Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108

Contents lists available at ScienceDirect

Progress in Neuropsychopharmacology
& Biological Psychiatry
journal homepage: www.elsevier.com/locate/pnp

Detrimental effects of a high-dose alcohol intoxication on sequential T

cognitive flexibility are attenuated by practice
⁎
Nicolas Zink1, Rui Zhang1, Witold X. Chmielewski, Christian Beste, Ann-Kathrin Stock
Cognitive Neurophysiology, Department of Child and Adolescent Psychiatry, Faculty of Medicine of the TU Dresden, Germany

A R T I C LE I N FO A B S T R A C T

Keywords: It is well-known that alcohol impairs behavioral control and motor response inhibition, but it has remained
Alcohol intoxication rather unclear whether it also impairs cognitive inhibition. As automatized behavior is less vulnerable towards
Backward inhibition the detrimental effects of alcohol than cognitive control processes, potential cognitive inhibition deficits might
Binge drinking however improve with training.
RIDE
We investigated the effect of an acute, binge-like alcohol intoxication in a balanced within-subjects design,
Switching
asking n=32 healthy young males to perform a backward inhibition paradigm once sober and once while in-
toxicated (~1.1 ‰). To identify the underlying neurophysiological mechanisms, we analyzed stimulus- and
response selection-related processes in neurophysiological data after Residue Iteration Decomposition (RIDE).
Alcohol generally impaired behavioral task performance (accuracy and response times) during task switching.
This was associated with impaired attentional processing of the task-relevant cue (reflected by reduced P1 and
N1 amplitudes), which likely resulted in a larger need for reactive control at the later stage of response selection
and control (reflected by increased fronto-central theta power). Without prior practice (~30 minutes), the in-
toxicated participants further struggled to overcome the cognitive inhibition of a previously relevant task set
(reflected by a larger backward inhibition effect). This was linked to reduced posterior theta power, which
reflects alcohol-induced impairments in working memory capacity and task set-relevant memory retrieval. As
individuals with ~30 min task practice did not show the same alcohol-related deficit, it may be deduced that
(partial) task set automatization via stimulus-response associations may help to reduce the detrimental effects of
alcohol on cognitive inhibition during task switching.

1. Introduction (like automatic processes or implicit learning) seem to be less impaired,

Binge drinking is a prevalent problem, especially in western coun-
tries (Montgomery et al., 2012). Occasional binge drinkers seem to have
a much higher risk of developing alcohol dependence or abuse com-
pared to the non-heavy episodic drinkers (Knight et al., 2002). Regular
binge drinking may furthermore be associated with slight to moderate
long-term decreases in cognitive functioning, but studies investigating
this aspect have so far yielded rather mixed results (Montgomery et al.,
2012). By comparison, findings on the acute detrimental effects of
binge-like high dose alcohol (i.e. ethanol) intoxications on cognition are
much more consistent, but also quite specific (Montgomery et al., 2012;
Stock and Beste, 2014): While executive functions and other top-down
cognitive control processes are usually strongly affected by an acute
alcohol intoxication, processes which do not require conscious control
if not even resistant to high-dose alcohol intoxication (Balodis et al.,
2007; Beaton et al., 2018; Lister et al., 1991). In this context, the ex-
ecutive function of inhibitory control has been demonstrated to be
particularly vulnerable to the effects of alcohol (Beaton et al., 2018;
Stock et al., 2016a, 2016b; Wolff et al., 2018b). As it may also play a
crucial role in the development and maintenance of substance abuse
(Aragues et al., 2011; Smith et al., 2014; Stock, 2017), there are many
studies focusing on response inhibition deficits during acute alcohol
intoxication (e.g. Stock et al., 2014; Wolff et al., 2018b). Other forms of
inhibitory control, like cognitive inhibition (of inappropriate thoughts,
action plans and task sets, which are incongruent with an individual’s
intentions or goals, see Diamond, 2013), could also be critical to the
persistence of alcohol abuse, but have received much less attention.
Deficits in this cognitive faculty may impede cognitive flexibility, i.e.

⁎
Corresponding author at: Cognitive Neurophysiology, Department of Child and Adolescent Psychiatry, Faculty of Medicine of the TU Dresden, Schubertstraße 42,
D-01309 Dresden, Germany
E-mail address: ann-kathrin.stock@uniklinikum-dresden.de (A.-K. Stock).
1
Both authors contributed equally.

https://doi.org/10.1016/j.pnpbp.2018.08.034
Received 3 May 2018; Received in revised form 13 August 2018; Accepted 31 August 2018
Available online 05 September 2018
0278-5846/ © 2018 Elsevier Inc. All rights reserved.
N. Zink et al.
efficient switching to new appropriate patterns of thought or behavior
in adaption to changes in environmental conditions, action outcomes,
or goals (Everitt and Robbins, 2005; Stalnaker et al., 2009). Impair-
ments in cognitive inhibition and the resulting changes in cognitive
flexibility might also aggravate alcohol abuse or addiction. Yet, the
effects alcohol on cognitive inhibition/inhibitory control during task
switching have remained largely unknown.
We therefore set out to experimentally investigate the effects of an
acute, binge-like alcohol intoxication on cognitive inhibition during
task switching using a backward inhibition (BI) paradigm. In order to
efficiently switch between different tasks, it is usually not sufficient to
merely activate the appropriate task set, as the task set of the previous
(i.e. now irrelevant) task needs to be inhibited in order to avoid or at
least minimize the conflict between the currently active and the pre-
viously relevant task set (Dajani and Uddin, 2015; Mayr and Keele,
2000). The BI effect is measured by assessing the time cost of over-
coming the inhibition of a recently abandoned task set that has just
become relevant again (Mayr and Keele, 2000). For this, BI paradigms
compare the behavioral performance in task sequences in which a given
task is repeated from n - 2 trials (BI condition), to when that particular
task has no n – 2 trial sequence history and thus re-occurred only after
at least two different trials in between (BASE condition). As the BI effect
reflects the cost to overcome the previous inhibition of a task set, a
larger BI effect is commonly thought to indicate a larger (cognitive)
inhibition of the previous task. With respect to alcohol, it could how-
ever also reflect larger difficulties in overcoming the inhibition of a
previous task. Given that the BI effect has been reported to strongly
depend on attentional selection mechanisms (Sinai et al., 2007; Zhang
et al., 2017, 2016a, 2016b), which can be impaired by an acute alcohol
intoxication (Stock et al., 2017a), it may be more difficult to reactivate
the recently suppressed task rules and to overcome the BI when in-
toxicated. We therefore expected that alcohol increases the BI effect, as
compared to the sober state.
