day (SD = 4.15, range 10-25),
had a mean score of 3.43 (SD = 1.83) on the Fagerström Test for Nicotine Dependence
(FTND; Vink et al., 2005), indicating low to moderate levels of nicotine dependence,
and had a mean carbon monoxide (CO) level of 10.21 parts per million (Ppm;
SD = 6.96). Light smokers smoked 16.14 cigarettes per week on average (SD = 9.87),
5.05 cigarettes per day (SD =3.02, range 1-13), 3.36 days per week (SD = 1.27),
and had a mean score of 0.18 (SD = 0.50) on the FTND, indicating an absence of
nicotine dependence, and a CO level of 2.95 parts per million (SD = 2.59). Regular
smokers and light smokers significantly differed on FTND score, t(48) = 8.05,
p < 0.001 and CO level, t(48) = 4.64, p < 0.001. With regard to smoking duration,
regular smokers (mean number of years = 5.66, SD = 3.58) and light smokers (mean
number of years = 5.09, SD = 1.94) did not show significant differences, t(48) = 0.67,
p > 0.10. Furthermore, no group differences were found for age, t(48) = 0.88,
p > 0.10, or sex ratio, χ2(1) = 0.19, p > 0.10.
Self-report measures
Smoking history and demographic data were self-reported (sex, age, smoking
duration, number of cigarettes/day, number of days/week). Smoking dependence
was assessed with the Dutch version of the Fagerström Test for Nicotine Dependence
(FTND; Vink et al., 2005). This questionnaire has good reliability and holds a
significant correlation with number of cigarettes smoked per day. The FTND is
153
Chapter 8
scored according to the scoring system described in Heatherton et al. (Heatherton
et al., 1991) and scores range from 0 - 10. Subjective craving was measured with
the brief Questionnaire on Smoking Urges (QSU-brief; L. S. Cox et al., 2001). This
questionnaire was adapted from the Questionnaire on Smoking Urges (QSU;
Tiffany & Drobes, 1991) and consists of two subscales: ‘desire and intention to
smoke’ (reward-craving) and ‘reduction of negative affect and withdrawal craving’
(withdrawal-craving). A Dutch translation of the QSU-brief was administered,
which has adequate psychometric properties (Littel et al., 2011).
Procedure
All participants were instructed to refrain from smoking for at least one hour prior
to the experiment in order to avoid direct effects of nicotine on task performance
and ERP signals. This was checked at the beginning of the experiment with the
EC50 Micro III Smokerlyzer® (Bedfont Scientific, Medford, NJ, USA), a portable
device which measures breath carbon monoxide levels (CO Ppm). After providing
informed consent, participants filled out questionnaires on demographics,
smoking history, subjective craving, and nicotine dependence. After completion,
electrodes were attached and the first instructions were given. All participants
were tested alone in a sound and light attenuated room. They were all tested by the
same experimenter and received the same instructions.
Firstly, the participants were presented with 40 smoking-related and 40 neutral
control pictures. They were instructed to watch these pictures closely without
employing distracting thoughts. After passive picture viewing the reappraisal
part of the task was started. The participants were presented with one of three
cognitive reappraisal blocks consisting solely of smoking-related stimuli. They
first received a reappraisal instruction and then practiced two pictures with the
experimenter. Subsequently, picture presentation was started. After that, the
second and the third block were explained, practiced and presented. The order of
the three blocks was counterbalanced across participants.
For the pleasant condition, participants were given the following verbatim
instructions: “During this block you will see only smoking-related pictures. You are
instructed to imagine how pleasant and delicious it would be to smoke the cigarettes
depicted in the pictures or to smoke like the persons in the pictures. Even if you do not
like the picture, try to imagine how pleasant and delicious it would be to smoke the
presented cigarettes. Hold on to this thought for as long as the picture is presented
154
Intentional modulation of processing bias
on the screen. Before each picture presentation you will receive a reminder of the
instruction, which in this case will be ‘delicious’.”
For the distraction condition, participants were given the following verbatim
instructions: “During this block you will see only smoking-related pictures. You are
instructed to actively think of the most prominent color in the picture, that is, the
color that pops out for you. If no particular color pops out, then pick a color from the
picture and keep that in mind. Think of this color for as long as the picture is presented
on the screen. Before each picture presentation you will receive a reminder of the
instruction, which in this case will be ‘color’.”
For the rational condition, participants were given the following verbatim
instructions: “During this block you will see only smoking-related pictures. You
are instructed to make up a short story about the content of the picture. Think of
something that is not directly visible in the picture. For example some background
information or something you can easily infer from the picture. The story has to be
completely rational, which means that it may not consist of your feelings about the
picture, such as ‘this looks nice’. Hold on to the rational thought for as long as the
picture is presented on the screen. Before each picture presentation you will receive a
reminder of the instruction, which in this case will be ‘rational’.”
The practice phase consisted of the presentation of two pictures that were the
same for all participants. The experimenter asked the participants to respond out
loud to the picture according to the reappraisal instruction that was given. After
that, the experimenter provided the participants with some alternative options.
Most extensively practiced was the rational strategy. Participant responses that
included emotions or feelings were strictly corrected (e.g., “she looks pretty” were
to be replaced by “she wears make-up because she has a date” or “she probably
dyed her hair”). Picture presentation was not started until the experimenter felt
that the participants completely understood the instructions.
Each reappraisal block consisted of 40 smoking-related pictures. They were
randomly selected from a list of 120 pictures. This list included the smoking-related
pictures that were presented in the passive viewing condition. Random selection
took place without replacement. This means that there was no overlap between
pictures presented in the three reappraisal blocks; there were no pictures that
were reappraised with more than one strategy. Pictures were presented for 2000
155
Chapter 8
ms. Prior to each picture, a reminder of the reappraisal strategy appeared on the
screen for 1000 ms. Between the reminder and the picture an interval of 500 ms
was used. Intertrial interval was 1000 ms.
At the end of the task participants were asked to fill out another craving
questionnaire. Then electrodes were removed. E-prime® software (Psychology
Software Tools, Pittsburgh, PA, USA) was used for picture presentation.
Neutral stimuli (mean valence level = 5.00, SD = 0.40, range = 4.38-6.21; mean
arousal level = 2.65, SD = 0.57, range 1.55-4.71) were selected from the International
Affective Picture System (IAPS; Lang, 1995). Smoking-related stimuli were
downloaded from public online sources and consisted of cigarettes and people
holding and smoking cigarettes (mean valence level = 5.97, SD = 0.74, range = 5-8;
mean arousal level = 3.30, SD = 1.55, range 1-8). The final stimulus set consisted of
40 neutral pictures and 120 smoking-pictures.
Electroencephalogram (EEG) recording and signal processing
The electroencephalogram (EEG) was recorded using a digital Active-Two system
(BioSemi, Amsterdam, the Netherlands), with active Ag/AgCl electrodes at
34 scalp sites according to the International 10/10 system (32 standard channels
mounted in an elastic cap and two mastoid locations, which were used for off-line
re-referencing; ACNS, 2006). The vertical electro-oculogram (VEOG) was recorded
with two active Ag/AgCl electrodes located above and underneath the left eye. The
horizontal electro-oculogram (HEOG) was recorded with two Ag/AgCl electrodes
located at the outer canthus of each eye. An additional active electrode (CMS –
common mode sense) and a passive electrode (DRL – driven right leg) were used
to comprise a feedback loop for amplifier reference. All signals were digitized with
a sampling rate of 512 Hz, a 24-bit A/D conversion, and a low pass filter of 134 Hz.
Offline, data were processed with BrainVision Analyzer 2 (Brain products GmbH,
Munich, Germany).
The EEG signals were referenced to the mathematically linked mastoids and EEG
and EOG were phase-shift-free filtered using a 0.01–35 Hz (24 dB/Octave roll off)
band-pass filter. EEG and EOG recordings were segmented in 2100 ms epochs,
including 100 ms pre-stimulus baseline. For correction of vertical and horizontal
eye movements and eye blinks we applied automatic processing algorithms,
i.e., Gratton and Coles algorithm (Gratton et al., 1983). All ERPs were baseline
156
Intentional modulation of processing bias
corrected. Artifact rejection criteria were minimum and maximum baseline-topeak −75 to +75 μV, and a maximum allowed voltage skip (gradient) of 50 μV for
each sample point. Epochs were averaged across trials.
Overall grand averages were obtained for each condition, yielding five conditions:
passive-neutral, passive- smoking, reappraisal-pleasant, reappraisal-distraction,
and reappraisal-rational. Numbers of artifact-free epochs were respectively
20.36 (SD = 9.99), 20.42 (SD = 9.46), 20.56 (SD = 9.24), 21.34 (SD = 9.62), and 20.78
(SD = 9.55), and did not differ between stimulus conditions, F(4,196) = 0.58, p > 0.10.
Analyses
Visual inspection of resulting ERPs led to the identification of a clear LPP in the
600-2000 ms time frame. Because previous studies have shown that the scalp
topography of the LPP shifts around 1000 ms (Foti et al., 2009) and because
a cross-over of waves was visually detected around 1000 ms (see figure 1), the
LPP was divided into two components: an early LPP (600-1000 ms) and a late LPP
(1000-2000 ms). This was done in order to investigate the attentive processing
over time. For both LPP components, mean activities (average amplitude in the
time window) were computed per group and stimulus category.
Because the LPP is typically maximal at posterior and parietal electrode sites
(Hajcak et al., 2009), but group differences in ERP studies of addiction are typically
maximal at frontal electrode sites (Franken et al., 2004; Littel & Franken, 2007),
ERP effects were assessed by performing repeated-measurement analyses of
variance (ANOVA) on all four midline electrode sites (Fz, Cz, Pz, and Oz), resulting in
4 (electrode) x 5 (condition) x 2 (group) repeated measures ANOVAs for both the early
LPP and the late LPP. To examine exact differences for the significant interaction
and main effects, pairwise follow-up analyses with Bonferroni correction were
applied to all ANOVAs (Bonferroni corrected p-values are reported). GreenhouseGeisser correction was applied to all ANOVAs if appropriate (uncorrected df’s are
reported). Increases in craving between pre- and post measure, i.e., changes in
QSU-brief scores after all conditions were administered, were calculated with
independent t-tests. An alpha-level of 0.05 was used for all statistical tests.
157
Chapter 8
µV
µV
Cz
Fz
µV
ms
Pz
µV
ms
Oz
ms
ms
Neutral pictures - passive viewing
Smoking pictures - passive viewing
Reappraisal - pleasant
Reappraisal - distraction
Reappraisal - rational
Figure 1. Average event-related potentials (ERPs) in response to the passively viewed
neutral and smoking cues and the cognitively reappraised smoking cues for the pooled
smokers group at midline electrodes (Fz, Cz, Pz, Oz). Left panels depict the early Late
Positive Potential time window (early LPP; 600-1000 ms). Right panels depict the late Late
Positive Potential time window (late LPP; 1000-2000 ms).
158
Intentional modulation of processing bias
Results
Craving
Regular smokers had a mean score of 3.00 (SD = 0.85) on the pre-measure of the
QSU-brief, and a mean score of 4.44 (SD = 0.84) on the post-measure (after all
conditions were administered). This increase reached significance, t(26) = 11.70,
p < 0.001, and appeared to be mainly driven by an increase in scores on the QSUbrief subscale ‘reduction of negative affect and withdrawal craving’, t(26) = 11.76,
p < 0.001. Light smokers had a mean score of 1.69 (SD = 0.74) on the pre-measure,
and a mean score of 2.31 (SD = 1.07) on the post-measure. This increase was
significant, t(21) = 4.09, p < 0.01, and also appeared to be driven by an increase in
withdrawal-craving, t(21) = 4.86, p < 0.001. Increases in QSU-brief scores, both for
the total QSU-brief and the subscales, were significantly larger for regular smokers
than for light smokers, all t’s > 3.11, all p’s < 0.01. See table 1 for all mean scores on
the QSU-brief and its subscales.
Table 1. Mean (SD) craving scores for regular smokers, light smokers, and across all
smokers on the QSU-brief and its subscales
Regular smokers
Light smokers
All smokers
Pre
Post
Pre
Post
Pre
Post
QSU-Total
3.00 (0.85)
4.44 (0.84)
1.69 (0.74)
2.31 (1.07)
2.43 (1.03)
3.49 (1.43)
QSU-Withdrawal
2.97 (0.94)
4.61 (0.89)
1.86 (0.87)
2.59 (1.21)
2.48 (1.06)
3.70 (1.45)
QSU-Desire
3.04 (0.87)
4.28 (0.93)
1.52 (0.73)
2.04 (1.00)
2.37 (1.11)
3.27 (1.48)
Electrophysiological data
Early LPP
A significant main effect was found for Condition, F(4,192) = 12.49, p < 0.001.
See figure 1 (left panels) for average early LPP amplitudes in response to all five
conditions. Post-hoc comparisons revealed that all smokers, both the regular
and the light smokers, displayed enlarged early LPP amplitudes in response to
smoking cues (passive viewing condition) compared to neutral cues, t(48) = 5.20,
p < 0.001. Furthermore, early LPP amplitudes were larger for the three reappraisal
conditions than for the neutral condition, all t’s > 4.01, p’s < 0.01. There were no
significant differences between the reappraisal-distraction condition and the
passive-smoking condition, or between the reappraisal-rational condition and
the passive-smoking condition, both t’s < 1.40, p’s > 0.10. However, the difference
159
Chapter 8
between the reappraisal-pleasant condition and the passive-smoking condition
was significant, t(48) = 3.17, p < 0.05; all smokers displayed more enhanced early
LPP amplitudes in response to smoking pictures that were reappraised utilizing
the pleasant strategy than in response to smoking pictures that were passively
viewed. The main effect for Condition was not moderated by Group (Condition
x Group interaction, F(4,192) = 0.78, p > 0.10, Electrode x Condition x Group
interaction, F(12,576) = 1.06, p > 0.10), indicating that ERP responding to the five
conditions did not differ between regular smokers and light smokers. The main
effect for Condition was accompanied by an Electrode x Condition interaction,
F(12,576) = 10.77, p < 0.001. At Fz, Cz, and Pz electrodes, LPP amplitudes in
response to all smoking stimuli, both passively viewed and reappraised, were
more enlarged than the LPP amplitude in response to neutral cues, all t’s > 3.94,
p’s < 0.01. No effects were found at Oz. The early LPP amplitude in response to the
reappraisal-pleasant condition was more enlarged than the early LPP amplitude
in response to the passive-smoking condition at both Fz and Cz electrodes, both t’s
> 3.57, p’s < 0.01. At Fz, the LPP in response to the reappraisal-rational condition
was larger than the LPP in response to the passive-smoking condition, t(48) = 2.96,
p < 0.05. Furthermore, there was a trend for the LPP in response to the reappraisaldistraction condition to be smaller than the LPP in response to the reappraisalpleasant condition at this electrode, t(48) = 2.71, p < 0.10. See table 2 for all mean
early LPP amplitudes.
Table 2. Mean early LPP amplitudes (SD) in microvolt per electrode and condition
Neutral
Smoking
Reappraisal – Pleasant
Reappraisal – Distraction
Reappraisal – Rational
Fz
-5.45 (6.63)
-0.85 (4.39)
2.39 (5.46)
-0.01 (5.56)
1.68 (5.84)
Cz
-2.14 (4.80)
1.38 (4.58)
4.95 (5.45)
2.99 (5.89)
3.65 (5.71)
Pz
0.62 (4.91)
3.90 (3.96)
5.83 (5.03)
4.70 (5.98)
5.18 (6.18)
Oz
3.35 (4.97)
3.94 (4.80)
2.91 (5.01)
2.25 (8.31)
1.96 (5.52)
Late LPP
A significant main effect of Condition was observed on the late LPP, F(4,192) = 3.90,
p < 0.01. See figure 1 (right panels) for average late LPP amplitudes in response
to all five conditions. Post-hoc analyses showed that the LPP difference between
passively viewed smoking pictures and neutral pictures remained significant,
t(48) = 3.21, p < 0.05, as well as the LPP difference between smoking pictures that
were reappraised with the pleasant strategy and the neutral pictures, t(48) = 3.83,
160
Intentional modulation of processing bias
p < 0.01. Furthermore, the late LPP in response to the passive-smoking condition
did not differ from all three reappraisal conditions (passive-smoking versus
reappraisal-pleasant, t(48) = 0.27, p > 0.10, passive-smoking versus reappraisalrational, t(48) = 1.16, p > 0.10), and passive-smoking versus reappraisaldistraction, t(48) = 2.57, p > 0.10). Importantly, the late LPP in response to the
reappraisal-distraction condition and the late LPP in response to the reappraisalrational condition did not differ anymore from the late LPP in response to the
neutral condition, both t’s < 1.50, p’s > 0.10. This result suggests that after one
second the LPP in response to the reappraised smoking cues was reduced to the
processing level of neutral cues. There was no significant LPP difference between
the reappraisal-pleasant condition and the reappraisal-rational condition, or
between the reappraisal-rational condition and the reappraisal-distraction
condition, both t’s < 1.42, p’s > 0.10. However, the LPP amplitude was significantly
smaller for the reappraisal-distraction condition than for the reappraisal-pleasant
condition, t(48) = 3.07, p < 0.05. The main effect for Condition was not moderated
by Group or Electrode (Condition x Group interaction, F(4,192) = 0.68, p > 0.10;
Electrode x Condition x Group interaction, F(12,576) = 1.07, p > 0.10). However, the
main effect for Condition was moderated by Electrode, F(4,192) = 8.29, p < 0.001.
Post-hoc analyses showed that at both Pz and Oz there were trends for the
reappraisal-distraction strategy to elicit smaller LPP amplitudes than the
passive-smoking condition, t(48) = 2.77, p < 0.10 and t(48) = 2.87, p < 0.10. At
Oz, the late LPP amplitude in response to the reappraisal-distraction strategy
was also significantly smaller than the late LPP amplitude in response to the
neutral condition, t(48) = 4.26, p < 0.01. Furthermore, at Pz, the LPP amplitude
was significantly smaller for the reappraisal-distraction strategy than for the
reappraisal-pleasant strategy, t(48) = 3.11, p < 0.05. These results indicate that at
occipital-parietal sites the reappraisal-distraction strategy not only reduced LPP
responding to the processing level of neutral cues (or beyond), but also reduced
LPP responding compared to the passively viewed smoking cues. However, at these
sites the difference between LPPs elicited by the passive smoking and neutral
cues was not significant, both t’s < 2.70, p’s > 0.10. The late LPP elicited by the
reappraisal-rational strategy was significantly smaller than the late LPP elicited
by the passive-smoking condition at Oz, t(48) = 3.54, p < 0.01, but not at other
electrodes. At Fz and Cz there were no significant differences between conditions,
except for significant differences between reappraised and passive smoking cues
and neutral cues (smoking > neutral). At Cz, however, the reappraisal-distraction
161
Chapter 8
condition was the only condition that did not significantly differ from the neutral
condition, t(48) = 2.29, p > 0.10. See table 3 for all mean late LPP amplitudes.
Table 3. Mean late LPP amplitudes (SD) in microvolt per electrode and condition
Neutral
Smoking
Reappraisal – Pleasant
Reappraisal – Distraction
Reappraisal – Rational
Fz
-2.65 (5.75)
1.41 (4.15)
2.29 (4.58)
1.02 (4.63)
1.80 (5.99)
Cz
-1.17 (4.94)
1.78 (4.93)
3.01 (5.37)
1.19 (5.20)
1.75 (6.08)
Pz
-0.48 (6.56)
2.35 (4.07)
2.60 (5.15)
0.32 (5.04)
1.53 (7.04)
Oz
2.51 (6.69)
1.88 (4.83)
0.08 (5.04)
-1.75 (8.63)
-1.49 (5.85)
Note that findings suggest that the distraction strategy is the best strategy
to reduce late LPP responding to smoking pictures. However, not all tests yield
significant results, probably due to insufficient power. Therefore, an explorative
analysis with fewer conditions was performed, namely a 2 (group) x 4 (electrode)
x 3 (condition; passive neutral, passive smoking, reappraisal-distraction) RM
ANOVA. As expected, this analysis resulted in a significant main effect for
Condition, F(2,96) = 5.51, p < 0.01, with the late LPP amplitude in response to
distraction strategy being significantly reduced compared to the LPP in response
to passively viewing smoking pictures, t(48) = 2.57, p < 0.05.
Discussion
The present study investigated the deliberate, cognitive modulation of attentive
processing of smoking cues in smokers as measured with event-related potentials
(ERPs). The effects of three different reappraisal strategies on LPP magnitude,
an ERP component associated with the allocation of motivated attentional
processes (Hajcak et al., 2009), were investigated. Early and late LPP components
in response to reappraised smoking pictures were compared with early and late
LPP components in response to passively viewed neutral and smoking pictures.
Furthermore, the present study investigated whether regular smokers differed
from light smokers concerning enhanced processing of smoking cues and their
ability to modulate this processing by cognitive reappraisal.
162
Intentional modulation of processing bias
The effects of reappraisal on the electrophysiological processing
of smoking cues
Results indicate that early and late LPP amplitudes in response to smoking
pictures are differentially modulated by different reappraisal strategies. In the
first reappraisal strategy participants were instructed to actively imagine how
pleasant and delicious it would be to smoke the depicted cigarettes (pleasant
strategy). Employing this strategy resulted in more enhanced LPP amplitudes in
response to smoking pictures at fronto-central midline electrodes (Fz, Cz) than
employing no strategy (passively viewing) within 600-1000 ms after picture
presentation. Because this cue-evoked electrophysiological responding has been
associated with enhanced attention toward motivationally relevant stimuli (Hajcak
et al., 2009), or motivated attention, it can be carefully inferred that when smokers
actively imagine how pleasant it would be to smoke, their motivated attention for
smoking cues increases. The second reappraisal strategy (a distraction strategy
in which participants had to focus on the main color in the picture) and the third
reappraisal strategy (a rational strategy in which participants were instructed
to make up a short, rational story about the content of the picture) did not
significantly alter LPP responding in this time window. In other words, early LPP
amplitudes in response to both smoking pictures reappraised with the distraction
strategy and the rational strategy were not enhanced or decreased compared to
early LPP amplitudes evoked by passively viewed smoking pictures. Therefore,
it appears that actively increasing attention for smoking cues is relatively easier
than decreasing attention for smoking cues.
However, within 1000-2000 ms after picture presentation, both the distraction
and the rational strategy were shown to reduce LPP amplitudes. In contrast to
the early timeframe, no significant differences were observed anymore between
LPP amplitudes elicited by smoking pictures that were reappraised utilizing the
distraction and the rational strategy and LPP amplitudes elicited by passively
viewed neutral pictures within the late timeframe. In addition, at occipito-parietal
sites, the LPP amplitudes in response to the distraction strategy showed a trend
to be decreased as compared to the LPP amplitudes elicited by passively viewed
smoking pictures. It must be noted, however, that at occipito-parietal sites there
were no significant differences between the passive viewing conditions (no
enhanced attentive processing of smoking cues compared to neutral cues) and
that across electrodes both strategies did not reduce LPP amplitudes beyond the
processing level of passively viewed smoking pictures. Therefore, there exist two
163
Chapter 8
different, opposing interpretations and implications of the late LPP results which
will be discussed below.
Although the distraction and rational strategies led to a reduction of
electrophysiological processing to the level of neutral stimuli, the late LPP
amplitudes in response to these strategies were not significantly smaller than the
late LPP amplitudes in response to passively viewed smoking pictures. Therefore
it could be argued that the strategies did not decrease the attentive processing of
smoking cues. This would implicate that it is best to apply no strategy at all. However,
the LPP differences between passively viewed smoking pictures and neutral
pictures remained significant throughout the entire timeframe (600-2000 ms).
This means that without employing a cognitive strategy, the LPP amplitude in
response to smoking pictures does not decrease to the same values as the LPP
amplitude in response to neutral pictures. The LPPs in response to pictures
that were reappraised, on the other hand, are reduced to these values; after one
second there is no significant difference between reappraised smoking pictures
and neutral pictures anymore. Therefore it could be argued that the strategies
did decrease the enhanced processing or processing bias (difference in attentive
processing between smoking cues and neutral cues) that is normally observed in
smokers. This would implicate that cognitive strategies might be useful in reducing
attentive processing of smoking-related stimuli.
Although no significant differences were found between late LPP amplitudes
in response to pictures reappraised with the distraction strategy and late LPP
amplitudes in response to pictures reappraised with the rational strategy, there
was a tendency for the distraction strategy to be somewhat more successful in
reducing LPP amplitudes than the rational strategy (see also figure 1, right panels).
The distraction strategy led to significantly smaller late LPP amplitudes than the
pleasant strategy, whereas the rational strategy did not. Furthermore, the LPP
amplitudes in response to the distraction strategy showed a trend to be decreased
as compared to the LPP amplitudes elicited by passively viewed smoking pictures
at both Pz and Oz. For the rational strategy, this finding was only obtained at Oz.
Limitations and recommendations for future research
Overall, the results suggest that electrophysiological responding to smoking
cues can be both enhanced and reduced by intentional, cognitive regulation.
However, in contrast to the pleasant strategy, which significantly increased
164
Intentional modulation of processing bias
electrophysiological responding to smoking cues compared to passively viewing
smoking cues, the rational and distraction strategies did not decrease responding
compared to passively viewing smoking cues, at least not across all electrode sites.
Therefore it is questionable whether these strategies can modulate motivated
attention for smoking cues. As noted above, there are some indications that this
is the case, especially for the distraction strategy. However, replication studies
need to confirm this finding using a slightly different design which can resolve the
interpretational issues encountered in the present study. First of all, we suggest
that replication studies increase the presentation duration of the stimuli. As can be
seen in figure 1, it might be that the LPP in response to the strategies will eventually
(> 2000 ms) be reduced beyond the level of passively viewed smoking pictures.
Secondly, we suggest that replication studies include repetitions of pictures in
both passive conditions. In the present design all reappraisal conditions contained
a number of pictures that were already presented in the passive viewing condition.
This could have increased overall electrophysiological responding to reappraisal
strategies as compared to the passive viewing conditions in which no pictures
were repeated.
Although no significant early LPP differences were observed between the
distraction and rational conditions and passive viewing of smoking cues,
suggesting no significant influence of picture overlap, picture overlap could still
have modulated LPP amplitudes in the late timeframe causing differences between
conditions to fail to reach significance. Finally, we suggest that in replication
research the number of conditions should be reduced. Because of its explorative
nature, the present study yielded five different conditions, thereby reducing
overall power of statistical tests. Results from an additional, explorative analysis
on present data showed that when only three conditions were compared (passive
neutral, passive smoking, reappraisal-distraction), LPP amplitudes in response to
the distraction strategy were significantly reduced as compared to LPP amplitudes
elicited by passively viewing smoking cues. This result implies that the distraction
strategy might be able to reduce enhanced attentive processing of smoking-related
cues. Therefore, future studies should further investigate this specific strategy in
relation to drug-related motivated attention.
Furthermore, although a large body of literature suggests that enlarged LPP
amplitudes are associated with the allocation of attentional resources to
motivationally relevant stimuli (Cuthbert et al., 2000; Hajcak et al., 2009; Schupp
165
Chapter 8
et al., 2000; Schupp, Junghöfer, Weike, & Hamm, 2003b), inferential caution is
warranted. In the present study no behavioral measures of attention were included
and self-reported craving was not administered before and after each reappraisal
block. Therefore one cannot be certain whether the observed enhanced LPP
amplitudes indeed reflect enhanced attentive processing or motivated attention
for the presented smoking cues.
Although we measured craving, which was shown to increase during the task,
it was beyond the scope of our study to examine the associations between the
specific reappraisal strategies and changes in craving levels. It is important
that future studies investigate these relationships in order to shed light on the
relationship between (enhanced/modulated) electrophysiological processing and
motivated attention as well as to be able to draw conclusions about causality. From
the present study, it cannot be inferred whether craving levels were influenced
by successfully and unsuccessfully employing reappraisal strategies or whether
the capability of employing the reappraisal strategies was influenced by craving.
However, there already have been some studies investigating the causal impact
of different cognitive regulation strategies on craving. In two studies by Kober
et al. (Kober et al., 2009; Kober et al., 2010) it was demonstrated that craving
increased as a result of thinking of the direct consequences of smoking a cigarette,
but decreased as a result of picturing the long-term consequences of repeatedly
smoking. Furthermore, in both smokers and cocaine users it was observed that
cognitive regulation strategies decreased activity in the nucleus accumbens and
the orbitofrontal cortex which are implicated in craving and respectively process
the predictive and motivational value of drug-associated stimuli (Kober et al.,
2010; Volkow et al., 2010). These studies adopted strategies in which participants
had to think of negative consequences or actively inhibit craving without specific
instructions and did not investigate drug-related processing biases. In order to
find the most successful reappraisal strategies for reducing both craving and
attentional bias, both associated with drug use and relapse, future studies need
to shed light on whether the abovementioned strategies can additionally lead
to successful reductions in attentional bias and whether distraction can also be
successful is reducing craving. Furthermore, future studies should investigate the
effectiveness of employing laboratory-studied cognitive regulation strategies in
addiction treatment.
166
Intentional modulation of processing bias
Attentive processing of smoking cues and cognitive reappraisal
in light smokers
The present findings show that light smokers and regular smokers do not differ in
LPP amplitude magnitude enhancement in response to smoking cues compared to
neutral cues or reappraised smoking cues compared to passively viewed smoking
and neutral cues. This indicates that regular smokers and light smokers have
comparable levels of motivated attention for smoking-related stimuli. This finding
is consistent with results from a study by Sayette et al. (Sayette et al., 2001) in
which regular smokers and light smokers were equally distracted by smokingrelated material while performing a reaction time task. It is also consistent with
results from a study by Mogg et al. (Mogg et al., 2005) in which low dependent
smokers did not differ from moderately dependent smokers in attentional bias as
measured with a dot probe task. Moreover, present results indicate that regular
smokers and light smokers do not differ in their ability to intentionally modulate
attentive processing of smoking cues, which is in line with observations by Kober
et al. (Kober et al., 2009), who found that light smokers and regular smokers
were equally successful at deliberately regulating craving levels in response to
smoking stimuli.
Despite of the similarities between present and previous results, the finding that
very low to non-dependent smokers do not differ from moderately dependent
smokers with regard to smoking cue-elicited electrophysiological responding
contradicts addiction models of attentional bias and the incentive sensitization
theory of addiction (Franken, 2003; Robinson & Berridge, 1993) in which
attentional bias is perceived as an incentive sensitization mechanism that plays an
important role in maintaining and exacerbating drug dependence. Following these
models one would have predicted that more dependent smokers display more
enhanced LPP amplitudes reflecting more attentive processing of smoking cues.
There have been some indications that low dependent or light smokers even show
increased attentional bias as compared to moderately dependent or heavy smokers
(B. P. Bradley et al., 2003; Hogarth et al., 2003; Mogg et al., 2005; Waters et al., 2003).
These results have been explained by the ‘incentive-habit’ theory of addiction (Di
Chiara, 2000), in which it is hypothesized that when addiction progresses, drug
use behaviors become more automatic and consequently the role of incentive
motivational processes in the maintenance of drug use becomes less important. In
other words, after longer periods of dependence, incentive responding to drug cues
167
Chapter 8
decreases; while at the same time habit responding increases (Mogg et al., 2005).
This incentive-habit theory of addiction also provides a conceivable explanation
for the absence of ERP differences between the regular smokers and light smokers
in the present study. It stands apart that targeting cognitive processing biases in
the treatment of smoking addiction remains important, since increased attention
for drug cues has been associated with increased relapse rates (e.g., Marissen et
al., 2006), and both daily smoking and occasional smoking have been associated
with higher mortality rates and health risks than non-smoking (Coggins, Murrelle,
Carchman, & Heidbreder, 2009).
Clinical relevance
There is mounting evidence suggesting that attentional bias is associated with
drug use (Field et al., 2007; Waters & Feyerabend, 2000) and that its strength
predicts relapse risk (Carpenter et al., 2006; W. M. Cox et al., 2002; W. M. Cox et al.,
2007; Marissen et al., 2006; Streeter et al., 2008; Waters et al., 2003). Therefore,
decrement of attentive processing of drug cues could be an important factor in
cessation. There already have been some indications that processing biases can
be influenced by using cognitive retraining strategies. For example, Fadardi and
Cox (2009) showed that the retraining of attentional bias utilizing a modified
alcohol Stroop task led to a decreased attentional bias for alcohol stimuli and a
reduced alcohol intake for at least three months after retraining. More recently,
Schoenmakers et al. (2010) showed that a visual probe based attentional bias
modification training successfully increased the ability to disengage from alcoholrelated cues, and more importantly, that this effect generalized to new, untrained
stimuli. With regard to smokers, results have been somewhat inconsistent. Two
studies demonstrated positive effects of attentional retraining (Attwood et al.,
2008; Field, Duka, Tyler, & Schoenmakers, 2009), whereas one study failed to find
any effects (McHugh, Murray, Hearon, Calkins, & Otto, 2010). All in all, these results
are promising and imply that attentional retraining sessions might be valuable in
the treatment of addiction. However, these techniques all target attentional bias in
an implicit way.
To the best of our knowledge, present data are the first to indicate that drug-related
attentional processing can also be modulated by explicit cognitive strategies. The
use of explicit strategies to regulate processing bias might be complementary to
or even advantageous over attentional retraining. For attentional retraining to
be successful in clinical practice, the implicitly trained disengagement from drug
168
Intentional modulation of processing bias
cues must not only be observed in the laboratory but also last and generalize
to new and real-life situations. However, it is exactly the literature on the
generalizability that is inconclusive, with the main body of research showing no or
limited generalizability (Field et al., 2007; Field et al., 2007; Schoenmakers, Wiers,
Jones, Bruce, & Jansen, 2007). A possible advantage of utilizing explicit strategies
might be that one can engage in cognitive reappraisal every time one encounters
attention-grabbing drug-related stimuli. Furthermore, cognitive coping strategies,
although not explicitly targeting attentional bias, have already been found to
reduce craving as well as instances of relapse in clinical practice (McCrady &
Ziedonis, 2001; O’Connell, Hosein, Schwartz, & Leibowitz, 2007; Shiffman et al.,
1996). Present results are the first to show that the functionality of these coping
strategies might work through reductions in attention for drug cues which can
be measured on the electrophysiological level and that additionally implementing
explicit strategies of processing bias regulation in the treatment of addiction
might be of clinical relevance.
Conclusions
The present study shows that smoking cue-elicited LPP amplitudes can be
modulated by cognitive strategies, suggesting that attentive processing of smoking
cues can be intentionally regulated. The present findings of LPP modulation fit well
within a larger body of work that has been done in the field of emotion research
(Foti & Hajcak, 2008; Hajcak & Nieuwenhuis, 2006; Krompinger et al., 2008;
MacNamara et al., 2009; MacNamara et al., 2010; Moser et al., 2006). Within this
field it has been repeatedly demonstrated that the LPP is sensitive to interactions
between enhanced processing of emotions and cognitive processes comparable to
the regulation strategies employed in the present study. Instructions to reappraise
or experience emotionally valenced stimuli less or more intensely lead to
significantly reduced or enhanced LPP amplitudes which, moreover, are associated
with reduced or enhanced self-reported valence ratings of the stimuli. Similar to
the finding that people are capable of regulating their attention for motivationally
relevant stimuli, smokers might be able to intentionally regulate their attention
for stimuli that are motivationally relevant to them, i.e., smoking-related stimuli.
The present study is the first to indicate that this is perhaps possible and that the
application of cognitive strategies might be valuable in the treatment of addiction.
There are clear indications that attention for smoking cues can be enhanced by
cognitive strategies. However, it must be noted that it is less clear whether cognitive
strategies are also successful in reducing smoking-related motivated attention.
169
Chapter 8
Although findings do point in this direction, the present study is best considered
preliminary and a starting point for other research on this topic. Future studies
should specifically investigate deliberate distraction as a possible strategy to
reduce motivated attention for smoking cues, increase presentation times to study
later effects of cognitive reappraisal, control for repeated picture presentation
and increase statistical power. Furthermore, the present study shows that regular
smokers with moderate dependence levels (14 cig/day on average) do not differ
from light smokers with very low to absent levels of dependence (5 cig/day,
3 days/week on average) with regard to attention for smoking-related stimuli as
measured on the electrophysiological level. This might be explained by smokers’
decreased incentive responding, or enhanced habit responding, as proposed by
the incentive-habit theory of addiction.
170
Chapter
9
Reduced cognitive processing of alcohol
cues in alcohol-dependent patients
seeking treatment: an ERP study
Littel, M., Field, M., van de Wetering, B.J.M., & Franken, I.H.A.
(submitted for publication). Reduced cognitive processing of alcohol
cues in alcohol-dependent patients seeking treatment: an ERP study.
Chapter 9
Abstract
Substance-dependent individuals have been shown to display increased P3
amplitudes in response to substance-related stimuli. The P3 component of the
event-related potential (ERP) has been associated with attention allocation
and motivational engagement, or ‘motivated attention’ for substance-related
cues. Enhanced processing of substance cues has been observed in heroin- and
cocaine users, cannabis- and cigarette smokers, but has not been unequivocally
demonstrated in alcohol-dependent patients. The main goal of the present
study was to further investigate electrophysiological processing of alcohol and
non-alcohol (soft drink) cues in alcohol-dependent patients and controls. In
addition, it was examined whether groups differed in the processing of positive
emotional cues and whether cue processing was different under implicit and
explicit attention conditions. An oddball paradigm was used in which each
of the stimulus categories served as a target (explicit attention; counting) or
as a non-target (implicit attention; oddball) category. ERPs were recorded to
all stimuli and conditions. Results showed that, across attention conditions,
alcohol-dependent patients did not respond with more enlarged P3 amplitudes
to alcohol cues than soft drink cues. At fronto-central sites they even showed
reduced alcohol cue-elicited P3 amplitudes as compared to controls. These
results are in line with results from studies using behavioral measures of
cognitive processing and might be explained by the use of avoidance strategies,
i.e., patients’ effort to remain abstinent or control their alcohol use at the time
of the experiment. There were no differences between groups regarding the
processing of positive emotional cues. Interpretations and implications of the
findings are discussed.
174
Biased processing in alcohol dependence
Introduction
Substance-dependent individuals show increased P3 and Late Positive Potential
(LPP) amplitudes in response to substance-related stimuli. This finding has been
obtained in heroin- (Franken, Stam et al., 2003) and cocaine users (Franken et al.,
2008), cannabis- (Wölfling et al., 2008) and cigarette smokers (Littel & Franken,
2007). These late component s of the event-related potential (ERP) have been
associated with attention allocation and motivational engagement (Cuthbert
et al., 2000; Hajcak et al., 2010; Schupp et al., 2000), and therefore substance
users’ enhanced late ERP amplitudes reflect their ‘motivated attention’ towards
substance-related cues. This idea is supported by the finding that increased late
ERP amplitudes are associated with self-reported craving (Field et al., 2009) as
well as self-reported valence and arousal of the cues (Franken, Stam et al., 2003;
Herrmann et al., 2000; Littel & Franken, 2007). Motivated attention for substancerelated stimuli is hypothesized to arise from a sensitization of dopaminergic
neurotransmission in the brain through repeated substance use (Robinson &
Berridge, 1993). Because of this sensitization substance cues become extremely
salient and ‘wanted’, i.e., they acquire incentive motivational properties,
and accordingly, an increased amount of cognitive resources is allocated to
these stimuli.
Increased P3 amplitudes to substance cues have been demonstrated to be
present both when attention is explicitly focused on the cues, and also when
attention is directed elsewhere, for example when participants do not need to
pay attention to the cues in order to perform an unrelated task (Fehr et al., 2007;
Littel & Franken, 2010; Lubman et al., 2007; Sokhadze et al., 2008). Therefore,
substance-related stimuli presumably capture the implicit, involuntary attention
of substance users in addition to the possible employment of explicit, voluntary
viewing strategies (Lubman et al., 2007). Furthermore, substance-related
processing bias is presumably selective and specific and is not caused by some
sort of hyperresponsivity to motivationally relevant stimuli in general. In smokers
and abstinent cocaine users, electrophysiological differences between substance
use and control groups have been observed specifically in response to substance
cues, that is, substance users and controls did not differ with regard to their
electrophysiological processing of motivationally relevant (positive, negative) cues
in general (Dunning et al., 2011; Littel & Franken, 2010). However, in both heroin
and current cocaine users absences of differences between electrophysiological
175
Chapter 9
responding to emotional and neutral cues have been observed, implying that
attention to emotional cues might be reduced in substance-dependent populations
(Dunning et al., 2011; Lubman et al., 2008).