To investigate this, we employed a balanced intra-individual study
design, where each participant performed the BI task once while in-
toxicated and once in a sober state. This means that half of the parti-
cipants were intoxicated during their first appointment and sober
during their second appointment, while the other half of the partici-
pants were sober during their first appointment and intoxicated during
their second appointment. Importantly, balancing the order of ap-
pointments in this manner does not only reduce variability between the
two investigated states, but also allows to test for potential training
effects: Executive functions, including cognitive flexibility, have re-
peatedly been shown to improve with training on the task and it has
been suggested that within-task practice may even help to alleviate
performance impairments and raise behavioral performance to an
“unimpaired” level in some groups that normally display executive
deficits (Sabah et al., 2018). One of the main reasons why training
might help to conceal executive performance deficits on the behavioral
level is that it establishes rather automated stimulus-response (S-R)
mappings, which reduces the strain on limited cognitive control capa-
cities. In the context of our study, it is furthermore important to note
that the degree of (task) automaticity may modulate the effects of binge
drinking (Stock et al., 2016a), as automated processes and implicit
learning seem to remain relatively intact even during high-dose in-
toxications (Balodis et al., 2007; Beaton et al., 2018; Lister et al., 1991;
Stock et al., 2016a). Training and the resulting automaticity may hence
diminish or at least reduce the detrimental effects of high-dose ethanol
intoxication on executive functions, including BI. We hence hypothe-
sized that extended practice (as defined by already having performed
the task for ~30 minutes) increases the automatization of response
selection and may therefore reduce the strain on working memory ca-
pacity and other cognitive control functions of intoxicated participants.
We therefore expected to find smaller intoxication effects (i.e. a smaller
increase in BI effect size) in those who were intoxicated during the
second of their two identical study appointments, as compared to those
98
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
who were intoxicated during their first appointment.
To be able to identify the cognitive sub-processes underlying the
expected behavioral effects, we recorded an EEG during task perfor-
mance.
In order to deal with the alcohol-induced increase in response
variability and to further be better able to dissociate stimulus- and re-
sponse-selection associated processes in the neurophysiological data,
we ran a residue iteration decomposition (RIDE) (Ouyang et al., 2011,
2015a; Verleger et al., 2014) before quantifying all neurophysiological
measures (for details, please refer to the methods section). In short,
RIDE decomposes EEG data into several component clusters (Mückschel
et al., 2017; Ouyang et al., 2011): The S-cluster reflects stimulus-related
processes like perception and attention, while the C-cluster depicts in-
termediate processes between stimulus and response like response se-
lection and the R-cluster depicts response-related processes including
motor preparation and execution (Ouyang et al., 2011). Previous stu-
dies had shown that the size of the BI effect seems to strongly depend on
early attentional stimulus processing and attentional selection processes
(Wolff et al., 2018a; Zhang et al., 2016b, 2017), as reflected by the
visual P1 and N1 event-related potentials (ERPs) (Luck et al., 2000). As
several other studies with comparable intoxication levels (but other
experimental paradigms) have shown that alcohol may decrease at-
tentional processing as reflected by smaller amplitudes in one or both
ERPs (Stock et al., 2017a; Wolff et al., 2018b) and impair the top-down
allocation of attention (Stock et al., 2017b), we expected alcohol-re-
lated changes in the size of the BI effect to be reflected by decreases in
the P1 and/or N1 amplitude in the S-cluster. In contrast to this, (the size
of) the BI effect has usually not been found to be associated with
modulations of the fronto-central N2, which is thought to reflect con-
flict monitoring and cognitive effort (Larson et al., 2014), or the par-
ietal P3 found in this paradigm, which is commonly believed to reflect
the process of linking stimuli to appropriate responses/response selec-
tion (Polich, 2007; Verleger et al., 2005). Due to the lack of BI effects on
those response selection-associated components (Zhang et al., 2016c),
we did not expect these C-cluster ERPs to reflect alcohol-associated or
any other changes of BI size. As detrimental effect of an acute alcohol
intoxication on various cognitive control functions, including motor
response inhibition, have however repeatedly been associated with
decreases in N2 and P3 ERPs (e.g. Rohrbaugh et al., 1987; Stock et al.,
2017a, 2016b, 2014), we nevertheless decided to also quantify and
analyze the N2 and P3 ERPs in the C-cluster in order to provide a more
complete picture.
Another neurophysiological measure, which has repeatedly been
linked to cognitive control and shown to reflect modulations of (re-
sponse) inhibition, are theta frequency band oscillations (Beste et al.,
2011; Cavanagh and Frank, 2014; Dippel et al., 2016; Mückschel et al.,
2017). Both stimulus- and response-related aspects of inhibitory control
can be found in theta frequency oscillations (Beste et al., 2011; Dippel
et al., 2016; Mückschel et al., 2017): Response-related theta frequency
oscillations are mainly observed over frontal areas (Beste et al., 2011;
Cavanagh and Frank, 2014; Dippel et al., 2016; Mückschel et al., 2017),
closely associated with the N2 and N450 ERPs (Larson et al., 2014), and
thought to reflect rather “reactive” aspects of cognitive control such as
conflict detection and monitoring or cognitive effort (Botvinick et al.,
2001; Cavanagh and Frank, 2014; Cooper et al., 2017). Stimulus-related
theta is however more often observed over posterior regions
(Freunberger et al., 2007; Gladwin and de Jong, 2005; Sauseng et al.,
2006) and thought to reflect more “proactive” aspects of cognitive
control (Cooper et al., 2017). Posterior theta has been found at parietal
and/or occipital electrodes during task switching (Freunberger et al.,
2007; Gladwin and de Jong, 2005) and is supposed to reflect top-down
regulation processes in recalling the task set memory, as its power is
enhanced when the previous irrelevant visual task must be reactivated
(Freunberger et al., 2007; Gladwin and de Jong, 2005). Likewise, pos-
terior theta has been reported to increase when cognitive content needs
to be suppressed (Depue et al., 2013). Further matching this, (Cooper
N. Zink et al.
et al., 2017) reported “anticipatory” parietal theta to emerge prior to
frontal theta during a cue period and to be linked to goal-updating.
They also reported that both enhanced power and phase of parietal
theta were associated with better behavioral task switching perfor-
mance and predictive of cognitive control efficiency. Based on this
finding, (Cooper et al., 2017) suggested that proactive control processes
reflected by parietal theta may also contribute to cognitive control ef-
ficiency. On this ground, we expected the effects of alcohol intoxication
on the BI effect to be reflected by posterior stimulus-related theta
quantified in the S-cluster. Given that alcohol intoxication impairs top-
down control processes (Day et al., 2015; Garbusow et al., 2014; Stock
et al., 2016a, 2017b), an acute intoxication should decrease posterior
theta power, particularly when the inhibition of a recently performed
task needs to be conquered (i.e. in the BI condition). Considering that
automaticity is supposed to reduce the strain on the residual top-down
cognitive control capacities of intoxicated participants, it might allow
for better recall of task memory/task rules, as reflected by larger pos-
terior theta power (Gladwin and de Jong, 2005). We hence expected
larger theta power in the well-practiced participants who were able to
use rather automatized S-R associations when intoxicated (during their
second appointment) than participants who did not have comparable
automatisms when intoxicated (during their first appointment). Given
that the BI effect emerges from early attentional processes, frontal re-
sponse-related theta in the C-cluster, which indicates the reactive need
for response control (Cavanagh and Frank, 2014; Cooper et al., 2017),
may reflect a general effect of alcohol intoxication rather than specific
effect of alcohol intoxication on BI effect.
Taken together, we hypothesized that the detrimental effects of an
acute, binge-like alcohol intoxication on inhibition are not limited to
the domain of motor response inhibition, but may also be found in the
domain of the cognitive backward inhibition. But opposing what has
been observed for motor response inhibition, we expected to find al-
cohol-induced increases in the size of the BI effect to be reflected by
cognitive processes associated with early attentional stimulus proces-
sing reflected by the visual P1/N1 ERPs and not by later, response-
selection associated processes reflected by the N2 and P3 ERPs.
Likewise, we expected to find alcohol-induced increases in the size of
the BI effect to be reflected by decreases in proactive control and top-
down attentional processes including task set reactivation, as reflected
by posterior theta, and not by reactive conflict monitoring processes
reflected by frontal theta. Lastly, we expected that the detrimental ef-
fects of alcohol on BI should be attenuated by prior task practice/per-
formance, which allows to establish (partly) automatized stimulus-re-
sponse mappings that should be less vulnerable towards alcohol than
top-down control.
2. Experimental procedures
2.1. Participants
N=32 young, male and right handed subjects (age 19 to 33, mean
age 25.37 ± 2.79), who reported no physical, neurological or mental
illness, were included in the study. Subjects were recruited via online
adverts and flyers at the local university. Applicants were only included
in our sample if they reported a stable pattern of moderate, non-risky
alcohol consumption habits as indicated by AUDIT scores between 1
and 15 points (mean 6.76; SD 2.53) (Babor et al., 2001), and reported
drinking eight or more standard units of alcohol on a single occasion/
evening no more than once a month and no less than once a year.
Women were not included in this study due to the potential risk of
undetected pregnancy and to avoid any confounding effects related to
hormonal changes/the menstrual cycle. All participants gave written
informed consent and were treated in accordance with the declaration
of Helsinki. The study was approved by the Ethics commission of the
Medical Faculty of the TU Dresden (ID number: EK293082014).
99
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
2.2. Experimental design and alcohol administration
The study design and alcohol administration procedure were iden-
tical to the protocol used in previous studies of our group (Stock et al.,
2017b): We employed a within-subject design where each subject was
tested two times (once sober and once intoxicated). The order of ap-
pointments was balanced between subjects. There was a delay of at
least 48 hours but no more than 7 days between the sober and the
intoxicated appointment. All participants were asked to refrain from the
use of caffeine, nicotine, guanine and all other stimulant or sedative
substances within the last 4 hours before the start of each experimental
session and to stop eating at least 3 hours prior to the intoxication
session.
A version of the equation by Widmark (1932) and Watson et al.
(1980) was used to calculate an individual amount of vodka (40 percent
alcohol by volume) for each participant. The individual amount was
targeting a maximal possible breath/blood alcohol concentration (BAC)
of 1.5‰ (mg/g) and an approximate mean peak BAC of 1.2 ‰. When
assuming an absorption deficit of 20% on an empty stomach, this means
that a BAC of up to 1.2‰ can be reached without exceeding a maximal
possible BAC of 1.5 ‰. The tool that was used for the calculation of
individual beverage amounts is provided in the supplementary mate-
rial. The individually determined amount of vodka was then diluted
with an equal proportion of orange juice at room temperature. Irre-
spective of their individual amount, participants were asked to consume
their drink within 30 minutes. During consumption and a subsequent
30-minutes waiting period, subjects watched three Big Bang Theory
episodes. These had been chosen for their lack of extensive discourses
on drinking and helped to prevent negative mood swings (e.g. Friedman
et al., 2007). Prior to working on the task described in this publication,
the participants spent 30 minutes performing a modified Simon task the
results of which are currently still unpublished. Hence, they started
working on the paradigm reported in this study 60 minutes after the
end of their consumption. Breath alcohol levels were measured im-
mediately before and after task performance using the “Alcotest 3000”
breath analyzer following the instructions of the manufacturer (Drä-
gerwerk, Lübeck, Germany). As shown in previous publications (Stock
et al., 2014b; Stock and Beste, 2014), analyzing breath alcohol levels
this way essentially yields the same results as the analysis of venous
blood using a capillary gas chromatography-headspace technique. The
urine of all participants was tested for illicit drugs (amphetamines,
methamphetamine, morphine and THC) on both appointments using
“nal von minden Drug-Screen” tests (nal von minden GmbH, Regens-
burg, Germany). The detection of any substance at any appointment
would have led to the exclusion from the sample.