Only few studies measured electrophysiological processing of alcohol cues in alcohol
dependence. These studies either found no effects on the P3 or significant effects
were limited to one or two electrode sites. Namkoong et al. (2004) demonstrated
enlarged P3 amplitudes in response to alcohol pictures as compared to neutral
pictures in alcoholics but not controls at Cz and Pz electrodes. Herrmann et al.
(2000) found that alcoholics displayed enhanced electrophysiological responding
to alcohol-related words between 176 and 305 ms and at Pz only. No effects were
found in a timeframe corresponding with the traditional P3 timeframe (> 300 ms).
In a study by Herrmann et al. (2001) significantly enhanced P3 amplitudes were
found for heavy drinkers as compared to light drinkers in response to alcohol
words relative to neutral words at electrode Fz. No P3 differences were observed
between alcoholics and controls in a study by Hansenne et al. (2003).
The main goal of the present study was to investigate electrophysiological
processing of alcohol, non-alcohol (soft drink), and positive cues in alcoholics and
controls under both implicit and explicit attention conditions across a large range
of electrode sites using an oddball paradigm. It was hypothesized that alcoholics
show increased P3 amplitudes in response to alcohol cues compared to soft drink
cues under both attention conditions, whereas control participants show no
discrepancies between these two conditions. In line with studies investigating the
electrophysiological processing of emotional cues (Hajcak et al., 2010), enhanced
P3 amplitudes were expected for positive cues compared to soft drink and alcohol
cues in the control group. No specific predictions were made for alcoholics with
regard to the processing of positive cues relative to soft drink and alcohol cues.
Methods
Participants
Thirty alcohol-dependent patients were recruited at the Bouman addiction
clinic (Rotterdam, the Netherlands). They were treated in out-patient settings
and, according to a medical examination, fulfilled the DSM-IV criteria of alcohol
dependence. Upon arrival in the lab this was confirmed with the Mini International
176
Biased processing in alcohol dependence
Neuropsychiatric Interview (MINI) semi-structured interview (Sheehan et al.,
1998). Exclusion criteria were severe somatic, neurological (e.g., Korsakoff), or
psychiatric (e.g., psychosis, but not mild depression/anxiety) diseases, suicidal
tendencies, aggression, impaired vision, and insufficient command of the Dutch
language. Fifteen patients used medication (anti-psychotics (n = 2), anti-craving
medication (n = 1), anti-depressants (n = 6), sedatives (n = 2), or alcohol withdrawal
medication (n = 4)). Sixty-two percent were abstinent at the time of testing
(5.3 months on average, SD = 3.0, range 0.3-12 months).
The control group (n = 34) was recruited from the local community by
advertisement. Exclusion criteria were lifetime alcohol abuse, lifetime alcohol
dependence, being teetotal (they were required to be familiar with alcohol and its
effects), impaired vision, and language difficulties. They drank 3.8 units of alcohol
(1 unit = 10-13 grams) per week on average (SD = 5.5, range 0-20) and are therefore
considered to be light drinkers. All participants were sober at the time of testing,
which was confirmed with a breath analyzer (Dräger Alcoholtest® 6510, Dräger
Safety AG & Co. KgaA, Luebeck, Germany). Alcoholics and controls did not differ in
age, t(62) = 1.58, ns, years of education, t(62) = 1.18, ns, or gender, χ2(1) =0.48, ns.
See table 1 for all demographics.
The study was conducted in accordance with the Declaration of Helsinki and
all procedures were carried out with the adequate understanding and written
informed consent of the participants. The study was approved by the Medical
Ethical Committee of the Erasmus Medical Center, Rotterdam.
Table 1. Characteristics of alcohol-dependent patients and controls.
Age
Gender (% male)
Years of education
(Prior) Number of drinks/week
MINI
AUDIT
DAQ
Alcoholics (n = 30)
Mean (SD)
49.5 (11.5)
70 %
13.3 (5.0)
67.7 (52.0)
5.8 (1.6)
24.1 (6.8)
2.5 (0.8)
Controls (n = 34)
Mean (SD)
45.6 (7.9)
61.8 %
14.5 (2.8)
3.8 (5.5)
5.2 (3.8)
1.8 (0.7)
Note. MINI = Mini International Neuropsychiatric Interview; AUDIT = Alcohol Use
Disorders Identification Test; DAQ = Desire for Alcohol Questionnaire; 1 drink = 10-13
grams of pure alcohol
177
Chapter 9
Self-report measures
Alcohol craving was measured with the Dutch translation (Franken, Hendriks, &
van den Brink, 2002) of the Desire for Alcohol Questionnaire (DAQ; Love, James,
& Willner, 1998). The DAQ is a 14-item questionnaire and consists of 4 subscales:
(1) strong desires and intentions to drink; (2) negative reinforcement; (3) control
over drinking; and (4) mild desires to drink. As a response format 7-point Likert
scales were used, ranging from “strongly disagree” to “strongly agree”. As
outcome, the mean score across all 14 items was used. Subscale 3 was reversed
before computing the mean.
Alcohol consumption and alcohol-related problems were measured with the
10-item Alcohol Use Disorders Identification Test (AUDIT; Saunders, Aasland,
Babor, de la Fuente, & Grant, 1993). Responses to each question are scored from
0 to 4, giving a maximum possible score of 40.
Patients’ alcohol consumption and dependence were measured with the MINI
(Sheehan et al., 1998). The MINI is a structured diagnostic psychiatric interview
for DSM-IV and ICD-10 criteria. In the present study, only criteria for alcoholdependence were questioned. This part of the MINI is comprised of nine questions.
The first two questions assess alcohol consumption, whereas the last seven
questions assess different criteria of dependence which can be answered with ‘yes’
or ‘no’. Number of ‘yes’ responses were counted, resulting in scores ranging from 0-7.
Arousal and valence properties of the positive, alcohol, and soft drink pictures
were assessed using a computerized Self Assessment Manikin (SAM; M. M. Bradley
& Lang, 1994), which is a non-verbal pictorial assessment technique that directly
measures the pleasure and arousal associated with a person’s affective reaction
to stimuli. The arousal scale ranged from a relaxed, sleepy figure to an excited,
wide-eyed figure; the valence scale ranged from a frowning, unhappy figure to a
smiling, happy figure.
Procedure
Stimuli consisted of 150 neutral pictures and 66 oddball/target pictures: 22 positive
pictures (animals), 22 alcohol pictures (alcoholic beverages), and 22 matched nonalcohol pictures (soft drinks). Neutral and positive pictures were selected from the
International Affective Picture System (IAPS; Lang, 1995). Alcohol and soft drink
pictures were obtained from Pieters et al. (2011).
178
Biased processing in alcohol dependence
The experiment consisted of three separate stimulus conditions. In each condition,
subjects were asked to silently count the (1) animal, (2) alcoholic, or (3) soft
drink pictures within a stream of neutral pictures, whilst the other two oddball
stimuli were also presented, but not counted. Counted oddball pictures were
considered to receive explicit attention (explicit attention condition). Non-counted
oddball pictures were considered to receive implicit attention (implicit attention
condition). ERPs were recorded to both explicitly attended (counted) oddball
stimuli and implicitly attended (non-counted) stimuli. See figure 1 for a visual
representation of the study design.
1HXWUDO
PV
1HXWUDO
PV
1HXWUDO
PV
$OFRKRO &
PV
1HXWUDO
PV
1HXWUDO
PV
3RVLWLYH
PV
1HXWUDO
Figure 1. Study design. Participants were presented with three blocks of frequent neutral
pictures and infrequent (oddball) alcohol, soft drink and animal pictures, all presented for
333 ms. In each block they had to count pictures from one of the three categories (C).
The order of the three stimulus conditions was counterbalanced across subjects.
Within each condition, pictures from every category were repeated three times.
Pictures were presented for 333 ms in a continuous stream without perceivable
inter-stimulus intervals. This fast-stimulus presentation procedure was adopted
from Schupp et al. (2007). The pictures were presented in a perceptually random
order. There were no successions of two or more targets. Afterwards, all target
pictures were rated on their valence and arousal properties on a 7-point scale.
Both for stimulus presentation and valence and arousal ratings e-prime® software
(Psychology Software Tools, Pittsburgh, PA, USA) was used.
179
Chapter 9
ERP recording and analysis
The electroencephalogram (EEG) was recorded with a digital Active-Two system
(BioSemi, Amsterdam, the Netherlands), using active Ag/AgCl electrodes at 34
scalp sites according to the International 10/10 system (32 standard channels
mounted in an elastic cap and two mastoid locations, which were used for off-line
re-referencing; ACNS, 2006). The vertical electro-oculogram (VEOG) was recorded
with two active Ag/AgCl electrodes located above and underneath the left eye. The
horizontal electro-oculogram (HEOG) was recorded with two Ag/AgCl electrodes
located at the outer canthus of each eye. An additional active electrode (CMS –
common mode sense) and a passive electrode (DRL – driven right leg) were used
to comprise a feedback loop for amplifier reference. All signals were digitized with
a sampling rate of 512 Hz, a 24-bit A/D conversion, and a low pass filter of 134 Hz.
Offline, data was processed with BrainVision Analyzer 2 (Brain products GmbH,
Munich, Germany).
First of all, the EEG signals were referenced to the mathematically linked mastoids
and EEG and EOG were phase-shift-free filtered using a 0.1–30 Hz (24 dB/Octave roll
off) band-pass filter. EEG and EOG recordings were segmented in 800 ms epochs,
including 100 ms pre-stimulus baseline. For correction of vertical and horizontal
eye movements and eye blinks we applied automatic processing algorithms, i.e.,
Gratton and Coles algorithm (Gratton et al., 1983). After ocular correction, the
ERPs were baseline corrected. Artifact rejection criteria were minimum and
maximum baseline-to-peak −75 to +75 μV, and a maximum allowed voltage skip
(gradient) of 50 μV. Epochs were averaged across trials. Overall grand averages
were obtained for each attention condition and picture category in the two groups,
yielding six conditions per group: alcohol-implicit (participants were counting
other stimuli), alcohol-explicit (participants were counting alcohol stimuli), soft
drink-implicit, soft drink-explicit, positive-implicit, and positive-explicit. A clear
P3 was identified between 450-650 ms. See figures 2-3. For this component, mean
activities (average amplitude in the time-window) were computed per group,
attention and stimulus category.
ERP effects were assessed with a repeated-measurement analyses of variance
(ANOVA) on crossed lateral and caudal sites, including 15 electrodes (F7, F3, Fz, F4,
F8, T7, C3, Cz, C4, T8, P7, P3, Pz, P4, P8), resulting in a 5 (laterality) x 3 (caudality)
x 2 (attention) x 3 (stimulus) x 2 (group) repeated measures ANOVA. Arousal and
valence ratings of the pictures were analyzed using two 3 (stimulus) x 2 (group)
180
Biased processing in alcohol dependence
repeated-measurement ANOVA’s. Greenhouse-Geisser correction was applied to
all ANOVAs (uncorrected df’s are reported). An alpha-level of 0.05 was used for all
statistical tests.
)]
—9
&]
—9
PV
PV
3]
—9
PV
Figure 2. Event-related potentials (ERPs) of alcohol-dependent patients in response to
alcohol cues (black) and soft drink cues (grey) and controls in response to alcohol cues
(black-dashed) and soft drink cues (grey-dashed) across attention conditions at electrodes
Fz, Cz, and Pz
181
Chapter 9
3]
—9
3]
—9
PV
PV
Figure 3. Event-related potentials (ERPs) in response to counted (explicit), alcohol cues
(black), soft drink cues (grey), and positive cues (light-grey) in controls (left) and alcoholdependent patients (right) at electrode Pz
Results
Self-report data
Alcohol patients showed higher craving levels (pre-test DAQ scores) than
controls, t(62) = 5.07, p < 0.001. With regard to the arousal ratings of the pictures,
a significant main effect of Stimulus was observed, F(2,124) = 19.21, p < 0.001, with
participants rating positive pictures as more arousing than soft drink pictures
(p < 0.001) and alcohol pictures (p < 0.001). There was a trend for soft drink pictures
to be rated as more arousing than alcohol pictures (p = 0.06). No significant Group x
Stimulus effects were observed, F(2,124) = 0.72, ns. A similar pattern was observed
for the valence ratings. There was a significant effect main effect of Stimulus,
F(2,124) = 65.95, p < 0.001. Across participants positive pictures were rated as more
positive than alcohol and soft drink stimuli (both p’s < 0.001). Soft drink pictures
were rated as more positive than alcohol pictures (p < 0.001). No significant Group
x Stimulus effects were observed, F(2,124) = 0.07, ns.
ERP data
On the P3 a significant Group x Stimulus x Caudality interaction was found,
F(4,248) = 3.21, p < 0.05. Post-hoc tests showed that alcoholics show reduced
P3 amplitudes in response to alcohol stimuli as compared to controls at frontal
and central sites (p < 0.05 and p = 0.06). No P3 differences were found between
182
Biased processing in alcohol dependence
groups in response to soft drink or positive pictures. Within the control group
there were no significant differences between alcohol and soft drink stimuli. At
parietal sites, the P3 in response to positive pictures was enlarged as compared
to alcohol and soft drink pictures (both p’s < 0.05). Within the alcohol group there
were also no significant differences between alcohol and soft drink stimuli. At
frontal sites, the P3 in response to positive pictures was reduced as compared to
the P3 in response to soft drink pictures (p < 0.05), whereas at parietal sites the P3
in response to positive pictures was enlarged as compared to both alcohol and soft
drink pictures. No other significant interactions involving Group were observed.
Furthermore, there was a significant main effect of Group, F(1,62) = 3.99, p = 0.05
(controls > alcoholics), a main effect of Stimulus, F(2,124) = 5.05, p < 0.05 (positive
> alcohol, soft drink), a main effect of attention, F(1,62) = 305.18, p < 0.001 (explicit
> implicit), and a significant Stimulus x Attention interaction, F(2,124) = 37.11,
p < 0.001 (explicit attention: positive > alcohol, soft drink; implicit attention: soft
drink > positive, alcohol).
Discussion
Alcohol-dependent patients did not respond with enlarged P3 amplitudes to alcohol
cues as compared to soft drink cues. At fronto-central sites they even showed reduced
alcohol cue-elicited P3 amplitudes as compared to controls, indicating reduced
processing bias or motivated attention for alcohol-related stimuli. Furthermore,
there were no differences between groups in valence- and arousal ratings of the
three picture types. Across groups, alcohol and soft drink pictures were rated as
equally pleasant and arousing. For the control group these results were predicted.
However, for the alcohol-dependent groups present results contradict our
hypotheses as well as results previously obtained in smokers, cannabis-, heroin-,
and cocaine-dependent persons. Furthermore, present results are difficult to
interpret in the light of the incentive-sensitization theory of addiction (Robinson &
Berridge, 1993), since this theory predicts processing bias is directly proportional
to the quantity and frequency of the substance use (Field & Cox, 2008).
As noted before, there has been one study that observed enhanced P3 amplitudes
to alcohol cues in alcohol dependence (Namkoong et al., 2004). However, there also
have been two studies in alcohol-dependent patients that failed to find an effect
on the P3 (Hansenne et al., 2003; Herrmann et al., 2000). The present study even
183
Chapter 9
tends to show opposite results, with alcohol-dependent patients showing reduced
P3 amplitudes in response to alcohol cues as compared to controls. These reduced
P3 amplitudes might indicate that the alcohol-dependent patients in the present
study avoid or disengage attention from alcohol cues.
Inconsistencies with regard to cognitive processing of alcohol-related stimuli in
alcohol dependence have not been confined to ERP measurements, but have also
been obtained in studies using behavioral measures (Stroop-, visual probe-, and
stimulus-response compatibility tasks) of alcohol-related cognitive processing
(attentional bias). For example, a number of studies demonstrated no attentional
bias differences between alcohol-dependent patients and controls (W. M. Cox et
al., 2002; Stetter, Chaluppa, Ackerman, Straube, & Mann, 1994; Vollstädt‐Klein,
Loeber, von der Goltz, Mann, & Kiefer, 2009) and one study observed no differences
between alcoholics and controls with regard to the approach of alcohol (Barkby,
Dickson, Roper, & Field, 2011). Furthermore, in several studies using alcoholrelated visual probe tasks, a vigilance-avoidance pattern of attentional bias was
identified. Attentional bias was demonstrated when stimuli were presented for
a short time interval (50 or 100 ms; Noël et al., 2006; Stormark, Field, Hugdahl, &
Horowitz, 1997; Vollstädt‐Klein et al., 2009), indicating greater initial orienting
toward alcohol cues. However, when stimuli were presented for longer presentation
intervals, attentional bias appeared to be absent (Noël et al., 2006) or even reversed
(Stormark et al., 1997; Townshend & Duka, 2007; Vollstädt‐Klein et al., 2009). This
reversal of attentional bias indicates that, after initial orienting, alcohol-dependent
patients tend to disengage or direct their attention away from alcohol stimuli,
suggesting that alcohol-dependent patients use avoidance strategies to overcome
their bias. Avoidance of alcohol cues might be caused the motivation to remain
abstinent at the time of testing or it could be a consequence of treatment, in which
patients are made explicitly aware of their inability to control their alcohol use.
Vollstadt-Klein et al. (2009) demonstrated that the reversal of attentional bias was
most pronounced in patients with a longer duration of abstinence, i.e., in patients
that presumably received most treatment and/or have the strongest motivation to
become or remain abstinent at the time of testing (or strictly control/limit alcohol
use to a certain maximum). Heavy social drinkers have not been observed to
show avoidance of alcohol cues at longer presentation intervals (Field et al., 2004;
Townshend & Duka, 2001), i.e., they show maintained attention for alcohol cues,
probably because they do not regard their alcohol consumption as problematic and
have no intentions to become abstinent.
184
Biased processing in alcohol dependence
Because in the present study relatively long presentation intervals were used
(333 ms) and because later timeframes of the ERP (> 300 ms) are believed to
reflect more top-down controlled processing (Carretie et al., 2004), the observed
reduced P3 amplitudes might reflect this alcohol-associated avoidance. Although
not all patients in the present study were abstinent, they all underwent treatment
and were motivated to remain abstinent or to limit their drinking. Therefore, it is
plausible that the alcohol-dependent patients used avoidance or other cognitive
strategies to overcome their bias.
Because duration of abstinence, treatment and motivation to quit appear to play
a role in biased cognitive processing in alcohol dependence, differences between
results from Namkoong et al. (2004) and the present study might be explained by
differences between characteristics of the patient samples. For example, patients
in the study by Namkoong et al. were all abstinent, but only for a short duration
(one month on average), whereas in the present study a more heterogeneous
group was studied (non-abstinent patients as well as patients with short and
long durations of abstinence; 5.3 months on average). In addition, patients in the
study by Namkoong et al. did not use medication, whereas fifty percent of the
participants in the present study did. Both these sample characteristics could
have reduced cognitive processing of alcohol cues or strengthened avoidance.
A different explanation for the differences between study results might lie in
the task design. Namkoong et al. (2004) used a blocked design whereas the
present study used an event-related design. Prior studies have demonstrated
that emotional cues become harder to ignore when they are grouped together
into blocks than when they are intermixed with neutral cues (e.g., Holle, Neely,
& Heimberg, 1997). Furthermore, Namkoong et al. used pictures of only one type
of alcoholic drink (Korean traditional liquor), whereas in the present study more
diverse pictures were used (e.g., beer, wine, liquor), perhaps making some of the
pictures less relevant to some of the participants.
We acknowledge that it might be possible that, because of the fast stimulus
presentation rate and perceptual similarities between stimuli, participants
experienced difficulties in discriminating between alcohol and soft drink stimuli.
This could have caused an absence of processing bias in both groups. However,
we believe that this is unlikely for several reasons. If stimuli were indiscernible
then the reversal of processing bias in patients could not have been observed.
In addition, the electrophysiological results were compatible with the arousal-
185
Chapter 9
and valence ratings of the stimuli (in which stimuli were visible until a choice
was made). Furthermore, the task design has been proven effective in studying
attention for emotion (Schupp et al., 2007) and biased attention for smokingrelated stimuli in smokers. Utilizing the same paradigm, Littel and Franken (2010)
demonstrated significant P3 differences between smokers and non-smokers, with
smokers showing larger P3 amplitudes in both explicit and implicit attention
conditions. Finally, the alcohol and soft drink pictures adapted from Pieters et al.
(2011) have been shown to be successful in eliciting attentional bias in high-risk
drinkers using a visual probe paradigm.
Other results
There were no P3 differences between groups for positive cues, indicating no
differences between groups with regard to the processing of motivationally
relevant stimuli in general. This is in line with the study by Littel and Franken
(2010), in which smokers and non-smokers were shown not to differ in their
electrophysiological responses to positive or negative emotional stimuli. In
addition, across conditions patients showed reduced P3 amplitudes as compared
to controls. This result is consistent with the well established findings of previous
ERP studies showing that patients with alcohol dependence as well as their relatives
display smaller P3 amplitudes in response to oddball stimuli than controls (e.g.,
Elmasian, Neville, Woods, Schuckit, & Bloom, 1982; Porjesz et al., 1998).
Conclusions and implications
In the present study no alcohol-associated processing bias was observed in
alcohol-dependent patients. Alcohol-dependent patients even showed reduced
cognitive processing of alcohol cues as compared to light drinking controls. These
results might be explained by the use of avoidance strategies, i.e., patients’ effort
to remain abstinent or control alcohol use at the time of the experiment. This
explanation might also account for the inconsistent results obtained in studies
using behavioral measures to examine cognitive processing of alcohol-related
stimuli in alcohol dependence.
It must be noted that it is remarkable that findings of absent or reversed processing
bias, i.e., avoidance of substance cues, have only been obtained in alcoholdependent patients and not in cocaine- or heroin dependent-patients receiving
treatment (e.g., Franken, Stam et al., 2003; Franken et al., 2008) . Although there
might be a role for publication bias here, differences might also be explained by
186
Biased processing in alcohol dependence
certain characteristics of the different substance-dependent populations (e.g.,
demographics, personality, IQ) or differences in treatment they receive (e.g., focus
on cognitive control). In a recent study by Garland et al. (2011) it was shown that,
among a sample of recovering alcohol-dependent patients, disengagement from
alcohol cues was associated with the trait of mindfulness, which encompasses nonjudgmental awareness of one’s thoughts and actions, attentional and inhibitory
control and cognitive flexibility. Differences between cognitive processing in
alcohol dependence and other types of substance dependence, as well as their
association with specific patient characteristics, need to be addressed in future
studies.
Furthermore, because processing biases in addiction are shown to be associated
with substance use and relapse (Carpenter et al., 2006; W. M. Cox et al., 2002; Garland
et al., 2011; Marissen et al., 2006; Powell, Dawkins, West, Powell, & Pickering, 2010;
Waters et al., 2003) it would be of major importance to further investigate the
relations between treatment, motivation to remain/become abstinent, processing
biases and relapse, for example by investigating processing bias in alcoholdependent persons that have not received treatment yet. Longitudinal studies
might clarify the development or shifts (i.e., waxes and wanes) of processing bias
over time. Furthermore, they might reveal whether alcohol-related processing bias,
despite its observed absence or reversal in experimental settings, still precipitates
relapse in alcohol-dependent patients.
187
Chapter
10
Error-processing and response
inhibition in excessive computer
game players: an ERP study
Littel, M., van den Berg, I., Luijten, M., van Rooij, A.J., Keemink, L.M.,
& Franken, I.H.A. (submitted). Error-processing and response
inhibition in excessive computer game players: an ERP study.
Chapter 10
Abstract
Excessive computer gaming has recently been proposed as a possible
pathological illness. However, research on this topic is still in its infancy and
underlying neurobiological mechanisms have not yet been determined. The
determination underlying mechanisms of excessive gaming might be useful for
the identification of those at-risk, a better understanding of the behavior, and
the development of interventions. Excessive gaming has been often compared
to pathological gambling and substance use disorder. Both disorders are
characterized by high levels of impulsivity, which incorporates deficits in errorprocessing and response inhibition. Because emerging research has shown
a variety of neurobiological similarities between the three disorders, the
present study aimed to investigate error-processing and response inhibition
in excessive gamers and controls using a Go/NoGo paradigm combined with
event-related potential (ERP) recordings. Results indicated that excessive
gamers show substantially reduced fronto-central ERN amplitudes in response
to incorrect trials relative to correct trials, implying poor error-processing in
this population. Furthermore, excessive gamers display higher levels of selfreported impulsivity as well as more impulsive responding as reflected by
less behavioral inhibition on the Go/NoGo task. The present study indicates
that excessive gaming partly parallels impulse control and substance use
disorders regarding impulsivity measured on the self-reported, behavioral
and electrophysiological level. Although the present study does not allow
drawing firm conclusions on causality, it might be that trait impulsivity, poor
error-processing and diminished behavioral response inhibition underlie the
excessive gaming patterns observed in certain individuals. They might be less
sensitive to negative consequences of gaming and therefore continue their
behavior despite adverse consequences.
190
Error-processing and inhibition in excessive gaming
Introduction
Excessive computer game playing
With our increasing use of computers and the internet, the concept of computer
game addiction has recently been proposed as a possible new pathological illness
(Block, 2008). Within the past ten years, both the media and health care have
reported possible problems that can go with excessive computer game playing
(e.g., Desai, Krishnan-Sarin, Cavallo, & Potenza, 2010; Meerkerk, Van den Eijnden,
Vermulst, & Garretsen, 2009; Van Rooij, Schoenmakers, Van den Eijnden, & Van de
Mheen, 2010). In addition, a growing body of research suggests that excessive game
playing is associated with negative outcomes, such as obesity, sleep abnormalities,
job loss, decreased academic achievement, stress, lower psychosocial well-being,
depression and anxiety (Kuss & Griffiths, 2011; Sublette & Mullan, 2010; Weinstein
& Lejoyeux, 2010). While the majority of computer game players experience little
to no disruption of their psychosocial functioning (King, Delfabbro, & Griffiths,
2010), a subset of individuals even develops behaviors characteristic of substance
dependence, including craving, mood modification, withdrawal, tolerance,
preoccupation, conflict, loss of control and relapse (e.g., Charlton & Danforth,
2007; Griffiths & Meredith, 2009; Tejeiro Salguero & Morán, 2002).
Researchers have frequently made the comparison between excessive computer
game playing and pathological gambling, which is currently classified as impulse
control disorder in the Diagnostic and Statistical Manual of Mental Disorders 4th
Edition (DSM-IV; American Psychiatric Association, 1994). Although no psychoactive
substance is ingested, pathological gambling also shares many characteristics with
regular substance use disorders. For example, both types of disorder are characterized
by impaired inhibition and decision making, increased reward seeking, high levels
of self-reported impulsivity and sensation-seeking, and similar neurotransmitter
systems appear to be involved in both disorders (for reviews, see Grant, Potenza,
Weinstein, & Gorelick, 2010; Van Holst, Van den Brink, Veltman, & Goudriaan, 2010).
Accordingly, pathological gambling is considered a behavioral addiction (Van Holst
et al., 2010) and hence it is proposed to be listed under ‘substance use and addictive
disorders’ in the upcoming DSM-V (APA, 2010b). However, excessive computer game
playing, another putative addictive behavior (Grant et al., 2010), is not recognized
as a psychiatric disorder in the DSM-IV (APA, 1994). The American Psychiatric
Association considered it for inclusion in the DSM-V, but decided that there is
not sufficient research at the moment warranting such measures (APA, 2010a).
191
Chapter 10
Although still scarce and fragmented, scientific research in the field of computer
gaming is emerging. Recently, researchers have begun to investigate the
cognitive, neurobiological and neuropsychological processes underlying excessive
computer gaming. Identifying the underlying mechanisms of excessive gaming
is important for a better understanding of the behavior, and might be useful for
the identification of those at-risk for excess and the development of specifically
tailored interventions. It must be noted that a great proportion of research focuses
on excessive internet use in general, which mainly includes online gaming (Van
Rooij, Schoenmakers, Van den Eijnden et al., 2010) among other online activities.
Because of scarcity of research and overlap in studied populations, we will also
refer to studies investigating in excessive internet use in general.
Up until now several interesting results have been obtained. For example, it has
been shown that excessive internet users show abnormal resting state glucose
metabolism in brain regions implicated in reward processing, craving, and
impulsivity (striatum, insula, orbito-frontal cortex; Park et al., 2010), as well as
reduced gray matter density in regions associated with craving and performance
monitoring (insula, anterior cingulate cortex; Y. Zhou et al., 2009). In addition, it
has been demonstrated that excessive gamers and excessive internet users exhibit
reduced levels of dopamine D2 receptor availability in parts of the striatum (Kim
et al., 2011) as well as increased frequency of the DRD2 Taq1A1 allele (Han et al.,
2007), which has been associated with reduced dopamine D2 receptor density.
Both abnormalities have been associated with decreased reward sensitivity and
high susceptibility to impulsive, addictive, and compulsive behaviors (Blum et al.,
2000). Similar abnormalities in brain activity and both a higher prevalence of the
Taq1A1 allele and hypodopaminergic activity have been consistently observed in
substance use and pathological gambling (Blum et al., 1990; Comings et al., 1996;
Goldstein et al., 2007; Volkow, Fowler, Wang, Baler, & Telang, 2009). Therefore
it appears that there also exist parallels between excessive gaming and impulse
control/ substance use disorders on the neurobiological level.
To summarize, although excessive gaming research is still in its infancy, there exist
some preliminary indications that excessive computer gamers show alterations
in brain structure and function implicated in impulsivity as well as heightened
biological susceptibility to impulsive behaviors. Therefore, impulsivity might be
considered as a plausible underlying mechanism of excessive gaming behavior. The
present study aims to further investigate the role of impulsivity in excessive gaming.
192
Error-processing and inhibition in excessive gaming
Impulsivity
Impulsivity has been defined as behavior without adequate thought, although
many other definitions exist (Dom, De Wilde, Hulstijn, & Sabbe, 2007; Moeller,
Barratt, Dougherty, Schmitz, & Swann, 2001; Verdejo-García, Lawrence, & Clark,
2008). It is often viewed as a hallmark personality characteristic in pathological
gambling and substance use disorder. Impulsivity is considered to be a complex
multidimensional concept that incorporates personality, behavioral and biological
components. These various underlying dimensions of impulsivity are assessed with
respectively self-report measures (e.g., impulsivity questionnaires), behavioral
measures (e.g., accuracy and reaction time tasks), and imaging techniques (e.g.,
electroencephalography, fMRI).
With regard to the behavioral dimension, impulsivity is used to describe
maladaptive behaviors including deficits in response inhibition, i.e., the ability
to adaptively suppress behavior when environmental contingencies demand this
(Groman, James, & Jentsch, 2009), and deficits in error-processing, i.e., the ability to
monitor ongoing performance in order to detect and correct errors (Ridderinkhof,
Van den Wildenberg, Segalowitz, & Carter, 2004). It is proposed that in impulsive
behaviors such as drug use, diminished response inhibition and poor errorprocessing contribute to difficulties to resist the consumption of a substance
(Dawe, Gullo, & Loxton, 2004) and the continuation of the behavior despite negative
consequences (Lubman, Yücel, & Pantelis, 2004). Several behavioral studies
in substance-dependent patients have confirmed that substance use is indeed
associated with decreased sensitivity to adverse consequences (Schoenbaum &
Setlow, 2005), making less behavioral adjustments after errors (Luijten, Van Meel,
& Franken, 2011) and comprised response inhibition (Verdejo-García, Perales, &
Pérez-García, 2007).
With regard to the biological dimension of impulsivity, research has focused on
brain reactivity accompanying maladaptive behaviors such as compromised errorprocessing and response inhibition. A frequently used index for measuring these
processes is the measurement of Event-Related Potentials (ERPs). There are two
components of the ERP that have been associated with error-processing, i.e., the
error-related negativity (ERN) and the error positivity (Pe). The ERN is a negative
peak that arises approximately 50-80 ms after making incorrect behavioral
responses and is assumed to reflect fast and automatic initial error detection
(Bernstein, Scheffers, & Coles, 1995; Falkenstein, Hohnsbein, Hoormann, & Blanke,
193
Chapter 10
1991). The ERN is maximal at fronto-central sites and is thought to be generated
in the anterior cingulate cortex (Holroyd & Coles, 2002; Miltner et al., 2003;
Ridderinkhof et al., 2004). The ERN is usually followed by the Pe, which is thought
to reflect the conscious recognition of and reflection on the error (Nieuwenhuis,
Ridderinkhof, Blom, Band, & Kok, 2001; Overbeek, Nieuwenhuis, & Ridderinkhof,
2005) or the motivational significance attributed to the error (Falkenstein,
Hoormann, Christ, & Hohnsbein, 2000; Overbeek et al., 2005; Ridderinkhof,
Ramautar, & Wijnen, 2009). The Pe is maximal at midline parieto-central sites
and emerges approximately 300 ms after the incorrect response (Falkenstein et
al., 2000). It has been observed that during speeded response tasks substancedependent patients respond to errors with reduced ERN and/ or Pe amplitudes. Up
until now, this finding has been obtained in cocaine-dependent patients (Franken,
van Strien, Franzek, & van de Wetering, 2007; Sokhadze, Stewart, Hollifield, &
Tasman, 2008) and in smokers (Franken, van Strien, & Kuijpers, 2010; Luijten, Van
Meel et al., 2011).
Response inhibition is often measured with Go/NoGo tasks, in which participants
have to respond as quickly as possible to frequently occurring ‘Go’ stimuli, but
to inhibit responses to infrequent ‘NoGo’ stimuli. Two major ERP components
have been found to differentiate between these two types of stimuli. The first is
the NoGo N2, a frontally distributed negative waveform emerging approximately
200-300 ms after stimulus onset. The second is the NoGo P3, which is a frontocentrally distributed positive waveform arising 300-500 ms after stimulus onset
(Falkenstein, Hoormann, & Hohnsbein, 1999). Both components are typically much
larger in response to NoGo trials than Go trials. They are thought to be generated
by a distributed network involving the prefrontal areas, the anterior cingulate
cortex, the motor areas, and the parietal regions (Kamarajan et al., 2005). Although
the exact functions of the NoGo N2 and NoGo P3 are still debated, it is proposed
that the N2 reflects an inhibition mechanism active on NoGo trials, whereas the
P3 presumably reflects the reset or closure of a preceding inhibition process
(Falkenstein et al., 1999). Both components have been found to be decreased in
impulsive populations, such as ADHD-patients and impulsive violent offenders (e.g.,
Johnstone, Barry, Markovska, Dimoska, & Clarke, 2009), and mounting evidence
suggests these components to be decreased in substance-dependent patients as
well. Reduced NoGo P3 amplitudes have been demonstrated in smokers (Evans,
Park, Maxfield, & Drobes, 2009), ecstasy users (Gamma, Brandeis, Brandeis, &
Vollenweider, 2005) and alcohol-dependent patients (e.g., Fallgatter, Wiesbeck,
194
Error-processing and inhibition in excessive gaming
Weijers, Boening, & Strik, 1998; Kamarajan et al., 2005). NoGo N2 amplitudes have
been found to be reduced in smokers (Luijten et al., 2011) and heroin-dependent
patients (Yang et al., 2009).
Impulsivity in excessive computer game players
Recent studies using self-report measures have found impulsivity to be positively
correlated with excessive computer game playing and excessive internet use in
general (e.g., Cao, Su, Liu, & Gao, 2007; Meerkerk, Van den Eijnden, Franken, &
Garretsen, 2010). No studies have investigated error-processing and response
inhibition in excessive computer gaming. However, two recent studies have
sought to investigate response inhibition in pathological internet use in general
(Dong, Lu, Zhou, & Zhao, 2010; Z. H. Zhou, Yuan, Yao, Li, & Cheng, 2010). Both
studies utilized a Go/NoGo task and measured behavioral performance as well as
electrophysiological responding to NoGo stimuli. In the study by Dong et al. (2010)
it was shown that in response to NoGo trials excessive internet users displayed
reduced NoGo N2 amplitudes as compared to controls, indicating reduced response
inhibition in this population. However, excessive internet users and controls did
not differ from each other with regard to behavioral performance. In addition,
excessive internet users responded to NoGo trials with increased P3 amplitudes.
Although the authors explain these P3 increases by compromised response
inhibition, the results contradict reduced NoGo P3 amplitudes usually reported in
impulsive and substance-dependent populations. In the study by Zhou et al. (2010)
only the NoGo N2 was investigated. This component was also found to be reduced
in excessive internet users relative to controls. In addition, differences were found
with regard to behavioral performance. Excessive internet users showed higher
false alarm rates (button presses in response to NoGo trials) than controls, again
indicating deficient inhibitory control in this population.
The present study
In the present study both error-processing and response inhibition was investigated
in excessive computer gamers as identified with the Videogame Addiction Test
(VAT; Van Rooij, Schoenmakers, Van den Eijnden, Vermulst, & Van de Mheen,
submitted) and control participants using a Go/NoGo task. ERPs were measured
in response to Go trials and NoGo trials, as well as in response to hits (correct
responses) and false alarms (errors). Since excessive computer gamers have been
found to exhibit higher levels of impulsivity, and since both error-processing and
response inhibition have been found to be comprised in impulsive populations, we
195
Chapter 10
hypothesize excessive gamers to show reduced NoGo N2 and NoGo P3 amplitudes
in response to NoGo trials and reduced ERN and Pe amplitudes in response to
errors. In addition, we predict excessive gamers to show compromised behavioral
performance (reaction times, errors) as well as increased levels of self-reported
impulsivity.
Methods
Instruments
For in- and exclusion of participants, the VAT was used, with its phrasing slightly
adjusted to fit young adult life. This questionnaire is adapted from the adolescent
version of the Compulsive Internet Use Scale or CIUS (Meerkerk et al., 2009) with
items rephrased into the 14 proposed items for the VAT. The VAT was validated
using a sample of adolescents: the scale demonstrated a high reliability (Cronbach’s
α = 0.93) and good construct validity, comparable to the CIUS parent scale (Van
Rooij et al., submitted). VAT scale items are averaged across the scale, resulting
in an answer range from 1 to 5 providing a general index of problematic gaming.
Items include: ‘How often do you find it difficult to stop gaming’ and ‘How often do
you think about gaming, even when you’re not online?’ and answer options range
from ‘never’ to ‘very often’ on a five-point scale.
In order to measure impulsivity, a Dutch translation of The Eysenck Impulsiveness
Questionnaire (I7; Eysenck, Pearson, Easting, & Allsopp, 1985; Lijffijt, Caci, &
Kenemans, 2005) was used. The I7 consists of 54 items assessing impulsivity,
venturesomeness, and empathy. Cronbach’s alpha coefficients range from 0.54 to
0.84 for the subscale ‘Impulsiveness’ (Luengo, Carrillo-De-La-Peña, & Otero, 1991),
which was of primary interest for the present study.
Because heightened impulsivity and compromised error-processing and response
inhibition are also associated with substance use, alcohol consumption and drug
use were additionally assessed. Quantity and frequency of alcohol consumption
were measured utilizing the Quantity-Frequency-Variability Index (QFV-Index;
Lemmens, Tan, & Knibbe, 1992; Meerkerk, Njoo, Bongers, Trienekens, & Van Oers,
1999). In this questionnaire three items are employed in order to determine the
drinking quantity (number of glasses), frequency (drinking days), and variability
(binge drinking) during the last six months.
196
Error-processing and inhibition in excessive gaming
Smoking and drug use were assessed by means of a short questionnaire on substance
use. This questionnaire consisted of simple questions about whether participants
had used various substances or not. Apart from smoking, listed substances were
cannabis, cocaine, amphetamine, ecstasy, hallucinogens (mushrooms), and opiates.
Answers to the smoking question as well as the overall measure of drug use, i.e.,
number of illicit substances ever used, were considered as the variables of interest.