2.3. Task
In order to investigate effects of alcohol intoxication on cognitive
(i.e. non-motor) aspects of inhibition and the underlying neurophysio-
logical mechanisms, we used a backward inhibition paradigm as pro-
posed by Mayr and Keele (2000) and adapted by Koch et al. (2004)
(compare also (Zhang et al., 2016a, 2016c). During the experiment,
participants were seated in front of a 17 inch CRT computer monitor at
a viewing distance of 57 cm. White cue and target stimuli were pre-
sented on black background at the center of the screen. Participants had
to respond on a regular USB keyboard by pressing either the left or right
Ctrl buttons using their left and right index fingers, respectively. “Pre-
sentation” software (version 14.9. by Neurobehavioral Systems, Inc.)
was used for stimulus presentation, response recording, and EEG trig-
gers.
The task is illustrated in Fig. 1. Each trial started with a 100 ms
presentation of one of three cues that indicated the task rule in effect: A
square cue was used to indicate the odd/even rule (“task A”), a dia-
mond cue was used to indicate the smaller/larger than five rule (“task
B”), and a triangle cue was used to indicate the double-press rule (“task
N. Zink et al.
Fig. 1. Schematic illustration of the backward inhibition paradigm. Each trial
began with the central presentation of a task cue for 100 ms. A square cue
indicated the odd/even task (left button press for odd numbers, right button
press for even numbers). A diamond cue (see bottom left) indicated the smaller/
larger rule (left for smaller than five, right for larger than five). A triangle cue
(see bottom left) indicated the double press rule (simultaneous button press
within the first 1000 ms after target onset). The target stimulus (any digit from
1 to 9, except for 5) was presented within the cue frame until a response was
made. Only in double press trials, a speedup sign (“Schneller!”, translating to
“Faster!”) appeared above the cue frame in case no response was given within
the 1000 ms after target onset. During the inter-trial interval of 2000 ms, there
was a 500 ms feedback for incorrect trials (“Falsch!”, translating to “Wrong!”),
but no feedback/only a fixation cross was presented in correct trials.
D”). After a stimulus onset asynchrony (SOA) of 100 ms, the target
stimulus was presented in the center of the cue frame. The target sti-
mulus set consisted of a single digit from 1 to 9 except for 5. Cue and
target stimulus were presented together until a response was made.
After the response, a fixation cross was presented at the center of the
screen for 1500 ms until the cue of the next trial was presented. During
the odd/even and smaller/larger trials, participants had to perform a
parity (“Is the number presented odd or even?”) or a magnitude jud-
gement (“Is the number presented smaller or larger than 5?”), where
either the left key (number is odd/smaller than 5) or the right key
(number is even/larger than 5) had to be pressed within 2000 ms after
the target stimulus onset. In the double-press task, participants were
asked to press both left and right key simultaneously within 1000 ms
after stimulus onset. A speed-up sign (German Word “Schneller!”,
translating to “Faster!”) was presented, if participants did not respond
within the respective time windows. When the participants responded
after the speed-up sign appeared, they also received a feedback with the
German words “zu langsam!” (translating to “too slow!”) at the center
of the screen. Both too slow responses and non-simultaneous key
presses in task D (when the two key presses were separated by more
than 50 ms) were coded as errors. An error feedback (German word
“falsch!” translating to “wrong!”) was presented for 500 ms when
participants did not give the right response in any of the cue conditions.
The experiment consisted of 768 trials divided into 8 equally sized
blocks. After each block, participants received feedback about their
mean response time (RT) during the last block (Fig. 1). Altogether, the
participants took between 32 and 40 minutes to complete the task.
Task/cue sequence and stimulus sequence were randomized within
each block, except for n-1 repetitions. This means that neither cues nor
stimuli could be repeated in any two consecutive trials. Furthermore, no
stimulus was directly repeated upon reoccurrence of a given cue (i.e. if
the square cue was presented with stimulus “7”, that particular stimulus
could not be presented in the next trial that started with a square cue).
Each possible cue and stimulus combination was presented equally
often. Within each block, each trial (except for the first two trials, of
course) built a triplet with the last two preceding trials. All twelve
possible triplet-combinations (ABA; ADA; BAB; BDB; DAD; DBD; DBA;
100
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
BDA; DAB; ADB; BAD; ABD) were presented equally often ( ± 1 triplet
for two triplet conditions in each block). Triplets where the last of the
three trials had the same cue as the n-2 trial were categorized as
backward inhibition (BI) triplets while triplets without that n-2 cue
repetition were categorized as baseline (BASE) triplets. “ABA” and
“BAB” were used as BI triplets, while “DBA” and “DAB” were used as
BASE triplets, as previous studies using the same paradigm indicated
that those four trials best depict the BI effect (Koch et al., 2004; Zhang
et al., 2016b).
Participants were asked to respond as quickly and accurately as
possible and they were asked not to keep track of previous trials. To
make sure that the participants understood the instructions and kept
the rules in mind, each participant started with a practice block con-
sisting of 12 trials. The RT and errors for each condition were collected
for behavioral analyses. In this context, please note that responses were
only rated as correct when the correct buttons were pressed within 100
and 2500 ms after target onset.
2.4. On the issue of component overlap
As already noted in (Zhang et al., 2016a), the overlap of ERP
components evoked by temporally close stimuli needs to be carefully
considered as it may hamper the comparison of different task condi-
tions (e.g. Keil et al., 2014). Many ERP studies investigate condition
differences of experimental paradigms and an overlap of ERPs in dif-
ferently composed or evoked conditions may flaw the experimental
design in such a case. The reason for this is that it is often hardly
possible to assess the origin of the observed condition differences in this
context. Yet, this does not apply to the analyses of the BI effect. While
the SOA between cue and target onset is very short and might therefore
cause some overlap of ERPs that reflect early attentional visual pro-
cessing, component overlap does not cause any problems in the inter-
pretation of the analyzed ERPs: When analyzing the BI effect in the last
trial of a triplet, we do not directly compare different conditions or
trials because there are no differences in the condition of the last trial of
the triplets used to compare the BASE and BI conditions. The BI con-
dition comprises the two triplets ABA and BAB, so that ERPs obtained
from both A and B trials will be averaged to form the BI condition. The
BASE condition comprises the two triplets DBA and DAB, ERPs obtained
from both A and B trials will again be averaged to form the BASE
condition. Therefore, comparing the BI and BASE conditions means
comparing an average of A and B trials to an average of A and B trials.
The differences between BI and BASE trials are evoked by the effect of
the n-2 and n-1 trials the analyzed nth trial. Given that the RSI lasted
1500 ms, there can however be no overlap between the ERPs evoked by
the n-1 or n-2 trial and the trial we quantified for analyses. (Zhang
et al., 2016a) Hence, it is entirely impossible that any of the reported
effects caused by either BI condition or appointment are caused by an
overlap of ERP components.
2.5. EEG recording and preprocessing
EEG recordings were made using 60 sintered Ag/AgCl ring elec-
trodes located at standard equidistant scalp positions with a sampling
rate of 500 Hz. Electrode Fpz was used as reference and the ground
electrode was located at θ = 58, ф = 78 (customized BrainCap Fast‘n
Easy sub-inion model EEG caps). During recording, all electrode im-
pedances were kept below 5 kΩ. After recording with Brain Vision
Recorder (Brain Products Inc.), the data was average-referenced, down-
sampled to 256 Hz and a band-pass filter (IIR filter from 0.5 Hz to 20 Hz
at a slope of 48 db/oct each) was applied. Subsequently, a raw data
inspection was conducted to manually remove pauses and rare tech-
nical artifacts. An independent component analysis (ICA; Infomax al-
gorithm) was applied to remove periodically recurring artifacts such as
pulse artifacts and EOG artifacts like horizontal and vertical eye
movements. Trials were segmented for all experimental conditions and
N. Zink et al.
ranged from -2000 ms before to 2000 ms after the onset of the cue. Of
note, the length of the segments was chosen to allow for time-frequency
decomposition/wavelet analyses. All segments underwent an automatic
artifact rejection (rejection criteria allowed for a maximal value dif-
ference of 200 μV in a 200 ms interval and excluded activity below
0.5μV in a 100 ms interval). A current source density (CSD) transfor-
mation was run to obtain a reference-free evaluation of the electro-
physiological data (Perrin et al., 1989; Tenke and Kayser, 2012). The
CSD transformation additionally serves as a spatial filter which helps to
identify electrodes that best reflect activity related to cognitive pro-
cesses because it accentuates the scalp topography (Nunez and Pilgreen,
1991; Tenke and Kayser, 2012). Subsequently, a baseline correction
was applied to the interval from -300 ms to 0 ms before cue onset.
For the quantification of non-decomposed ERP components, we
extracted the mean amplitude of brief time intervals centered around
the respective peaks. The choice of electrodes was based on the scalp
topography of the different ERP components. The cue P1 and cue N1
ERP amplitudes were quantified at electrodes P7 and P8 by extracting
the average voltage in the time window ranging from 85-100 ms (cue
P1) and 145-170 ms (cue N1). The cue N2 ERP amplitudes were
quantified at electrode Cz by extracting the average voltage in the time
window ranging from 245-270 ms. The target P1 and target N1 ERP
amplitudes were quantified at electrodes P7 and P8 by extracting the
average voltage in the time window ranging from 280-300 ms (target
P1) and 340-355 ms (target N1). The target N2 and target P3 ERP
amplitudes were quantified at electrode Cz (target N2) and Pz (target
P3) by extracting the average voltage in the time window ranging from
415-440 ms (target N2) and 560-630 ms (target P3). As we however
expected the non-decomposed ERPs to be affected by the increase in RT
variation under the influence of alcohol, the results of those analyses
are reported in the supplementary material.
2.6. Residue Iteration Decomposition (RIDE)
To account for the alcohol-induced increase response time varia-
bility and to dissociate stimulus-related and response selection-related
processes in the EEG signal, we performed a RIDE decomposition of the
EEG signal in accordance with established procedures and previous
studies (Friedrich et al., 2018; Ouyang et al., 2011; Verleger et al.,
2014; Wolff et al., 2017) using the RIDE toolbox (available on http://
cns.hkbu.edu.hk/RIDE.htm) for MATLAB (MATLAB 12.0; Mathworks
Inc.).
In the following, the RIDE method is briefly described (as already
done in (Bluschke et al., 2017) and (Friedrich et al., 2018): RIDE de-
composes ERP components by applying L1-norm minimization (i.e.,
obtaining median waveforms), which minimizes the residual error due
to temporal variability in single trials (Ouyang et al., 2015a). RIDE
decomposition is based on latency variability and is separately applied
to each electrode, regardless of scalp distributions or waveforms
(Ouyang et al., 2015b). The previously applied CSD transformation
does therefore not affect the results. RIDE decomposes the ERP signal
into clusters that are either correlated to the stimulus onset (S-cluster)
or to the response time (R-cluster), as well as a central C-cluster with
variable latency, which is initially estimated and then iteratively im-
proved. RIDE uses a self-optimized iteration scheme for latency esti-
mation through which the latency estimation of the C-cluster is im-
proved. The initial latency of the C-cluster is estimated using a time
window function. In an iterative procedure, the S-cluster is removed
and the latency of the C-cluster is re-estimated based on a template-
matching approach until convergence of the initial latency estimation
and the S- and C-cluster. Additionally, the latest RIDE algorithm uses a
time window confinement for each cluster. Each time window is as-
sumed to cover the range within which each component is supposed to
occur and the specific values should be adjusted to fit the data on ap-
plication (Ouyang et al., 2015a). -For further details on the method,
please see Ouyang et al. (2011, 2015a).
101
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
For the current study, the time window for the S-cluster was set to
−200 to 600 ms around stimulus onset. The initial time window for the
estimation of the C-cluster was set to 0–900 ms after stimulus onset. The
time window for the R-cluster was set around the response trigger
(−300 to 300 ms). As we were mainly interested in the S- and C-cluster,
the R-cluster was not further analyzed. Just like for traditional ERP
analyses, we applied a data-driven approach based on the scalp topo-
graphies and latencies of the ERP components of interest to quantify
stimulus-related and response selection-related processes in the S-
cluster and C-cluster, respectively: In the S-cluster, the cue-elicited
RIDE-P1 and RIDE-N1 amplitudes were quantified at electrodes P7 and
P8 by extracting the average voltage in the time window ranging from
85 to 100 ms (cue-P1) and from 145 to 170 ms (cue-N1). The target-
elicited RIDE-P1 and RIDE-N1 amplitudes were also quantified in the S-
cluster at electrodes P7 and P8 by extracting the average voltage in the
time window ranging from 255 to 290 ms (target-P1 at electrode P7),
280-300 ms (target-P1 at electrode P8) and from 340 to 355 ms (target-
N1 at electrodes P7 and P8). In the C-cluster, the target-elicited RIDE-
N2 and RIDE-P3 amplitudes were quantified at electrode Cz (target-N2)
and Pz (target-P3) by extracting the average voltage in the time win-
dows ranging from 385 to 460 ms (target-N2) and from 560 to 630 ms
(target-P3). The choice of time windows and electrodes quantified for
the RIDE decomposition was also validated using the statistical ap-
proach outlined in (Mückschel et al., 2014).