Participants
Fifty-two students participated in the present study. They were recruited at the
Erasmus University Rotterdam and Albeda College Rotterdam and were selected
for screening if they considered themselves to be frequent World of Warcraft
players or non-gamers/very casual gamers (no game specified). They were selected
for participation if they scored > 2.5 (excessive gamers) or < 1.5 (controls) on the
VAT. While no reliable cut-off scores exist yet for the VAT or CIUS, exploratory
analyses suggest that the average VAT scores for problematic ‘addicted’ gamers
(as identified in Van Rooij, Schoenmakers, Vermulst, Van den Eijnden, & Van de
Mheen, 2010) center around 3.0 (95% CI 2.7–3.3, SD = 1.0 n = 46, N = 1352), while
the scores for the largest, non-problematic group center around 1.64 (95%%20CI
1.6-1.7, SD = 0.60, n = 679, N = 1352; Van Rooij, Schoenmakers, Vermulst et al., 2010;
Van Rooij et al., submitted). Given this information, and the aim to distinguish a
non-problematic and a problematic group for comparison, we opted to divide our
sample conservatively in a group scoring < 1.5 and > 2.5 on the VAT.
Excessive gamers (N = 25; 23 males) had a mean age of 20.52 (SD = 2.95). They had
a mean VAT score of 3.05 (SD = 0.42) and played on average 5.05 days per week
(SD = 1.80) and 4.67 hours per day (SD = 2.29). Control participants (N = 27,
10 males) had a mean age of 21.42 (SD = 2.59), a mean VAT score of 1.09 (SD = 0.16),
and played on average 0.28 days/week (SD = 0.58) and 0.47 hours/day (SD = 1.20).
Groups did not significantly differ in age, t(50) = 1.16, p > 0.05. However, the gender
difference was significant, χ2(1) = 16.91, p < 0.001, with the excessive gamers group
consisting of more males and less females than the control group. Excessive gamers
and controls did not differ with regard to smoking behavior, χ2(1) = 0.38, p > 0.05,
illicit drug use, t(50) = 0.60, p > 0.05, and alcohol use; frequency, t(50) = 0.44,
p > 0.05, quantity, t(49) = 0.16, p > 0.05, variability, t(49) = 1.29, p > 0.05. All participants
received either course credit or financial compensation for participation. The
study was conducted in accordance with the Declaration of Helsinki and all
procedures were carried out with the adequate understanding and written
197
Chapter 10
informed consent of the subjects. The study protocol was approved by the ethics
committee of the Institute of Psychology of the Erasmus University Rotterdam.
Task paradigm
Participants were presented with a Go/NoGo paradigm, consisting of four blocks
of 159 letters (636 in total) that appeared one by one on the screen (e.g., A B C D). In
total, 74 letters (11.6%) were repetitions of the previously presented letter (e.g., A
B C C). Participants had to press a button with the right index finger for all letters
(Go trials) but withhold their response for repeated letters (NoGo trials). Letters
were presented for 700 ms, each preceded by a fixation cross, which was displayed
for 300 ms. NoGo trials were presented unpredictably by introducing jitter in the
number of intermitted Go trials. NoGo trials were never presented in succession.
Between blocks, participants received breaks of 60 seconds.
Procedure
After providing informed consent, participants filled out questionnaires on
demographics, play time, drug use, alcohol consumption and impulsivity. Then
electrodes were attached and a gambling task was performed. After this task, the
Go/NoGo task was explained. Participants were instructed to sit still and remain
quiet. It was told that making errors was inevitable, but that it was important
to maintain accuracy throughout the task. First, participants were given the
opportunity to practice in 30 trials. Then the actual task was started. All participants
were tested alone in a sound and light attenuated room. E-prime® software
(Psychology Software Tools, Pittsburgh, PA, USA) was used for letter presentation.
ERP recording and data reduction
The electroencephalogram (EEG) was recorded using a digital Active-Two system
(BioSemi, Amsterdam, the Netherlands), with active Ag/AgCl electrodes at 36
scalp sites according to the International 10/10 system (32 standard channels
mounted in an elastic cap and two mastoid locations, which were used for off-line
re-referencing; ACNS, 2006). The vertical electro-oculogram (VEOG) was recorded
with two active Ag/AgCl electrodes located above and underneath the left eye. The
horizontal electro-oculogram (HEOG) was recorded with two Ag/AgCl electrodes
located at the outer canthus of each eye. An additional active electrode (CMS –
common mode sense) and a passive electrode (DRL – driven right leg) were used
to comprise a feedback loop for amplifier reference. All signals were digitized with
a sampling rate of 512 Hz, a 24-bit A/D conversion, and a low pass filter of 134 Hz.
198
Error-processing and inhibition in excessive gaming
Offline, data were processed with BrainVision Analyzer 2 (Brain products GmbH,
Munich, Germany).
EEG signals were referenced to the mathematically linked mastoids and EEG and
EOG were phase-shift-free filtered using a 0.1–35 Hz (24 dB/Octave roll off) bandpass filter. EEG and EOG recordings were segmented in 1000 ms epochs, including
200 ms pre-response or pre-stimulus baseline. For correction of vertical and
horizontal eye movements and eye blinks Gratton and Coles algorithm (Gratton et
al., 1983) was applied. All ERPs were baseline corrected. Artifact rejection criteria
were minimum and maximum baseline-to-peak −75 to +75 μV, and a maximum
allowed voltage skip (gradient) of 50 μV for each sample point. Epochs were
averaged across trials.
For response inhibition, overall grand averages were obtained from the correct Go
and NoGo trials (segments with incorrect responses, i.e., miss for Go and false alarm
for NoGo, were excluded). Number of artifact-free epochs was 594.00 (SD = 85.34)
for Go trials and 38.46 (SD = 11.01) for NoGo trials. Two participants (one gamer,
one non-gamer) were excluded from Go/NoGo ERP analyses because of less than
10 artifact free epochs in the NoGo condition. These participants were included in
all remaining data analyses, including ERP analyses of error-processing. For errorprocessing, overall grand averages were obtained from correct and incorrect
button presses. Number of artifact-free epochs was 469.81 (SD = 95.30) for correct
trials and 26.46 (SD = 11.09) for incorrect trials.
Statistical analysis
Visual inspection of resulting stimulus-locked ERPs (Go/NoGo) led to the
identification of a clear NoGo N2 in the 220-320 ms time frame as well as a clear
NoGo P3 in the 320-500 ms time frame (see Figure 1). Response-locked ERPs
(correct response/error) revealed a clear ERN in the 0-75 ms time frame and a Pe in
the 200-400 ms time frame (see Figure 2). Because both the NoGo N2 and the ERN
have been found to be maximal at midline fronto-central sites in previous studies
(Falkenstein et al., 1999; Miltner et al., 2003), both these components were studied
at electrodes Fz, FCz, Cz, and CPz. Because both the NoGo P3 and the Pe have been
found to be most pronounced at midline centro-parietal sites in previous studies
(Falkenstein et al., 1999; Nieuwenhuis et al., 2001; Overbeek et al., 2005), these
components were studied at electrodes FCz, Cz, CPz and Pz. However, because
the current source density (CSD) maps for response inhibition indicate additional
199
Chapter 10
activation of parietal brain regions in response to NoGo trials in both groups (see
Figure 3), and because these deviant brain activations to NoGo stimuli have been
observed in fMRI studies in substance-dependent patients before (e.g., Yücel &
Lubman, 2007), the NoGo N2 was additionally and exploratively investigated at
electrode sites PO3, PO4, Pz, P3, P4, CP5, and CP6. These results will be discussed
in Appendix 2.
For all components, mean activities (average amplitude in the time-window) were
computed per group and condition. Repeated Measures Analyses of Variance (RMANOVA) with Greenhouse-Geisser corrected p-values were applied to analyze
the ERP indices of inhibition (N2, P3), error-processing (ERN, Pe), and behavioral
performance measures (number of false alarms, reaction times). This resulted
in a Group (excessive gamers, non-gamer) x Condition (Go, NoGo) x Electrode
ANOVAs for N2 and P3 amplitudes related to inhibition, Group x Correctness
(correct response, error) x Electrode ANOVAs for ERN and Pe amplitudes related
to error-processing, a Group x Error (number of false alarms, misses) and a
Group x Reaction Time (RT Go, RT NoGo) ANOVA for behavioral measures. Posthoc tests for interactions were only performed for interactions including the
between subjects factor Group. Bonferroni corrections for multiple comparisons
were applied to all post-hoc analyses. Questionnaire data were analyzed using
independent t-tests. Because of the significant gender difference between groups,
it was first determined whether Gender was significantly related to the dependent
variables. Therefore, data were first analyzed with RM ANOVAs in which Gender
served as between-subjects variable. These analyses yielded no significant results.
Therefore, only analyses with Group as between subjects factor will be described
(see Appendix for analyses with Gender as between subjects factor). Spearman
correlation coefficients were calculated between number of false alarms, selfreported impulsivity and NoGo N2 and P3 amplitudes in response to NoGo trials
as well as ERN and Pe amplitudes in response to errors. An alpha-level of 0.05 was
used for all statistical tests.
200
Error-processing and inhibition in excessive gaming
Results
Self-reported data
No significant Group differences were found on the I7 subscales ‘venturesomeness’,
t(50) = 1.05, p > 0.05, and ‘empathy’, t(50) = 0.02, p > 0.05. However, on the
impulsivity subscale, excessive gamers were shown to exhibit higher scores
(M = 7.24, SD = 5.00) than non-gamers (M = 4.85, SD = 3.16), t(50) = 2.08, p = 0.04.
Behavioral data
A robust main effect was found for Error, F(1,50)= 125.10, p < 0.001, showing that
participants in general made more errors on NoGo than on Go trials. Furthermore,
a significant Group x Error interaction was found, F(1,50) = 11.44, p < 0.001.
Excessive gamers made more errors in response to NoGo trials (54%, M = 40.12,
SD = 11.93) than non-gamers (41%, M = 30.67, SD = 12.17; p < 0.01), whereas the
Groups did not differ on number of errors in response to Go trials (p = 0.29).
The main effect for Reaction Time (RT) was significant, F(1,50) = 246.96, p < 0.001,
indicating that participants in general were faster on incorrect NoGo trials (false
alarms) than on correct Go trials. A significant Group x RT interaction was found,
F(1,50) = 5.85, p < 0.05. Post hoc tests revealed that excessive gamers were faster
than non-gamers on Go trials (p < 0.05), but not on NoGo trials (false alarms;
p = 0.28).
Significant correlations were found between self-reported Impulsivity and
number of false alarms, ρ = 0.36, p < 0.01, and between number of false alarms
and RT to Go trials, ρ = -0.46, p < 0.01, indicating that both trait impulsivity and
fast response times were associated with making more errors on NoGo trials (i.e.,
speed accuracy trade off).
201
Chapter 10
)]
µV
)&]
µV
ms
&3]
µV
&]
µV
ms
ms
3]
µV
Controls - Error
Controls - Correct response
Gamers - Error
ms
Gamers - Correct response
ms
Figure 1. Average event-related potentials (ERPs) in response to correct Go and NoGo
trials for excessive gamers and controls at Fz, FCz, Cz, CPz, and Pz.
)]
µV
µV
&]
µV
)&]
µV
ms
&3]
µV
ms
ms
3]
Controls - NoGo
Controls - Go
ms
Gamers - NoGo
Gamers - Go
ms
Figure 2. Average event-related potentials (ERPs) in response to errors and correct
responses for excessive gamers and controls at Fz, FCz, Cz, CPz, and Pz.
202
Error-processing and inhibition in excessive gaming
Figure 3. Current Source Density (CSD) maps for error-processing (upper) and response
inhibition (lower).
203
Chapter 10
ERP indices of error-processing
ERN
As expected, a significant main effect was found for Correctness, F(1,50) = 23.13,
p < 0.001, indicating that the ERN was significantly enlarged for errors compared to
correct responses. More important, a significant Group x Correctness interaction
was found, F(1,50) = 10.93, p < 0.01. Post-hoc analyses revealed that excessive
gamers relative to non-gamers show a significantly reduced ERN in response to
errors (p < 0.001), but that both groups do not differ in ERP amplitude elicited by
correct responses (p = 0.08). In addition, it was revealed that excessive gamers
show no significant differences between ERP amplitudes elicited by incorrect and
correct trials (p = 0.30), whereas non-gamers show enhanced ERNs in response to
errors compared to correct responses (p < 0.001).
Pe
A significant main effect was found for Correctness, F(1,50) = 109.05, p < 0.001,
indicating that the Pe was enhanced for errors compared to correct responses.
However, no significant Group x Correctness interaction was found, F(1,50) = 0.83,
p > 0.05. No correlations were found between Impulsivity or number of false
alarms and ERP indices of error-processing (ERN/Pe).
ERP indices of response inhibition
N2
When analyzed at the traditional, fronto-central electrode sites, the main effect
for Inhibition was not significant, F(1,48) = 0.20, p > 0.05, indicating that the NoGo
N2 amplitude did not differ between Go and NoGo trials. Furthermore, no Group x
Inhibition interaction was found, F(1,48) = 3.10, p > 0.05. However, the explorative
N2 analysis at parietal electrode sites yielded several significant results. See
appendix 2.
P3
A significant main effect was found for Inhibition, F(1,48) = 112.43, p < 0.001, with
NoGo trials eliciting larger NoGo P3 amplitudes than Go trials. No significant Group
x Inhibition interaction effect was found, F(1,48) = 0.06, p > 0.05. No correlations
were found between Impulsivity or number of false alarms and ERP indices of
response inhibition (N2/P3).
204
Error-processing and inhibition in excessive gaming
Discussion
The present study aimed to investigate error-processing and response inhibition
in excessive computer gamers and controls without excessive gaming patterns,
both on the behavioral and electrophysiological level, using a Go/NoGo paradigm
combined with ERP recordings. As observed in previous studies (e.g., Meerkerk
et al., 2010), excessive gamers exhibited higher levels of self-reported impulsivity
than controls. In addition, excessive gamers showed faster response rates and
made more errors than controls on the Go/NoGo task. This is in line with results
from a majority of studies in impulsive populations (e.g., Lijffijt, Kenemans,
Verbaten, & Van Engeland, 2005), substance use disorder (Verdejo-García et al.,
2008), pathological gambling (e.g., Kertzman et al., 2008), and excessive internet
use (Z. H. Zhou et al., 2010). It is considered indicative of impulsive responding
prior to the complete processing of the stimulus (Rabbitt & Vyas, 1981) which is
underscored by the finding that self-reported impulsivity correlated significantly
with number of errors in the present study. Most importantly, excessive gamers
showed reduced error-processing as compared to controls. This was reflected
by substantially reduced fronto-central ERN amplitudes in response to incorrect
trials (erroneous button presses) relative to correct trials (correct button
presses). The ERN is an electrophysiological measure associated with errors and
is assumed to reflect fast and automatic initial error detection (Bernstein et al.,
1995; Falkenstein et al., 1991). Compromised error-processing has been observed
in impulsivity-associated externalizing psychopathology (Hall, Bernat, & Patrick,
2007) as well as in substance use disorder (e.g., Luijten, Van Meel et al., 2011).
Therefore, electrophysiological, self-reported and behavioral data of the present
study indicate that excessive gaming might share some similar neuropsychological
and personality characteristics with substance use disorders.
In the present study no differences were found between groups on the Pe. This
component typically follows the ERN and has been argued to be related to error
awareness (Nieuwenhuis et al., 2001; Overbeek et al., 2005) and the motivational
salience of an error (Falkenstein et al., 2000; Overbeek et al., 2005; Ridderinkhof et
al., 2009). Thus, although the early and unconscious stages of error-processing are
compromised in excessive gamers, it appears that the later, more conscious stages
of error-processing are still intact, or at least do not differ from controls. Although
these results might seem counterintuitive, it must be noted that the functional
significance of the Pe is still debated. As reviewed by Overbeek et al. (2005), the
205
Chapter 10
ERN and the Pe show more dissociations than associations, i.e., they appear to be
differentially sensitive to experimental manipulations and individual differences.
In addition to error-processing, response inhibition was investigated. As described
above, behavioral results indicated that excessive gamers show faster reaction
times and made more errors than non-gamers implying deficient inhibitory control.
On the electrophysiological level, stimulus-locked fronto-central NoGo N2 and
centro-parietal NoGo P3 amplitudes in response to NoGo trials relative to Go trials
did not differ between excessive gamers and controls. This finding is incongruent
with results obtained in substance-dependent persons, showing either reduced
NoGo N2 amplitudes (Luijten et al., 2011; Yang et al., 2009), or reduced NoGo P3
amplitudes (Evans et al., 2009; Gamma et al., 2005; Kamarajan et al., 2005) in
response to NoGo trials. It also contradicts results from two studies in excessive
internet users (Dong et al., 2010; Z. H. Zhou et al., 2010) in which an association
was shown between excessive internet use and reduced NoGo N2 amplitudes.
The incongruence between present and previous ERP results as well as the
discrepancy between behavioral and electrophysiological data might have been
caused by certain characteristics of the Go/NoGo task employed in this study.
After all, no main effect of inhibition was found, that is, overall participants did not
respond with increased fronto-central N2 amplitudes to NoGo trials as compared
to Go trials. In traditional ERP studies of response inhibition, NoGo percentages
between 25 and 50 are employed (e.g., Dong et al., 2010; Falkenstein et al., 1999;
Luijten et al., 2011). However, in order to optimize the task for measuring errorprocessing, only 12% of the trials in the current study were NoGo trials. This
probably increased task difficulty, which is endorsed by the finding that both
excessive gamers and controls made many errors (respectively 54% and 41%).
Poor response inhibition in all participants could have led to an absence of frontocentral N2 modulation by Go and NoGo trials.
To our knowledge, this is the first study to show reduced error-processing in
excessive gamers compared to controls. Furthermore, the present study shows
that excessive gamers display higher levels of self-reported impulsivity as well as
more impulsive responding, i.e., less behavioral inhibition on the Go/NoGo task.
Although the present study does not allow drawing conclusions on causality, it
might be that trait impulsivity, poor error-processing and diminished behavioral
response inhibition underlie the pathological gaming patterns observed in excessive
206
Error-processing and inhibition in excessive gaming
gamers. These individuals might be less sensitive to the negative consequences
of gaming and therefore continue their behavior despite adverse consequences.
On the other hand, it cannot be ruled out that reduced error-processing as well
as impulsive responding are the results of excessive gaming or consequences of
excessive gaming (e.g., social isolation, depressive symptoms).
The present findings have several theoretical and clinical implications. With
regard to theoretical implications, results suggest that excessive computer gaming
(partly) parallels impulse control and substance use disorders with regard to
impulsivity as measured on the self-reported, behavioral and electrophysiological
level. Although it is not possible to draw any firm conclusions, present results
provide a good starting point for future research on this topic and might
contribute to the ongoing discussion of excessive gaming or “internet addiction”
as a potentially new psychiatric disorder. With regard to clinical implications,
the present study indicates that behavioral and electrophysiological measures
of impulsivity can be useful for the identification of at-risk populations or atrisk individuals. Subsequently, these populations and individuals can be better
informed or receive specifically tailored interventions. Furthermore, the present
findings might be valuable for the development of prevention and intervention
strategies as well as treatments. For example, excessive gamers seeking treatment
might be offered training programs that focus on the enhancement of inhibitory
control. As noted before, more studies on this topic are necessary. These studies
might use other tools, such as functional magnetic resonance imaging (fMRI), to
further investigate the relationship between the various dimensions of impulsivity
and excessive computer gaming. We also suggest that future research focuses on
direct comparisons between disorders (substance use, gambling, gaming) and
replicate findings in excessive gamers seeking treatment.
207
Chapter 10
Appendix 1
No significant Gender differences were found for smoking, χ2(1) = 0.16, p = 0.69,
illicit drug use, t(50) = 0.50, p = 0.62, and alcohol use (frequency, quantity, and
variability), all t’s < 1.04, all p’s > 0.30. No significant Gender differences were
found on the I7 subscales, all t’s < 1.15, all p’s > 0.15.
With regard to behavioral performance data, the Gender x Error interaction
showed a trend to significance F(1,50) = 3.67, p = 0.06, but post-hoc tests revealed
no significant differences between men and women on number of false alarms
(p = 0.39) and number of misses (p = 0.23). The Gender x Reaction Time interaction
did not reach significance, F(1,50) = 2.61, p = 0.11.
Regarding the ERN, the Gender x Correctness interaction was not significant,
F(1,50) = 0.10, p = 0.75. In addition, no significant Gender x Correctness interaction
was observed for the Pe, F(1,50) = 0.10, p = 0.75.
No significant Gender x Inhibition interaction was observed for the NoGo N2 when
Gender served as between subjects variable, F(1,48) = 0.273, p = 0.60. At parietal
electrode sites, no significant Gender x Inhibition interaction was found either,
F(1,48) = 1.95, p = 0.17. For the NoGo P3 the Gender x Inhibition interaction was
not significant, F(1,48) = 0.03, p = 0.85.
208
Error-processing and inhibition in excessive gaming
Appendix 2
Results parietal N2 analysis
At parietal electrode sites (PO3, PO4, Pz, P3, P4, CP5, and CP6), a significant main
effect for Inhibition was found F(1,48) = 13.26, p < 0.001, indicating that for all
participants the NoGo N2 was enhanced in response to NoGo trials as compared
to Go trials. In addition, a significant Group x Inhibition interaction was found,
F(1,48) = 6.45, p < 0.01. Post-hoc tests revealed that the parietally distributed N2 in
response to NoGo trials was more enhanced for excessive gamers than for controls
(p = 0.04), whereas groups did not differ in their N2 response to Go trials (p = 0.74).
In addition, excessive gamers showed significantly more enhanced parietally
distributed N2 amplitudes in response to NoGo trials than in response to Go trials
(p < 0.001), whereas this difference was not observed in controls (p = 0.43).
Discussion of parietal N2 results
Participants did not respond with increased fronto-central N2 amplitudes to NoGo
trials as compared to Go trials. However, present results show that all participants
responded with increased parietal N2 amplitudes to NoGo trials relative to Go trials.
It has been argued that with attenuated frontal activity, additional recruitment of
compensatory networks, including parietal regions, might be involved in successful
inhibitory control (Bolla et al., 2004; Bunge, Dudukovic, Thomason, Vaidya, &
Gabrieli, 2002; Yücel & Lubman, 2007; Yücel et al., 2007). Indeed, hypo-activation
of fronto-central regions combined with hyperactivation of parietal regions has
been frequently observed in fMRI studies of response inhibition (Chambers,
Garavan, & Bellgrove, 2009; Yücel et al., 2007). In the present study, excessive
gamers showed more parietal activation to NoGo trials than controls. This is in
line with results from fMRI studies in cannabis users (Eldreth, Matochik, Cadet,
& Bolla, 2004), alcoholics (e.g., Desmond et al., 2003), cocaine users (e.g., Hester &
Garavan, 2004) and heroin users (e.g., Yücel et al., 2007), and presumably indicates
stronger and/or less adequate recruitment of compensatory networks in excessive
gamers than in controls. Although interpretive caution is warranted, it might be
that this parietal overcompensation in excessive gamers is caused by neuronal
alterations in fronto-central brain regions, including the ACC, which would be in
line with the deficits in error-processing observed in the present study.
209
Chapter
11
Summary and discussion
Chapter 11
Background
Substance use disorders (SUD) are characterized by biases in the cognitive
processing of substance-related stimuli such as the sight or smell of drugs and drug
paraphernalia. It is hypothesized that SUD individuals automatically detect and
orient their attention toward substance-related stimuli, which in turn diminishes
attentional resources left for alternative cues, enhances drug-related cognitions,
and causes subjective craving (Franken, 2003). These processes are thought to
have mutual excitatory relationships with each other. Consequently the drug user
gets caught in a vicious circle of increasing attention and craving until eventually
the substance is sought out and consumed. The emergence of processing biases
can be explained by the incentive-sensitization theory of addiction (Robinson &
Berridge, 1993), which posits that repeated administration of a substance causes
a sensitization of dopaminergic neurotransmission in the brain. Subsequently,
both the substance itself and the substance-related stimuli acquire incentive
motivational properties. In other words, the sensitized dopaminergic system
causes the substance and substance-related stimuli to be perceived as particularly
salient, reinforcing, and ‘wanted’, which in turn leads to a greater allocation of
attentional resources to them.
Experimental research confirms that SUD individuals exhibit an excessive
attentional focusing on substance-related cues and memorize these at the cost
of other stimuli. Utilizing attention tasks such as the emotional Stroop, dualtask procedures, the flicker-induced change blindness paradigm, and visual
probe and attentional cuing tasks (Field & Cox, 2008), and memory tasks such
as free recall and word-stem completion tasks (Franken et al., 2003; McCusker &
Gettings, 1997), processing biases have been demonstrated in various addiction
disorders. Furthermore, meta-analytic research confirms that these biases are
associated with self-reported craving (Field et al., 2009). Mounting evidence
suggests that cognitive processing biases are important in the development and
maintenance of addiction. Biases in the attentive processing of substance cues
have been associated with poor treatment outcome, relapse following treatment,
and substance consumption behaviors (e.g., Marissen et al., 2006). These findings
emphasize the clinical relevance of investigating in these biases.
A relatively new method to asses cognitive processing of substance cues is the
measurement of Event-Related Potentials (ERPs) using electroencephalography
212
Summary and discussion
(EEG) techniques. ERP methodology provides a potentially more direct assessment
of attentional processing than conventional behavioral (reaction time/ accuracy)
data relying on indirect motor-responses. A recent study shows that the majority
of behavioral measures assessing biased cognitive processing in addiction have
poor internal reliability (Ataya et al., 2011).
ERPs are manifestations of brain activities that occur in preparation for, or in
response to discrete events (Fabiani et al., 2000). They consist of several peaks and
troughs that tend to co vary in response to experimental manipulations. Positive
and negative deflections that have been associated with specific informationprocessing operations are called components (Coles & Rugg, 1995). Components
are often labeled after their polarity (e.g., positive) and relative latency (e.g.,
300 ms) and vary in amplitude which presumably depicts the extent to which a
processing operation is engaged (Kok, 1990). Two late positive ERP components,
i.e., the P3 or P300 and the Late Positive Potential (LPP), have been demonstrated
to be adequate indices of the processing of substance-related stimuli (Franken,
Stam et al., 2003; Van de Laar et al., 2004). In SUD individuals, these components
are typically enhanced in response to substance-related stimuli relative to neutral
stimuli, whereas this discrepancy is not observed in healthy control participants.
Although there is still some debate on the exact meaning of the P3/LPP components,
it is widely believed that they reflect attentive processing as well as the activation
of motivational and arousal systems in the brain (Cuthbert et al., 2000; Hajcak et
al., 2010; Lang et al., 1997; Olofsson et al., 2008; Schupp et al., 2000). In SUD they
are therefore assumed to reflect ‘motivated attention’, or the enhanced value or
motivational significance of substance-related stimuli for SUD individuals, with
enhanced attentional allocation as the underlying mechanism.
Cognitive processing bias in substance use
disorder
In Chapter 1 a meta-analytic investigation was conducted in order to summarize
and quantitatively integrate the increasing amount of knowledge that is gained
from empirical studies addressing the relationship between late ERP amplitudes
and substance-related processing bias in SUD individuals. The primary objective
of this chapter was to compute the overall effect size of late ERP amplitude
213
Chapter 11
differences between substance-related and neutral stimulus conditions and
SUD and control groups. Because of the topographic divergence across studies
addressing electrophysiological processing of substance cues, a second aim of
this chapter was to compute whether the effect was most pronounced at frontal
(Fz) or parietal (Pz) electrode sites. Furthermore, additional stratified moderator
analyses by specific sample and study characteristics, i.e., type of substance used
(stimulants vs. depressants), substance use status (non-abstinent vs. abstinent
for 10-30 days), age, gender and task requirements (active vs. passive paradigms),
were conducted in order to assess potential moderating effects of those variables
on ERP amplitude outcomes and to provide an empirical basis for future ERP
studies in SUD individuals.
Results demonstrate enhanced electrophysiological processing of substancerelated stimuli in substance users relative to controls as reflected by enlarged
P300 and LPP amplitudes. This enhanced processing resembles processing of
stimuli motivationally relevant to all individuals (e.g., emotional stimuli) and
stimuli relevant to motivations that differ between individuals (e.g., disorderspecific stimuli). Furthermore, this enhanced processing has been associated with
self-reported craving and valence- and arousal ratings of the stimuli in several
studies. Therefore, substance users’ enhanced electrophysiological processing of
substance cues can be explained by substance users’ motivated attention. Additional
stratified moderator analyses revealed that both P300 and LPP amplitudes were
not moderated by electrode site (Fz vs. Pz), type of substance used (stimulants
vs. depressants), substance use status (10-30 days abstinent vs. non-abstinent),
age, gender and task requirements (active vs. passive paradigms). These results
indicate that enhanced electrophysiological processing of substance cues appears
characteristic of SUD in general, might be independent of recency of substance use,
age and gender, and presumably occurs irrespective of task demands.
The main conclusion that can be drawn from Chapter 1 is that SUDs are
characterized by enhanced cognitive processing of substance-related stimuli
relative to neutral stimuli or ‘motivated attention for substance cues’ as reflected
by enlarged P300 and LPP amplitudes. As outlined in Chapter 2, the remaining
chapters of the present dissertation were written to further enhance insight in
these electrophysiological correlates of substance-related cognitive processing.
Firstly, however, the validation of the Dutch translation of the frequently used
brief Questionnaire on Smoking Urges (QSU-brief) was described.
214
Summary and discussion
The validation of the Dutch version of the
QSU-brief
In Chapter 3 the factor structure, internal consistency, and validity of the
translated version of the QSU-Brief was examined in a Dutch smokers population
(N= 208) utilizing a cross-sectional design with random subject selection. This
was necessary because it was still unknown whether the questionnaire had
acceptable psychometric properties similar to its original English version (L. S.
Cox et al., 2001) and whether the questionnaire and its subscales could be scored in
a similar fashion. Results showed that the Dutch QSU-Brief displays good internal
consistency, and that scores on the questionnaire are strongly correlated with
three other rating scales for measuring craving, urge, and desire for cigarettes,
and moderately linked to questionnaires tapping related constructs, i.e., cigarette
dependence and number of cigarettes smoked per day. An exploratory factor
analysis revealed a two factor structure slightly deviating from the original, English
version. The first factor is best described by ‘the relief from nicotine withdrawal
or negative affect with an urgent and overwhelming desire to smoke’, whereas the
second factor reflects ‘the desire and intention to smoke’. It was concluded that the
Dutch translation of the QSU-Brief offers a reliable, valid and multi-dimensional
assessment of craving for cigarettes in a general population of young adults and
that this questionnaire is suitable for use in laboratory settings.
Smoking-related processing bias after
prolonged abstinence
In chapter 4 and 5 substance-related cognitive processing after prolonged
abstinence was investigated. The incentive-sensitization theory (Robinson &
Berridge, 1993) not only predicts that repeated substance consumption causes a
sensitization of the dopaminergic system of the brain (causing the substance and
substance-related stimuli to acquire enhanced motivational significance and to
automatically attract attention), but it also implicates that this sensitization is an
irreversible feature of addiction. This would explain why so many SUD individuals
relapse after periods of abstinence. However, it is still unclear whether enhanced
215
Chapter 11
electrophysiological processing of substance cues is a more or less permanent
feature of addiction.
Chapter 4 addressed the question whether smoking-associated processing bias
is still present in ex-smokers after prolonged abstinence (1.4 years on average).
For this purpose, smokers, ex-smokers and never-smokers were compared on
their electrophysiological brain responses (P300 and LPP) to passively viewed
smoking-related and neutral pictures. Furthermore, self-report measures of
nicotine craving and pleasantness ratings of smoking stimuli were obtained.
Results showed that both P300 and LPP amplitudes in response to smoking-related
pictures were significantly more enhanced for smokers than for ex-smokers and
never-smokers at frontal and central sites, whereas the magnitude of the P300
and SPW amplitudes in response to neutral pictures did not differ between the
three groups. Importantly, ex-smokers and never-smokers did not differ in their
electrophysiological response to smoking stimuli. In other words, smokers show
more bias for smoking-related pictures than ex-smokers, whereas smokers and
ex-smokers display the same (low) level of processing bias as never-smokers. In
addition, nicotine-craving ratings and pleasantness ratings of smoking stimuli
were higher in smokers compared to ex-smokers. It was concluded that the
smoking-related craving, pleasantness of smoking stimuli, and smoking-related
processing bias decreases after a period of prolonged abstinence.
Whereas the previously described ERPs can provide us with information about
discrete cortical activations in response to specific events, EEG spectrum measures
can enhance insight into broader changes in brain activation (i.e., mental states)
during prolonged stimulus exposure, such as changes in alertness, vigilance, and
arousal. In Chapter 5 changes in the electroencephalographic (EEG) spectrum
were examined in smokers during a 30-s exposure to a neutral (pen) and a
smoking-related cue (burning cigarette). In order to determine whether these EEG
changes were still present after prolonged abstinence, smokers’ EEG spectrum
changes were compared to those of ex-smokers who quit smoking for 1.4 years
on average. It was demonstrated that smokers, but not ex-smokers, responded
with both a significant self-reported craving increase and a significant beta power
increase between the neutral condition and the smoking condition. Since the beta
band is associated with arousal, attention, alertness, and enhancement of the P300
component of the ERP (Egner & Gruzelier, 2001; Egner & Gruzelier, 2004; Haenschel
et al., 2000; Singer, 1993; Wrobel, 2000), it is suggested that the beta increase in
216
Summary and discussion
response to the smoking cue might reflect an enhanced allocation of cognitive
resources to smoking-related stimuli, i.e., a processing bias, which is an important
feature of substance abuse. Since ex-smokers did not respond to the smoking cue
with either self-reported craving or beta activity enhancement, it is preliminarily
concluded that smoking cues do not arouse ex-smokers (anymore) or capture their
attention as much (anymore) as they do in smokers. Because no control group was
tested in this study, no conclusions can be drawn about similarities between exsmokers and non-smokers with regard to EEG spectrum activity.
Both chapters 4 and 5 indicate that the cognitive processing of smoking-related
stimuli is smaller in ex-smokers than in smokers, and therefore appears to decrease
after prolonged abstinence. Importantly, this finding cannot be explained by
nicotine dependence differences between groups, since ex-smokers and smokers
had similar levels of (retrospectively assessed) nicotine dependence.
The nature and specificity of smoking-related
processing bias
Most studies up until now used passive viewing paradigms in which participants
were instructed to simply view presented stimuli. There were no additional task
demands and stimuli were presented with equal probabilities. An advantage of this
design is that effects can neither be caused by high or low probability of certain
stimuli compared to others nor by specific task instructions that dictate which
stimuli are relevant. However, this passive paradigm also has a disadvantage.
Although results from studies employing this paradigm have all been explained
in terms of substance users’ motivated attention toward substance-related cues,
one cannot be completely sure whether the effects are caused by the motivational
significance of substance stimuli or merely (or additionally) by participants’
intentional viewing strategies (Lubman et al., 2007). In other words, it cannot
be inferred whether the effects arise because of an implicit, involuntary capture
of attention or because of an explicit, voluntary choice to focus attention and
elaborate on substance stimuli.
It Chapter 6 was investigated whether the increased electrophysiological
processing of substance-related stimuli arises because of an implicit attention
217
Chapter 11
capture, an explicit, voluntary choice to elaborate on the content of the stimuli,
or both. Furthermore it was investigated whether substance users’ enhanced
ERP response is uniquely triggered by substance-related cues, i.e., whether there
is a selective bias for substance cues, or whether substance users show this bias
to motivational relevant stimuli in general, including positively or negatively
valenced pictures with certain arousing properties.
Smokers and non-smokers were presented with an oddball task, i.e., a rapid
stream of predominantly neutral pictures, in which attention for oddball stimuli,
i.e., less frequently occurring smoking and other motivationally relevant cues, was
manipulated by instructions. During three different blocks, each of the stimulus
categories once served as a target (explicit attention; counting) and twice as a
non-target (implicit attention; no counting) category. ERPs were recorded to all
targets and non-targets. Results showed that, across attention conditions, the
P300 amplitude to smoking-related cues was more enhanced in smokers than in
non-smokers, thereby replicating the effects obtained in chapter 4. When looking
at the two attention conditions separately, it was observed that smokers, relative
to controls, showed increased P300 amplitudes to smoking cues in both the
explicit condition (when they had to count the smoking pictures) and the implicit
attention condition (when they had to count one of the other two pictures). This
implicates that smoking-related stimuli automatically capture smokers’ attention,
even when attention is explicitly focused on other cues. It also indicates that with
the instruction to focus on the smoking-related stimuli, smokers mobilize more
attentional resources to process these stimuli than non-smokers. These results
show that smoking-related processing bias is both implicit and explicit in nature.
Furthermore, smokers and non-smokers did not differ in their P300 amplitude in
response to positive or negative stimuli. This means that smokers do not have a
bias is the cognitive processing of motivationally relevant stimuli in general, but
only and selectively in the processing of stimuli relevant to their tobacco-addicted
states: the smoking-related stimuli.
To recapitulate, smokers are shown to have a unique and specific bias for smokingrelated material which appears to be both implicit and explicit in nature. These
results are important in the light of influential theories of addiction, which predict
that attention is automatically captured by substance-related stimuli because
of their motivational relevance to SUD individuals (Franken, 2003; Robinson &
Berridge, 1993).
218
Summary and discussion
Higher-order conditioning of smoking-related
processing bias
In all preceding chapters we talked about ‘substance-related stimuli’, that is, stimuli
that are associated with a substance. This association is believed to arise from the
process of classical conditioning. Classical conditioning theory predicts that with
repeated substance use, substance-related stimuli or contexts (conditioned stimuli,
CS) become associated with substance intake (unconditioned stimulus, UCS), and
consequently, in the course of time, these stimuli acquire motivational significance
and evoke conditioned substance responses or cue reactivity, such as subjective
craving, substance seeking behaviors, or changes in physiological measures (e.g.,
Carter & Tiffany, 1999; Drummond et al., 2000; Drummond, 2000; Thewissen et al.,
2005). Once the learning process has taken place and the CS are able to elicit the
conditioned responses, the CS can be paired with new neutral stimuli or contexts,
which will also acquire associative strength and elicit conditioned responses or
cue reactivity. This process is called second-order conditioning (higher-order
conditioning; CS-CS learning) and can lead to unlimited sequences of associations
that presumably contribute to substance-seeking in real world environments
(Everitt et al., 2001; Gewirtz & Davis, 2000; Schindler et al., 2002).
The main goal of Chapter 7 was to examine higher-order learning processes
associated with smoking addiction and its electrophysiological correlates. For this
purpose, a second-order smoking conditioning task utilizing ERP methodology was
designed. Smoking-related and neutral pictures were paired with two geometrical
figures (pyramid, cube). Both smokers’ and non-smokers’ ERPs in response to the
pictures and the preceding figures were recorded throughout the experiment.
Afterwards, participants were asked to rate the figures on craving, arousal, and
valence properties.
First of all, results replicated the previously observed finding that smokers
exhibit a processing bias for smoking-related stimuli (see chapters 4 and 6). At
frontal, central and parietal sites, P3 components of the ERP were larger in
response to smoking pictures than in response to neutral pictures for smokers
compared to non-smokers. This effect was most pronounced during the first half
of the experiment. Furthermore, results showed that, during this first half of the
experiment, smokers responded with more enlarged electrophysiological activity
219
Chapter 11
in response to the figure associated with smoking cues than in response to the
figure associated with neutral cues. This was not only reflected by increased P300
amplitudes of the ERP, but also by increased P200 amplitudes. Although the P200
is reported less often than the P300 in picture processing, there are indications
that this earlier component is also sensitive to automatic attention capture and
could be modulated by stimulus valence (Carretie et al., 2004; Delplanque et al.,
2004; Potts, 2004). Non-smokers did not respond with P300 or P200 amplitude
differences between the smoking-associated figure and the neutral figure during
the first half of the experiment. Therefore, it is concluded that smokers, relative
to non-smokers, show more enhanced associative learning for smoking cues than
for neutral cues even though these cues were never paired directly with an UCS
(i.e., smoking). Besides the electrophysiological evidence, we also observed selfreported evidence of enhanced second-order conditioning in smokers: smokers
reported more cue-elicited craving for the smoking-associated figure compared
to the neutral figure. Furthermore, smokers found the smoking-associated figure
more arousing and more pleasurable than non-smokers.