2.7. Time frequency transformation
The time–frequency (TF) analysis was conducted on the RIDE-de-
composed single trial S- and C- cluster data by means of a continuous
wavelet transformation (CWT) employing Morlet wavelets (ⱳ) in the
time domain to different frequencies (ƒ):
ⱳ (t , ƒ) = A exp(−t 2/2σ 2t ) exp(2iπƒt )
where t=time, A=(σt π )-1/2, σt=wavelet duration, and i = −1 .
For analysis and TF-plots, a ratio of ƒ0/ σƒ=5.5 was used, where σƒ is the
width of the Gaussian shape in the frequency domain and ƒ0 is the
central frequency. The analysis was conducted in the frequency range
from 0.5 to 20 Hz, and a central frequency at 0.5 Hz intervals was
employed. For different ƒ0, time and frequency resolutions (or wavelet
duration and spectral bandwidth; Tallon-Baudry et al., 1997) can be
calculated as 2σt and 2σƒ respectively. σt and σƒ are related by the
equation σt=1/(2πσƒ). For example, for ƒ0 = 1 Hz, 2σt = 1770 ms and
2σƒ = 0.36 Hz; for ƒ0 = 3 Hz, 2σt = 580 ms and 2σƒ = 1.09 Hz; for ƒ0 =
5 Hz, 2σt = 350 ms and 2σƒ = 1.82 Hz. The identification of relevant
electrodes and time windows for theta band activity quantification was
based on TF plots and scalp topographies. The evoked power in the
theta frequency band in the S-cluster was quantified at 5.5 Hz for
electrode P7 within the time window of 180-240 ms after cue onset. As
the visual inspection of the TF plot at electrode P7 also showed dif-
ferences in the alpha band, we additionally ran an add-on quantifica-
tion of the alpha frequency band at 9 Hz in the time window from 160
to 200 ms. In the C-cluster, it was quantified at 6 Hz at electrode Fz
within the time window of 800-900 ms after cue presentation. The theta
power in the S- and C-cluster was analyzed by normalizing wavelet
power to the baseline from -750 to -450 ms prior to cue onset.
2.8. Statistical analyses
Separate repeated-measures ANOVAs were used to analyze the ob-
tained data. For the behavioral data, we separately analyzed accuracy
(i.e. the percentage of triplets with correct responses between 100 and
2500 ms RT in all three trials) and the RTs of the last trials in each
entirely correct triplet. In this context, it should be noted that due to the
prerequisite of three consecutive correct trials, the chance level is at p
= .53 = 12.5 %, not 50 %. Given the aforementioned exclusion criteria
(i.e., requiring three consecutive correct trials for RT analysis), the
N. Zink et al.
mean number of the remaining trials for each triplet lay between 60 and
65 trials for each condition of the sober appointment (ABA: 62.9 ± 3.8;
DBA: 62.4 ± 3.9; ADA: 61.5 ± 4.2; BDA: 63.6 ± 4.2; DBD: 62.8 ± 3.9;
BAD: 63.7 ± 4.5; ABD: 62.3 ± 4.1; ADB: 63.3 ± 4.2; BAB: 61.7 ± 4.3;
BDB: 61.7 ± 4.4; DAB: 63.8 ± 4.4; DAD: 61.4 ± 4.3) and the in-
toxicated appointment (ABA: 63.5 ± 5.3; DBA: 61.7 ± 5.2; ADA:
61.7 ± 4.8; BDA: 62.5 ± 5.1; DBD: 63.1 ± 5.3; BAD: 61.6 ± 4.2; ABD:
61.1 ± 5.1; ADB: 62.8 ± 5.1; BAB: 63.1 ± 4.1; BDB: 61.8 ± 4.8; DAB:
61.2 ± 4.2; DAD: 63.0 ± 4.4).
For the neurophysiological data, we quantified and analyzed both
ERPs and theta oscillations in the S- and C-clusters. The amplitudes of
the cue- and target-elicited P1 and N1 ERPs, as well as the amplitudes of
the N2 and parietal P3 were analyzed both in regular EEG segments and
in RIDE-decomposed segments. Furthermore, we quantified the evoked
power of the theta frequency band in the RIDE S- and C-clusters.
In all analyses, we used “intoxication status” (sober vs. intoxicated)
and “experimental condition” (BI vs. BASE triplet) as within-subject
factors. “Drinking sequence” (intoxicated on the first appointment vs.
intoxicated on the second appointment) was used as a between-subjects
factor. Whenever applicable, we also used “electrode” as an additional
within-subject factor in the neurophysiological analyses.
All reported values underwent Greenhouse-Geisser correction and
post-hoc tests were Bonferroni-corrected, whenever necessary. For all
descriptive statistics, the standard error of the mean (SEM) is given as a
measure of variability.
3. Results
3.1. Intoxication results
Right before performing the experimental paradigm, the partici-
pants presented with a mean BAC of 1.07 ‰ ± .03. After completing
the paradigm, the participants presented with a mean BAC of 1.01
‰ ± .03.
3.2. Behavioral data
Behavioral effects related to the intoxication status are illustrated in
Fig. 2. For the accuracy in percent, the mixed effects ANOVA revealed a
main effect of “intoxication status” (F(1,38) = 26.82; p < .001; ηp2=
.414) showing that participants had a lower accuracy (i.e. made more
errors) during their acute ethanol intoxication (65.8 % ± 2.5) than
while sober (76.1 % ± 2.1). Furthermore, a main effect of “experi-
mental condition” (F(1,38) = 15.47; p < .001; ηp2 = .289) showed that
participants responded less accurately in the BI condition (69.2
% ± 2.2) than in the BASE condition (72.7 % ± 2.0). No other main
effects or interactions were significant (all F ≤ 1.48; p ≥ .232).