However, during the second half of the experiment these group differences
disappeared. During the second half, non-smokers’ electrophysiological response
to smoking figure increased as compared to the first half, whereas smokers’
electrophysiological response to the smoking figure decreased (and increased in
response to the neutral figure). Therefore it appears that non-smokers to show a
slow, perhaps ‘normal’, learning pattern when learning the association between
neutral and smoking-related stimuli, whereas smokers show a steep learning
pattern, which ceases after a certain amount of time and is replaced by learning the
associations between neutral stimuli. This change over time might be explained by
the idea that after some time smokers lose their interest in the smoking-associated
figure or the presented stimuli in general. After all, none of the presented stimuli
truly predicted smoking and its pleasurable effects. This explanation is in line
with the observed processing of smoking-related pictures. The P300 amplitude in
response to smoking-related pictures also decreased over time. However, findings
could also be explained by boredom. Because of the many repetitions of the same
pictures and figures participants might eventually become distracted and pay
less attention to the task causing the ERP differences between both groups and all
stimuli to become smaller.
220
Summary and discussion
Overall, results indicate that smokers are able to learn smoking-related highorder associations faster, easier and better than non-smokers, at least during
a short period of time. This is reflected in both self-report and enhanced
electrophysiological processing of neutral stimuli that have been paired with
smoking-related stimuli. Furthermore, results underscore the idea that addiction
affects basic learning and memory systems, and that their neural substrates,
normally involved in obtaining more conventional goals, are recruited by the
substance of abuse (Carretie et al., 2004; Robbins et al., 2008). This is the first
study to directly show the contribution of higher-order conditioning to smoking
addiction in humans. Replication studies are warranted, ideally using a design
in which actual smoking is paired with certain neutral cues, which are in turn
paired with other neutral cues. Although results are preliminary, they may help
in understanding the etiology of smoking addiction and its persistence. Craving
and relapse might not be triggered by concrete cues and contexts only, but also,
or predominantly, by more complex and divergent cues and contexts which do not
necessarily have intrinsic motivational value, but have motivational value that is
acquired through the processes of higher-order conditioning.
Modulation of smoking-related processing bias
by cognitive strategies
Because substance-related processing bias is associated with relapse, it is of
major clinical relevance to investigate ways to reduce or modulate the cognitive
processing of substance-related cues. In emotion research it has been shown that
the later components of the ERP, especially the LPP, can be enhanced and reduced
by the employment of cognitive regulation- or reappraisal strategies (e.g., Hajcak
& Nieuwenhuis, 2006; Krompinger et al., 2008; Moser et al., 2006). Examples are
actively reinterpreting the emotional content of presented stimuli, perceiving the
stimuli from an uninvolved, detached perspective, or seeking distraction in more
neutral aspects of the stimuli.
In Chapter 8 it was investigated whether it is possible for smokers to modulate the
electrophysiological indices of smoking-related cognitive processing (early and
late LPP amplitudes) by the employment of cognitive regulation strategies. The
effects of three different strategies were examined: a) a pleasant strategy, in which
221
Chapter 11
participants were asked to actively imagine how pleasant it would be to smoke the
depicted cigarettes; b) a distraction strategy, in which participants had to actively
focus on alternative (color) cues in the presented pictures; and c) a rational strategy,
in which participants were asked to think of a rational, uninvolved interpretation
of the depicted situation. Early and late LPPs in response to smoking pictures
that were reappraised with three different reappraisal strategies were compared
to LPP components in response to passively viewed, non-regulated neutral and
smoking pictures, which were presented at the beginning of the experiment.
Furthermore, it was tested whether enhanced attentive processing as reflected by
enhanced LPP amplitudes as well as the cognitive modulation of these amplitudes
differs between regular smokers and light smokers.
First of all, results from previous studies were replicated. Smokers responded
with more enhanced P300 and LPP amplitudes to smoking-related stimuli
compared to neutral stimuli. Most importantly, results showed that early and
late LPP amplitudes in response to smoking pictures are differentially modulated
by different reappraisal strategies. Employing the pleasant strategy resulted in
more enhanced LPP amplitudes in response to smoking pictures than employing
no strategy (passively viewing) within 600-1000 ms after picture presentation.
Thus, when smokers actively imagine how pleasant it would be to smoke, their
motivated attention for smoking cues increases relatively quick. Furthermore,
within 1000-2000 ms after picture presentation, both the distraction and the
rational strategy were shown to reduce LPP amplitudes. No significant differences
were observed anymore between late LPP amplitudes elicited by smoking pictures
that were reappraised utilizing the distraction and the rational strategy and late
LPP amplitudes elicited by passively viewed neutral pictures. In other words,
using these two strategies, smoking cues were processed to the same extent as
neutral cues.
Although this is a very promising result, it must be noted that the late LPP amplitudes
in response to the distraction- and rational strategy were not significantly smaller
than the late LPP amplitudes in response to passively viewed smoking pictures.
Therefore it could be argued that the strategies did not decrease the cognitive
processing of smoking cues and that it might be best to apply no strategy at all.
However, the LPP differences between passively viewed smoking pictures and
neutral pictures remained significant throughout the entire timeframe (600-2000
ms). This means that without employing a cognitive strategy, the LPP amplitude
222
Summary and discussion
in response to smoking pictures does not decrease to the same values as the
LPP amplitude in response to neutral pictures. The LPPs in response to pictures
that were reappraised, on the other hand, are reduced to these values; after one
second there is no significant difference between reappraised smoking pictures
and neutral pictures anymore. Therefore it could be argued that the strategies
did decrease the enhanced processing or processing bias (difference in processing
between smoking cues and neutral cues) that is usually observed in smokers. This
would implicate that cognitive strategies still might be useful in reducing cognitive
processing of smoking-related stimuli.
Because the distraction strategy did show a trend towards significance, an extra,
exploratory analysis with fewer conditions (to enhance power) was conducted.
In this analysis, LPP amplitudes in response to the distraction strategy were
compared with the LPP amplitudes in response to the passively viewed neutral
and smoking cues. As predicted, the LPP amplitudes elicited by the regulated
smoking cues were now significantly reduced as compared to the passively viewed
smoking cues.
It can be concluded that, similar to the finding that people are capable of regulating
their attention for motivationally relevant stimuli, smokers might be able to
intentionally regulate their attention for stimuli that are motivationally relevant
to them, i.e., smoking-related stimuli. The present study is the first to indicate that
this is perhaps possible and that the application of cognitive strategies might be
valuable in the treatment of addiction. There are clear indications that attention for
smoking cues can be enhanced by cognitive strategies. However, it must be noted
that it is less clear whether cognitive strategies are also successful in reducing
smoking-related motivated attention. Although findings do point in this direction,
the present study is best considered preliminary and a starting point for other
research on this topic. Future studies should specifically investigate deliberate
distraction as a possible strategy to reduce motivated attention for smoking cues,
since this strategy is proven to be most promising.
Furthermore, results indicate that that regular smokers with moderate dependence
levels (14 cig/day on average) do not differ from light smokers with very low to
absent levels of dependence (5 cig/day, 3 days/week on average) with regard to
attention for smoking-related stimuli as measured on the electrophysiological
level. This finding poses a problem for the incentive-sensitization theory of
223
Chapter 11
addiction, which predicts that dependence levels are positively associated with
processing bias. However, these results can be explained by smokers’ decreased
incentive responding, or enhanced habit responding, as proposed by the incentivehabit theory of addiction (Di Chiara, 2000; Mogg et al., 2005).
Alcohol-related processing bias in alcoholdependent patients
Electrophysiological processing of substance cues is not unequivocally
demonstrated in alcohol-dependent patients. There exists only one study that
observed enhanced P300 amplitudes to alcohol cues as compared to neutral cues
(Namkoong et al., 2004), whereas there are two studies that failed to find an effect
on the P300 (Hansenne et al., 2003; Herrmann et al., 2000). Therefore, Chapter 9
aims to further examine biases in the electrophysiological processing of alcoholrelated stimuli in alcohol dependence. The main goal was to replicate the oddball
study described in Chapter 6. The task was almost identical, except that alcoholrelated stimuli were used instead of smoking-related stimuli. Furthermore, negative
pictures were replaced by neutral (soft drink) pictures. Participants were alcoholdependent patients recruited at several addiction out-patient clinics of the Bouman
GGZ Rotterdam and age-, education-, and gender matched controls. After the task
was finished, all oddball stimuli were rated on arousal and valence properties.
In contrast to the results obtained in individuals dependent on other types
of substances, alcohol-dependent patients did not respond with enlarged P3
amplitudes to alcohol cues as compared to soft drink cues. At fronto-central sites
they even showed reduced alcohol cue-elicited P3 amplitudes as compared to
controls, indicating reduced processing bias or motivated attention for alcoholrelated stimuli. Furthermore, there were no differences between groups in
valence- and arousal ratings of the three picture types. Across groups, alcohol
and soft drink pictures were rated as equally pleasant and arousing. For the
control group these results were predicted. For the alcohol-dependent patients
these results are remarkable and difficult to interpret in the light of the incentivesensitization theory of addiction (Robinson & Berridge, 1993), since this theory
predicts processing bias is directly proportional to the quantity and frequency of
the substance use (Field & Cox, 2008).
224
Summary and discussion
It must be noted that inconsistencies with regard to cognitive processing of alcoholrelated stimuli in alcohol dependence have also been obtained in studies using
behavioral measures (especially visual probe tasks) of alcohol-related cognitive
processing. In several studies, attentional bias was demonstrated when stimuli
were presented for a short time interval (50 or 100 ms; Noël et al., 2006; Stormark
et al., 1997; Vollstädt‐Klein et al., 2009), indicating greater initial orienting toward
alcohol cues. However, when stimuli were presented for longer presentation
intervals, attentional bias appeared to be absent (Noël et al., 2006) or even reversed
(Stormark et al., 1997; Townshend & Duka, 2007; Vollstädt‐Klein et al., 2009). This
reversal of attentional bias indicates that, after initial orienting, alcohol-dependent
patients tend to disengage or direct their attention away from alcohol stimuli,
suggesting that alcohol-dependent patients use avoidance strategies to overcome
their bias. Avoidance of alcohol cues might be caused the motivation to remain
abstinent at the time of testing or it could be a consequence of treatment, in which
patients are made explicitly aware of their inability to control their alcohol use.
All patients in the present study underwent treatment and were motivated to quit
drinking or keep control over their limited alcohol use. Therefore, it is plausible
that they used avoidance or other cognitive strategies to overcome their bias.
Because in the present study relatively long presentation intervals were used (333
ms) and because later timeframes of the ERP (> 300 ms) are believed to reflect
more top-down controlled processing (Carretie et al., 2004), the observed reduced
P3 amplitudes might reflect this alcohol-associated avoidance.
However, it remains quite remarkable that findings of absent or reversed
processing bias, i.e., avoidance of substance cues, have only been obtained in
alcohol-dependent patients and not in cocaine- or heroin-dependent patients
receiving treatment (e.g., Franken, Stam et al., 2003; Franken et al., 2008). Although
there might be a role for publication bias here, differences might also be explained
by certain characteristics of the different substance-dependent populations (e.g.,
demographics, personality, IQ) or differences in treatment they receive (e.g., focus
on cognitive control). In a recent study by Garland et al. (2011) it was shown that,
among a sample of recovering alcohol-dependent patients, disengagement from
alcohol cues was associated with the trait of mindfulness, which encompasses nonjudgmental awareness of one’s thoughts and actions, attentional and inhibitory
control and cognitive flexibility. Differences between cognitive processing in
alcohol dependence and other types of substance dependence, as well as their
225
Chapter 11
association with specific patient characteristics, need to be addressed in future
studies. Furthermore, it would be of major importance to investigate the relations
between treatment, motivation to remain/become abstinent, processing biases
and relapse, for example by investigating processing bias in alcohol-dependent
persons that have not received treatment yet.
Cognitive control in excessive gaming
Not only has it been suggested that in addiction substance-related stimuli become
especially salient and receive more attention than other cues; it is has also been
proposed that at the same time the ability to control behavior is decreased
(Goldstein & Volkow, 2002). According to this view, enhanced motivation and related
cognitive processing biases as well as decreased cognitive control contribute to
substance use behaviors. Research has indeed confirmed that SUD persons show
deficits on different indices of cognitive control, including response inhibition and
error-processing. The same results have been obtained in pathological gambling
(which will be listed under ‘addiction’ in the upcoming DSM-V). These cognitive
control deficits might contribute to difficulties to resist the consumption of a
substance and the continuation of the behavior despite negative consequences. We
recently demonstrated reduced cognitive control in smokers using a Go/NoGo task
combined with ERP measurements (Luijten et al., 2011). The NoGo-N2 component
of the ERP elicited by correctly inhibiting motor responses was observed to be
reduced in smokers relative to non-smokers.
Recently, scientific interest has grown for a putative new behavioral addiction:
excessive computer gaming. Although there are many indications that excessive
gaming can have adverse (psychological, physical, and social) consequences,
research on this topic is still in its infancy and underlying neurobiological
mechanisms have not yet been identified. Identifying underlying mechanisms of
excessive gaming might be useful for a better understanding of the behavior, but
also for the identification of those at-risk and the development of interventions. It
has been demonstrated in a recent study that excessive gamers, like SUD individuals
(see chapters 1, 4-8) show enhanced cognitive processing of game-related stimuli.
As compared to casual gamers, excessive gamers demonstrated enhanced P300
amplitudes in response to game-related cues relative to neutral cues (Thalemann
et al., 2007), which makes this study one of the first to indicate that there might be
226
Summary and discussion
similarities between underlying neurobiological and/or cognitive mechanisms of
excessive gaming and SUD.
In Chapter 10 mechanisms of cognitive control in excessive gaming are examined.
Both error-processing and response inhibition were studied in excessive gamers
and casual or non-playing controls using a Go/NoGo paradigm combined with
ERP recordings. ERPs were compared between errors and correct responses
(error-processing) as well as between Go stimuli and correctly inhibited NoGo
stimuli (response inhibition) in both groups. With regard to error-processing,
ERP components of interest were the error-related negativity (ERN) and the
Positivity associated with errors (Pe), which are believed to reflect error detection
and elaboration on errors, respectively (Bernstein et al., 1995; Falkenstein et al.,
1991; Nieuwenhuis et al., 2001; Overbeek et al., 2005). With regard to response
inhibition, components of interest were the NoGo-N2 and the NoGo-P3, which have
been observed to be enhanced in response to successful response inhibition and
the deeper evaluation of this process, respectively (Falkenstein et al., 1999). In
addition to ERP measurements, the number of errors on the Go/NoGo task was
registered (behavioral measure) as well as self-reported impulsivity.
Results demonstrated that excessive gamers exhibited significantly reduced
ERN amplitudes in response to errors as compared to controls. This indicates
that excessive gamers are less capable to detect their errors, which is believed to
be an indication for reduced sensitivity to adverse consequences. Furthermore,
excessive gamers made more errors on the Go/NoGo task than controls, which
is believed to be an indication for impulsive behavior or reduced response
inhibition. Finally, excessive gamers were shown to have higher levels of selfreported impulsivity. These results suggest that excessive gaming parallels SUD
and pathological gambling with regard to cognitive control as measured on the
self-report-, behavioral-, and electrophysiological level.
No differences between excessive gamers and control were observed on the Pe,
which is difficult to explain. Furthermore, no differences were observed on the
electrophysiological indices of response inhibition (NoGo-P2 and NoGo-P3). The
absence of these latter differences might be explained by certain characteristics of
the task. After all, no main effect of inhibition was found, that is, overall participants
did not respond with increased N2 amplitudes to NoGo trials as compared to
Go trials. In traditional ERP studies of response inhibition much higher NoGo
227
Chapter 11
percentages are employed than in the present study (e.g., Falkenstein et al., 1999;
Luijten et al., 2011). The small number of NoGo trials probably increased task
difficulty, which is endorsed by the finding that both excessive gamers and controls
made many errors. Poor response inhibition in all participants could have led to an
absence of N2 modulation by Go and NoGo trials. When tasks are difficult there is
often compensation by other brain areas in order to successfully complete the task
(Bolla et al., 2004; Bunge et al., 2002; Yücel & Lubman, 2007; Yücel et al., 2007). In the
present study such compensation is observed in the form of increased activation on
parietal electrode sites. However, excessive gamers showed more of this parietal
activation to NoGo trials than controls. This presumably indicates stronger and/or
less adequate recruitment of compensatory networks in excessive gamers than in
controls which in turn might still point to deficits in response inhibition.
Of course, the abovementioned findings and interpretations are quite speculative
in nature. However, the differences between excessive gamers and controls on
three other measures of cognitive control, i.e., the ERN, the number of errors on
the Go/NoGo task and self-reported impulsivity, are relatively large. Although the
present study does not allow drawing conclusions on causality, it might be that
heightened impulsivity, reduced error-processing and diminished behavioral
response inhibition underlie excessive gaming patterns. Excessive gamers might
be less sensitive to the negative consequences of gaming and therefore continue
their behavior despite adverse consequences. On the other hand, it cannot be ruled
out that reduced error-processing as well as impulsive responding are the results
of excessive gaming or consequences of excessive gaming. Longitudinal studies
are warranted, in which gamers are followed over longer periods of time.
Main conclusions
The main conclusion of the present dissertation is that SUDs are characterized
by enhanced electrophysiological processing of substance-related stimuli, that
is, enhanced P300 and LPP amplitudes in response to substance cues compared
to neutral cues. Since these components of the ERP are associated with cognitive
processes (attention) and motivation (craving, arousal, and valence), it can be
additionally concluded that -utilizing a direct measure of mental operations, it is
confirmed that SUDs are characterized by biases in the cognitive processing of
substance cues or ‘motivated attention’.
228
Summary and discussion
Furthermore, it can be concluded from this dissertation that enhanced processing
of substance-related stimuli: a) can be measured on both frontal and parietal
electrode sites; b) is not moderated by type of substance (stimulants vs.
depressants), abstinence (non-abstinent vs. abstinent for 10-30 days), gender, age,
and type of task employed (passive vs. active); c) tends to disappear after prolonged
abstinence (in smokers); d) is both explicit and implicit in nature and therefore
reflects both involuntary attention capture and voluntary deeper elaboration (in
smokers); e) is specific for substance-related motivational stimuli (in smokers);
f) can be higher-order conditioned (in smokers); g) can be intentionally modulated
by cognitive regulation strategies (in smokers); and h) is absent or even reversed
in alcohol-dependent patients receiving treatment.
Furthermore it can be concluded that the Dutch translation of the QSU-brief has good
psychometric properties and is suitable for use in experimental settings. Finally,
it can be concluded that the potentially new psychological disorder or behavioral
addiction ‘excessive gaming’ displays certain similarities with SUD with regard to
biases in the cognitive processing of game-related cues (previous research) as well
as deficits in more general cognitive control (present dissertation).
Discussion and suggestions for future research
As noted before, biases in the cognitive processing of substance-related stimuli
are associated with substance consumption and relapse. This indicates that
investigating in substance-related processing biases is of major clinical importance.
However, these associations have only been obtained using behavioral measures
of cognitive processing (attention tasks). Up until now it has never been directly
demonstrated that enhanced electrophysiological processing of substance-related
stimuli in SUD, as described in this dissertation, is associated with substance
consumption and relapse. Because there is evidence that the behavioral and
electrophysiological measures both reflect motivated attention, such an
association is implicitly assumed. However, for this assumption to hold it should be
directly demonstrated that behavioral and electrophysiological measures indeed
measure (approximately) the same attentional processes.
Unfortunately there exist only two studies that combined behavioral attention
tasks with ERP recordings in the same sample of SUD individuals (Fehr et al.,
229
Chapter 11
2006; Fehr et al., 2007). In both of these studies discrepancies were observed
between groups on electrophysiological measures, while no group differences
were found on the behavioral measures. This might implicate that the two
measures are indicative of different processes. On the other hand, it might be that
electrophysiological measures are more sensitive and/or tap into broader aspects
of cognitive functioning. Indeed, results from the meta-analysis by Field et al.
(2009) demonstrate that ERP measures show a stronger correlation with craving
than behavioral measures, suggesting that ERP measures are most sensitive
in capturing motivated attention for substance cues. However, future studies
combining tasks need to shed light on these issues, for example by recording ERPs in
response to the stimuli presented within an attention task or by correlating results
of behavioral and electrophysiological measurements in the same participants.
Furthermore, it is important to investigate whether enhanced electrophysiological
processing of substance-related stimuli is directly associated with substance
consumption and relapse. In chapter 4 it was demonstrated that ex-smokers with
a mean quit duration of 1.4 years display significantly reduced P300 amplitudes
in response to substance cues relative to current smokers. Moreover, there were
no P300 amplitude differences between ex-smokers and non-smokers anymore.
However, as noted before, it cannot be inferred from this study whether the
observed reduced processing bias is a cause or a consequence of current smoking
status. Does the electrophysiological reactivity return to ‘normal’ after quitting?
Or is reduced electrophysiological reactivity predictive of successful quitting
(and is enhanced electrophysiological reactivity predictive of smoking)? Two ERP
cue-reactivity studies have been conducted providing indications for the latter
(Bartholow et al., 2007; Bartholow et al., 2010). In these studies P300 amplitudes
in response to alcohol cues relative to neutral cues were investigated in individuals
low in alcohol sensitivity (LS) and individuals high in alcohol sensitivity (HS). Low
sensitivity to the acute effects of alcohol is considered a genetically mediated risk
factor for the development of alcohol use disorders. Results demonstrated that
LS individuals displayed more enhanced P300 amplitudes in response to alcohol
cues than HS individuals. Importantly, this pattern remained significant when
differences in recent alcohol use, family history or alcohol dependence levels were
controlled for. Furthermore, Bartholow et al. (2007) observed that alcohol cueelicited P300 amplitude predicted drinking prospectively. These findings indicate
that enhanced P300 cue-reactivity may precede the onset of substance use
disorders, may be predictive of future substance use, and therefore might serve as
230
Summary and discussion
a marker or endophenotype for substance use and dependence. It is necessary that
future studies further investigate this possibility, for example by using substance
cue-elicited P300/LPP amplitudes measured among abstinent substance users as
a predictor of relapse, or by using substance cue-elicited P300/LPP amplitudes
among casual or social substance users as predictor of dependency. This requires
the use of longitudinal research designs.
Although the findings in this dissertation are generally in line with addiction
models of attentional bias (Franken, 2003) and the incentive-sensitization theory
of addiction (Robinson & Berridge, 1993), the results from chapter 4, 8 and 9 pose a
potential problem for certain aspects of these theories. Apart from the presence of
cognitive processing biases in SUD, the theories predict that enhanced substancerelated processing is: a) a permanent feature of addiction; and b) directly
proportional to the quantity and frequency of the substance use. However, results
from this dissertation indicate that substance-related processing decreases after
prolonged abstinence (irrespective of dependence levels), that substance-related
processing bias is of equal magnitude in low and medium dependent smokers,
and that it is absent or even reversed in alcohol-dependent patients receiving
treatment (whereas it has been shown to be present in heavy social drinkers as
opposed to light drinkers in previous research). Therefore, instead of a linear
pattern, cognitive processing bias in SUD perhaps follows an inverted-U function,
with non-dependent persons showing no bias, light-dependent persons showing
increasing bias, until an optimum is reached in medium to heavy dependence, after
which the bias decreases to (or beyond) the level of non-dependent persons.
The proposed inverted-U function model of substance-related processing bias would
be in accordance with the ‘incentive-habit’ theory of addiction (Di Chiara, 2000), in
which it is hypothesized that when addiction progresses, substance use behaviors
become more automatic and consequently the role of incentive motivational
processes in the maintenance of substance use becomes less important. In other
words, after longer periods of dependence, incentive responding to substance
cues decreases; while at the same time habit responding increases (Mogg et al.,
2005). Furthermore, heavier users, as opposed to light users, might consider
their consumption as problematic and therefore may be trying to suppress their
subjective craving or to disengage their attention from substance-related stimuli.
This might be especially true for SUD individuals seeking treatment; they probably
have the strongest motivation to become or remain abstinent (or control or limit
231
Chapter 11
their use) and might have learned to control their biases in treatment programs. As
we observed in chapter 8, the active employment of cognitive regulation strategies
can indeed reduce (electrophysiological) processing of substance cues as well as
subjective craving (Kober et al., 2009; Kober et al., 2010).
The exact shape as well as the optimum of the inverted-U function probably differs
between different types of SUD. For example, whereas research has consistently
shown that heavy drinkers show more bias than light drinkers (Field et al., 2004;
Townshend & Duka, 2001), there are studies that indicate that regular and light
smokers have comparable levels of processing bias (see chapter 8, but also Mogg et
al., 2005; Sayette et al., 2001), or even that regular smokers have reduced bias as
compared to light smokers (B. P. Bradley et al., 2003; Hogarth et al., 2003; Mogg et
al., 2005; Waters et al., 2003). Furthermore, the reversal of processing bias after
prolonged dependence has, up until now, only been observed in alcohol dependence.
Substance-related processing bias differences between different types of SUD
might be explained by certain characteristics of the specific SUD populations (e.g.,
demographics, personality, IQ), the treatment they receive (e.g., focus on cognitive
control), or specific substance action, addictiveness and availability (and therefore
the degree or strength of habit-learning).
Processing bias and relapse have been observed to be associated in several studies.
However, if processing bias decreases after prolonged dependence and prolonged
abstinence, then what is the exact role of processing bias in relapse? What is the
nature of the association? It remains plausible that, despite an observed reduction
or reversal of substance-related processing biases in laboratory settings, these
biases wax and wane over time in real life and therefore still precipitate relapse
in heavier users or patients trying to remain abstinent. It would be of major
importance to further investigate the associations between processing biases and
relapse, its exact nature, as well as the roles of specific patient characteristics,
dependence levels, treatment and motivation to remain or become abstinent.
Finally, as can be inferred from chapter 8 of this dissertation, it seems fruitful to
further investigate the effects of cognitive reappraisal strategies on substancerelated cognitive processing biases. Replication studies in other SUDs (other than
smoking) are necessary as well as studies investigating in the effects on the long
term. Because the distraction strategy was observed to be the most promising
strategy, future studies should especially focus on the effects of distraction on
232
Summary and discussion
substance-related processing. Previous studies have demonstrated that cognitive
regulation strategies can also be used to modulate subjective craving (Kober et
al., 2009; Kober et al., 2010). Participants in these studies were instructed to think
about the (positive) short-term effects of consuming the substance (increase of
craving) or to think about the (negative) long-term effects of consuming the
substance (decrease of craving). It would be interesting to see whether these two
strategies also have modulating effects on cognitive processing bias. The other
way around, it would be interesting to see whether the strategies used in chapter 8
are also suitable to modulate subjective craving. Eventually, the effects of the
cognitive strategies that appear most successful in reducing processing bias and
craving should be examined in clinical practice.
233
References
References
ACNS (2006). American clinical neurophysiology society guideline 5: Guidelines for
standard electrode position nomenclature. Journal of Clinical Neurophysiology, 23, 107-110.
Adan, A., Prat, G., & Sanchez-Turet, M. (2004). Effects of nicotine dependence on diurnal
variations of subjective activation and mood. Addiction, 99(12), 1599-1607.
Allen, S. S., Bade, T., Hatsukami, D., & Center, B. (2008). Craving, withdrawal, and smoking
urges on days immediately prior to smoking relapse. Nicotine & Tobacco Research, 10(1),
35-45.
American Psychiatric Association (1994). Diagnostic and statistical manual of mental
disorders (4th ed.). Washington, DC: American Psychiatric Association.
American Psychiatric Association (2010a). APA announces draft diagnostic criteria for DSM5: New proposed changes posted for leading manual of mental disorders., June 15, 2011, from
http://www.dsm5.org/Newsroom/Documents/Diag%20%20Criteria%20General%20
FINAL%202.05.pdf
American Psychiatric Association (2010b). Substance use and addictive
disorders., June 15, 2011, from http://www.dsm5.org/proposedrevision/Pages/
SubstanceUseandAddictiveDisorders.aspx
Arends, L. R., Voko, Z., & Stijnen, T. (2003). Combining multiple outcome measures in a
meta-analysis: An application. Statistics in Medicine, 22(8), 1335-1353.
Ataya, A. F., Adams, S., Mullings, E., Cooper, R. M., Attwood, A. S., & Munafò, M. R. (2011).
Internal reliability of measures of substance-related cognitive bias. Drug and Alcohol
Dependence,
Attwood, A. S., O’Sullivan, H., Leonards, U., Mackintosh, B., & Munafò, M. R. (2008). Attentional
bias training and cue reactivity in cigarette smokers. Addiction, 103(11), 1875-1882.
Baer, J. S., & Lichtenstein, E. (1988). Classification and prediction of smoking relapse
episodes: An exploration of individual differences. Journal of Consulting and Clinical
Psychology, 56(1), 104-110.
Baker, T. B., Morse, E., & Sherman, J. E. (1986). The motivation to use drugs: A psychobiological
analysis of urges. Nebraska Symposium on Motivation, 34, 257-323.
Barkby, H., Dickson, J. M., Roper, L., & Field, M. (2011). To approach or avoid alcohol?
automatic and self-reported motivational tendencies in alcohol dependence. Alcoholism,
Clinical and Experimental Research, DOI: 10.1111/j.1530-0277.2011.01620.x
Bartholow, B. D., Henry, E. A., & Lust, S. A. (2007). Effects of alcohol sensitivity on P3 eventrelated potential reactivity to alcohol cues. Psychology of Addictive Behaviors, 21(4), 555-563.
235
References
Bartholow, B. D., Lust, S. A., & Tragesser, S. L. (2010). Specificity of P3 event-related potential
reactivity to alcohol cues in individuals low in alcohol sensitivity. Psychology of Addictive
Behaviors, 24(2), 220-228.
Bauer, D., & Cox, W. M. (1998). Alcohol-related words are distracting to both alcohol abusers
and non-abusers in the stroop colour-naming task. Addiction, 93(10), 1539-42.
Bernstein, P. S., Scheffers, M. K., & Coles, M. G. H. (1995). “Where did I go wrong?” A
psychophysiological analysis of error detection. Journal of Experimental Psychology: Human
Perception and Performance, 21(6), 1312-1322.
Bevins, R. A., & Palmatier, M. I. (2004). Extending the role of associative learning processes
in nicotine addiction. Behavioral and Cognitive Neuroscience Reviews, 3, 143-158.
Block, J. J. (2008). Issues for DSM-V: Internet addiction. American Journal of Psychiatry,
165(3), 306-307.
Blum, K., Braverman, E. R., Holder, J. M., Lubar, J. F., Monastra, V. J., Miller, D., et al. (2000).
Reward deficicency syndrome: A biogenetic model for the diagnosis and treatment of
impulsive, addictive, and compulsive behaviors. Journal of Psychoactive Drugs, 32(1-68)
Blum, K., Noble, E. P., Sheridan, P. J., Montgomery, A., Ritchie, T., Jagadeeswaran, P., et al.
(1990). Allelic association of human dopamine D2 receptor gene in alcoholism. Journal of
the American Medical Association, 263(15), 2055-2060.
Bolla, K. I., Ernst, M., Kiehl, K., Mouratidis, M., Eldreth, D. A., Contoreggi, C., et al. (2004).
Prefrontal cortical dysfunction in abstinent cocaine abusers. Journal of Neuropsychiatry
and Clinical Neurosciences, 16(4), 456-464.
Boon, M. T. G., & Peeters, F. P. M. L. (1999). Affectieve dimensies bij depressie en angst.
Tijdschrift Voor Psychiatrie, 41(2), 109-113.
Bradley, B. P., Field, M., Healy, H., & Mogg, K. (2008). Do the affective properties of smokingrelated cues influence attentional and approach biases in cigarette smokers? Journal of
Psychopharmacology, 22(7), 737-745.
Bradley, B. P., Field, M., Mogg, K., & De Houwer, J. (2004). Attentional and evaluative biases
for smoking cues in nicotine dependence: Component processes of biases in visual orienting.
Behavioural Pharmacology, 15(1), 29-36.
Bradley, B. P., Mogg, K., Wright, T., & Field, M. (2003). Attentional bias in drug dependence:
Vigilance for cigarette-related cues in smokers. Psychology of Addictive Behaviors, 17(1),
66-72.
Bradley, M. M., Hamby, S., Low, A., & Lang, P. J. (2007). Brain potentials in perception:
Picture complexity and emotional arousal. Psychophysiology, 44(3), 364-373.
Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: The self-assessment manikin and
the semantic differential. Journal of Behavior Therapy and Experimental Psychiatry, 25(1),
49-59.
236
References
Bunge, S. A., Dudukovic, N. M., Thomason, M. E., Vaidya, C. J., & Gabrieli, J. D. E. (2002).
Immature frontal lobe contributions to cognitive control in children: Evidence from fMRI.
Neuron, 33(2), 301-311.
Caggiula, A. R., Donny, E. C., White, A. R., Chaudhri, N., Booth, S., Gharib, M. A., et al. (2001).
Cue dependency of nicotine self-administration and smoking. Pharmacology Biochemistry
and Behavior, 70(4), 515-530.
Cao, F., Su, L., Liu, T., & Gao, X. (2007). The relationship between impulsivity and internet
addiction in a sample of chinese adolescents. European Psychiatry, 22(7), 466-471.
Cappelleri, J. C., Bushmakin, A. G., Baker, C. L., Merikle, E., Olufade, A. O., & Gilbert, D. G.
(2007). Multivariate framework of the brief questionnaire of smoking urges. Drug and
Alcohol Dependence, 90(2-3), 234-242.
Carpenter, K. M., Schreiber, E., Church, S., & McDowell, D. (2006). Drug stroop performance:
Relationships with primary substance of use and treatment outcome in a drug-dependent
outpatient sample. Addictive Behaviors, 31(1), 174-81.
Carretie, L., Hinojosa, J. A., Martin-Loeches, M., Mercado, F., & Tapia, M. (2004). Automatic
attention to emotional stimuli: Neural correlates. Human Brain Mapping, 22(4), 290-9.
Carter, B. L., & Tiffany, S. T. (1999). Meta-analysis of cue-reactivity in addiction research.
Addiction, 94(3), 327-40.
Cepeda-Benito, A., Henry, K., Gleaves, D. H., & Fernandez, M. C. (2004). Cross-cultural
investigation of the questionnaire of smoking urges in american and spanish smokers.
Assessment, 11, 152-159.
Cepeda-Benito, A., & Reig-Ferrer, A. (2004). Development of a brief questionnaire of
smoking urges--spanish. Psychological Assessment, 16(4), 402-407.
Chambers, C. D., Garavan, H., & Bellgrove, M. A. (2009). Insights into the neural basis of
response inhibition from cognitive and clinical neuroscience. Neuroscience & Biobehavioral
Reviews, 33(5), 631-646.
Chanon, V. W., Sours, C. R., & Boettiger, C. A. (2010). Attentional bias toward cigarette cues
in active smokers. Psychopharmacology (Berl), 212(3), 309-320.
Charlton, J. P., & Danforth, I. D. W. (2007). Distinguishing addiction and high engagement in
the context of online game playing. Computers in Human Behavior, 23(3), 1531-1548.
Chase, H. W., Eickhoff, S. B., Laird, A. R., & Hogarth, L. (2011). The neural basis of drug
stimulus processing and craving: An activation likelihood estimation meta-analysis.
Biological Psychiatry, 70(8), 785-793.
Chiamulara, C. (2005). Cue reactivity in nicotine and tobacco dependence: A “multipleaction” model of nicotine as a primary reinforcement and as an enhancer ofthe effects of
smoking-associated stimuli. Brain Research. Brain Research Reviews, 48, 74-97.
237
References
Church, R. (1989). Smoking and the human EEG. In T. Ney, & A. Gale (Eds.), Smoking and
human behavior (pp. 115-140). Chichester: John Wiley & Sons.
Coan, J. A., & Allen, J. J. B. (2004). Frontal EEG asymmetry as a moderator and mediator of
emotion. Biological Psychology, 67(1-2), 7-50.
Coggins, C. R., Murrelle, E. L., Carchman, R. A., & Heidbreder, C. (2009). Light and intermittent
cigarette smokers: A review (1989-2009). Psychopharmacology (Berl), 207(3), 343-63.
Coles, M. G. H., Gratton, G., & Fabiani, M. (1990). Event-related brain potentials. In J. T.
Cacioppo, & I. G. Tassinary (Eds.), Principles of psychophysiology: Physical, social, and
inferential elements (pp. 413-455). Cambridge, England: Cambridge University Press.
Coles, M. G. H., & Rugg, M. D. (1995). Event-related brain potentials: An introduction. In
M. D. Rugg, & M. G. H. Coles (Eds.), Electrophysiology of mind (pp. 1-26). Oxford: Oxford
University Press.
Comings, D. E., Rosenthal, R. J., Lesieur, H. R., Rugle, L. J., Muhleman, D., Chiu, C., et al. (1996).
A study of the dopamine D2 receptor gene in pathological gambling. Pharmacogenetics,
6(3), 223-234.
Conrin, J. (1980). The EEG effects of tobacco smoking - a review. Clinical
Electroencephalography, 11(4), 180-187.
Cox, L. S., Tiffany, S. T., & Christen, A. G. (2001). Evaluation of the brief questionnaire of
smoking urges (QSU-brief) in laboratory and clinical settings. Nicotine & Tobacco Research,
3(1), 7-16.
Cox, W. M., Fadardi, J. S., & Pothos, E. M. (2006). The addiction-stroop test: Theoretical
considerations and procedural recommendations. Psychological Bulletin, 132(3), 443-476.
Cox, W. M., Hogan, L. M., Kristian, M. R., & Race, J. H. (2002). Alcohol attentional bias as a
predictor of alcohol abusers’ treatment outcome. Drug and Alcohol Dependence, 68(3), 237-43.
Cox, W. M., Pothos, E. M., & Hosier, S. G. (2007). Cognitive-motivational predictors of
excessive drinkers’ success in changing. Psychopharmacology (Berl), 192(4), 499-510.
Cuthbert, B. N., Schupp, H. T., Bradley, M. M., Birbaumer, N., & Lang, P. J. (2000). Brain
potentials in affective picture processing: Covariation with autonomic arousal and
affective report. Biological Psychology, 52(2), 95-111.
David, S. P., Munafò, M. R., Johansen-Berg, H., Smith, S. M., Rogers, R. D., Matthews, P. M., et al.
(2005). Ventral striatum/nucleus accumbens activation to smoking-related pictorial cues
in smokers and nonsmokers: A functional magnetic resonance imaging study. Biological
Psychiatry, 58(6), 488-494.
Davidson, R. J., Ekman, P., Saron, C. D., Senulis, J. A., & Friesen, W. V. (1990). Approachwithdrawal and cerebral asymmetry: Emotional expression and brain physiology. I. Journal
of Personality and Social Psychology, 58(2), 330-41.
238
References
Davies, G. M., Willner, P., & Morgan, M. J. (2000). Smoking-related cues elicit craving
in tobacco “chippers”: A replication and validation of the two-factor structure of the
questionnaire of smoking urges. Psychopharmacology, 152(3), 334-342.
Davis, J. A., & Gould, T. J. (2008). Associative learning, the hippocampus, and nicotine
addiction. Current Drug Abuse Reviews, 1(1), 9-19.
Dawe, S., Gullo, M. J., & Loxton, N. J. (2004). Reward drive and rash impulsiveness as
dimensions of impulsivity: Implications for substance misuse. Addictive Behaviors, 29(7),
1389-1405.
De Cesarei, A., & Codispoti, M. (2006). When does size not matter? effects of stimulus size
on affective modulation. Psychophysiology, 43(2), 207-215.
Delplanque, S., Lavoie, M. E., Hot, P., Silvert, L., & Sequeira, H. (2004). Modulation of cognitive
processing by emotional valence studied through event-related potentials in humans.
Neuroscience Letters, 356(1), 1-4.
DerSimonian, R., & Laird, N. (1986). Meta-analysis in clinical trials. Controlled Clinical
Trials, 7(3), 177-188.
Desai, R. A., Krishnan-Sarin, S., Cavallo, D., & Potenza, M. N. (2010). Video-gaming among
high school students: Health correlates, gender differences, and problematic gaming.
Pediatrics, 126(6), 1414-1424.
Desmond, J. E., Chen, S. H. A., DeRosa, E., Pryor, M. R., Pfefferbaum, A., & Sullivan, E. V.
(2003). Increased frontocerebellar activation in alcoholics during verbal working memory:
An fMRI study. NeuroImage, 19(4), 1510-1520.
Di Chiara, G. (2000). Role of dopamine in the behavioural actions of nicotine related to
addiction. European Journal of Pharmacology, 393(1-3), 295-314.
Dien, J., & Santuzzi, A. M. (2005). Application of repeated measures ANOVA to high-density
ERP data-sets: A review and tutorial. In T. C. Handy (Ed.), Event-related potentials: A methods
handbook. (pp. 57-82). Cambridge: The MIT Press.
Dijkstra, A., & Borland, R. (2003). Residual outcome expectations and relapse in ex-smokers.
Health Psychology, 22(4), 340-346.
Dillon, D. G., & Labar, K. S. (2005). Startle modulation during conscious emotion regulation
is arousal-dependent. Behavioral Neuroscience, 119(4), 1118-24.
Doherty, K., Kinnunen, T., Militello, F., & Garvey, A. (1995). Urges to smoke during the first
month of abstinence: Relationship to relapse and predictors. Psychopharmacology, 119(2),
171-178.
Dols, M., Hout, M. v. d., Kindt, M., & Willems, B. (2002). The urge to smoke depends on the
expectation of smoking. Addiction, 97(1), 87-93.
Dols, M., Willems, B., van den Hout, M., & Bittoun, R. (2000). Smokers can learn to influence
their urge to smoke. Addictive Behaviors, 25(1), 103-108.
239
References
Dom, G., De Wilde, B., Hulstijn, W., & Sabbe, B. (2007). Dimensions of impulsive behaviour in
abstinent alcoholics. Personality and Individual Differences, 42(3), 465-476.
Domino, E. F. (2003). Effects of tobacco smoking on electroencephalographic, auditory
evoked and event related potentials. Brain and Cognition, 53(1), 66-74.
Donchin, E. (1981). Presidential address, 1980. surprise!...surprise? Psychophysiology,
18(5), 493-513.
Dong, G., Lu, Q., Zhou, H., & Zhao, X. (2010). Impulse inhibition in people with internet
addiction disorder: Electrophysiological evidence from a Go/NoGo study. Neuroscience
Letters, 485(2), 138-142.
Drummond, D. C. (2000). What does cue-reactivity have to offer clinical research? Addiction,
95(8s2), 129-144.
Drummond, D. C., Litten, R. Z., Lowman, C., & Hunt, W. A. (2000). Craving research: Future
directions. Addiction, 95(8s2), 247-255.
Duka, T., & Townshend, J. M. (2004). The priming effect of alcohol pre-load on attentional
bias to alcohol-related stimuli. Psychopharmacology (Berl), 176(3-4), 353-61.
Duncan-Johnson, C. C., & Donchin, E. (1977). On quantifying surprise: The variation of
event-related potentials with subjective probability. Psychophysiology, 14(5), 456-467.
Dunning, J. P., & Hajcak, G. (2009). See no evil: Directing visual attention within unpleasant
images modulates the electrocortical response. Psychophysiology, 46(1), 28-33.
Dunning, J. P., Parvaz, M. A., Hajcak, G., Maloney, T., Alia-Klein, N., Woicik, P. A., et al. (2011).
Motivated attention to cocaine and emotional cues in abstinent and current cocaine users-an ERP study. European Journal of Neuroscience, 33(9), 1716-1723.
Egner, T., & Gruzelier, J. H. (2001). Learned self-regulation of EEG frequency components
affects attention and event-related brain potentials in humans. NeuroReport, 12(18), 4155-9.
Egner, T., & Gruzelier, J. H. (2004). EEG biofeedback of low beta band components:
Frequency-specific effects on variables of attention and event-related brain potentials.
Clinical Neurophysiology, 115(1), 131-9.
Ehrman, R. N., Robbins, S. J., Bromwell, M. A., Lankford, M. E., Monterosso, J. R., & O’Brien, C.
P. (2002). Comparing attentional bias to smoking cues in current smokers, former smokers,
and non-smokers using a dot-probe task. Drug and Alcohol Dependence, 67(2), 185-91.
Eldreth, D. A., Matochik, J. A., Cadet, J. L., & Bolla, K. I. (2004). Abnormal brain activity in
prefrontal brain regions in abstinent marijuana users. NeuroImage, 23(3), 914-920.
Elmasian, R., Neville, H., Woods, D., Schuckit, M., & Bloom, F. (1982). Event-related brain
potentials are different in individuals at high and low risk for developing alcoholism.
Proceedings of the National Academy of Sciences of the United States of America, 79(24),
7900-7903.
240
References
Evans, D. E., Park, J. Y., Maxfield, N., & Drobes, D. J. (2009). Neurocognitive variation in
smoking behavior and withdrawal: Genetic and affective moderators. Genes, Brain and
Behavior, 8(1), 86-96.
Everitt, B. J., Dickinson, A., & Robbins, T. W. (2001). The neuropsychological basis of
addictive behaviour. Brain Research Reviews, 36(2-3), 129-138.
Eysenck, S. B. G., Pearson, P. R., Easting, G., & Allsopp, J. F. (1985). Age norms for impulsiveness,
venturesomeness and empathy in adults. Personality and Individual Differences, 6(5),
613-619.
Fabiani, M., Gratton, G., & Coles, M. G. H. (2000). Event-related brain potentials: Methods,
theory and applications. In J. T. Cacioppo, L. Tassinary & G. Berntson (Eds.), Handbook of
psychophysiology (pp. 53-84). New York: Cambridge University Press.
Fadardi, J. S., & Cox, W. M. (2009). Reversing the sequence: Reducing alcohol consumption
by overcoming alcohol attentional bias. Drug and Alcohol Dependence, 101(3), 137-45.
Fagerström, K. O., & Furberg, H. (2008). A comparison of the fagerström test for nicotine
dependence and smoking prevalence across countries. Addiction, 103(5), 841-845.
Falkenstein, M., Hohnsbein, J., Hoormann, J., & Blanke, L. (1991). Effects of crossmodal
divided attention on late ERP components. II. error processing in choice reaction tasks.
Electroencephalography and Clinical Neurophysiology, 78(6), 447-455.
Falkenstein, M., Hoormann, J., Christ, S., & Hohnsbein, J. (2000). ERP components on
reaction errors and their functional significance: A tutorial. Biological Psychology, 51(2-3),
87-107.
Falkenstein, M., Hoormann, J., & Hohnsbein, J. (1999). ERP components in Go/Nogo tasks
and their relation to inhibition. Acta Psychologica, 101(2-3), 267-291.
Fallgatter, A. J., Wiesbeck, G. A., Weijers, H. G., Boening, J., & Strik, W. K. (1998). Eventrelated correlates of response suppression as indicators of novelty seeking in alcoholics.
Alcohol and Alcoholism, 33(5), 475-481.
Fehr, T., Wiedenmann, P., & Herrmann, M. (2006). Nicotine stroop and addiction memory-an ERP study. International Journal of Psychophysiology, 62(2), 224-32.
Fehr, T., Wiedenmann, P., & Herrmann, M. (2007). Differences in ERP topographies during
color matching of smoking-related and neutral pictures in smokers and non-smokers.
International Journal of Psychophysiology, 65(3), 284-93.
Ferguson, S. G., Shiffman, S., & Gwaltney, C. J. (2006). Does reducing withdrawal severity
mediate nicotine patch efficacy? A randomized clinical trial. Journal of Consulting and
Clinical Psychology, 74(6), 1153-1161.
Ferrari, V., Codispoti, M., Cardinale, R., & Bradley, M. M. (2008). Directed and motivated
attention during processing of natural scenes. Journal of Cognitive Neuroscience, 20(10),
1753-1761.
241
References
Field, M., & Cox, W. M. (2008). Attentional bias in addictive behaviors: A review of its
development, causes, and consequences. Drug and Alcohol Dependence, 97, 1-20.
Field, M., & Duka, T. (2001). Smoking expectancy mediates the conditioned responses to
arbitrary smoking cues. Behavioural Pharmacology, 12(3), 183-94.
Field, M., & Duka, T. (2004). Cue reactivity in smokers: The effects of perceived cigarette
availability and gender. Pharmacology Biochemistry and Behavior, 78(3), 647-652.
Field, M., Duka, T., Eastwood, B., Child, R., Santarcangelo, M., & Gayton, M. (2007).
Experimental manipulation of attentional biases in heavy drinkers: Do the effects
generalise? Psychopharmacology, 192(4), 593-608.
Field, M., Duka, T., Tyler, E., & Schoenmakers, T. (2009). Attentional bias modification in
tobacco smokers. Nicotine & Tobacco Research, 11(7), 812-22.
Field, M., & Eastwood, B. (2005). Experimental manipulation of attentional bias increases
the motivation to drink alcohol. Psychopharmacology (Berl), 183(3), 350-7.
Field, M., Mogg, K., & Bradley, B. P. (2006). Attention to drug-related cues in drug abuse and
addiction: Component processes. In R. W. Wiers, & A. W. Stacy (Eds.), Handbook of implicit
cognition and addiction (pp. 151-163). Thousand Oaks: Sage Publications.
Field, M., Mogg, K., Zetteler, J., & Bradley, B. P. (2004). Attentional biases for alcohol cues
in heavy and light social drinkers: The roles of initial orienting and maintained attention.
Psychopharmacology (Berl), 176(1), 88-93.
Field, M., Munafò, M. R., & Franken, I. H. A. (2009). A meta-analytic investigation of
the relationship between attentional bias and subjective craving in substance abuse.
Psychological Bulletin, 135(4), 589-607.
Foti, D., & Hajcak, G. (2008). Deconstructing reappraisal: Descriptions preceding arousing
pictures modulate the subsequent neural response. Journal of Cognitive Neuroscience, 20(6),
977-88.
Foti, D., Hajcak, G., & Dien, J. (2009). Differentiating neural responses to emotional pictures:
Evidence from temporal-spatial PCA. Psychophysiology, 46(3), 521-530.
Fox, N. A. (1991). If it’s not left, it’s right. electroencephalograph asymmetry and the
development of emotion. American Psychologist, 46(8), 863-72.
Franken, I. H. A. (2003). Drug craving and addiction: Integrating psychological and
neuropsychopharmacological approaches. Progress in Neuro-Psychopharmacology and
Biological Psychiatry, 27(4), 563-79.
Franken, I. H. A., Dietvorst, R. C., Hesselmans, M., Franzek, E. J., van de Wetering, B. J., &
Van Strien, J. W. (2008). Cocaine craving is associated with electrophysiological brain
responses to cocaine-related stimuli. Addiction Biology, 13(3-4), 386-92.
242
References
Franken, I. H. A., Hendriks, V. M., Stam, C. J., & Van den Brink, W. (2004). A role for
dopamine in the processing of drug cues in heroin dependent patients. European
Neuropsychopharmacology, 14(6), 503-8.
Franken, I. H. A., Hendriks, V. M., & van den Brink, W. (2002). Initial validation of two opiate
craving questionnaires the obsessive compulsive drug use scale and the desires for drug
questionnaire. Addictive Behaviors, 27(5), 675-685.
Franken, I. H. A., Hulstijn, K. P., Stam, C. J., Hendriks, V. M., & van den Brink, W. (2004).
Two new neurophysiological indices of cocaine craving: Evoked brain potentials and cue
modulated startle reflex. Journal of Psychopharmacology, 18(4), 544-52.
Franken, I. H. A., Kroon, L. Y., & Hendriks, V. M. (2000). Influence of individual differences in
craving and obsessive cocaine thoughts on attentional processes in cocaine abuse patients.
Addictive Behaviors, 25(1), 99-102.
Franken, I. H. A., Kroon, L. Y., Wiers, R. W., & Jansen, A. (2000). Selective cognitive processing
of drug cues in heroin dependence. Journal of Psychopharmacology, 14(4), 395-400.
Franken, I. H. A., Rassin, E., & Muris, P. (2007). The assessment of anhedonia in clinical and
non-clinical populations: Further validation of the snaith-hamilton pleasure scale (SHAPS).
Journal of Affective Disorders, 99(1-3), 83-89.
Franken, I. H. A., Rosso, M., & van Honk, J. (2003). Selective memory for alcohol cues in
alcoholics and its relation to craving. Cognitive Therapy and Research, 27(4), 481-488.
Franken, I. H. A., Stam, C. J., Hendriks, V. M., & van den Brink, W. (2003). Neurophysiological
evidence for abnormal cognitive processing of drug cues in heroin dependence.
Psychopharmacology (Berl), 170(2), 205-12.
Franken, I. H. A., Van Strien, J. W., Bocanegra, B. R., & Huijding, J. (2011). The P3 eventrelated potential as an index of motivational relevance: A conditioning experiment. Journal
of Psychophysiology, 25, 32-39.
Franken, I. H. A., van Strien, J. W., Franzek, E. J., & van de Wetering, B. J. (2007). Errorprocessing deficits in patients with cocaine dependence. Biological Psychology, 75(1), 45-51.
Franken, I. H. A., van Strien, J. W., & Kuijpers, I. (2010). Evidence for a deficit in the salience
attribution to errors in smokers. Drug and Alcohol Dependence, 106(2-3), 181-185.
Gamma, A., Brandeis, D., Brandeis, R., & Vollenweider, F. X. (2005). The P3 in ‘ecstasy’
polydrug users during response inhibition and execution. Journal of Psychopharmacology,
19(5), 504-512.
García-Larrea, L., & Cézanne-Bert, G. (1998). P3, positive slow wave and working memory
load: A study on the functional correlates of slow wave activity. Electroencephalography
and Clinical Neurophysiology, 108(3), 260-273.
Garland, E. L., Franken, I. H. A., & Howard, M. O. (2011). Cue-elicited heart rate variability
and attentional bias predict alcohol relapse following treatment. Psychopharmacology, DOI:
10.1007/s00213-011-2618-4
243
References
Gehring, W. J., Gratton, G., Coles, M. G., & Donchin, E. (1992). Probability effects on stimulus
evaluation and response processes. Journal of Experimental Psychology: Human Perception
and Performance, 18(1), 198-216.
Geier, A., Mucha, R. F., & Pauli, P. (2000). Appetitive nature of drug cues confirmed with
physiological measures in a model using pictures of smoking. Psychopharmacology (Berl),
150(3), 283-91.
Genkina, O. A., & Shostakovich, G. S. (1983). Elaboration of a conditioned reflex in
chronic alcoholics using an unrecognizable motivationally significant word. [Vyrabotka
uslovnoi sviazi u bol’nykh khronicheskim alkogolizmom s pomoshch’iu neosoznavaemogo
motivatsionno-znachimogo slova] Zhurnal Vysshei Nervnoi Deiatelnosti Imeni I P Pavlova,
33(6), 1010-1018.
Gewirtz, J. C., & Davis, M. (2000). Using pavlovian higher-order conditioning paradigms to
investigate the neural substrates of emotional learning and memory. Learning & Memory,
7(5), 257-266.
Gilbert, D. G., McClernon, F. J., Rabinovich, N. E., Dibb, W. D., Plath, L. C., Hiyane, S., et al.
(1999). EEG, physiology, and task-related mood fail to resolve across 31 days of smoking
abstinence: Relations to depressive traits, nicotine exposure, and dependence. Experimental
and Clinical Psychopharmacology, 7(4), 427-43.
Gilbert, D. G., McClernon, J., Rabinovich, N. E., Sugai, C., Plath, L. C., Asgaard, G., et al. (2004).
Effects of quitting smoking on EEG activation and attention last for more than 31 days and
are more severe with stress, dependence, DRD2 A1 allele, and depressive traits. Nicotine &
Tobacco Research, 6(2), 249-67.
Gilbert, D. G., Sugai, C., Zuo, Y., Rabinovich, N. E., McClernon, F. J., & Froeliger, B. (2007).
Brain indices of nicotine’s effects on attentional bias to smoking and emotional pictures
and to task-relevant targets. Nicotine & Tobacco Research, 9(3), 351-63.
Goldberg, S. R., Spealman, R. D., & Goldberg, D. M. (1981). Persistent behavior at high rates
maintained by intravenous self-administration of nicotine. Science, 214(4520), 573-575.
Goldstein, R. Z., Tomasi, D., Rajaram, S., Cottone, L. A., Zhang, L., Maloney, T., et al. (2007).
Role of the anterior cingulate and medial orbitofrontal cortex in processing drug cues in
cocaine addiction. Neuroscience, 144(4), 1153-1159.
Goldstein, R. Z., & Volkow, N. D. (2002). Drug addiction and its underlying neurobiological
basis: Neuroimaging evidence for the involvement of the frontal cortex. American Journal of
Psychiatry, 159(10), 1642-1652.
Grabus, S. D., Martin, B. R., Batman, A. M., Tyndale, R. F., Sellers, E., & Damaj, M. I.
(2005). Nicotine physical dependence and tolerance in the mouse following chronic oral
administration. Psychopharmacology, 178(2), 183-192.
Grant, J. E., Potenza, M. N., Weinstein, A., & Gorelick, D. A. (2010). Introduction to behavioral
addictions. American Journal of Drug and Alcohol Abuse, 36(5), 233-241.
244
References
Gratton, G., Coles, M. G., & Donchin, E. (1983). A new method for off-line removal of ocular
artifact. Electroencephalography and Clinical Neurophysiology, 55(4), 468-84.
Griffiths, M. D., & Meredith, A. (2009). Videogame addiction and its treatment. Journal of
Contemporary Psychotherapy, 39(4), 247-253.
Groman, S. M., James, A. S., & Jentsch, J. D. (2009). Poor response inhibition: At the nexus
between substance abuse and attention deficit/hyperactivity disorder. Neuroscience &
Biobehavioral Reviews, 33(5), 690-698.
Gross, J. J. (1998). Antecedent- and response-focused emotion regulation: Divergent
consequences for experience, expression, and physiology. Journal of Personality and Social
Psychology, 74(1), 224-37.
Gross, J. J., & Levenson, R. W. (1997). Hiding feelings: The acute effects of inhibiting negative
and positive emotion. Journal of Abnormal Psychology, 106(1), 95-103.
Gross, T. M., Jarvik, M. E., & Rosenblatt, M. R. (1993). Nicotine abstinence produces contentspecific stroop interference. Psychopharmacology (Berl), 110(3), 333-6.
Haenschel, C., Baldeweg, T., Croft, R. J., Whittington, M., & Gruzelier, J. (2000). Gamma
and beta frequency oscillations in response to novel auditory stimuli: A comparison of
human electroencephalogram (EEG) data with in vitro models. Proceedings of the National
Academy of Sciences of the United States of America, 97(13), 7645-50.
Hajcak, G., Dunning, J. P., & Foti, D. (2007). Neural response to emotional pictures is
unaffected by concurrent task difficulty: An event-related potential study. Behavioral
Neuroscience, 121(6), 1156-1162.
Hajcak, G., Dunning, J. P., & Foti, D. (2009). Motivated and controlled attention to emotion:
Time-course of the late positive potential. Clinical Neurophysiology, 120(3), 505-10.
Hajcak, G., MacNamara, A., & Olvet, D. M. (2010). Event-related potentials, emotion, and
emotion regulation: An integrative review. Developmental Neuropsychology, 35(2), 129-155.
Hajcak, G., Moser, J. S., & Simons, R. F. (2006). Attending to affect: Appraisal strategies
modulate the electrocortical response to arousing pictures. Emotion, 6(3), 517-22.
Hajcak, G., & Nieuwenhuis, S. (2006). Reappraisal modulates the electrocortical response
to unpleasant pictures. Cognitive, Affective, and Behavioral Neuroscience, 6(4), 291-7.
Hajcak, G., & Olvet, D. M. (2008). The persistence of attention to emotion: Brain potentials
during and after picture presentation. Emotion, 8(2), 250-5.
Hall, J. R., Bernat, E. M., & Patrick, C. J. (2007). Externalizing psychopathology and the
error-related negativity. Psychological Science, 18(4), 326-333.
Han, D. H., Lee, Y. S., Yang, K. C., Kim, E. Y., Lyoo, I. K., & Renshaw, P. F. (2007). Dopamine
genes and reward dependence in adolescents with excessive internet video game play.
Journal of Addiction Medicine, 1(3), 133-138.
245
References
Hansenne, M., Olin, C., Pinto, E., Pitchot, W., & Ansseau, M. (2003). Event-related potentials
to emotional and neutral stimuli in alcoholism. Neuropsychobiology, 48(2), 77-81.
Harmon-Jones, E. (2004). Contributions from research on anger and cognitive dissonance to
understanding the motivational functions of asymmetrical frontal brain activity. Biological
Psychology, 67(1-2), 51-76.
Heatherton, T. F., Kozlowski, L. T., Frecker, R. C., & Fagerström, K. O. (1991). The fagerström
test for nicotine dependence: A revision of the fagerström tolerance questionnaire. British
Journal of Addiction, 86(9), 1119-27.
Heinze, M., Wölfling, K., & Grüsser, S. M. (2007). Cue-induced auditory evoked potentials in
alcoholism. Clinical Neurophysiology, 118(4), 856-862.
Herning, R. I., Glover, B. J., Koeppl, B., Phillips, R. L., & London, E. D. (1994). Cocaineinduced increases in EEG alpha and beta activity: Evidence for reduced cortical processing.
Neuropsychopharmacology, 11(1), 1-9.
Herning, R. I., Jones, R. T., Hooker, W. D., Mendelson, J., & Blackwell, L. (1985). Cocaine
increases EEG beta: A replication and extension of hans berger’s historic experiments.
Electroencephalography and Clinical Neurophysiology, 60(6), 470-7.
Herrmann, M. J., Weijers, H. G., Wiesbeck, G. A., Aranda, D., Böning, J., & Fallgatter, A. J.
(2000). Event-related potentials and cue-reactivity in alcoholism. Alcoholism: Clinical and
Experimental Research, 24(11), 1724-1729.
Herrmann, M. J., Weijers, H. G., Wiesbeck, G. A., Böning, J., & Fallgatter, A. J. (2001). Alcohol
cue-reactivity in heavy and light social drinkers as revealed by event-related potentials.
Alcohol and Alcoholism, 36(6), 588-593.
Hester, R., Dixon, V., & Garavan, H. (2006). A consistent attentional bias for drug-related
material in active cocaine users across word and picture versions of the emotional stroop
task. Drug and Alcohol Dependence, 81(3), 251-257.
Hester, R., & Garavan, H. (2004). Executive dysfunction in cocaine addiction: Evidence
for discordant frontal, cingulate, and cerebellar activity. Journal of Neuroscience, 24(49),
11017-11022.
Hillyard, S. A., Hink, R. F., Schwent, V. L., & Picton, T. W. (1973). Electrical signs of selective
attention in the human brain. Science, 182(108), 177-180.
Hogarth, L., Dickinson, A., & Duka, T. (2003). Discriminative stimuli that control
instrumental tobacco-seeking by human smokers also command selective attention.
Psychopharmacology (Berl), 168(4), 435-45.
Hogarth, L., Dickinson, A., & Duka, T. (2005). Explicit knowledge of stimulus-outcome
contingencies and stimulus control of selective attention and instrumental action in human
smoking behaviour. Psychopharmacology (Berl), 177(4), 428-37.
246
References
Hogarth, L., Dickinson, A., & Duka, T. (2009). Detection versus sustained attention to drug
cues have dissociable roles in mediating drug seeking behavior. Experimental and Clinical
Psychopharmacology, 17(1), 21-30.
Hogarth, L., Dickinson, A., & Duka, T. (2010). The associative basis of cue-elicited drug
taking in humans. Psychopharmacology (Berl), 208(3), 337-51.
Hogarth, L., Dickinson, A., Janowski, M., Nikitina, A., & Duka, T. (2008). The role of
attentional bias in mediating human drug-seeking behaviour. Psychopharmacology (Berl),
201(1), 29-41.
Hogarth, L., Dickinson, A., Wright, A., Kouvaraki, M., & Duka, T. (2007). The role of drug
expectancy in the control of human drug seeking. Journal of Experimental Psychology:
Animal Behavior Processes, 33(4), 484-96.
Hogarth, L., & Duka, T. (2006). Human nicotine conditioning requires explicit contingency
knowledge: Is addictive behaviour cognitively mediated? Psychopharmacology, 184(3),
553-566.
Hogarth, L., Mogg, K., Bradley, B. P., Duka, T., & Dickinson, A. (2003). Attentional orienting
towards smoking-related stimuli. Behavioural Pharmacology, 14(2), 153-60.
Holle, C., Neely, J. H., & Heimberg, R. G. (1997). The effects of blocked versus random
presentation and semantic relatedness of stimulus words on response to a modified stroop
task among social phobics. Cognitive Therapy and Research, 21(6), 681-697.
Holroyd, C. B., & Coles, M. G. H. (2002). The neural basis of human error processing:
Reinforcement learning, dopamine, and the error-related negativity. Psychological Review,
109(4), 679-709.
Houlihan, M. E., Pritchard, W. S., & Robinson, J. H. (1996). Faster P300 latency after smoking
in visual but not auditory oddball tasks. Psychopharmacology (Berl), 123(3), 231-8.
Hudmon, K. S., Pomerleau, C. S., Brigham, J., Javitz, H., & Swan, G. E. (2005). Validity of
retrospective assessments of nicotine dependence: A preliminary report. Addictive
Behaviors, 30(3), 613-7.
Hyman, S. E. (2005). Addiction: A disease of learning and memory. American Journal of
Psychiatry, 162(8), 1414-1422.
Ioannidis, J. P. (2011). An epidemic of false claims. competition and conflicts of interest
distort too many medical findings. Scientific American, 304(6), 16.
Jackson, D. C., Malmstadt, J. R., Larson, C. L., & Davidson, R. J. (2000). Suppression and
enhancement of emotional responses to unpleasant pictures. Psychophysiology, 37(4),
515-22.
Jang, K. W., Lee, J. S., Yang, B. H., & Lee, J. H. (2007). Changes of brain potentials in response
to smoking-induced stimuli in smokers. Cyberpsychology & Behavior, 10(3), 460-463.
247
References
Jarvik, M. E. (1991). Beneficial effects of nicotine. British Journal of Addiction, 86(5), 571-5.
Johnsen, B. H., Thayer, J. F., Laberg, J. C., & Asbjornsen, A. E. (1997). Attentional bias in active
smokers, abstinent smokers, and nonsmokers. Addictive Behaviors, 22(6), 813-817.
Johnson, R.,Jr. (1986). A triarchic model of P300 amplitude. Psychophysiology, 23(4),
367-384.
Johnstone, S. J., Barry, R. J., Markovska, V., Dimoska, A., & Clarke, A. R. (2009). Response
inhibition and interference control in children with AD/HD: A visual ERP investigation.
International Journal of Psychophysiology, 72(2), 145-153.
Jurcak, V., Tsuzuki, D., & Dan, I. (2007). 10/20, 10/10, and 10/5 systems revisited: Their
validity as relative head-surface-based positioning systems. NeuroImage, 34(4), 1600-1611.
Kamarajan, C., Porjesz, B., Jones, K. A., Choi, K., Chorlian, D. B., Padmanabhapillai, A., et al.
(2005). Alcoholism is a disinhibitory disorder: Neurophysiological evidence from a Go/Nogo task. Biological Psychology, 69(3), 353-373.
Kertzman, S., Lowengrub, K., Aizer, A., Vainder, M., Kotler, M., & Dannon, P. N. (2008). Go-nogo performance in pathological gamblers. Psychiatry Research, 161(1), 1-10.
Killen, J. D., & Fortmann, S. P. (1997). Craving is associated with smoking relapse: Findings
from three prospective studies. Experimental and Clinical Psychopharmacology, 5(2),
137-142.
Killen, J. D., Fortmann, S. P., Newman, B., & Varady, A. (1991). Prospective study of
factors influencing the development of craving associated with smoking cessation.
Psychopharmacology, 105(2), 191-196.
Kim, S. H., Baik, S. -., Park, C. S., Kim, S. J., Choi, S. W., & Kim, S. E. (2011). Reduced striatal
dopamine D2 receptors in people with internet addiction. NeuroReport, 22(8), 407-411.
King, D. L., Delfabbro, P. H., & Griffiths, M. D. (2010). Recent innovations in video game
addiction research and theory. Global Media Journal, 4, 1-13.
Knott, V. J. (2001). Electroencephalographic characterization of cigarette smoking behavior.
Alcohol, 24(2), 95-7.
Knott, V. J., Cosgrove, M., Villeneuve, C., Fisher, D. J., Millar, A., & McIntosh, J. F. (2008). EEG
correlates of imagery-induced cigarette craving in male and female smokers. Addictive
Behaviors, 33(4), 616-621.
Knott, V. J., Harr, A., Ilivitsky, V., & Mahoney, C. (1998). The cholinergic basis of the smokinginduced EEG activation profile. Neuropsychobiology, 38(2), 97-107.
Knott, V. J., Raegele, M., Fisher, D. J., Robertson, N., Millar, A., McIntosh, J. F., et al. (2005).
Clonidine pre-treatment fails to block acute smoking-induced EEG arousal/mood in
cigarette smokers. Pharmacology Biochemistry and Behavior, 80(1), 161-171.
248
References
Knott, V. J., & Venables, P. H. (1977). EEG alpha correlates of non-smokers, smokers, smoking,
and smoking deprivation. Psychophysiology, 14(2), 150-156.
Kober, H., Kross, E. F., Mischel, W., Hart, C. L., & Ochsner, K. N. (2009). Regulation of craving
by cognitive strategies in cigarette smokers. Drug and Alcohol Dependence, 106(1), 52-5.
Kober, H., Mende-Siedlecki, P., Kross, E. F., Weber, J., Mischel, W., Hart, C. L., et al. (2010).
Prefrontal-striatal pathway underlies cognitive regulation of craving. Proceedings of the
National Academy of Sciences of the United States of America, 107(33), 14811-6.
Koenig, S., & Mecklinger, A. (2008). Electrophysiological correlates of encoding and
retrieving emotional events. Emotion, 8(2), 162-173.
Kok, A. (1990). Internal and external control: A two-factor model of amplitude change of
event-related potentials. Acta Psychologica, 74(2-3), 203-236.
Kok, A. (1997). Event-related-potential (ERP) reflections of mental resources: A review and
synthesis. Biological Psychology, 45(1-3), 19-56.
Kok, A. (2001). On the utility of P3 amplitude as a measure of processing capacity.
Psychophysiology, 38(03), 557-577.
Kooi, K., Tucker, R. P., & Marshall, R. E. (1978). Fundamentals of electroencephalography (2nd
ed.). New York: Harper and Row.
Krompinger, J. W., Moser, J. S., & Simons, R. F. (2008). Modulations of the electrophysiological
response to pleasant stimuli by cognitive reappraisal. Emotion, 8(1), 132-7.
Kuhn, S., & Gallinat, J. (2011). Common biology of craving across legal and illegal drugs
- a quantitative meta-analysis of cue-reactivity brain response. European Journal of
Neuroscience, 33(7), 1318-1326.
Kuss, D. J., & Griffiths, M. D. (2011). Internet gaming addiction: A systematic review of
empirical research. International Journal of Mental Health and Addiction, DOI: 10.1007/
s11469-011-9318-5.
Lang, P. J. (1995). The emotion probe. studies of motivation and attention. American
Psychologist, 50(5), 372-85.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1997). Motivated attention: Affect, activation,
and action. New Jersey: Lawrence Erlbaum Associates Inc.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1998). Emotion, motivation, and anxiety: Brain
mechanisms and psychophysiology. Biological Psychiatry, 44(12), 1248-63.
LaRowe, S. D., Saladin, M. E., Carpenter, M. J., & Upadhyaya, H. P. (2007). Reactivity to
nicotine cues over repeated cue reactivity sessions. Addictive Behaviors, 32(12), 2888-2899.
Lazev, A. B., Herzog, T. A., & Brandon, T. H. (1999). Classical conditioning of environmental
cues to cigarette smoking. Experimental and Clinical Psychopharmacology, 7(1), 56-63.
249
References
Le Foll, B., & Goldberg, S. R. (2006). Nicotine as a typical drug of abuse in experimental
animals and humans. Psychopharmacology, 184(3), 367-381.
Lemmens, P., Tan, E. S., & Knibbe, R. A. (1992). Measuring quantity and frequency of
drinking in a general population survey: A comparison of five indices. Journal of Studies on
Alcohol, 53(5), 476-486.
Lijffijt, M., Caci, H., & Kenemans, J. L. (2005). Validation of the dutch translation of the I7
questionnaire. Personality and Individual Differences, 38(5), 1123-1133.
Lijffijt, M., Kenemans, J. L., Verbaten, M. N., & Van Engeland, H. (2005). A meta-analytic
review of stopping performance in attention-deficit/ hyperactivity disorder: Deficient
inhibitory motor control? Journal of Abnormal Psychology, 114(2), 216-222.
Littel, M., & Franken, I. H. A. (2007). The effects of prolonged abstinence on the processing
of smoking cues: An ERP study among smokers, ex-smokers and never-smokers. Journal of
Psychopharmacology, 21(8), 873-882.
Littel, M., & Franken, I. H. A. (2010). Implicit and explicit selective attention to smoking cues
in smokers indexed by brain potentials. Journal of Psychopharmacology, 25(4), 503-513.
Littel, M., & Franken, I. H. A. (2011). Intentional modulation of the late positive potential in
response to smoking cues by cognitive strategies in smokers - Public Library of Science.
Littel, M., & Franken, I. H. A. (2012). Electrophysiological correlates of associative learning
in smokers: A higher-order conditioning experiment BMC Neuroscience, 13(8)
Littel, M., Franken, I. H. A., & Muris, P. (2011). Psychometric properties of the brief
questionnaire on smoking urges (QSU-brief) in a dutch smoker population. Netherlands
Journal of Psychology, 66, 44-49.
Littel, M., Franken, I. H. A., & Van Strien, J. W. (2009). EEG spectrum changes in response to
smoking cues in smokers and ex-smokers. Neuropsychobiology, 59(1), 43-50.
Liu, X., Vaupel, D. B., Grant, S., & London, E. D. (1998). Effect of cocaine-related
environmental stimuli on the spontaneous electroencephalogram in polydrug abusers.
Neuropsychopharmacology, 19(1), 10-7.
Love, A., James, D., & Willner, P. (1998). A comparison of two alcohol craving questionnaires.
Addiction, 93(7), 1091-1102.
Lubman, D. I., Allen, N. B., Peters, L. A., & Deakin, J. F. (2007). Electrophysiological evidence
of the motivational salience of drug cues in opiate addiction. Psychological Medicine, 37(8),
1203-9.
Lubman, D. I., Allen, N. B., Peters, L. A., & Deakin, J. F. (2008). Electrophysiological evidence
that drug cues have greater salience than other affective stimuli in opiate addiction. Journal
of Psychopharmacology, 22(8), 836-42.
Lubman, D. I., Peters, L. A., Mogg, K., Bradley, B. P., & Deakin, J. F. (2000). Attentional bias for
drug cues in opiate dependence. Psychological Medicine, 30(1), 169-75.
250
References
Lubman, D. I., Yucel, M., Kettle, J. W., Scaffidi, A., Mackenzie, T., Simmons, J. G., et al. (2009).
Responsiveness to drug cues and natural rewards in opiate addiction: Associations with
later heroin use. Archives of General Psychiatry, 66(2), 205-12.
Lubman, D. I., Yücel, M., & Pantelis, C. (2004). Addiction, a condition of compulsive
behaviour? neuroimaging and neuropsychological evidence of inhibitory dysregulation.
Addiction, 99(12), 1491-1502.
Luengo, M. A., Carrillo-De-La-Peña, M. T., & Otero, J. M. (1991). The components of
impulsiveness: A comparison of the I.7 impulsiveness questionnaire and the barratt
impulsiveness scale. Personality and Individual Differences, 12(7), 657-667.
Luijten, M., Littel, M., & Franken, I. H. A. (2011). Deficits in inhibitory control in smokers
during a Go/Nogo task: An investigation using event-related brain potentials. Public Library
of Science ONE, 6(4), 1-7.
Luijten, M., Van Meel, C. S., & Franken, I. H. A. (2011). Diminished error processing in
smokers during smoking cue exposure. Pharmacology Biochemistry and Behavior, 97(3),
514-520.
Luijten, M., Veltman, D. J., van den Brink, W., Hester, R., Field, M., Smits, M., et al. (2010).
Neurobiological substrate of smoking-related attentional bias. NeuroImage, 54(3), 2374-81.
MacNamara, A., Foti, D., & Hajcak, G. (2009). Tell me about it: Neural activity elicited by
emotional pictures and preceding descriptions. Emotion, 9(4), 531-43.
MacNamara, A., Ochsner, K. N., & Hajcak, G. (2010). Previously reappraised: The lasting
effect of description type on picture-elicited electrocortical activity. Social Cognitive and
Affective Neuroscience,
Marissen, M. A., Franken, I. H. A., Blanken, P., van den Brink, W., & Hendriks, V. M. (2007).
Cue exposure therapy for the treatment of opiate addiction: Results of a randomized
controlled clinical trial. Psychotherapy and Psychosomatics, 76(2), 97-105.
Marissen, M. A., Franken, I. H. A., Waters, A. J., Blanken, P., van den Brink, W., & Hendriks, V.
M. (2006). Attentional bias predicts heroin relapse following treatment. Addiction, 101(9),
1306-12.
Matousek, M., & Petersen, I. (1983). A method for assessing alertness fluctuations from EEG
spectra. Electroencephalography and Clinical Neurophysiology, 55(1), 108-13.
McChargue, D. E., Cohen, L. M., & Cook, J. W. (2004a). Attachment and depression
differentially influence nicotine dependence among male and female undergraduates: A
preliminary study. Journal of American College Health, 53(1), 5-10.
McChargue, D. E., Cohen, L. M., & Cook, J. W. (2004b). The influence of personality and affect
on nicotine dependence among male college students. Nicotine & Tobacco Research, 6(2),
287-294.
McCrady, B. S., & Ziedonis, D. (2001). American psychiatric association practice guideline
for substance use disorders. Behavior Therapy, 32(2), 309-336.
251
References
McCusker, C. G., & Gettings, B. (1997). Automaticity of cognitive biases in addictive
behaviours: Further evidence with gamblers. British Journal of Clinical Psychology, 36(4),
543-554.
McDonough, B. E., & Warren, C. A. (2001). Effects of 12-h tobacco deprivation on event-related
potentials elicited by visual smoking cues. Psychopharmacology (Berl), 154(3), 282-91.
McHugh, R. K., Murray, H. W., Hearon, B. A., Calkins, A. W., & Otto, M. W. (2010). Attentional
bias and craving in smokers: The impact of a single attentional training session. Nicotine &
Tobacco Research, 12(12), 1261-4.
McRae, K., Hughes, B., Chopra, S., Gabrieli, J. D. E., Gross, J. J., & Ochsner, K. N. (2010).
The neural bases of distraction and reappraisal. Journal of Cognitive Neuroscience, 22(2),
248-262.
Meerkerk, G. J., Njoo, K. H., Bongers, I. M. B., Trienekens, P., & Van Oers, J. A. M.
(1999). Comparing the diagnostic accuracy of carbohydrate-deficient transferrin,
y-glutamyltransferase, and mean cell volume in a general practice population. Alcoholism:
Clinical and Experimental Research, 23(6), 1052-1059.
Meerkerk, G. J., Van den Eijnden, R. J., Franken, I. H. A., & Garretsen, H. F. L. (2010). Is
compulsive internet use related to sensitivity to reward and punishment, and impulsivity?
Computers in Human Behavior, 26(4), 729-735.
Meerkerk, G. J., Van den Eijnden, R. J., Vermulst, A. A., & Garretsen, H. F. L. (2009). The
compulsive internet use scale (CIUS): Some psychometric properties. Cyberpsychology &
Behavior, 12(1), 1-6.
Miltner, W. H. R., Lemke, U., Weiss, T., Holroyd, C. B., Scheffers, M. K., & Coles, M. G. H. (2003).