For the hit RTs, the mixed effects ANOVA revealed a main effect of
“intoxication status” (F(1,38) = 55.05; p < .001; ηp2 = .592) showing
that participants responded more slowly when they were intoxicated
(804.2 ms ± 24.0) than when they were sober (666.5 ms ± 18.4). A
main effect of “experimental condition” (F(1,38) = 39.98; p < .001; ηp2
= .513) showed that RTs were larger in the BI condition (752.9
ms ± 20.1), as compared to the BASE condition (717.8 ms ± 18.7).
Moreover, a main effect of “drinking sequence” (F(1,38) = 11.89; p =
.001; ηp2 = .238) revealed that participants who were sober during
their first appointment had on average faster responses (669.0
ms ± 28.6) than those who were intoxicated during their first ap-
pointment (801.7 ms ± 25.8). Importantly, we also found an interac-
tion of “intoxication status x drinking sequence” (F(1,38) = 64.52;
p < .001; ηp2 = .629), as well as an interaction of “intoxication status x
drinking sequence x experimental condition” (F(1,38) = 12.20; p =
.001; ηp2 = .243). Post-hoc analyses showed that participants who were
intoxicated during the first of their two appointments displayed a sig-
nificant interaction of “experimental condition x intoxication status” (F
(1,21) = 9.15; p = .006; ηp2 = .304). Specifically, this group had a
102
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
larger BI effect (i.e. the condition differences assessed as BI-BASE) in
their intoxication session (51.5 ms ± 9.4) than in their sober session
(20.9 ms ± 9.7) (t(21) = 3.03; p = .006). In participants who were
intoxicated during their second appointment/sober on the first we
neither found the interaction of “experimental condition x intoxication
status” (F(1,17) = 3.91; p = .065; ηp2 = .187), nor a main effect of
“intoxication status” (F(1,17) = 0.249; p = .624; ηp2= .014). Further
analyzing the interaction of “intoxication status x drinking sequence x
experimental condition” also revealed an interaction of “drinking se-
quence x experimental condition” during intoxication (F(1,38) = 4.13;
p = .049; ηp2 = .098). Here, the BI effect (condition difference BI-
BASE) was larger when the intoxication session was the first appoint-
ment/performed before the sober session (51.5 ms ± 9.4) than when it
was the second session/performed after the sober appointment (25.4
ms ± 9.4) (t(38) = 2.03; p = .049). No such interaction of “drinking
sequence x experimental condition” could be found for the sober state
(F(1,38) = 2.67; p = .110; ηp2 = .066). No other main effects or in-
teractions were significant (all F ≤ .99; p ≥ .326)
Taken together, high-dose alcohol intoxication reduced the general
response accuracy and slowed down the responses. Yet, the factor
drinking sequence also had a marked effect: Intoxication-related RT
deficits/an increase in the size of the BI effect could only be observed in
participants who were intoxicated the first time they performed the
task. Participants who had been given the opportunity to perform and
therefore practice the task before (due to being sober during the first of
two otherwise identical appointments), did not show any slowing of
responses or an increased BI effect. This strongly suggests that prior
practice partly protected our subjects from the detrimental effects of a
high-dose ethanol intoxication on task performance. One of the po-
tential explanations for this effect could be a stronger automatization of
stimulus-response mapping processes after ~30 min of training, as in-
creased automatization likely reduces cognitive control requirements.
3.3. Neurophysiological data
To dissociate stimulus-related processes und response selection
processes with the help of the S- and C-clusters, RIDE was applied.
Please refer to the supplementary information for the results of regular
(i.e. non-decomposed) ERPs.
3.3.1. Event-related potentials
All attentional ERPs derived from the S-cluster are illustrated in
Fig. 3. For the cue-elicited P1 quantified in the S-cluster at electrodes P7
and P8, a main effect of “intoxication status” (F(1,33) = 11.77; p =
.002; ηp2 = .263) showed that P1 amplitude was significantly smaller
during the intoxication (15.1 μV/m2 ± 1.6) than during the sober ap-
pointment (20.3 μV/m2 ± 1.9). A main effect of “experimental condi-
tion” (F(1,33) = 5.09; p = .031; ηp2 = .117) showed decreased cue P1
amplitudes in the BI condition (16.9 μV/m2 ± 1.6) compared to the
BASE condition (18.4 μV/m2 ± 1.5). A main effect of “electrode” (F
(1,33) = 5.80; p = .022; ηp2 = .150) revealed larger amplitudes at
electrode P8 (20.0 μV/m2 ± 2.2) than at P7 (15.3 μV/m2 ± 1.4μV/m2).
A main effect of “drinking sequence” (F(1,33) = 5.85; p = .021; ηp2 =
.150) was also detected, showing that P1 amplitudes were on average
smaller in participants who were intoxicated during their first ap-
pointment (13.9 μV/m2 ± 2.1) than in those who were intoxicated
during their second appointment (21.5 μV/m2 ± 2.3). An interaction of
“electrode x drinking sequence” (F(1,33) = 4.87; p = .034; ηp2 = .129)
revealed that the aforementioned effect of drinking sequence was only
found at electrode P8 (t(33) = -2.70; p = .011; intoxicated in first
session: 14.1 μV/m2 ± 2.0; sober in first session: 26.0 μV/m2 ± 4.2) but
not at electrode P7 (t(33) = -1.16; p = .253). No other main effects or
interactions were significant (all F ≤ 1.00; p ≥ .326).
For the cue-elicited N1 quantified in the S-cluster at electrodes P7
and P8, a main effect of “intoxication status” (F(1,33) = 23.52;
p < .001; ηp2 = .416) revealed that the N1 amplitude was smaller
N. Zink et al. Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108

Fig. 2. Illustration of the behavioral performance data. Absolute values are depicted in the left graphs, the backward inhibition effect/condition difference is depicted
in the right graphs. In each graph, a dotted line separates the two groups: The two appointments/states of the sober first group are always depicted on the left, while
the two appointments/states of intoxicated first group are always depicted on thr right of each graph. A) In the accuracy measure depicted in the top row, we found
significantly higher accuracy values for participants when they were sober (compared to intoxicated) and in BASE trials (compared to BI trials). B) In the hit RTs
depicted in the bottom row, we found significantly slower responses for participants who were intoxicated (compared to sober), who were intoxicated on their first
appointment (compared to those who were sober on their first appointment) and in BI trials (compared to BASE trials). We further found that the BI effect (i.e. the
condition difference BI – BASE) was larger when intoxicated than when sober, but only in the group that was intoxicated during the first of their two appointments
(marked with an asterisk). The group that was intoxicated on the second appointment did not show this alcohol-dependent effect (n.s.). In line with this, we found
that the BI effect was larger in the first INTOX group than in the first SOBER group, but only during the intoxicated appointment (also marked with an asterisk). The
size of the BI effect did not differ between groups during the sober appointment (not marked). The error bars denote the standard error of the mean (SEM).