Implementation of error-processing in the human anterior cingulate cortex: A source
analysis of the magnetic equivalent of the error-related negativity. Biological Psychology,
64(1-2), 157-166.
Moeller, F. G., Barratt, E. S., Dougherty, D. M., Schmitz, J. M., & Swann, A. C. (2001). Psychiatric
aspects of impulsivity. American Journal of Psychiatry, 158(11), 1783-1793.
Mogg, K., & Bradley, B. P. (2002). Selective processing of smoking-related cues in smokers:
Manipulation of deprivation level and comparison of three measures of processing bias.
Journal of Psychopharmacology, 16(4), 385-92.
Mogg, K., Bradley, B. P., Field, M., & De Houwer, J. (2003). Eye movements to smoking-related
pictures in smokers: Relationship between attentional biases and implicit and explicit
measures of stimulus valence. Addiction, 98(6), 825-36.
Mogg, K., Field, M., & Bradley, B. P. (2005). Attentional and approach biases for smoking
cues in smokers: An investigation of competing theoretical views of addiction.
Psychopharmacology (Berl), 180(2), 333-41.
Moser, J. S., Hajcak, G., Bukay, E., & Simons, R. F. (2006). Intentional modulation of emotional
responding to unpleasant pictures: An ERP study. Psychophysiology, 43(3), 292-296.
252
References
Mucha, R. F., Pauli, P., & Angrilli, A. (1998). Conditioned responses elicited by experimentally
produced cues for smoking. Canadian Journal of Physiology and Pharmacology, 76(3), 259-68.
Muller, B. W., Stude, P., Nebel, K., Wiese, H., Ladd, M. E., Forsting, M., et al. (2003). Sparse
imaging of the auditory oddball task with functional MRI. NeuroReport, 14(12), 1597-1601.
Munafò, M. R., Mogg, K., Roberts, S., Bradley, B. P., & Murphy, M. (2003). Selective processing
of smoking-related cues in current smokers, ex-smokers and never-smokers on the modified
stroop task. Journal of Psychopharmacology, 17(3), 310-316.
Namkoong, K., Lee, E., Lee, C. H., Lee, B. O., & An, S. K. (2004). Increased P3 amplitudes
induced by alcohol-related pictures in patients with alcohol dependence. Alcoholism:
Clinical and Experimental Research, 28(9), 1317-1323.
Neidermeyer, E. (1999). The normal EEG of the waking adult. In E. Neidermeyer, & F. Lopes
da Silva (Eds.), Electroencephalography: Basic principles, clinical applications and related
fields (pp. 149-173). Baltimore: Lippincott Williams and Wilkins.
Niaura, R., Abrams, D. B., Pedraza, M., Monti, P. M., & Rohsenow, D. J. (1992). Smokers’
reactions to interpersonal interaction and presentation of smoking cues. Addictive
Behaviors, 17(6), 557-566.
Niaura, R., Rohsenow, D. J., Binkoff, J. A., Monti, P. M., Pedraza, M., & Abrams, D. B. (1988).
Relevance of cue reactivity to understanding alcohol and smoking relapse. Journal of
Abnormal Psychology, 97(2), 133-152.
Niaura, R., Shadel, W. G., Abrams, D. B., Monti, P. M., Rohsenow, D. J., & Sirota, A. (1998).
Individual differences in cue reactivity among smokers trying to quit: Effects of gender and
cue type. Addictive Behaviors, 23(2), 209-224.
Nickerson, L. D., Ravichandran, C., Lundahl, L. H., Rodolico, J., Dunlap, S., Trksak, G. H.,
et al. (2011). Cue reactivity in cannabis-dependent adolescents. Psychology of Addictive
Behaviors, 25(1), 168-173.
Nieuwenhuis, S., Ridderinkhof, K. R., Blom, J., Band, G. P. H., & Kok, A. (2001). Error-related
brain potentials are differentially related to awareness of response errors: Evidence from
an antisaccade task. Psychophysiology, 38(5), 752-760.
Nijs, I. M., Franken, I. H., & Muris, P. (2008). Food cue-elicited brain potentials in obese and
healthy-weight individuals. Eating Behaviors, 9(4), 462-470.
Nijs, I. M., Muris, P., Euser, A. S., & Franken, I. H. (2010). Differences in attention to food
and food intake between overweight/obese and normal-weight females under conditions
of hunger and satiety. Appetite, 54(2), 243-254.
Noël, X., Colmant, M., Van Der Linden, M., Bechara, A., Bullens, Q., Hanak, C., et al. (2006).
Time course of attention for alcohol cues in abstinent alcoholic patients: The role of initial
orienting. Alcoholism, Clinical and Experimental Research, 30(11), 1871-1877.
O’Brien, C. P., O’Brien, T. J., Mintz, J., & Brady, J. P. (1975). Conditioning of narcotic abstinence
symptoms in human subjects. Drug and Alcohol Dependence, 1(2), 115-123.
253
References
Ochsner, K. N., & Gross, J. J. (2008). Cognitive emotion regulation. Current Directions in
Psychological Science, 17(2), 153-158.
O’Connell, K. A., Hosein, V. L., Schwartz, J. E., & Leibowitz, R. Q. (2007). How does coping
help people resist lapses during smoking cessation? Health Psychology, 26(1), 77-84.
Ohman, A., Flykt, A., & Esteves, F. (2001). Emotion drives attention: Detecting the snake in
the grass. Journal of Experimental Psychology: General, 130(3), 466-478.
Olofsson, J. K., Nordin, S., Sequeira, H., & Polich, J. (2008). Affective picture processing: An
integrative review of ERP findings. Biological Psychology, 77(3), 247-265.
Orleans, C. T., Rimer, B. K., Cristinzio, S., Keintz, M. K., & Fleisher, L. (1991). A national survey
of older smokers: Treatment needs of a growing population. Health Psychology, 10(5),
343-351.
Overbeek, T. J. M., Nieuwenhuis, S., & Ridderinkhof, K. R. (2005). Dissociable components
of error processing: On the functional significance of the pe vis-à-vis the ERN/Ne.
Psychophysiology, 19(4), 319-329.
Park, H. S., Kim, S. H., Bang, S. A., Yoon, E. J., Cho, S. S., & Kim, S. E. (2010). Altered regional
cerebral glucose metabolism in internet game overusers: A 18F-fluorodeoxyglucose
positron emission tomography study. CNS Spectrums, 15(3), 159-166.
Piasecki, T. M., Kenford, S. L., Smith, S. S., Fiore, M. C., & Baker, T. B. (1997). Listening to
nicotine: Negative affect and the smoking withdrawal conundrum. Psychological Science,
8(3), 184-189.
Pieters, S., Van Der Vorst, H., Burk, W. J., Schoenmakers, T. M., Van Den Wildenberg, E.,
Smeets, H. J., et al. (2011). The effect of the OPRM1 and DRD4 polymorphisms on the
relation between attentional bias and alcohol use in adolescence and young adulthood.
Developmental Cognitive Neuroscience, 1(4), 591-599.
Pizzagalli, D. A., Sherwood, R. J., Henriques, J. B., & Davidson, R. J. (2005). Frontal brain
asymmetry and reward responsiveness. A source-localization study. Psychological Science,
16(10), 805-813.
Polich, J., & Criado, J. R. (2006). Neuropsychology and neuropharmacology of P3a and P3b.
International Journal of Psychophysiology, 60(2), 172-185.
Polich, J., & Kok, A. (1995a). Cognitive and biological determinants of P300: An integrative
review. Biological Psychology, 41(2), 103-146.
Polich, J., & Kok, A. (1995b). Cognitive and biological determinants of P300: An integrative
review. Biological Psychology, 41(2), 103-146.
Pontifex, M. B., Hillman, C. H., & Polich, J. (2009). Age, physical fitness, and attention: P3a
and P3b. Psychophysiology, 46(2), 379-387.
254
References
Porjesz, B., Begleiter, H., Reich, T., Van Eerdewegh, P., Edenberg, H. J., Foroud, T., et al. (1998).
Amplitude of visual P3 event-related potential as a phenotypic marker for a predisposition
to alcoholism: Preliminary results from the COGA project. collaborative study on the
genetics of alcoholism. Alcoholism, Clinical and Experimental Research, 22(6), 1317-1323.
Potts, G. F. (2004). An ERP index of task relevance evaluation of visual stimuli. Brain and
Cognition, 56(1), 5-13.
Poulos, C. X., Hinson, R. E., & Siegel, S. (1981). The role of pavlovian processes in drug
tolerance and dependence: Implications for treatment. Addictive Behaviors, 6(3), 205-211.
Powell, J., Dawkins, L., West, R., Powell, J., & Pickering, A. (2010). Relapse to smoking during
unaided cessation: Clinical, cognitive and motivational predictors. Psychopharmacology,
212(4), 537-549.
Pritchard, W., Sokhadze, E., & Houlihan, M. (2004). Effects of nicotine and smoking on
event-related potentials: A review. Nicotine & Tobacco Research, 6(6), 961-84.
Rabbitt, P., & Vyas, S. (1981). Processing a display even after you make a response to it.
how perceptual errors can be corrected. The Quarterly Journal of Experimental Psychology
Section A, 33(3), 223-239.
Reid, M. S., Flammino, F., Howard, B., Nilsen, D., & Prichep, L. S. (2006). Topographic
imaging of quantitative EEG in response to smoked cocaine self-administration in humans.
Neuropsychopharmacology, 31(4), 872-84.
Reid, M. S., Prichep, L. S., Ciplet, D., O’Leary, S., Tom, M., Howard, B., et al. (2003).
Quantitative electroencephalographic studies of cue-induced cocaine craving. Clinical
Electroencephalography, 34(3), 110-23.
Ridderinkhof, K. R., Ramautar, J. R., & Wijnen, J. G. (2009). To PE or not to PE: A P3-like ERP
component reflecting the processing of response errors. Psychophysiology, 46(3), 531-538.
Ridderinkhof, K. R., Van den Wildenberg, W. P. M., Segalowitz, S. J., & Carter, C. S. (2004).
Neurocognitive mechanisms of cognitive control: The role of prefrontal cortex in action
selection, response inhibition, performance monitoring, and reward-based learning. Brain
and Cognition, 56(2), 129-140.
Robbins, T. W., Ersche, K., & Everitt, B. (2008). Drug addiction and the memory systems of
the brain. Annals of the New York Academy of Sciences, 1141(1), 1-21.
Robbins, T. W., & Everitt, B. J. (1996). Neurobehavioural mechanisms of reward and
motivation. Current Opinion in Neurobiology, 6(2), 228-236.
Robinson, T. E., & Berridge, K. C. (1993). The neural basis of drug craving: An incentivesensitization theory of addiction. Brain Research. Brain Research Reviews, 18(3), 247-91.
Robinson, T. E., & Berridge, K. C. (2003). Addiction. Annual Review of Psychology, 54, 25-53.
Rusted, J. M., Caulfield, D., King, L., & Goode, A. (2000). Moving out of the laboratory: Does
nicotine improve everyday attention? Behavioural Pharmacology, 11(7-8), 621-629.
255
References
Ryan, F. (2002). Detected, selected, and sometimes neglected: Cognitive processing of cues
in addiction. Experimental and Clinical Psychopharmacology, 10(2), 67-76.
Salisbury, D. F., Rutherford, B., Shenton, M. E., & McCarley, R. W. (2001). Button-pressing
affects P300 amplitude and scalp topography. Clinical Neurophysiology, 112(9), 1676-1684.
Saunders, J. B., Aasland, O. G., Babor, T. F., de la Fuente, J. R., & Grant, M. (1993). Development
of the alcohol use disorders identification test (AUDIT): WHO collaborative project on early
detection of persons with harmful alcohol consumption--II. Addiction, 88(6), 791-804.
Sayette, M. A., & Hufford, M. R. (1994). Effects of cue exposure and deprivation on cognitive
resources in smokers. Journal of Abnormal Psychology, 103(4), 812-8.
Sayette, M. A., Martin, C. S., Wertz, J. M., Shiffman, S., & Perrott, M. A. (2001). A multidimensional analysis of cue-elicited craving in heavy smokers and tobacco chippers.
Addiction, 96(10), 1419-32.
Schindler, C., Panlilio, L., & Goldberg, S. R. (2002). Second-order schedules of drug selfadministration in animals. Psychopharmacology, 163(3), 327-344.
Schoenbaum, G., & Setlow, B. (2005). Cocaine makes actions insensitive to outcomes but
not extinction: Implications for altered orbitofrontal-amygdalar function. Cerebral Cortex,
15(8), 1162-1169.
Schoenmakers, T. M., de Bruin, M., Lux, I. F., Goertz, A. G., Van Kerkhof, D. H., & Wiers, R.
W. (2010). Clinical effectiveness of attentional bias modification training in abstinent
alcoholic patients. Drug and Alcohol Dependence, 109(1-3), 30-6.
Schoenmakers, T. M., Wiers, R. W., Jones, B. T., Bruce, G., & Jansen, A. T. M. (2007). Attentional
re-training decreases attentional bias in heavy drinkers without generalization. Addiction,
102(3), 399-405.
Schupp, H. T., Cuthbert, B. N., Bradley, M. M., Cacioppo, J. T., Ito, T., & Lang, P. J. (2000).
Affective picture processing: The late positive potential is modulated by motivational
relevance. Psychophysiology, 37(2), 257-61.
Schupp, H. T., Cuthbert, B. N., Bradley, M. M., Hilman, C. H., Hamm, A. O., & Lang, P. J. (2004).
Brain processes in emotional perception: Motivated attention. Cognition & Emotion, 18(5),
593-611.
Schupp, H. T., Junghöfer, M., Weike, A. I., & Hamm, A. O. (2003a). Attention and emotion:
An ERP analysis of facilitated emotional stimulus processing. NeuroReport, 14(8), 1107-10.
Schupp, H. T., Junghöfer, M., Weike, A. I., & Hamm, A. O. (2003b). Emotional facilitation of
sensory processing in the visual cortex. Psychological Science, 14(1), 7-13.
Schupp, H. T., Junghöfer, M., Weike, A. I., & Hamm, A. O. (2004). The selective processing of
briefly presented affective pictures: An ERP analysis. Psychophysiology, 41(3), 441-9.
Schupp, H. T., Stockburger, J., Codispoti, M., Junghöfer, M., Weike, A. I., & Hamm, A. O. (2007).
Selective visual attention to emotion. Journal of Neuroscience, 27(5), 1082-9.
256
References
Shadel, W. G., Niaura, R., & Abrams, D. B. (2001). Does completing a craving questionnaire
promote increased smoking craving? an experimental investigation. Psychology of Addictive
Behaviors, 15(3), 265-267.
Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., et al. (1998).
The mini-international neuropsychiatric interview (M.I.N.I.): The development and
validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal
of Clinical Psychiatry, 59(20), 22-33.
Shiffman, S. (2005). Dynamic influences on smoking relapse process. Journal of Personality,
73(6), 1715-1748.
Shiffman, S., Engberg, J. B., Paty, J. A., Perz, W. G., Gnys, M., Kassel, J. D., et al. (1997). A day
at a time: Predicting smoking lapse from daily urge. Journal of Abnormal Psychology, 106(1),
104-116.
Shiffman, S., & Jarvik, M. E. (1976). Smoking withdrawal symptoms in two weeks of
abstinence. Psychopharmacology, 50(1), 35-39.
Shiffman, S., Kassel, J. D., Paty, J., Gnys, M., & Zettler-Segal, M. (1994). Smoking typology
profiles of chippers and regular smokers. Journal of Substance Abuse, 6(1), 21-35.
Shiffman, S., Paty, J. A., Gnys, M., Kassel, J. A., & Hickcox, M. (1996). First lapses to smoking:
Within-subjects analysis of real-time reports. Journal of Consulting and Clinical Psychology,
64(2), 366-79.
Shiffman, S., Paty, J. A., Gnys, M., Kassel, J. D., & Elash, C. (1995). Nicotine withdrawal in
chippers and regular smokers: Subjective and cognitive effects. Health Psychology, 14(4),
301-9.
Shiffman, S., Paty, J. A., Kassel, J. D., Gnys, M., & Zettler-Segal, M. (1994). Smoking behavior
and smoking history of tobacco chippers. Experimental and Clinical Psychopharmacology,
2(2), 126-142.
Shostakovich, G. S. (1987). Neurologic mechanism of the unconscious craving for alcohol
in chronic alcoholic patients. [O nervnom mekhanizme neosoznavaemogo vlecheniia k
alkogoliu u bol’nykh khronicheskim alkogolizmom] Zhurnal Nevropatologii i Psikhiatrii
Imeni S.S.Korsakova, 87(6), 898-902.
Siegel, S. (1983). Classical conditioning, drug tolerance, and drug dependence. In Y. Israel, F.
B. Glaser, H. Kalant, R. E. Popham & W. Schmidt (Eds.), Research advances in alcohol and drug
problems (pp. 207-246). New York: Plenum Press.
Singer, W. (1993). Synchronization of cortical activity and its putative role in information
processing and learning. Annual Review of Psychology, 55, 349-74.
Snaith, R. P., Hamilton, M., Morley, S., Humayan, A., Hargreaves, D., & Trigwell, P. (1995).
A scale for the assessment of hedonic tone: The snaith-hamilton pleasure scale. British
Journal of Psychiatry, 167, 99-103.
257
References
Sokhadze, E., Singh, S., Stewart, C., Hollifield, M., El-Baz, A., & Tasman, A. (2008). Attentional
bias to drug- and stress-related pictorial cues in cocaine addiction comorbid with PTSD.
Journal of Neurotherapy, 12(4), 205-225.
Sokhadze, E., Stewart, C., Hollifield, M., & Tasman, A. (2008). Event-related potential
study of executive dysfunctions in a speeded reaction task in cocaine addiction. Journal of
Neurotherapy, 12(4), 185-204.
Spealman, R. D., & Goldberg, S. R. (1982). Maintenance of schedule-controlled behavior
by intravenous injections of nicotine in squirrel monkeys. Journal of Pharmacology and
Experimental Therapeutics, 223(2), 402-408.
Stetter, F., Chaluppa, C., Ackerman, K., Straube, E. R., & Mann, K. (1994). Alcoholics’ selective
processing of alcohol related words and cognitive performance on a stroop task. European
Psychiatry, 9(2), 71-76.
Stewart, J., de Wit, H., & Eikelboom, R. (1984). Role of unconditioned and conditioned drug
effects in the self-administration of opiates and stimulants. Psychological Review, 91(2),
251-68.
Stockburger, J., Schmalzle, R., Flaisch, T., Bublatzky, F., & Schupp, H. T. (2009). The impact of
hunger on food cue processing: An event-related brain potential study. NeuroImage, 47(4),
1819-1829.
Stormark, K. M., Field, N. P., Hugdahl, K., & Horowitz, M. (1997). Selective processing of
visual alcohol cues in abstinent alcoholics: An approach-avoidance conflict? Addictive
Behaviors, 22(4), 509-519.
Stormark, K. M., Laberg, J. C., Nordby, H., & Hugdahl, K. (2000). Alcoholics’ selective attention
to alcohol stimuli: Automated processing? Journal of Studies on Alcohol, 61(1), 18-23.
Streeter, C. C., Terhune, D. B., Whitfield, T. H., Gruber, S., Sarid-Segal, O., Silveri, M. M., et al.
(2008). Performance on the stroop predicts treatment compliance in cocaine-dependent
individuals. Neuropsychopharmacology, 33(4), 827-36.
Sublette, V. A., & Mullan, B. (2010). Consequences of play: A systematic review of the effects
of online gaming. International Journal of Mental Health and Addiction, 10(1), 3-23.
Surwillo, W. W. (1961). Frequency of the “alpha” rhythm, reaction time and age. Nature,
191(4790), 823-824.
Surwillo, W. W. (1963). The relation of simple response time to brain-wave frequency and
the effects of age. Electroencephalography and Clinical Neurophysiology, 15, 105-14.
Swan, G. E., Ward, M. M., & Jack, L. M. (1996). Abstinence effects as predictors of 28-day
relapse in smokers. Addictive Behaviors, 21(4), 481-490.
Tejeiro Salguero, R. A., & Morán, R. M. B. (2002). Measuring problem video game playing in
adolescents. Addiction, 97(12), 1601-1606.
258
References
Teneggi, V., Squassante, L., Milleri, S., Polo, A., Lanteri, P., Ziviani, L., et al. (2004). EEG
power spectra and auditory P300 during free smoking and enforced smoking abstinence.
Pharmacology Biochemistry and Behavior, 77(1), 103-9.
Thalemann, R., Wölfling, K., & Grüsser, S. M. (2007). Specific cue reactivity on computer
game-related cues in excessive gamers. Behavioral Neuroscience, 121(3), 614-8.
Thewissen, R., Havermans, R. C., Geschwind, N., van den Hout, M., & Jansen, A. (2007).
Pavlovian conditioning of an approach bias in low-dependent smokers. Psychopharmacology
(Berl), 194(1), 33-39.
Thewissen, R., Snijders, S. J. B. D., Havermans, R. C., van den Hout, M., & Jansen, A. (2006).
Renewal of cue-elicited urge to smoke: Implications for cue exposure treatment. Behaviour
Research and Therapy, 44(10), 1441-1449.
Thewissen, R., van den Hout, M., Havermans, R. C., & Jansen, A. (2005). Context-dependency
of cue-elicited urge to smoke. Addiction, 100(3), 387-396.
Thiruchselvam, R., Blechert, J., Sheppes, G., Rydstrom, A., & Gross, J. J. (2011). The temporal
dynamics of emotion regulation: An EEG study of distraction and reappraisal. Biological
Psychology, 87(1), 84-92.
Thorberg, F. A., & Lyvers, M. (2006). Negative mood regulation (NMR) expectancies, mood,
and affect intensity among clients in substance disorder treatment facilities. Addictive
Behaviors, 31(5), 811-20.
Tiffany, S. T., & Drobes, D. J. (1991). The development and initial validation of a questionnaire
on smoking urges. British Journal of Addiction, 86(11), 1467-76.
Torregrossa, M. M., Corlett, P. R., & Taylor, J. R. (2011). Aberrant learning and memory in
addiction. Neurobiology of Learning and Memory, , doi:10.1016/j.nlm.2011.02.014.
Townshend, J. M., & Duka, T. (2001). Attentional bias associated with alcohol cues:
Differences between heavy and occasional social drinkers. Psychopharmacology (Berl),
157(1), 67-74.
Townshend, J. M., & Duka, T. (2007). Avoidance of alcohol-related stimuli in alcoholdependent inpatients. Alcoholism, Clinical and Experimental Research, 31(8), 1349-1357.
Ulett, J. A., & Itil, T. M. (1969). Quantitative electroencephalogram in smoking and smoking
deprivation. Science, 164(882), 969-70.
UNDCP/WHO (1992). Informal expert committee on the craving mechanism: Report No. V.
92-54439T)
Van de Laar, M. C., Licht, R., Franken, I. H. A., & Hendriks, V. M. (2004). Event-related
potentials indicate motivational relevance of cocaine cues in abstinent cocaine addicts.
Psychopharmacology (Berl), 177(1-2), 121-9.
259
References
Van Holst, R. J., Van den Brink, W., Veltman, D. J., & Goudriaan, A. E. (2010). Why gamblers
fail to win: A review of cognitive and neuroimaging findings in pathological gambling.
Neuroscience & Biobehavioral Reviews, 34(1), 87-107.
Van Houwelingen, H. C., Arends, L. R., & Stijnen, T. (2002). Advanced methods in metaanalysis: Multivariate approach and meta-regression. Statistics in Medicine, 21(4), 589-624.
Van Rooij, A. J., Schoenmakers, T. M., Van den Eijnden, R. J., Vermulst, A. A., & Van de Mheen,
D. (submitted). Videogame addiction test (VAT): Validity and psychometric characteristics.
Van Rooij, A. J., Schoenmakers, T. M., Van den Eijnden, R. J. J. M., & Van de Mheen, D. (2010).
Compulsive internet use: The role of online gaming and other internet applications. Journal
of Adolescent Health, 47(1), 51-7.
Van Rooij, A. J., Schoenmakers, T. M., Vermulst, A. A., Van den Eijnden, R. J., & Van de Mheen,
D. (2010). Online video game addiction: Identification of addicted adolescent gamers.
Addiction, 106(1), 205-12.
Van Strien, J. W. (1992). Classificatie van links- en rechtshandige proefpersonen. Nederlands
Tijdschrift Voor De Psychologie En Haar Grensgebieden, 47, 88-92.
Verdejo-García, A. J., Lawrence, A. J., & Clark, L. (2008). Impulsivity as a vulnerability
marker for substance-use disorders: Review of findings from high-risk research, problem
gamblers and genetic association studies. Neuroscience & Biobehavioral Reviews, 32(4),
777-810.
Verdejo-García, A. J., Perales, J. S., & Pérez-García, M. (2007). Cognitive impulsivity in
cocaine and heroin polysubstance abusers. Addictive Behaviors, 32(5), 950-966.
Versace, F., Lam, C. Y., Engelmann, J. M., Robinson, J. D., Minnix, J. A., Brown, V. L., et al.
(2011). Beyond cue reactivity: Blunted brain responses to pleasant stimuli predict longterm smoking abstinence. Addiction Biology, doi: 10.1111/j.1369-1600.2011.00372.x
Versace, F., Minnix, J. A., Robinson, J. D., Lam, C. Y., Brown, V. L., & Cinciripini, P. M. (2011).
Brain reactivity to emotional, neutral and cigarette-related stimuli in smokers. Addiction
Biology, 16(2), 296-307.
Vink, J. M., Willemsen, G., Beem, A. L., & Boomsma, D. I. (2005). The fagerström test for
nicotine dependence in a dutch sample of daily smokers and ex-smokers. Addictive
Behaviors, 30(3), 575-9.
Volkow, N. D., Fowler, J. S., Wang, G. J., Baler, R., & Telang, F. (2009). Imaging dopamine’s role
in drug abuse and addiction. Neuropharmacology, 56(1), 3-8.
Volkow, N. D., Fowler, J. S., Wang, G. J., Telang, F., Logan, J., Jayne, M., et al. (2010). Cognitive
control of drug craving inhibits brain reward regions in cocaine abusers. NeuroImage,
49(3), 2536-43.
Vollstädt-Klein, S., Loeber, S., von der Goltz, C., Mann, K., & Kiefer, F. (2009). Avoidance of
alcohol-related stimuli increases during the early stage of abstinence in alcohol-dependent
patients. Alcohol and Alcoholism, 44(5), 458-463.
260
References
Vuilleumier, P. (2005). How brains beware: Neural mechanisms of emotional attention.
Trends in Cognitive Sciences, 9(12), 585-594.
Warren, C. A., & McDonough, B. E. (1999). Event-related brain potentials as indicators of
smoking cue-reactivity. Clinical Neurophysiology, 110(9), 1570-84.
Waters, A. J., & Feyerabend, C. (2000). Determinants and effects of attentional bias in
smokers. Psychology of Addictive Behaviors, 14(2), 111-20.
Waters, A. J., & Sayette, M. A. (2006). Implicit cognition and tobacco addiction. In R. W.
Wiers, & A. W. Stacy (Eds.), Handbook of implicit cognition and addiction (pp. 309-338).
Thousand Oaks: Sage Publications.
Waters, A. J., Shiffman, S., Bradley, B. P., & Mogg, K. (2003). Attentional shifts to smoking
cues in smokers. Addiction, 98(10), 1409-17.
Waters, A. J., Shiffman, S., Sayette, M. A., Paty, J. A., Gwaltney, C. J., & Balabanis, M. H. (2003).
Attentional bias predicts outcome in smoking cessation. Health Psychology, 22(4), 378-87.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures
of positive and negative affect: The PANAS scales. Journal of Personality and Social
Psychology, 54(6), 1063-1070.
Weinstein, A., & Lejoyeux, M. (2010). Internet addiction or excessive internet use. American
Journal of Drug and Alcohol Abuse, 36(5), 277-283.
Wertz, J. M., & Sayette, M. A. (2001). A review of the effects of perceived drug use opportunity
of self-reported urge. Experimental and Clinical Psychopharmacology, 9(1), 3-13.
West, R., Hajek, P., & Belcher, M. (1989). Time course of cigarette widhdrawal symptoms
while using nicotine gum. Psychopharmacology, 99(1), 143-145.
White, N. M. (1996). Addictive drugs as reinforcers: Multiple partial actions on memory
systems. Addiction, 91(7), 921-950.
Wikler, A. (1980). Opoid dependence: Mechanisms and treatment. New York: Plenum Press.
Wilson, S. J., Sayette, M. A., Delgado, M. R., & Fiez, J. A. (2005). Instructed smoking expectancy
modulates cue-elicited neural activity: A preliminary study. Nicotine & Tobacco Research,
7(4), 637-645.
Wilson, S. J., Sayette, M. A., & Fiez, J. A. (2004). Prefrontal responses to drug cues: A
neurocognitive analysis. Nature Neuroscience, 7(3), 211-4.
Winkler, M. H., Weyers, P., Mucha, R. F., Stippekohl, B., Stark, R., & Pauli, P. (2010). Conditioned
cues for smoking elicit preparatory responses in healthy smokers. Psychopharmacology
(Berl),
Wise, R. A. (1988). The neurobiology of craving: Implications for the understanding and
treatment of addiction. Journal of Abnormal Psychology, 97(2), 118-32.
261
References
Wölfling, K., Flor, H., & Grüsser, S. M. (2008). Psychophysiological responses to drugassociated stimuli in chronic heavy cannabis use. European Journal of Neuroscience, 27(4),
976-83.
Wrobel, A. (2000). Beta activity: A carrier for visual attention. Acta Neurobiologiae
Experimentalis, 60(2), 247-60.
Yang, B., Yang, S. Y., Zhao, L., Yin, L. H., Liu, X., & An, S. S. (2009). Event-related potentials in
a Go/Nogo task of abnormal response inhibition in heroin addicts. Science in China, Series C:
Life Sciences, 52(8), 780-788.
Yücel, M., & Lubman, D. I. (2007). Neurocognitive and neuroimaging evidence of
behavioural dysregulation in human drug addiction: Implications for diagnosis, treatment
and prevention. Drug and Alcohol Review, 26(1), 33-39.
Yücel, M., Lubman, D. I., Harrison, B. J., Fornito, A., Allen, N. B., Wellard, R. M., et al. (2007). A
combined spectroscopic and functional MRI investigation of the dorsal anterior cingulate
region in opiate addiction. Molecular Psychiatry, 12(7), 691-702.
Zhou, Y., Lin, F. C., Du, Y. S., Qin, L. D., Zhao, Z. M., Xu, J. R., et al. (2009). Gray matter
abnormalities in internet addiction: A voxel-based morphometry study. European Journal
of Radiology, 79(1), 92-95.
Zhou, Z. H., Yuan, G. Z., Yao, J. J., Li, C., & Cheng, Z. H. (2010). An event-related potential
investigation of deficient inhibitory control in individuals with pathological internet use.
Acta Neuropsychiatrica, 22(5), 228-236.
Zinser, M. C., Baker, T. B., Sherman, J. E., & Cannon, D. S. (1992). Relation between selfreported affect and drug urges and cravings in continuing and withdrawing smokers.
Journal of Abnormal Psychology, 101(4), 617-629.
Zinser, M. C., Fiore, M. C., Davidson, R. J., & Baker, T. B. (1999). Manipulating smoking
motivation: Impact on an electrophysiological index of approach motivation. Journal of
Abnormal Psychology, 108(2), 240-54.
262
Samenvatting
Samenvatting
Achtergrond
Verslaving of middelenafhankelijkheid is een zeer moeilijk te behandelen stoornis
die gekenmerkt wordt door onweerstaanbare verlangens naar een middel (craving)
en frequente terugval. Daarnaast spelen bepaalde cognitieve processen (aandacht,
geheugen) een centrale rol in verslaving. Meer specifiek gaat het om verstoringen
(biases) in de cognitieve verwerking van met middelen geassocieerde prikkels
(middelgerelateerde stimuli of cues), zoals de geur en het aangezicht van het
middel zelf of attributen die vaak aanwezig zijn bij het consumeren van het middel.
Invloedrijke theorieën stellen dat middelafhankelijke personen automatisch hun
aandacht richten op middelgerelateerde cues; zij detecteren deze cues dus zonder
er bij stil te staan (Franken, 2003). Daardoor blijft er minder aandacht over voor
alternatieve cues in de omgeving, wordt men overspoeld door middelgerelateerde
gedachten en ontstaat er intense craving. Van deze processen wordt gedacht dat ze
een wederkerige relatie met elkaar hebben: meer craving leidt tot meer aandacht
en meer aandacht leidt tot meer craving. Dit zorgt ervoor dat de middelafhankelijke
persoon als het ware gevangen raakt in een vicieuze cirkel van steeds groter
wordende aandacht en steeds groter wordende craving, tot uiteindelijk het middel
wordt opgezocht en wordt geconsumeerd.
Het ontstaan van cognitieve verwerkingsbiases kan verklaard worden door de
incentieve-sensitisatie theorie van verslaving (Robinson & Berridge, 1993).
Deze theorie stelt dat herhaalde toediening van een middel ervoor zorgt dat
het dopaminesysteem in het brein, het systeem dat betrokken is bij beloning,
overgevoelig (of: gesensitiseerd) raakt. Als gevolg hiervan krijgen zowel het middel
zelf als de middelgerelateerde cues een bepaalde belonende, motivationele waarde
toegekend. Met andere woorden, het overgevoelige dopaminesysteem zorgt ervoor
dat het middel en de middelgerelateerde cues voor de middelafhankelijke persoon
zeer opvallend, belonend en ‘gewild’ worden, wat er toe leidt dat er automatisch
meer aandacht aan besteed wordt.
Experimenteel onderzoek heeft bevestigd dat middelafhankelijke personen
inderdaad excessief veel aandacht besteden aan middelgerelateerde cues. Zij
kunnen middelgerelateerde cues daarnaast ook beter onthouden dan andere
cues. Door gebruikmaking van aandachtstaken, zoals de Stroop taak en de visuele
probe taak, en geheugentaken, zoals free recall en de woord-completie taak, zijn
265
Samenvatting
cognitieve verwerkingsbiases vastgesteld in een grote verscheidenheid aan
verslavingen, varierend van heroïne-afhankelijkheid tot roken en pathologisch
gokken (Field & Cox, 2008; Franken et al., 2003; McCusker & Gettings, 1997).
Ook heeft onderzoek aangetoond dat er een significante relatie bestaat tussen
cognitieve verwerkingsbiases en craving (Field et al., 2009). Daarnaast is er steeds
meer bewijs dat cognitieve verwerkingsbiases belangrijk zijn bij het ontstaan en
het voortduren van verslaving. Zo komt er uit diverse onderzoeken naar voren dat
biases in aandacht voor middelgerelateerde cues geassocieerd zijn met een slechte
behandeluitkomst, terugval na behandeling en middelconsumptie (bv. Marissen
et al., 2006). Dus hoe sterker de bias, hoe lastiger men te behandelen is, hoe
eerder men terugvalt en hoe eerder en vaker men het middel consumeert. Deze
bevindingen onderstrepen de klinische relevantie van onderzoek naar de rol van
cognitieve processen in verslaving.
Een relatief nieuwe methode om de cognitieve verwerking van middelgerelateerde
cues te bestuderen is het meten van Event-Related Potentials (gebeurtenisgerelateerde hersengolven; ERPs) door gebruikmaking van elektro-encefalografie
(EEG). ERPs geven een potentieel directere maat van cognitieve verwerking dan
conventionele gedragstaken zoals de Stroop en visuele probe taak. Gedragstaken
hebben reactietijden of nauwkeurigheidspercentages als uitkomstmaat, oftewel:
uitkomstmaten die gebaseerd zijn op motorresponsen (het drukken op een knop)
en zijn daardoor slechts een indirecte weergave van mentale processen. Een recent
onderzoek laat daarnaast zien dat de meerderheid van gedragstaken die cognitieve
verwerkingsprocessen in verslaving beogen te meten een lage betrouwbaarheid
heeft (Ataya et al., 2011).
ERPs zijn weergaven van hersenactiviteit die plaatsvindt in voorbereiding of
reactie op afgebakende gebeurtenissen, zoals het maken van een fout of het zien
van een bepaald soort plaatje (Fabiani et al., 2000). Ze worden gemeten met zeer
gevoelige elektroden die verspreid over de schedel worden bevestigd. ERPs bestaan
uit verschillende pieken en dalen die fluctueren als gevolg van experimentele
manipulaties. Zo kunnen sommige pieken groter worden bij specifieke instructies
(bv. ‘besteed aandacht aan dit plaatje’) of bij specifieke inhoud van aangeboden
stimuli (bv. plezierige inhoud). Golven die geassocieerd zijn met specifieke mentale
processen (bv. het richten van aandacht of opslag in het geheugen) worden ook wel
componenten genoemd (Coles & Rugg, 1995). Componenten worden doorgaans
genoemd naar hun polariteit (positief of negatief) en latentietijd (bv. 300 ms). Ze
266
Samenvatting
varieren in amplitude, wat naar alle waarschijnlijkheid aangeeft in welke mate of
met welke sterkte een mentaal proces plaatsvindt (Kok, 1990).
Van twee late ERP componenten, namelijk de P300 en het Late Positieve Potentiaal
(LPP), is aangetoond dat ze een goede maat zijn voor de cognitieve verwerking van
middelgerelateerde stimuli (Franken, Stam et al., 2003; Van de Laar et al., 2004). In
middelafhankelijke personen zijn deze twee componenten doorgaans vergroot in
reactie op middelgerelateerde cues (bv. een foto van een pakje sigaretten) vergeleken
met neutrale cues (bv. een foto van een huishoudelijk object). Dit amplitudeverschil
in reactie op de twee soorten cues wordt niet gevonden in niet-afhankelijke, controle
proefpersonen. Ondanks dat er geen volledige duidelijkheid is over waar de P300
en LPP componenten precies voor staan, wordt er over het algemeen aangenomen
dat de late ERP componenten een weerspiegeling zijn van aandachtsprocessen,
alsmede de activatie van motivatie- en arousalgerelateerde mechanismen is het
brein (Cuthbert et al., 2000; Hajcak et al., 2010; Lang et al., 1997; Olofsson et al., 2008;
Schupp et al., 2000). In studies met middelafhankelijke personen wordt daarom
aangenomen dat vergrote P300 en LPP componenten staan voor ‘motivationele
aandacht’; ze weerspiegelen de verhoogde motivationele waarde, belangrijkheid
en opvallendheid van middelgerelateerde cues voor middelafhankelijke personen,
met verhoogde aandacht als onderliggend mechanisme.
Cognitieve verwerkingsbias in
middelenafhankelijkheid
In Hoofdstuk 1 is er een meta-analyse uitgevoerd met als doel de resultaten van
alle ERP studies samen te vatten en te integreren die de relatie hebben onderzocht
tussen de P300 en LPP amplitudes en middelgerelateerde verwerkingsbiases in
middelafhankelijke personen. Daarbij werd de algemene effect grootte (effect-size)
berekend van late ERP verschillen tussen middelgerelateerde en neutrale stimulus
condities en tussen middelafhankelijke personen en controle proefpersonen.
Eigenlijk werd er een poging gedaan de volgende vragen te beantwoorden: Laten
middelafhankelijke personen in het algemeen grotere late ERP amplitudes zien in
reactie op middelgerelateerde plaatjes dan in reactie op neutrale plaatjes? En is
dit verschil in het algemeen groter dan het eventuele verschil dat niet-afhankelijke
controle proefpersonen laten zien?
267
Samenvatting
Omdat er tussen de verschillende, individuele ERP studies verschillen zijn tussen
de locaties waar de effecten optreden (soms voor of frontaal in het brein, soms
meer naar achter of pariëtaal in het brein) is er in hoofdstuk 1 ook berekend waar
de effecten in het algemeen het grootst zijn. Daartoe werden effecten op een
frontale elektrode (Fz) vergeleken met effecten op een pariëtale elektrode (Pz).
Daarnaast werden er zogeheten ‘gestratificeerde moderator analyses’ uitgevoerd
voor een aantal specifieke kenmerken van de steekproeven en de studies, namelijk
voor het soort middel dat gebruikt werd (stimulerende versus kalmerende
middelen), de status van het middelgebruik (niet-abstinent versus abstinent voor
10 tot 30 dagen), geslacht, leeftijd en het soort taak dat proefpersonen moesten
doen (actieve versus passieve taken). Met de analyses kon vastgesteld worden of
de zojuist genoemde variabelen een potentieel modererend effect hebben op de
gevonden ERP effecten (bv. verschillen zijn significant groter/kleiner wanneer
middelafhankelijke personen abstinent zijn of ouder zijn), zodat daar in toekomstig
ERP onderzoek in middelenafhankelijkheid rekening mee gehouden kan worden.
De resultaten van de meta-analyse laten zien dat de elektrofysiologische
verwerking van middelgerelateerde cues ten opzichte van neutrale cues in het
algemeen groter is in middelafhankelijke personen dan in controle proefpersonen,
blijkens uit vergrote P300 en LPP amplitudes. Deze verschillen zijn voor beide
componenten significant en hebben een gemiddelde effect-size. Omdat deze
vergrote elektrofysiologische verwerking overeenkomsten vertoont met de
verwerking van stimuli die motivationeel belangrijk zijn voor iedereen (bv.
emotionele stimuli), en omdat eerder onderzoek heeft uitgewezen dat P300/LPP
amplitudes geassocieerd zijn met aandacht, craving en subjectieve valentie(plezierigheid) en arousalbeoordelingen van de stimuli, kan de huidige verhoogde
elektrofysiologische verwerking van middelgerelateerde stimuli verklaard
worden door de motivationele aandacht van middelafhankelijke personen.
De gestratificeerde moderator analyses laten zien dat de effecten niet verschillend
zijn op frontale en pariëtale elektrode locaties. Toekomstig onderzoek zou
daarom effecten moeten bestuderen op beide locaties. Effecten worden ook niet
gemodereerd door het soort middel, de status van gebruik, geslacht, leeftijd of
het soort taak dat gebruikt wordt. Deze resultaten geven aan dat de verhoogde
elektrofysiologische verwerking van middelgerelateerde cues, oftwel de
motivationele aandacht voor deze cues, kenmerkend is voor middelafhankelijkheid
in het algemeen, dat de verhoogde aandacht waarschijnlijk onafhankelijk is van
268
Samenvatting
geslacht, leeftijd en de recentheid van het gebruik (tot maximaal 30 dagen) en
dat de verhoogde aandacht waarschijnlijk aanwezig is gedurende allerlei soorten
taken, waaronder taken waarin expliciet geïnstrueerd wordt om op andere cues
dan de middelgerelateerde cues te letten.
De voornaamste conclusie die getrokken kan worden uit hoofdstuk 1 is dat
middelenafhankelijkheid gekenmerkt wordt door vergrote elektrofysiologische
verwerking (dat is: grotere P300 en LPP amplitudes) van middelgerelateerde cues.
Vanwege de associatie van deze elektrofysiologische activiteit met cognitieve
processen (aandacht) en motivatie (craving, valentie, arousal) wordt er tevens, en wel
door gebruikmaking van een directe maat, bevestigd dat middelenafhankelijkheid
gekenmerkt wordt door biases in de cognitieve verwerking van middelgerelateerde
cues (dat is: motivationele aandacht voor middelgerelateerde cues).
Zoals beschreven in Hoofdstuk 2 is het overgrote deel van de overige hoofdstukken
van dit proefschrift geschreven om inzicht te vergroten in de zojuist beschreven
elektrofysiologische correlaten van middelgerelateerde cognitieve verwerking.
Daartoe werden voornamelijk rokers met een gemiddelde afhankelijkheid getest. Er
werd onder andere gekeken of de cognitieve verwerkingbias blijft voortbestaan na
langere perioden van abstinentie, of de verwerkingsbias vooral impliciet of expliciet
van aard is, of de verwerkingbias exclusief betrekking heeft op middelgerelateerde
stimuli, of de verwerkingsbias geconditioneerd kan worden, of de verwerkingsbias
gemoduleerd (vergroot, verkleind) kan worden wanneer proefpersonen bepaalde
cognitieve denkstrategieën toepassen, en of de verwerkingsbias ook aanwezig is
in alcoholisten, een relatief onderbestudeerde groep (nb. er is slechts één studie
met alcoholisten meegenomen in de meta-analyse van hoofdstuk 1). Tevens is er
één hoofdstuk gewijd aan meer algemene (niet aan middelgerelateerde stimuligebonden) cognitieve processen die een rol spelen in het ontstaan en voortbestaan
van verslaving, namelijk responsinhibitie en foutverwerking; twee processen
waarvan gedacht wordt dat ze ten grondslag liggen aan controle over eigen gedrag
(zoals middelengebruik) en impulsiviteit. Deze processen werden bestudeerd in
een mogelijk nieuwe gedragsmatige verslaving, namelijk computer gaming. Er is
echter begonnen met het beschrijven van de validatie van de Nederlandse versie
van de brief Questionnaire on Smoking Urges (QSU-brief ), een vragenlijst die craving
voor sigaretten beoogt te meten en gebruikt is in de meeste hoofdstukken van dit
proefschrift.
269
Samenvatting
De validatie van de Nederlandse versie van de
QSU-brief
In Hoofdstuk 3 werden de factorstructuur, de interne consistentie
(betrouwbaarheid) en de validiteit bestudeerd van de naar het Nederlands
vertaalde QSU-brief. Dit werd gedaan in een Nederlandse populatie van rokers
(N = 208) door gebruikmaking van een cross-sectioneel design met random
proefpersoon selectie. Dit onderzoek was belangrijk omdat het tot op dat moment
nog onbekend was of de vragenlijst acceptabele psychometrische kenmerken had,
gelijkwaardig aan die van de originele, Engelse versie (L. S. Cox et al., 2001), en
of de vragenlijst (en de subschalen) op dezelfde wijze gescoord konden worden
als in de originele versie. De resultaten laten zien dat de Nederlandse QSU-brief
een goede interne consistentie bezit. Daarnaast blijken scores op de vragenlijst
sterk gecorreleerd te zijn met drie andere maten om craving voor sigaretten
mee vast te stellen, en gemiddeld gecorreleerd te zijn met vragenlijsten die
gerelateerde constructen meten, namelijk afhankelijkheid van sigaretten en het
aantal sigaretten per dag. Daarmee is aangetoond dat de Nederlandse QSU-brief
ook valide is. Een exploratieve factoranalyse duidt een twee-factor structuur aan
die net iets afwijkt van die van de originele versie. De eerste factor (subschaal 1)
kan het best omschreven worden als ‘bevrijding van ontwenningsverschijnselen
of negatief affect met een directe en overweldigende drang om te roken’. De
tweede factor (subschaal 2) weerspiegelt ‘het verlangen en de intentie om te
roken’. Geconcludeerd kan worden dat de Nederlandse vertaling van de QSU-brief
een betrouwbare, valide en multidimensionele assessment geeft van craving voor
sigaretten, tenminste, in een algemene populatie van jong volwassenen, en dat de
vragenlijst geschikt is voor gebruik in experimentele settings.
Rookgerelateerde verwerkingsbias na langere
abstinentie
In hoofdstuk 4 en 5 werd aandacht besteed aan de cognitieve verwerking van
middelgerelateerde stimuli na langere abstinentie. De incentieve-sensitisatie
theorie van verslaving (Robinson & Berridge, 1993) voorspelt niet alleen dat
herhaald middelgebruik leidt tot sensitisatie van het dopaminesysteem (waardoor
270
Samenvatting
het middel zelf en middelgerelateerde stimuli meer motivationele waarde en
aandacht krijgen), maar impliceert ook dat deze sensitisatie onomkeerbaar is.
Dit zou verklaren waarom mensen die volledig gestaakt zijn met middelgebruik
toch dikwijls terugvallen. Het is echter nog onduidelijk of verhoogde cognitieve
verwerking van middelgerelateerde cues een permanent kenmerk van verslaving is.
In Hoofdstuk 4 is onderzocht of middelgerelateerde verwerkingbias, zoals gemeten
met ERPs, nog steeds aanwezig is in ex-rokers na langere abstinentie (gemiddeld
1,4 jaar). Hiertoe werden de hersengolven (P300 en LPP amplitudes) van rokers,
ex-rokers en niet-rokers in reactie op rookgerelateerde en neutrale plaatjes
vergeleken. Daarnaast werd zelf-gerapporteerde nicotinecraving vastgesteld en
werden de valentie en arousal van de plaatjes gemeten. De resultaten laten zien
dat zowel P300 als LPP amplitudes in reactie ook rookstimuli significant groter
zijn in rokers dan in ex-rokers en niet-rokers, met name op frontale en centrale
elektrodeposities, terwijl de P300 en LPP amplitudes in reactie op neutrale stimuli
niet verschillen tussen de drie groepen. Een belangrijke bevinding is daarbij
dat ex-rokers en niet-rokers niet verschillen in hun elektrofysiologische reactie
op rookstimuli. Deze resultaten zijn niet het gevolg van verschillen in mate van
nicotine-afhankelijkheid; rokers en ex-rokers zijn/waren even afhankelijk van
roken. Samengevat hebben rokers een grotere verwerkingsbias dan ex-rokers,
terwijl ex-rokers dezelfde (lage) verwerkingsbias als niet-rokers laten zien.
Daarnaast hebben ex-rokers significant minder craving dan rokers en beoordelen
ze rookplaatjes als negatiever. Geconcludeerd kan worden dat rook-gerelateerde
verwerkingsbias, alsmede rookgerelateerde craving en zelf-gerapporteerde
plezierigheid van rookstimuli, kleiner wordt na een langere periode van abstinentie.
Waar ERPs ons met informatie kunnen verschaffen over afgebakende
hersenactiviteit in reactie op specifieke gebeurtenissen, kunnen EEG spectrum
metingen inzicht geven in meer algemene hersenactiviteit, oftewel de mentale
toestand, tijdens langere blootstelling aan stimuli. Hierbij valt te denken aan
veranderingen in alertheid, concentratie en arousal. In Hoofdstuk 5 werden de
veranderingen in het EEG spectrum van rokers bestudeerd tijdens 30 seconden
lange blootstelling aan een neutrale cue (pen) en een rookgerelateerde cue
(brandende sigaret). Om vast te kunnen stellen of eventuele EEG spectrum
veranderingen nog steeds aanwezig zijn na langere abstinentie, werd de activiteit
van rokers vergeleken met de activiteit van ex-rokers tijdens blootstelling aan
dezelfde twee cues. Daarnaast werden verschillen in zelf-gerapporteerde craving
271
Samenvatting
gemeten. Resultaten laten zien dat rokers tussen de twee condities (neutraal
-> rookgerelateerd) zowel een significante toename in craving vertonen als
een significante toename van activiteit in één van de frequentiebanden van het
EEG spectrum, namelijk in de bèta-band (14-30 Hertz). Omdat de bèta-band
geassocieerd is met arousal, alertheid en aandacht, evenals met vergroting van
de P300 amplitude in het ERP (Egner & Gruzelier, 2001; Egner & Gruzelier, 2004;
Haenschel et al., 2000; Singer, 1993; Wrobel, 2000), kan er gesteld worden dat de
geobserveerde bèta-toename een verhoogde toewijzing van cognitieve bronnen
aan rookgerelateerde stimuli reflecteert, oftewel een cognitieve verwerkingsbias,
zoals die ook in ERP onderzoek is gevonden. In tegenstelling tot rokers, reageren
ex-rokers noch met een toename van craving noch met een toename van bètaactiviteit. Ook hier geldt dat de gevonden verschillen niet het gevolg kunnen zijn van
verschillen in mate van nicotine-afhankelijkheid; rokers en ex-rokers zijn/waren
even afhankelijk van roken. Daarom kan er voorzichtig geconcludeerd worden dat
rookgerelateerde stimuli voor ex-rokers minder aantrekkelijk of arousing zijn en
dat ze niet meer zo sterk de aandacht trekken van ex-rokers als van rokers. Omdat
er in dit onderzoek geen controle groep van niet-rokers is getest, is het helaas niet
mogelijk uitspraken te doen over overeenkomsten tussen ex-rokers en niet-rokers
wat betreft EEG spectrum activiteit.
In ieder geval komt uit zowel hoofdstuk 4 als hoofdstuk 5 naar voren dat de bias
in de cognitieve verwerking van rookgerelateerde stimuli na langere abstinentie
kleiner wordt. Het is echter onduidelijk of de gereduceerde bias veroorzaakt wordt
door de abstinentie zelf of dat rokers met een lage bias makkelijker (langdurig)
abstinent kunnen worden. Recente studies geven indicaties voor de laatstgenoemde
verklaring (Bartholow et al., 2007; Bartholow et al., 2010), maar meer onderzoek
naar dit onderwerp is noodzakelijk.
De aard en specificiteit van rookgerelateerde
verwerkingsbias
In Hoofdstuk 6 is er meer ingezoomd op de kenmerken van verwerkingsbias
zelf. Er is onderzocht of de verhoogde elektrofysiologische verwerking van
middelgerelateerde stimuli het gevolg is van impliciete aandacht (aandacht
wordt automatisch gevangen), van een expliciete, vrijwillige keuze om de
272
Samenvatting
stimuli aandachtiger te verwerken, of van allebei. Daarnaast is er onderzocht
of de verwerkingsbias uniek is voor middelgerelateerde stimuli, of dat
middelafhankelijke personen ook een bias (d.w.z. een grotere verwerking dan
controle proefpersonen) laten zien in aandacht voor motivationele stimuli in het
algemeen. Daarmee worden positieve en negatieve stimuli bedoeld, die een zekere
mate van arousal teweegbrengen.
Rokers en niet-rokers kregen een oddball taak voorgelegd waarin de aandacht
voor rookgerelateerde, positieve (dieren) en negatieve (vuilnis) plaatjes werd
gemanipuleerd. Beide groepen kregen in rap tempo een lange reeks plaatjes te zien
(3 per seconde). Dit waren voornamelijk neutrale plaatjes, maar af en toe kwamen er
ook rook-, dier- en vuilnisplaatjes voorbij, de zogenaamde oddballs. Proefpersonen
werden steeds geïnstrueerd om één van de drie categorieën van plaatjes te tellen
(bv. rookplaatjes), terwijl alle andere plaatjes ook voorbij kwamen. Na dat een tijdje
gedaan te hebben dienden ze een volgende categorie te tellen (bv. dierplaatjes)
en uiteindelijk telden ze plaatjes van de laatste categorie (bv. vuilnisplaatjes).
Op deze manier was elk van de drie categorieën van plaatjes een keer target
(expliciete aandacht; tellen) en twee keer non-target (impliciete aandacht;
niet tellen). ERPs werden opgenomen in reactie op alle targets en non-targets.
In overeenstemming met de resultaten van hoofdstuk 4, laten de resultaten van
huidig hoofdstuk zien dat de P300 amplitude in reactie op rookgerelateerde
cues groter is voor rokers dan voor niet-rokers. Wanneer afzonderlijk gekeken
wordt naar de twee aandachtscondities blijkt dat rokers, in vergelijking met
controle proefpersonen, zowel een vergrote P300 voor rookstimuli hebben in
de expliciete conditie, dus wanneer ze rookplaatjes moesten tellen, als in de
impliciete conditie, dus wanneer ze één van de twee andere soorten plaatjes
moesten tellen. Dit impliceert dat de aandacht van rokers automatisch getrokken
wordt door rookgerelateerde cues, zelfs wanneer de aandacht bewust gericht is
op andere cues. Het betekent ook dat wanneer er een instructie is om wel op de
rookgerelateerde stimuli te letten, rokers hiervoor meer cognitieve (aandachts-)
bronnen mobiliseren dan niet-rokers. Rookgerelateerde verwerkingsbias
is daarmee aangetoond zowel impliciet als expliciet van aard te zijn. Taken
waarin proefpersonen alleen passief plaatjes kijken, dus waarin aandacht niet
gemanipuleerd wordt (zoals in hoofdstuk 3), konden hierover nooit volledig
uitsluitsel bieden. Rokers en niet-rokers verschillen verder niet in P300 amplitude
in reactie op positieve of negatieve plaatjes. Dat betekent dat rokers geen bias
273
Samenvatting
vertonen in de verwerking van motivationele stimuli in het algemeen, maar enkel
en alleen in de verwerking van stimuli relevant voor hun nicotine-afhankelijke
toestand: de rookgerelateerde stimuli. Rokers hebben dus een specifieke en
unieke bias voor rookgerelateerd materiaal die zowel impliciet als expliciet van
aard is. Deze bevindingen zijn belangrijk in het licht van de eerder genoemde
invloedrijke theorieën die voorspellen dat aandacht automatisch gericht wordt op
middelgerelateerde stimuli (Franken, 2003; Robinson & Berridge, 1993).
Hogere-orde conditionering van
rookgerelateerde verwerkingsbias
In alle voorgaande hoofdstukken is gesproken over ‘middelgerelateerde stimuli’,
oftewel stimuli die met een middel geassocieerd zijn. Deze associatie wordt
verondersteld voort te komen uit een proces van klassieke conditionering. De
klassieke conditioneringstheorie voorspelt dat door herhaald middelgebruik, de
stimuli of contexten die daarbij vaak aanwezig zijn (geconditioneerde stimuli,
CS) geassocieerd raken met de consumptie van het middel en bijkomende
plezierige effecten (ongeconditioneerde stimulus, UCS). Hierdoor zullen de
middelgerelateerde stimuli op den duur geconditioneerde responsen opwekken,
waaronder verhoogde psychologische (craving) en fysiologische (hartslag,
huidgeleiding) veranderingen (bv. Carter & Tiffany, 1999; Drummond et al., 2000;
Drummond, 2000; Thewissen et al., 2005). In samenspel met de eerder genoemde
incentieve-sensitisatie van het dopaminesysteem, zorgt de conditionering er
tevens voor dat de middelgerelateerde stimuli een verhoogde motivationele waarde
krijgen. Dit leidt dan weer tot de verhoogde elektrofysiologische verwerking
van middelgerelateerde stimuli (de verwerkingsbias) zoals beschreven in
dit proefschrift.
Wanneer het leerproces eenmaal heeft plaatsgevonden en de CS in staat zijn
geconditioneerde responsen op te wekken, kunnen de CS weer gekoppeld worden
aan nieuwe neutrale stimuli en contexten, die op hun beurt ook associatieve waarde
krijgen en geconditioneerde middelgerelateerde responsen opwekken. Dit proces
wordt tweede-orde conditionering genoemd (of: hogere-orde conditionering; CSCS leren). Door deze vorm van conditionering kunnen er oneindige reeksen van
associaties ontstaan; het zijn waarschijnlijk dit soort soort associatieve reeksen
274
Samenvatting
die bijdragen aan middelgebonden gedragingen (het opzoeken en gebruiken
van het middel) in de echte wereld (Everitt et al., 2001; Gewirtz & Davis, 2000;
Schindler et al., 2002). Bijvoorbeeld: een pakje sigaretten raakt geassocieerd met
roken en de prettige effecten ervan. Dit pakje sigaretten zal bij een roker al snel
allerlei geconditioneerde responsen opwekken, zoals craving, bepaalde fysieke
verschijnselen (bv. versnelling van de hartslag) en verhoogde rookgerelateerde
aandacht. Op den duur kan ook de winkel waar de roker zijn pakje sigaretten koopt
de stimulus worden die de geconditioneerde responsen opwekt. En na voldoende
herhaling kan dit ook gebeuren bij de stimuli die weer geassocieerd zijn met de
winkel, zoals de wandeling er naartoe of erlangs, de verkoper, het logo of de naam
van de winkel, winkelen in het algemeen etc. Hoewel in experimentele settings
eerste-orde klassieke conditionering van rookstimuli meermaals aangetoond
is, bestaan er geen onderzoeken naar de rol van tweede-orde conditionering in
nicotine-afhankelijkheid, tenminste niet in mensen.
In Hoofdstuk 7 werden hogere-orde leerprocessen in rookverslaving onderzocht
door gebruikmaking van ERPs en zelf-rapportage. Rokers en niet-rokers kregen
herhaaldelijk twee geometrische figuren (piramide, kubus) te zien. Deze figuren
werden steeds gepaard met ofwel rook-gerelateerde plaatjes ofwel neutrale
plaatjes. ERPs werden opgenomen in reactie op de plaatjes (zoals in hoofdstuk 4 en
6), maar ook in reactie op de figuren. Na afloop werd aan de proefpersonen gevraagd
de twee figuren te beoordelen op valentie (plezierigheid), arousal en craving.
Om te beginnen bevestigen de resultaten eerdere bevindingen, namelijk dat rokers
grotere P300 amplitudes laten zien dan niet-rokers in reactie op rookgerelateerde
plaatjes vergeleken met neutrale plaatjes. Dit verschil bleek het grootst gedurende
de eerste helft van het experiment. Resultaten laten tevens zien dat, gedurende
deze helft van het experiment, rokers een sterkere elektrofysiologische reactie
hebben in reactie op het figuur dat gepaard werd met rookplaatjes dan in reactie
op het figuur dat gepaard werd met neutrale plaatjes. Dit wordt niet alleen
gereflecteerd in grotere P300 amplitudes, maar ook in grotere P200 amplitudes.
De P200 wordt niet zo vaak gerapporteerd in studies van middelenafhankelijkheid,
maar er zijn er aanwijzingen dat deze vroegere component automatische aandacht
weerspiegelt evenals de mate van plezierigheid van de stimulus (Carretie et al.,
2004; Delplanque et al., 2004; Potts, 2004). Niet-rokers laten geen P300 en P200
amplitude verschillen zien tussen het rookfiguur en het neutrale figuur gedurende
de eerste helft van het experiment. Daarom kan er geconcludeerd worden dat
275
Samenvatting
rokers, in vergelijking met niet-rokers, makkelijker en sneller associaties leren
met rookgerelateerde stimuli dan met neutrale stimuli, ook al wordt geen van de
stimuli ooit direct gekoppeld aan roken zelf. De resultaten van de zelf-rapportages
zijn in lijn met deze bevindingen. Rokers rapporteren meer craving bij het zien
van het rookgerelateerde figuur dan bij het zien van de neutrale figuur. Daarnaast
beoordelen rokers de rookgerelateerde figuur als plezieriger en meer arousing dan
niet-rokers.
Tijdens de tweede helft van het experiment verdwijnen de verschillen tussen
de groepen echter. Tijdens de tweede helft laten niet-rokers een grotere
elektrofysiologische reactie op het rookfiguur zien dan tijdens de eerste helft
van het experiment, terwijl de elektrofysiologische reactie van rokers op het
rookfiguur afneemt (en toeneemt voor het neutrale figuur). Niet-rokers laten dus
een langzame, wellicht ‘normale’, leercurve zien bij het leren van de relatie tussen
neutrale stimuli en rookgerelateerde stimuli, terwijl rokers een steile leercurve
hebben, die op een gegeven moment verzwakt en plaatsmaakt voor het leren van
de relatie tussen het neutrale figuur en de neutrale stimuli. Deze verandering over
tijd zou verklaard kunnen worden doordat rokers op een gegeven moment hun
interesse verliezen in het rookfiguur of in alle stimuli gepresenteerd in de taak;
ze voorspellen immers geen van allen roken en de bijbehorende beloning. Deze
verklaring komt overeen met het geobserveerde verwerkingspatroon van de
rookgerelateerde plaatjes zelf. De P300 in reactie op rookgerelateerde plaatjes
neemt bij rokers immers ook af over tijd. Er kan echter ook sprake zijn van verveling.
Door de vele herhalingen van plaatjes en figuren besteden proefpersonen op een
gegeven moment wellicht minder aandacht aan de taak, waardoor de ERPs van alle
groepen en condities dichter bij elkaar komen te liggen.
Samengevat laten de resultaten zien dat rokers rookgerelateerde, hogereorde associates beter, sneller en makkelijker leren dan niet-rokers, in elk geval
gedurende een korte periode van tijd. Dit blijkt zowel uit zelf-rapportage als
verhoogde elektrofysiologische verwerking van aan rookgerelateerde stimuli
gekoppelde neutrale stimuli. Resultaten onderstrepen daarnaast het eerder
beschreven idee dat verslaving niet alleen invloed heeft op aandachtsprocessen,
maar ook op leer- en geheugenprocessen en hun onderliggende neurale systemen
(Carretie et al., 2004; Robbins et al., 2008).
276
Samenvatting
Modulatie van rookgerelateerde
verwerkingbias door cognitieve strategieën
Omdat middelgerelateerde verwerkingsbias geassocieerd is met terugval na
abstinentie, is het van groot klinisch belang om te onderzoeken of er mogelijkheden
bestaan om de cognitieve verwerking van middelgerelateerde stimuli te reguleren
of verminderen. Uit emotieonderzoek is naar voren gekomen dat de latere
componenten van het ERP, met name de LPP, vergroot en verkleind kunnen worden
door het toepassen van cognitieve regulatie- of herwaarderingsstrategieën (bv.
Hajcak & Nieuwenhuis, 2006; Krompinger et al., 2008; Moser et al., 2006), zoals
het actief herinterpreteren van de emotionele inhoud van de plaatjes (bv. “het is
niet zo erg als het lijkt” of “het is niet echt”), het niet al te betrokken, maar van
een afstandje naar de plaatjes kijken, of het zoeken van afleiding in neutralere
aspecten van de plaatjes.
In Hoofdstuk 8 is er onderzocht of het ook in rookverslaving mogelijk is om
elektrofysiologische maten van rookgerelateerde cognitieve verwerking (vroege en
late LPP amplitudes) te moduleren middels cognitieve herwaarderingsstrategieën.
De effecten van drie strategieën werden onderzocht, namelijk een ‘plezierige’
strategie (denken aan hoe lekker het zou zijn om de afgebeelde sigaret te roken),
een afleidingsstrategie (denken aan de voornaamste kleur in het plaatje), en een
rationele strategie (een kort, rationeel verhaaltje bedenken bij het afgebeelde
tafereel). Vroege (600-1000 ms) en late (1000-2000 ms) LPP amplitudes in reactie op
de cognitief gereguleerde rookplaatjes werden vergeleken met de LPP amplitudes
in reactie op rookplaatjes en neutrale plaatjes waar passief naar gekeken werd
(geen regulatie) en die eerst werden aangeboden. Daarnaast werd er gekeken of
verhoogde cognitieve verwerking zoals gereflecteerd in vergrote LPP amplitudes
alsmede de cognitieve modulatie van deze amplitudes verschilt tussen reguliere,
dagelijkse rokers en lichte (gelegenheids) rokers.
Om te beginnen laten resultaten zien dat rokers grotere P300 en LPP amplitudes
hebben in reactie op rookgerelateerde stimuli dan in reactie op neutrale stimuli.
Daarmee wordt bevestigd dat ook de rokers in dit onderzoek een verhoogde
motivationele aandacht, of verwerkingsbias hebben voor rookgerelateerd
materiaal. Daarnaast laten de resultaten verschillende effecten zien van de
verschillende cognitieve strategieën. Vroege LPP amplitudes in reactie op
277
Samenvatting
rookplaatjes die gereguleerd worden met de plezierige strategie zijn significant
vergroot in vergelijking met vroege LPP amplitudes in reactie op passief
geobserveerde rookgerelateerde stimuli. Denken aan de plezierige aspecten van
roken vergroot dus relatief snel de cognitieve verwerkingsbias. Late LPP amplitudes
in reactie op rookplaatjes die gereguleerd worden met de afleidingsstrategie en de
rationele strategie worden kleiner en wel dusdanig dat er geen significant verschil
meer is met de late LPP amplitudes in reactie op de passief bekeken neutrale
plaatjes. De afleidings- en rationele strategie reduceren de cognitieve verwerking
van rookstimuli dus tot het niveau van neutrale stimuli. Hoewel dit een zeer
betekenisvol resultaat is, dient er opgemerkt te worden dat de LPP amplitudes niet
significant kleiner worden dan de LPP amplitudes in reactie op de passief bekeken
rookplaatjes. Het nut van het gebruiken van de twee strategieën is daarmee niet
volledig aangetoond.
Omdat er voor de afleidingsstrategie een trend gevonden werd naar significante
LPP reductie ten opzichte van passief bekeken rookplaatjes, is er een extra,
exploratieve analyse gedaan waarin slechts drie condities werden meegenomen
(nb. reductie van het aantal condities vergoot statistische power), namelijk
passief bekeken neutrale plaatjes, passief bekeken rookplaatjes en met de
afleidingsstrategie gereguleerde rookplaatjes. Resultaten van deze analyse laten
zien dat de afleidingsstrategie nu een dusdanige reductie van de late LPP amplitude
veroorzaakt dat deze ook significant kleiner wordt dan de late LPP amplitude in
reactie op passief bekeken rookplaatjes.
Geconcludeerd kan worden dat cognitieve herwaarderingsstrategieën
elektrofysiologische maten van middelgerelateerde cognitieve verwerking
kunnen beinvloeden. Middels de plezierige strategie kan cognitieve verwerking
van rookstimuli binnen één seconde vergroot worden; middels de afleidings- en
rationele strategie kan cognitieve verwerking van rookstimuli na één seconde
verkleind worden. Daarbij zijn de effecten van de afleidingsstrategie het meest
veelbelovend. Deze resultaten geven aan dat de toepassing van cognitieve
herwaarderingsstrategieën in de behandeling van verslaving waardevol zou
kunnen zijn.
Verder laten de resultaten zien dat reguliere, dagelijkse rokers met een gemiddelde
nicotine-afhankelijkheid (gemiddeld 14 sigaretten per dag) niet verschillen van
lichte rokers met lage tot afwezige niveaus van afhankelijkheid (gemiddeld
278
Samenvatting
5 sigaretten per dag, 3 dagen per week) wat betreft de elektrofysiologische
verwerking van rookgerelateerde stimuli. Deze bevinding is moeilijk te interpreteren
in het licht van de incentieve-sensitisatie theorie van verslaving, die voorspelt dat
afhankelijkheidsniveaus positief geassocieerd zijn met verwerkingsbias (dus hoe
afhankelijker de persoon, hoe groter de bias). Deze resultaten kunnen echter wel
verklaard worden door een verhoogde gewoonte-respons in dagelijkse rokers,
die volgens de incentieve-gewoonte theorie van verslaving samengaat met een
verkleinde respons op de belonende waarde van middelgerelateerde stimuli
(Di Chiara, 2000; Mogg et al., 2005).
Alcoholgerelateerde verwerkingsbias in
alcoholpatienten
Elektrofysiologische verwerking van middelgerelateerde cues is tot op heden
niet vaak onderzocht in alcoholafhankelijkheid en resultaten die er wel zijn, zijn
erg tegenstrijdig. Er is slechts één studie in alcohol-afhankelijke patiënten die
vergrote P300 amplitudes heeft gevonden in reactie op alcoholstimuli (Namkoong
et al., 2004). Er zijn daarentegen twee studies die géén vergroting van de P300
component hebben gevonden (Hansenne et al., 2003; Herrmann et al., 2000).
Daarom is er in Hoofdstuk 9 onderzoek gedaan naar biases in de cognitieve
verwerking van alcoholgerelateerde stimuli in alcoholafhankelijkheid. Het
voornaamste doel was om de resultaten van de oddballstudie in rokers (hoofdstuk
6) te repliceren. De taak was vrijwel identiek, alleen werden er in plaats van
rookgerelateerde stimuli, alcoholgerelateerde stimuli gebruikt en werd de
negatieve stimuluscategorie vervangen door een neutrale categorie (frisdrank).
Proefpersonen waren patiënten van verschillende poliklinieken van Bouman
GGZ Rotterdam en gezonde, gemiddeld licht-drinkende controles met een gelijke
verdeling van leeftijd, opleidingsniveau en geslacht. Na afloop van de taak werden
alle oddballplaatjes door alle proefpersonen beoordeeld op arousal en valentie.
In strijd met resultaten van onderzoek in andere soorten middelafhankelijkheid,
laten de resultaten van dit hoofdstuk zien dat de alcoholafhankelijke patiënten
geen vergrote P300 amplitudes hebben in reactie op alcoholgerelateerde stimuli.
Op frontaal-centrale elektrodeposities laten alcoholpatiënten zelfs een verkleinde
P300 amplitude zien in vergelijking met controle proefpersonen. Daarnaast zijn
279
Samenvatting
er geen verschillen tussen de groepen in de valentie en arousalbeoordelingen van
de stimuli. Over de beide groepen heen worden alcohol- en frisdrankplaatjes als
even (on)plezierig en (laag) arousing beoordeeld. Voor de controle groep werden
deze resultaten verwacht, maar voor de groep alcoholpatiënten zijn de resultaten
opvallend of merkwaardig te noemen.
Nu is het zo dat ook studies die gebruik maken van gedragstaken om cognitieve
verwerkingsbiases te meten (m.n. de visuele probe taak) tegenstrijdige resultaten
hebben gevonden in alcoholafhankelijke patiënten. Een groot deel van deze studies
laat wel een verhoogde aandacht voor alcoholcues zien bij korte aanbiedingstijden
(50 of 100 ms: Noël et al., 2006; Stormark et al., 1997; Vollstädt‐Klein et al., 2009), wat
zou duiden op een vergrote initiële oriëntatie op de cues, maar juist een verkleinde
aandacht, of omgekeerde bias, bij langere aanbieding van alcoholgerelateerde cues
(Stormark et al., 1997; Townshend & Duka, 2007; Vollstädt‐Klein et al., 2009). Dat
wil zeggen dat bij langere, bewustere aanbieding van stimuli, alcoholpatiënten de
alcoholgerelateerde stimuli vermijden. Vermijding van alcoholcues kan het directe
gevolg zijn van behandeling, waarin patiënten doorgaans expliciet geconfronteerd
worden met hun gebrekkige controle over alcoholgebruik, maar kan ook puur
voortkomen uit de motivatie om abstinent te blijven of drinkgedrag onder
controle te houden op het testmoment. Zware drinkers laten geen vermijding
van alcoholcues zijn bij langere aanbiedingstijden, waarschijnlijk omdat zij hun
gebruik niet als problematisch beschouwen en geen intenties hebben om te
stoppen of minderen met drinken. In de huidige studie werden plaatjes ook relatief
lang aangeboden (333 ms), waardoor het gebruik van vermijdingsstrategieën een
plausibele verklaring biedt voor de gevonden omgekeerde resultaten.
Toch blijft het opmerkelijk dat bevindingen van afwezige en omgekeerde cognitieve
verwerkingsbias, oftewel vermijding, alleen in alcoholpatiënten worden gevonden
en bijvoorbeeld niet in cocaïne- of heroïne afhankelijke patiënten (bv. Franken,
Stam et al., 2003; Franken et al., 2008). Zij ontvangen immers ook behandeling en
zijn waarschijnlijk even gemotiveerd abstinent te blijven. Hoewel er hier sprake
kan zijn van een publicatiebias (dat is: positieve resultaten worden over het
algemeen vaker gepubliceerd dan negatieve of onduidelijke resultaten), kunnen
de verschillen wellicht ook verklaard worden door specifieke kenmerken van de
verschillende middelafhankelijke populaties. Daarbij valt te denken aan bepaalde
demografische kenmerken, persoonlijkheidskenmerken of IQ. Een recente
studie in herstellende alcoholpatiënten laat zien dat de persoonlijkheidstrek
280
Samenvatting
‘mindfulness’ (cognitieve flexibiliteit, inhiberend vermogen, sterk bewustzijn
van eigen gedachten en handelingen) succesvolle vermijding van alcoholcues
voorspelt (Garland et al., 2011). Mogelijk bezitten alcoholpatiënten deze trek, of
gerelateerde trekken, vaker of in grotere mate dan cocaïne- of heroïne patienten.
Toekomstig onderzoek zou in moeten zoomen op dit soort kenmerken zodat
verschillen tussen verschillende stoornissen in middelgebruik verklaard kunnen
worden. Daarnaast is het belangrijk meer onderzoek te doen naar de relaties
tussen behandeling, de motivatie om abstinent te worden of blijven, (omgekeerde)
verwerkingsbiases en terugval.
Cognitieve controle in excessief gamen
Niet alleen voorspellen invloedrijke theorieën van verslaving dat
middelafhankelijke personen verhoogde motivationele aandacht hebben voor
middelgerelateerde stimuli; er wordt tevens gehypothetiseerd dat -tegelijkertijdhun vermogen om gedrag onder controle te houden verminderd is (Goldstein &
Volkow, 2002). Volgens deze visie dragen dus zowél verhoogde motivatie en
cognitieve verwerkingsbiases áls verminderde algemene cognitieve controle bij
aan het ontstaan en voortbestaan van verslaving. Onderzoek heeft bevestigd
dat middelafhankelijke personen inderdaad gebreken vertonen in diverse
aspecten van cognitieve controle, waaronder in de onderdrukking van responsen
(responsinhibitie) en de verwerking van fouten. Dezelfde bevindingen zijn gedaan
in gokverslaafde personen. Gebreken in cogitieve controle zorgen er waarschijnlijk
voor dat het lastiger is de verleidingen van middelgebruik (of gokken) te weerstaan
en dat er doorgegaan wordt met het gebruik (of gedrag) ondanks negatieve
consequenties. Met behulp van een Go/NoGo taak gecombineerd met ERP metingen
hebben we recentelijk aangetoond dat ook rokers een significant verminderde
cognitieve controle laten zien (Luijten et al., 2011). De NoGo-N2 component van
het ERP, een component dat ontstaat in reactie op het correct onderdrukken van
motorresponsen, bleek gereduceerd in rokers in vergelijking met niet-rokers.
Sinds enige tijd is er een groeiende wetenschappelijke interesse voor een mogelijk
nieuwe gedragsmatige verslaving, namelijk het overmatig veel spelen van
computerspellen, oftewel: excessieve computer gaming. Hoewel er veel indicaties
zijn dat excessief gamen schadelijke (psychologische, fysieke, sociale) gevolgen
heeft, staat onderzoek naar deze gedraging nog in de kinderschoenen en zijn
281
Samenvatting
onderliggende neurobiologische mechanismen niet bekend. Identificatie van de
onderliggende mechanismen is belangrijk om het gedrag goed in kaart te kunnen
brengen en beter te kunnen begrijpen, maar ook om risicogroepen te kunnen
onderscheiden en om goede interventies en behandelingen te kunnen ontwikkelen.
Onlangs is aangetoond dat excessieve gamers, net als middelafhankelijke personen
(zie hoofdstuk 1, 4-8), biases laten zien in de cognitieve verwerking van gamegerelateerde cues (Thalemann et al., 2007). Ten opzichte van met mensen die
slechts af en toe gamen, lieten excessieve gamers een vergrote P300 amplitude
zien in reactie op game-gerelateerde cues vergeleken met neutrale cues. Deze
verhoogde cognitieve verwerking bleek uniek voor game-gerelateerde cues; er
waren geen verschillen tussen de groepen in reactie op positieve, negatieve en
alcoholgerelateerde cues. Dit onderzoek is één van de eerste onderzoeken die laat
zien dat er mogelijk overeenkomsten zijn tussen de onderliggende neurobiologische
en/of cognitieve mechanismen van excessief gamen en middelafhankelijkheid.
In Hoofdstuk 10 werden de mechanismen van cognitieve controle in excessief
gamen onder de loep genomen. Zowel responsinhibitie als foutverwerking
werden bestudeerd in excessieve gamers en weinig- tot niet-spelende controle
proefpersonen. Dit werd gedaan met behulp van een Go/NoGo taak in combinatie
met ERP metingen. In de Go/NoGo taak krijgen proefpersonen een reeks letters te
zien, die zeer kort en zeer snel achter elkaar gepresenteerd worden. De instructie
is om op een knop te drukken bij elke letter, behalve wanneer de letter hetzelfde
is als de vorige letter. Deze taak doet overduidelijk een beroep op het vermogen
responsen te onderdrukken. En omdat dit lastig is worden er doorgaans veel
fouten gemaakt. Ten behoeve van het meten van responsinhibitie werden ERPs
opgenomen in reactie op correcte NoGo trials (letters waarbij niet gedrukt mocht
worden en ook niet gedrukt werd) en Go trials (letters waarbij gedrukt moest
worden en waarbij ook gedrukt werd). Bestudeerde componenten in deze ERPs
waren de Error-Related Negativity (foutgerelateerde negativiteit; ERN) en de
Positivity associated with errors (positiviteit geassocieerd met fouten; Pe). Deze
componenten zijn doorgaans vergroot bij het maken van fouten en weerspiegelen
respectievelijk de registratie van een fout en de diepere verwerking van een fout
(Bernstein et al., 1995; Falkenstein et al., 1991; Nieuwenhuis et al., 2001; Overbeek
et al., 2005). Ten behoeve van het meten van foutverwerking werden ERPs
opgenomen in reactie op fouten (onterechte knopdrukken) en correcte responsen
(terechte knopdrukken). Bestudeerde componenten waren de NoGo-N2 en de
NoGo-P3. Deze componenten zijn doorgaans vergroot bij correcte onderdrukking
282
Samenvatting
van een respons en reflecteren daarom respectievelijk correcte responsinhibitie en
de diepere verwerking ervan (Falkenstein et al., 1999). Daarnaast werd het aantal
fouten op de taak geregistreerd (gedragsmaat) en werd de persoonlijkheidstrek
impulsiviteit gemeten aan de hand van een vragenlijst.
De resultaten laten zien dat excessieve gamers significant gereduceerde
ERN amplitudes laten zien in reactie op fouten in vergelijking met controle
proefpersonen. Dit geeft aan dat excessieve gamers fouten minder goed detecteren
en registreren, wat over het algemeen als indicatie wordt gezien voor verminderde
gevoeligheid voor negatieve consequenties. Daarnaast maken de excessieve gamers
meer fouten op de Go/NoGo taak dan controles, wat als indicatie gezien wordt voor
impulsief gedrag of slechte responsonderdrukking. Tenslotte hebben excessieve
gamers hogere niveaus van zelf-gerapporteerde impulsiviteit. Deze resultaten
tezamen doen suggereren dat excessief gamen zekere parallellen vertoont met
middelafhankelijkheid en pathologisch gokken wat betreft cognitieve controle
zoals gemeten op zelf-rapportage-, gedrags-, én elektrofysiologisch niveau.
Er worden echter geen verschillen gevonden tussen excessieve gamers en controle
proefpersonen op de fout-gerelateerde Pe, wat lastig te verklaren valt. Daarnaast
worden er geen verschillen gevonden op de elektrofysiologische maten van
responsinhibitie (NoGo-N2 en NoGo-P3). De afwezigheid van deze laatstgenoemde
verschillen zou verklaard kunnen worden door bepaalde kenmerken van de taak.
Er is namelijk, over beide groepen gezien, geen vergroting van NoGo-N2 amplitudes
in reactie op NoGo trials vergeleken met Go trials. Dat betekent eigenlijk dat
niemand goed in staat is responsen te onderdrukken. Mogelijk is de taak te
moeilijk; er wordt immers een veel lager percentage NoGo trials gehanteerd in het
huidige onderzoek dan in andere Go/NoGo onderzoeken en er worden door alle
proefpersonen meer fouten gemaakt dan in andere onderzoeken (bv. Falkenstein
et al., 1999; Luijten et al., 2011). Wanneer een taak lastig is zien we doorgaans
compensatie door andere hersengebieden om de taak alsnog tot een goed einde te
brengen (Bolla et al., 2004; Bunge et al., 2002; Yücel & Lubman, 2007; Yücel et al.,
2007). In reactie op de NoGo trials in de huidige studie vinden we een dergelijke
compensatie in de vorm van toegenomen pariëtale hersenactiviteit. Deze is echter
sterker aanwezig in excessieve gamers dan in controles, wat mogelijk aangeeft dat
de compensatie in excessieve gamers minder adequaat is en wat mogelijk alsnog
reden geeft om aan te nemen dat excessieve gamers ook op elektrofysiologisch
niveau een verminderde responsinhibitie laten zien.
283
Samenvatting
Ondanks dat de laatstgenoemde bevindingen vrij speculatief van aard zijn, zijn
de verschillen tussen excessieve gamers en controles op een groot aantal andere
maten van cognitieve controle, namelijk de ERN, de gedragsmaat en de vragenlijst,
wel erg sterk. Het is mogelijk dat de geobserveerde hoge impulsiviteit, beperkte
foutverwerking en verminderde gedragsmatige responsinhibitie ten grondslag
liggen aan excessief game-gedrag. Het kan een reden zijn voor het feit dat sommige
gamers de verleiding om te gamen maar moeilijk kunnen weerstaan, moeilijk mate
kunnen houden en doorgaan met gamen ondanks negatieve consequenties. We
kunnen echter niet uitsluiten dat de verminderde cognitieve controle het gevolg
(i.p.v. de oorzaak) is van excessief gamen. Daarvoor zijn longitudinale studies nodig
waarin gamers gevolgd worden over tijd. In elk geval suggereren de resultaten dat
excessief gamen gepaard gaat met een gebrekkige cognitieve controle, waarmee
deze mogelijk nieuwe gedragsverslaving, in elk geval gedeeltelijk, overeenkomsten
vertoont met middelafhankelijkheid dan wel pathologisch gokken.
Conclusies en suggesties voor toekomstig
onderzoek
In Hoofdstuk 11 worden de bevindingen van dit proefschrift uitgebreid
samengevat en bediscussiëerd. De hoofdconclusie van dit proefschrift is dat
middelenafhankelijkheid gekenmerkt wordt door verhoogde elektrofysiologische
verwerking van middelgerelateerde stimuli, oftewel grotere P300 en LPP
amplitudes in reactie op middelgerelateerde cues dan in reactie op neutrale cues.
Omdat deze componenten van het ERP geassocieerd zijn met cognitieve processen
(aandacht) en motivatie (craving, valentie, arousal) kan er tevens gesteld worden
dat er middels een zeer directe maat bevestigd wordt dat middelenafhankelijkheid
gekenmerkt wordt door biases in de cognitieve verwerking van middelgerelateerde
cues of verhoogde motivationele aandacht voor middelgerelateerde cues. Daarnaast
kan er uit dit proefschrift geconcludeerd worden dat verhoogde verwerking van
middelgerelateerde cues:
a) zowel op frontale als op pariëtale elektrodeposities gemeten kan worden
b) niet gemodereerd wordt door type middel (stimulerende vs. kalmerende
middelen), abstinentie (niet abstinent vs. abstinent voor 10-30 dagen),
geslacht, leeftijd, of het soort taak dat gebruikt wordt (passieve vs. actieve
taken)
284
Samenvatting
c) lijkt te verdwijnen na langere perioden van abstinentie (in rokers).
Ditzelfde geldt voor zelf-gerapporteerde craving en ervaren plezierigheid
van middelgerelateerde stimuli
d) zowel impliciet als expliciet van aard is (in rokers)
e) specifiek is voor middelgerelateerde motivationele stimuli (in rokers)
f) geconditioneerd kan worden en kan verschuiven naar in eerste instantie
neutrale stimuli (in rokers). Ditzelfde geldt voor zelf-gerapporteerde
craving en ervaren plezierigheid van middelgerelateerde stimuli.
g) gemoduleerd kan worden door het toepassen van cognitieve strategieën
(in rokers)
h) niet of zelfs omgekeerd aanwezig is in alcoholpatiënten
Daarnaast kan er geconcludeerd worden dat de Nederlandse vertaling van de
QSU-brief goede psychometrische kenmerken bezit en dat de potentieel nieuwe
psychologische stoornis ‘excessief gamen’ overeenkomsten vertoont met
middelafhankelijkheid wat betreft biases in de cognitieve verwerking van gamegerelateerde cues (eerder onderzoek) en gebrekkige algemene cognitieve controle
(dit proefschrift).
Zoals eerder beschreven zijn biases in de cognitieve verwerking van
middelgerelateerde stimuli geassocieerd met middelconsumptie en terugval. Dit
maakt onderzoek naar verwerkingsbiases belangrijk en klinisch relevant. Echter,
deze associaties zijn enkel en alleen gemeten met gedragsmaten van cognitieve
verwerking (aandachtstaken). Het is tot op heden nooit direct aangetoond dat
de verhoogde elektrofysiologische verwerking van middelgerelateerde cues in
middelenafhankelijkheid, zoals beschreven in dit proefschrift, geassocieerd is
met middelconsumptie en terugval. Hier wordt impliciet van uitgegaan, maar
dat kan eigenlijk alleen als gedrags- en elektrofysiologische maten met zekerheid
(ongeveer) dezelfde (aandachts)processen meten. En ook hier ligt een probleem: er
bestaan nauwelijks studies in middelenafhankelijkheid die in dezelfde steekproef
zowel ERPs registreren als de prestatie op aandachtstaken meten. De studies die er
wel zijn vinden significante ERP verschillen tussen groepen, maar geen verschillen
op de gedragsmaten (Fehr et al., 2006; Fehr et al., 2007). Dit zou kunnen betekenen
dat beide maten iets anders meten. Het zou echter ook kunnen betekenen dat
ERPs juist veel sensitiever zijn en/of bredere aspecten van cognitieve verwerking
meten. Resultaten van een recente meta-analyse van Field en collega’s (2009)
geven een indicatie voor het laatste. De studie toont aan dat verwerkingsbiases
285
Samenvatting
zoals gemeten met ERPs sterker correleren met craving dan verwerkingbiases
zoals gemeten met gedragstaken. Daarmee lijken ERPs inderdaad een betere
maat te zijn voor motivationele aandacht voor middelgerelateerde stimuli. Het is
belangrijk dat toekomstige studies onderzoek doen naar de precieze associatie
tussen ERPs en conventionele metingen van middelgerelateerde verwerkingsbias.
Dit kan door ERPs op te nemen in reactie op stimuli binnen een gedragsstaak of
door de resultaten van twee experimenten in dezelfde proefpersonen met elkaar
te correleren.
Daarnaast is het belangrijk dat er wordt onderzocht of de verhoogde
elektrofysiologische verwerking van middelgerelateerde stimuli zelf ook
geassocieerd is met middelconsumptie en terugval. In hoofdstuk 4 van dit
proefschrift is aangetoond dat ex-rokers met een gemiddelde stopduur van 1,4
jaar significant verkleinde middelgerelateerde P300 en LPP amplitudes laten zien
ten opzichte van rokers. Er worden daarnaast geen verschillen meer gevonden
tussen ex-rokers en niet-rokers. Zoals eerder opgemerkt is het onduidelijk of deze
verminderde verwerkingsbias de oorzaak of het gevolg is van het stoppen met
roken. Wordt de elektrofysiologische reactiviteit weer ‘normaal’ na het stoppen?
Of is verminderde elektrofysiologische reactiviteit een voorspeller van succesvol
stoppen (en is verhoogde elektrofysiologische reactiviteit een voorspeller van
roken)? Recentelijk zijn er twee ERP studies uitgevoerd die een indicatie geven
voor het laatste (Bartholow et al., 2007; Bartholow et al., 2010). Uit deze studies
kwam naar voren dat personen met een lage gevoeligheid voor alcohol (risicofactor
voor het ontwikkelen van alcoholafhankelijkheid) grotere P300 amplitudes lieten
zien in reactie op alcoholgerelateerde stimuli dan in reactie op neutrale stimuli.
Dit was niet het geval in proefpersonen met een hoge alcoholgevoeligheid. Dit
patroon bleef significant wanneer er gecontroleerd werd voor verschillen in
huidig alcoholgebruik, huidige alcoholafhankelijkheid en een familiegeschiedenis
van alcoholafhankelijkheid. Daarnaast bleek de grootte van de P300 in reactie
op alcoholstimuli later drinkgedrag te voorspellen (Bartholow et al., 2007).
Deze resultaten geven aan dat verhoogde elektrofysiologische verwerking
van middelgerelateerde cues mogelijk voorafgaat aan het ontwikkelen van
middelgerelateerde stoornissen en toekomstig middelengebruik kan voorspellen.
Verhoogde elektrofysiologische verwerking van middelgerelateerde cues zou
daarom wellicht een marker of endofenotype kunnen zijn voor middelgebruik en –
afhankelijkheid. Het is zeer belangrijk dat deze mogelijkheid in de toekomst verder
onderzocht wordt, bijvoorbeeld door te toetsen of de P300/LPP reactiviteit van
286
Samenvatting
abstinente middelafhankelijke personen op langere termijn terugval voorspelt
en of de P300/LPP reactiviteit van gelegenheidsgebruikers op langere termijn
afhankelijkheid voorspelt. Daartoe dienen longitudinale onderzoeksdesigns
gebruikt te worden.
Hoewel het gros van de bevindingen in dit proefschrift in overeenstemming is
met het verslavingsmodel van aandachtsbias en de incentieve-sensitisatie theorie
van verslaving, zijn de resultaten van hoofdstuk 4, 8 en 9 problematisch voor
sommige aspecten van deze theorieen. Naast de aanwezigheid van cognitieve
verwerkingsbiases in middelenafhankelijkheid, voorspellen de theorieen
namelijk dat de verhoogde middelgerelateerde verwerking: a) een permanent
kenmerk van verslaving is; en b) direct proportioneel is aan de kwantiteit en
frequentie van het middelengebruik. Echter, resultaten van dit proefschrift laten
zien dat middelgerelateerde verwerking vermindert na langdurige abstinentie
(onafhankelijk van niveaus van afhankelijkheid), dat middelgerelateerde
verwerkingsbias even groot is in laag- en gemiddeld afhankelijke rokers, en
dat middelgerelateerde verwerkingsbias afwezig of zelfs omgekeerd is in
alcoholafhankelijke patiënten die in behandeling zijn (terwijl eerder onderzoek
laat zien dat de verwerkingsbias wel aanwezig is in zware drinkers vergeleken met
lichte drinkers). Daarom zou het goed kunnen dat cognitieve verwerkingbias in
middelafhankelijkheid geen lineair patroon volgt (hoe groter de afhankelijkheid,
hoe groter de bias), maar een omgekeerde U-curve, waarbij niet-afhankelijke
personen geen bias laten zien en licht-afhankelijke personen een gemiddelde bias,
tot er in gemiddelde tot zware afhankelijkheid een optimum bereikt wordt waarna
de bias weer afneemt tot (of verder dan) het niveau van niet-afhankelijke personen.
De aanname dat verwerkingsbias een omgekeerde U-curve volgt zou in
overeenstemming zijn met de incentieve-gewoonte theorie van verslaving
(Di Chiara, 2000). Deze theorie stelt dat wanneer verslaving langer duurt,
middelgerelateerde gedragingen automatischer worden (dat is: ze worden een
gewoonte). Daardoor wordt de rol van de incentief-motivationele processen in
de verslaving (richten van aandacht vanwege motivationele, belonende waarde
van stimuli) minder belangrijk. Met andere woorden, na langere perioden van
afhankelijkheid nemen incentieve responsen op middelgerelateerde cues af, terwijl
tegelijkertijd gewoonte-reponsen toenemen (Mogg et al., 2005). Daarnaast kan
het zo zijn dat zwaardere gebruikers, in vergelijking tot lichtere gebruikers, hun
middelengebruik als problematisch beschouwen en daarom mogelijk hun craving
287
Samenvatting
proberen te onderdrukken of bewust hun aandacht proberen te onttrekken van
middelgerelateerde stimuli. Dit zou in het bijzonder het geval kunnen zijn in
middelafhankelijke personen die in behandeling zijn; zij hebben waarschijnlijk
de grootste motivatie om abstinent te worden of te blijven (of hun gebruik te
beperken dan wel te controleren) en hebben mogelijk tijdens de behandeling
geleerd meer controle te hebben over hun gedrag. Zoals we zagen in hoofdstuk
8 kan het actief toepassen van cognitieve regulatiestrategieën inderdaad de
(elektrofysiologische) verwerking van middelgerelateerde stimuli verminderen
alsmede zelfgerapporteerde craving (Kober et al., 2009; Kober et al., 2010).
De precieze vorm van deze omgekeerde U-curve (oftewel waar precies
het optimum ligt) verschilt waarschijnlijk tussen verschillende soorten
middelenafhankelijkheid. Waar is aangetoond dat zware drinkers over het
algemeen een grotere verwerkingsbias hebben dan lichte drinkers (Field et al.,
2004; Townshend & Duka, 2001), is er aangetoond dat dagelijkse en lichte rokers
juist gelijke niveaus van verwerkingsbias laten zien (zie hoofdstuk 8, maar ook
Mogg et al., 2005; Sayette et al., 2001) of dat dagelijkse rokers zelfs een kleinere
bias hebben dan lichte rokers (B. P. Bradley et al., 2003; Hogarth et al., 2003; Mogg
et al., 2005; Waters et al., 2003). Daarnaast is omgekeerde verwerkingsbias na
langere afhankelijkheid tot nu toe alleen aangetoond in alcoholafhankelijkheid.
Verschillen in de cognitieve verwerking van middelgerelateerde stimuli tussen
verschillende soorten middelenafhankelijkheid zouden verklaard kunnen worden
door specifieke kenmerken middelafhankelijke populaties (bv. demografische
kenmerken, persoonlijkheid, IQ), het soort behandeling dat men ondergaat
(bv. behandelingen gericht op cognitieve controle), of de directe effecten, de
verslavende werking en de verkrijgbaarheid van de middelen (en daardoor de
mogelijkheid tot of de mate van gewoonte-leren).
Meerdere studies tonen aan dat verwerkingsbiases geassocieerd zijn met terugval.
Echter, als verwerkingsbias afneemt na langdurige afhankelijkheid en langdurige
abstinentie, wat is dan de precieze rol van verwerkingsbias in terugval? Wat is de
precieze aard van de associatie? Het blijft mogelijk dat, ondanks de geobserveerde
afname of omkering van verwerkingbias in experimentele settings, de biases meer
fluctueren in het echte leven waardoor ze nog steeds voorafgaan aan terugval
in zwaardere gebruikers of patiënten die proberen abstinent te blijven. Het is
zeer belangrijk dat er meer onderzoek gedaan wordt naar de associatie tussen
verwerkingsbias en terugval evenals de precieze aard van deze associatie. Ook dient
288
Samenvatting
er meer onderzoek gedaan te worden naar de rol van specifieke patiëntkenmerken
hierin, alsmede afhankelijkheidsniveaus, behandeling en motivatie om abstinent
te worden of te blijven.
Tot slot lijkt het, naar aanleiding van hoofdstuk 8 van dit proefschrift, belangrijk om
meer onderzoek te doen naar de effecten van cognitieve herwaarderingsstrategieën
op middelgerelateerde cognitieve verwerkingsbiases. Daarbij valt te denken aan
replicatie-onderzoek in andere stoornissen in middelengebruik (anders dan roken)
en onderzoek naar effecten op langere termijn. Omdat de afleidingsstrategie
(aandacht richten op niet-middelgerelateerde aspecten van stimuli) het meest
veelbelovend bleek, zou toekomstig onderzoek met name de effecten van deze
strategie gedetailleerder moeten onderzoeken. Eerder onderzoek heeft laten zien
dat ook craving gemoduleerd kan worden door cognitieve regulatie (Kober et al.,
2009; Kober et al., 2010). Hiervoor werden echter andere strategieën gebruikt dan
in hoofdstuk 8, namelijk het denken aan de (positieve) effecten van het middel
op de korte termijn en het denken aan de (negatieve) gevolgen van het middel op
de lange termijn. Het zou interessant zijn om te zien of deze strategieën ook een
modulerend effect hebben op middelgerelateerde cognitieve verwerkingsbias.
Andersom zou het interessant zijn om te zien of de strategieën gebruikt in dit
proefschrift ook invloed hebben op craving. Voorts zou het interessant zijn om de
effecten van de meest effectieve cognitieve regulatiestrategieën te toetsen in de
klinische praktijk.
289
Dankwoord
Dankwoord
Het is af. Het is gedaan. Mijn proefschrift is klaar. Het fantastische resultaat van vier
jaar hard werken. Maar ik heb het niet alleen gedaan. Door de jaren heen hebben
zoveel mensen mij geholpen, gesteund, ondersteund, geïnspireerd, opgepept, afgeleid,
aangemoedigd, onderwezen, uitgedaagd, op ideeën gebracht en feedback gegeven.
Hierdoor werden de onderzoeken beter, de stukken werden beter en ikzelf werd beter.
Als onderzoeker én persoon. Graag wil ik alle mensen bedanken die op wat voor manier
dan ook hebben bijdragen.
Professor Franken, Ingmar, bedankt dat je mij de mogelijkheid hebt gegeven om dit
promotietraject te doorlopen. Het was me een groot genoegen jouw AiO te mogen
zijn. Bedankt voor alle brainstorm sessies, je gedegen feedback, je altijd snelle
reacties, je geduld, en de vrijheid die je me gaf. Bedankt ook voor je waardevolle
peptalks. Je had altijd vertrouwen in mij, zelfs wanneer ik dat zelf even niet had. In
zoveel verschillende opzichten heb ik zoveel van je geleerd. Ik had me geen betere
promotor kunnen wensen.
Leden van de promotiecommissie, Wim van de Brink, Jan van Strien, Rolf Zwaan,
Matt Field, Ernst Koster, Ben van de Wetering, en Elke Geraerts, bedankt dat
jullie kennis hebben genomen van mijn proefschrift, deze kritisch hebben gelezen
en hebben beoordeeld. Tevens bedankt voor jullie bereidheid om met mij van
gedachten te wisselen tijdens de verdediging van mijn proefschrift.
Verder wil ik graag mijn (ex-)collega’s van het Instituut voor Psychologie bedanken.
Om te beginnen iedereen van de klinische sectie: An, Angela, Anita, Anja, Arjan,
Birgit, Colin, Danielle, Elke, Eric, Freddy, Guus, Hans, Ilse, Ivo, Jorg, Katrien, Leonie,
Marlies, Peter, Renske, Reshmi, Sabine, Susan, Suzanne en Tim. Allemaal ontzettend
bedankt voor de prettige samenwerking, de praatjes, de goede ideeën, de feedback
op mijn onderzoek en natuurlijk de gezelligheid. Daarnaast ben ik een aantal andere
collega’s mijn dank verschuldigd. B&C’ers (en EAN-labbers) Inge, Jan, Karen, Kiki,
Lisa, Liselotte, Rolf en Sandra, bedankt voor jullie input, betrokkenheid en voor
alles wat jullie me hebben geleerd over EEG, ERPs en gedegen onderzoek doen in
het algemeen. Peter, Samantha en Marike, bedankt voor jullie hulp met statistische
uitdagingen. Gerrit-Jan, Christiaan, Marcel, Jeffrey en Freek, bedankt voor jullie
technische ondersteuning en alle (VELE) hulp met E-Prime en EEG. Hanny, Mirella
291
Dankwoord
en Angelique, de medewerkers van het onderwijsbureau en Psyweb, bedankt voor
alle praktische ondersteuning.
Een aantal collega’s wil ik in het bijzonder bedanken. Benjamin, dank voor alle
momentjes van ‘bezinning’. Jammer dat onze weddenschap maar 1 maand betrof ;).
Lisa, dank voor alle lol en funky reggaeton moves. Kiki, dank voor de fijne tijd en
het altijd vertrouwde gevoel; ruim 10 jaar (!) samen op de EUR. Angela, bedankt
voor je vriendschap, je adviezen en je steun op zo ongeveer alle vlakken van mijn
leven. Onze logeerpartijtjes waren altijd gezellig en houden we er natuuuurlijk in
(zie je eigen dankwoord ;)). Reshmi en Anja, lieve reisgenootjes, bedankt voor alle
ervaringen die ik met jullie heb mogen delen. Eerst Arizona en de aansluitende
roadtrip naar de Grand Canyon, Vegas en Hollywood; later Boston en de aansluitende
trips naar New York, de Niagara Falls en IJsland. Wat was het fantastisch!! Het
was fijn om julie (zooo goed) te leren kennen. Een grote dosis pret, veel goede
gesprekken, grappige blogs, maar ook flink wat (heerlijk) gekibbel nu en dan; laten
we vooral nog een keer met z’n drieën op vakantie gaan. Hertenkontfotografe Anja
en I’m-still-working-on-it-Reshmi, thanks for everything!
Lieve kamergenootjes, my roomies, Maartje, Mario en Anita, bedankt voor alles.
De goede en inspirerende gesprekken, de prietpraat, de mentale steun, de film- en
serietips, de slappe lach, de roddels, de lekkernijen, de slingers op m’n verjaardag,
de fruitmand toen ik ziek was, en het algehele delen in de AiO stress. Ik ga jullie
missen. Sterker nog, dat doe ik al. Maartje, het was fijn om met je te kunnen
sparren over ons onderzoek. Bedankt voor al je goede en buitengewoon slimme
feedback op m’n stukken en de oneindige andere hulp, zowel werkgerelateerd
als privé (ik geniet nog elke dag van die mooi geschilderde kozijnen). Mario, als
enige man op de kamer zorgde jij voor de balans. Het mannelijk perspectief op de
zaken was dikwijls een verademing. Grappig was dat als wij, meiden, druk aan het
kletsen waren, en het leek alsof jij keihard aan het werk was, je ineens, -BAM- , uit
het niets een rake opmerking kon maken over hetgeen wij aan bespreken waren
(zat je toch stiekem te luisteren ;)). Mario, bedankt voor al je input, de hulp met
E-prime, de filmrecensies, de leuke gesprekken en discussies. Anita, wij begonnen
precies tegelijk aan ons AiO-avontuur. Dat heeft een band geschept die nooit meer
verdwenen is. Bedankt voor al je steun, je luisterend oor, de fijne gesprekken, je
vriendschap en dat ik altijd kon en mocht zeggen wat er op mijn hart lag, al was
het 9 uur ’s morgens en al was ik er al 1000 keer eerder over begonnen. Het maakte
niet uit; je was er. Ik ben dankbaar dat jij tijdens mijn promotie achter me staat.
292
Dankwoord
Verder gaat mijn dank uit naar alle hulpverleners van de Bouman GGZ die zich
hebben ingezet voor het werven van de alcoholpatiënten voor mijn laatste
onderzoek, naar alle auteurs die bereid waren ruwe data op te sturen ten behoeve
van de meta-analyse, en naar alle co-auteurs van de diverse hoofdstukken van
dit proefschrift. Fijn dat jullie wilden meedenken, meeschrijven en me hebben
geholpen met de analyses. Marcus Munafò, thank you for all your help with our
meta-analysis. Matt Field, thank you for your help with the alcohol paper and
the interesting and inspiring conversations during the EPP meetings. Daarnaast
wil ik de studenten noemen die bij mij hun onderzoeksstage hebben gelopen.
Marina, Ana, Lianne, Felicity, Marit en Nargis, bedankt voor het werven van al die
proefpersonen en het verzamelen van al die data. Zonder jullie hulp had ik nooit
zo snel zoveel hoofdstukken kunnen schrijven. Ik wens jullie veel succes in jullie
verdere loopbaan!
Natuurlijk gaat mijn dank ook uit naar al mijn lieve vrienden uit DE vriendengroep.
Kasia (a.k.a. Kaas), Menno, Tunja, Yona, Tony, Jan, Jennie, Michiel (a.k.a. Wallie a.k.a.
Mic the Sing), Michiel (a.k.a. Cref a.k.a. CrefRymz), Willem (a.k.a. Wimmel-London),
Arnoud (a.k.a. The Ring), Rients, Roel, Daantje, Sven, Marleen, Tanne, Jessica,
Joram, Eefje, Luuk, Lennart, Fatima, Stephan, Ronald en de rest. Zonder af en toe
over jullie te mogen regeren als Grote Dalmutti had ik het nooit gered.
Ook erg inspirerend waren al die heerlijke avondjes Boudewijn, Pub, Cinema, Vibez,
Blender en 80’s-90’s, de tevergeefse, maar enerverende pogingen om Remco’s
singstar score te verbreken, de altijd spannende pokertoernooitjes, de fenomenale,
alles kan-mag-en zal gebeuren IJzerstraat feestjes en later de wat bescheidenere,
maar heus niet minder gezellige partijtjes op de Binnenweg, de festivals, de
vakanties, de road- en stedentrips, de bungalowweekendjes, de bierproeverij
(zonder c1000 bier), het paramidespel, de maddogs (yugh), de salsa, de sushi, de
high-teas, het padvinderen en als klap op de vuurpijl: het onlangs zelfbedachte
RisCassonneHydroTandenborstelReligiotuur-spel (daar had je bij moeten zijn.. :-S).
Hoewel ik er bij nader inzien enigszins aan twijfel of dit allemaal direct of indirect
bijgedragen heeft aan de succesvolle afronding van dit proefschrift, toch bedankt.
Een paar vrienden wil ik in het bijzonder bedanken. Om te beginnen Arnoud, mijn
grote neef en huisgenoot voor weet ik hoeveel jaar. Wat was het toch gezellig samen
op de IJzerstraat. En daar hoefde ik je niet eens echt voor te zien, nee, ik hoefde
enkel en alleen maar langs je kamerdeur te lopen. Daar hoorde ik namelijk altijd,
293
Dankwoord
ALTIJD (!), gelach achter vandaan komen. Dat deed mij ook altijd even
grinniken. Bedankt voor al je computerhulp, de goede gesprekken en
discussies, je buitengewone interesse voor mijn onderzoek en al je interessante
wetenswaardigheden over werkelijk ALLES!
Tony, oud-studiegenoot, danspartner voor vele jaren en boven alles: inspirator.
Niet alleen omdat jij jouw proefschrift INEENS (!?) afhad, maar ook vanwege je
nuchterheid, je legendarische sarcasme en je buitengewone intelligentie. En je
rob-dans ;). Bedankt voor alle goede gesprekken, wetenschappelijke sparsessies,
peptalks (“heb je het nou nog niet af?”) en de jarenlange gezonde portie wekelijks
bewegen. Ik ben dankbaar je als vriend te mogen hebben.
Kasia en Tunja. Zo lang ik me kan heugen zijn jullie er voor me geweest.
Onvoorwaardelijk. Tunja, altijd sterk, wijs, kritisch, liefdevol, begripvol en soms
zo direct en eerlijk dat het me weleens kwaad heeft gemaakt. Maar kwaad omdat
je gelijk had. Omdat je me liet zien wat ik soms niet wilde zien. Door jou heb ik
kunnen groeien. Ik voel me meer dan vereerd dat jij, zo’n sterke vrouw, tijdens
mijn promotie achter me wilt staan. Kasia, altijd liefdevol, begripvol, warm, open,
een luisterend oor, mijn steun en toeverlaat. Jij was en bent er altijd. Zelfs midden
in de nacht kan ik bij je aankloppen. Je deur en armen staan altijd open; het bier
staat altijd koud. Meiden, musketiers, jullie zijn goud waard. Bedankt voor jullie
dierbare vriendschap.
Jennie, ik ben blij dat ik je de afgelopen jaren steeds beter heb leren kennen. Ik
bewonder je nuchterheid en optimisme. Bedankt voor onze verse vriendschap. Jan,
friends forever, need I say more? Op jou kan ik altijd terugvallen. Rients, wijsheid en
whisky. Drinken we er snel weer een? Michiel, waar jij bent wordt gezongen! Dank
voor deze ouderwetse gezelligheid en voor je grenzeloze fantasie (Riscassonne?
LL-dalmutti-contest? PP-eurovisie songfestival?). Willem, ow ow ow, bedankt. Je
speeches zijn de bom (en ik heb er wel weer een verdiend dacht ik zo ;)).
Nelly, we zien elkaar niet zo vaak meer, maar ik grijp deze kans om ook jou
eindelijk eens echt goed te bedanken. Jij hebt me geleerd (zo goedkoop mogelijk ;))
op mezelf te wonen, m’n eigen boontjes te doppen, (ook letterlijk) gezond te koken,
altijd de fiets te pakken en je hebt me wegwijs gemaakt in Rotterdam. Je hebt me
gestimuleerd tijdens de studie psychologie, -wat waren we samen goed!-, maar ook
anderszins intellectueel uitgedaagd. Je hebt me opgevangen toen dat nodig was,
294
Dankwoord
me uit de put getrokken en met me samengewoond op 5 m2. Nelly, zonder jou was
ik misschien wel nooit zover gekomen. Ontzettend bedankt voor alles wat je ooit
voor me hebt gedaan.
Tot slot wil ik graag mijn familie bedanken. Opa en oma van Dommele en Oma
Littel, bedankt voor jullie trots en betrokkenheid. Ben en Jannie, ik ben blij dat
ik jullie heb mogen leren kennen en dat ik deel van jullie grote familie mag zijn.
Bedankt ook voor jullie mooie, lieve zoon, die zoveel voor mij betekent. Klein zusje
Annelotte, bedankt dat je op deze feestelijke dag voor me wilt optreden. Je bent een
knapperd, een kanjer en een wereldzangeres. Je gaat het maken!!! Lisette, Paul, en
kleine Jayden. Bedankt voor jullie eeuwige gastvrijheid, openheid en warmte. Ik
wens jullie alle sterkte, geluk en liefde van de wereld! Sis Lis, mooie, lieve meid,
geef nooit op. Jayden, mijn eerste neefje, mijn oogappeltje voor altijd.
Papa en mama, Willem en Elly, bedankt voor jullie liefde en stimulatie door het
leven heen, de vrijheid die jullie me hebben gegeven, de zelfstandigheid die jullie
me hebben bijgebracht, en jullie oneindige vertrouwen in mij. Zonder jullie had ik
dit alles nooit kunnen bereiken.
Daarom, papa en mama, is dit boekje is voor jullie.
Remco, liefde van mijn leven. Jij kent me door en door, al de gezichten die ik heb,
elke kronkel in m’n kop, al het positieve, al het negatieve, mijn passies, mijn
standpunten, mijn gedachtegoed, het vuur in m’n ogen, de tranen over m’n wangen,
de pieken, de dalen, het hele landschap van mijn gevoel. Maar jij accepteert me voor
wie ik ben en draagt me over elke kuil. Zonder jou waren de afgelopen vier jaar
een stuk hobbeliger geweest . Bedankt voor je sterke schouders, je vertrouwen,
je geduld, je humor en je enorme trots. Het was niet altijd makkelijk en dat zal het
in de toekomst ook niet zijn, maar “waar een wil is, is een weg”. En niets liever
bewandel ik die met jou. Mijn Remco, mijn liefste, mijn frikandel speciaal, bedankt
voor alles wat je me gegeven en geleerd hebt. Ik hou van je.
Marianne, februari 2012
295
Curriculum Vitae
Curriculum Vitae
Marianneke Littel was born in Willemstad, The Netherlands, on October 9th, 1983.
She completed grammar school (gymnasium) in 2001 at the Norbertus College
in Roosendaal, after which she started studying Psychology at the Erasmus
University Rotterdam. In 2004, she obtained her Bachelor’s degree in Clinical and
Health Psychology (cum laude), and in 2005, she received her Master’s degree in
Biological and Cognitive Psychology (cum laude). From 2005 to 2007 she worked
as a “wetenschappelijk docent/tutor” at the Institute of Psychology, Erasmus
University Rotterdam. From 2006 to 2007 she additionally worked as a research
assistant at the Rotterdam School of Management, Erasmus University Rotterdam.
In January 2008, she started her PhD research which formed the basis of this
dissertation. The studies in this project focused on cognitive processing biases
in addiction and were supervised by Professor Ingmar H. A. Franken. During
her PhD, she participated in the education program of the Dutch-Flemish postgraduate research school “Experimental Psychopathology”. She was also engaged
in teaching a number of psychology bachelor and master courses: she supervised
research projects of bachelor and master students, (guest) lectured on clinical
subjects, and co-coordinated a master course on Biological Psychopathology.
Furthermore, she reviewed empirical articles for international journals.
297
Publications
Publications
International peer-reviewed articles
Littel, M. & Franken, I.H.A. (2012). Electrophysiological correlates of associative
learning in smokers: A higher-order conditioning experiment. BMC Neuroscience,
13(8).
Littel, M. & Franken, I.H.A. (2011). Intentional modulation of the Late Positive
Potential in response to smoking cues by cognitive strategies in smokers. PLoS
ONE, 6(11): e27519.
Littel, M., Franken, I.H.A., & Muris, P. (2011). Psychometric properties of the brief
Questionnaire on Smoking Urges (QSU-Brief) in a Dutch smoker population.
Netherlands Journal of Psychology, 66, 44-49.
Luijten, M., Littel, M., & Franken, I.H.A. (2011). Inhibitory processing in smokers
during a Go/NoGo task with elements of smoking cue exposure: an investigation of
N2 and P3 amplitudes, PloS ONE, 6(4), e1889.
Littel, M. & Franken, I.H.A. (2011). Implicit and explicit selective attention to smoking
cues in smokers indexed by brain potentials. Journal of Psychopharmacology, 25(4),
503-513.
Littel, M., Franken, I.H.A., & Van Strien, J.W. (2009). Changes in the
electroencephalographic spectrum in response to smoking cues in smokers and
ex-smokers. Neuropsychobiology, 59(1), 43-50.
Littel, M., & Franken, I.H.A. (2007). The effects of prolonged abstinence on the
processing of smoking cues: an ERP study among smokers, ex-smokers and neversmokers. Journal of Psychopharmacology, 21(8), 873-882.
Muris, P. & Littel, M. (2005). Domains of childhood teasing and psychopathological
symptoms in Dutch adolescents. Psychological Reports, 96, 707-708.
299
Publications
Submitted manuscripts
Littel, M., van den Berg, I., Luijten, M., van Rooij, A.J., Keemink, L.M., & Franken,
I.H.A. (under revision). Error-processing and response inhibition in excessive
computer game players: an ERP study.
Littel, M., Euser, A.S., Munafò, M. R., & Franken, I.H.A. (submitted).
Electrophysiological indices of biased cognitive processing of substance-related
cues: a meta-analysis.
Littel, M., Field, M., van de Wetering, B.J., & Franken, I.H.A. (submitted). Reduced
cognitive processing of alcohol cues in alcohol-dependent patients seeking
treatment: an ERP study.
Dutch book chapter
Littel, M., Luijten, M., & Franken, I.H.A. (submitted). Nicotine afhankelijkheid. In:
Franken, I.H.A., Muris, P., & Denys, D. Basisboek Psychopathologie.
Presentations
Oral presentations
Littel, M. (2011). Electrophysiological correlates of biased cognitive processing in
addiction. Research school of experimental psychopathology (EPP): Symposium
“Food and drugs: two of a kind? Recent developments in eating disorders and
addictions”. Heeze, the Netherlands.
Littel, M. (2011). Intentional modulation of the Late Positive Potential in
response to smoking cues by cognitive strategies in smokers. Nationaal Instituut
voor Psychologen (NIP): Symposium “Tussen kennis en coping”. Utrecht,
the Netherlands.
Littel, M (2010).Electrophysiological correlates of associative learning in smokers:
A higher-order conditioning experiment. Forum Alcohol en Drugs Onderzoek
(FADO). Utrecht, the Netherlands.
300
Publications
Littel, M. (2009). Implicit and explicit selective attention to smoking cues in smokers
indexed by brain potentials. Forum Alcohol en Drugs Onderzoek (FADO). Utrecht,
the Netherlands.
Poster presentations
Littel, M. & Franken, I.H.A. (2011). Intentional modulation of the Late Positive
Potential in response to smoking cues by cognitive strategies in smokers. Society for
Psychophysiologal Research (SPR). Boston, MA, USA, 2011 annual meeting.
Littel, M. & Franken, I.H.A. (2010). Implicit and explicit selective attention to smoking
cues in smokers indexed by brain potentials. Donders Discussions. Nijmegen,
the Netherlands.
Littel, M., van den Berg, I., Luijten, M., van Rooij, A.J., Keemink, L.M., & Franken,
I.H.A. (2010). Error-processing and response inhibition in excessive computer game
players: an ERP study. Graduate Research Day, Erasmus University. Rotterdam,
the Netherlands.
Littel, M. & Franken, I.H.A. (2010). Implicit and explicit selective attention to smoking
cues in smokers indexed by brain potentials. National Institute on Drug Abuse
(NIDA), International Forum. Scottsdale, AZ, USA.
Littel, M. & Franken, I.H.A. (2010). Implicit and explicit selective attention to
smoking cues in smokers indexed by brain potentials. College on Problems of Drug
Dependence (CPDD). Scottsdale, AZ, USA, 72nd Annual meeting.
Littel, M. & Franken, I.H.A. (2009). Implicit and explicit selective attention to smoking
cues in smokers indexed by brain potentials. Social and Affective Neuroscience
Society (SANS). New York, NY, USA, 2nd Annual meeting.
301