Fig. 3. Illustration of the cue- and target-associated P1 and N1 ERPs in the RIDE
S-cluster at electrodes P7 and P8 (averaged). Time point zero denotes the onset
of the cue stimulus. The target stimulus was added to the visual array at 100 ms.
ERPs assessed while intoxicated are denoted in red line color, ERPs assessed
while sober are denoted in green line color. The BASE condition is denoted in a
lighter color than the BI condition. The four quantified ERP peaks are denoted
by arrows. Topographic plots of each ERP within the time interval used for ERP
quantification are provided right next to the ERP labels. Alcohol effects could
only be observed for the cue-associated P1 and N1 amplitudes, which were
smaller when intoxicated than when sober. The amplitudes of the target-asso-
ciated P1 and N1 did not show this effect. For a more detailed description of
non-intoxication-related significant effects differences, please refer to the re-
sults section.
during intoxication (-5.0 μV/m2 ± 2.6) than during the sober state
(-16.2 μV/m2 ± 3.3μV/m2). Interestingly, an interaction of “intoxica-
tion status x experimental condition” (F(1,33) = 8.80; p = .006; ηp2 =
.211) was found. Post hoc tests showed that during the intoxicated
appointment, the N1 amplitude was larger in the BI condition (-5.8 μV/
m2 ± 2.6) than in the BASE condition (-4.3 μV/m2 ± 2.6) (t(34) =
-2.38; p = .023). This condition difference was not found in the sober
appointment (t(34) = 1.59; p = .121). No other main effects or in-
teractions were significant (all F ≤ 2.25; p ≥ .143).
For the target-elicited P1 quantified in the S-cluster at electrodes P7
and P8, an interaction of “experimental condition x drinking sequence”
(F(1,33) = 4.15; p = .050; ηp2 = .112) was found. Post-hoc tests re-
vealed that the BI effect (BI-BASE) was on average larger in participants
who were intoxicated during their first appointment (-2.3 μV/
m2 ± 1.2) than in those who were intoxicated during their second ap-
pointment (1.1 μV/m2 ± 1.1) (t(33) = -2.04; p = .050). No other main
effects or interactions were significant (all F ≤ 2.11; p ≥ .156).
For the target-elicited N1 quantified in the S-cluster at electrodes P7
and P8, no significant effects were detected (all F ≤ 2.33; p ≥ .136).
For the N2 and the P3 quantified in the C-cluster at electrodes Cz
and Pz, respectively, no main effects or interactions were significant
(N2: all F ≤ 3.88; p ≥ .057; P3: all F ≤ 3.43; p ≥ .073). The ERPs are
illustrated in Fig. 4 and Fig. 5.
3.3.2. Time-frequency decomposition
The evoked power of the theta frequency band was quantified in the
S-cluster at 5.5 Hz for electrode P7. In the C-cluster, it was quantified at
6 Hz at electrode Fz. The choice of frequencies and electrodes was
based on TF plots and scalp topographies.
Concerning the power of theta oscillations (5.5 Hz) at electrode P7,
the ANOVA revealed a main effect of “intoxication status” (F(1,33) =
11.32; p = .002; ηp2= .255), showing that theta power was larger

103
N. Zink et al.
Fig. 4. Illustration of the N2 ERP in the RIDE C-cluster at electrode Cz. Time
point zero denotes the onset of the cue stimulus. The target stimulus was added
to the visual array at 100 ms. ERPs assessed while intoxicated are denoted in
red line color, ERPs assessed while sober are denoted in green line color. The
BASE condition is denoted in a lighter color than the BI condition. Topographic
plots of the N2 within the time interval used for ERP quantification are pro-
vided for the intoxicated and sober state. No statistically significant effects were
found.
Fig. 5. Illustration of the P3 ERP in the RIDE C-cluster at electrode Pz. Time
point zero denotes the onset of the cue stimulus. The target stimulus was added
to the visual array at 100 ms. ERPs assessed while intoxicated are denoted in
red line color, ERPs assessed while sober are denoted in green line color. The
BASE condition is denoted in a lighter color than the BI condition. Topographic
plots of the P3 within the time interval used for ERP quantification are provided
for the intoxicated and sober state. No statistically significant effects were
found.
during the sober appointment (124.9 μV2 ± 17.0) than during the in-
toxication appointment (70.4 μV2 ± 9.2). Moreover, a main effect of
“drinking sequence” was detected (F(1,33) = 10.24; p = .003; ηp2 =
.237), showing that participants who were sober during their first ap-
pointment had on average larger theta power (132.9 μV2 ± 16.3) than
those who were intoxicated during their first appointment (62.3
μV2 ± 14.9). Most importantly, an interaction of “intoxication status x
drinking sequence x experimental condition” (F(1,33) = 5.81; p =
.022; ηp2 = .150) was detected. Post hoc analyses showed an interac-
tion of “drinking sequence x experimental condition” for the intoxica-
tion appointment (F(1,33) = 10.05; p = .003; ηp2 = .234), but not for
the sober appointment (F(1,33) =.04; p = .839; ηp2 = .001): Only
during acute ethanol intoxication, theta power in the BI condition was
significantly larger in the participants who were intoxicated in their
second appointment (102.2 μV2 ± 19.2) than in those who were in-
toxicated in their first appointment (39.0 μV2 ± 8.3) (t(33) = 3.20; p
= .003). This effect could not be found in BASE condition (t(33) =
1.78; p = .084). No other significant main effects or interactions were
found (all F ≤ 2.76; p ≥ .106). The posterior theta power is illustrated
in Fig. 6. Interestingly, add-on correlation analyses showed that in the
entire sample, the BI effect size (BI minus BASE) of hit RTs and parietal
104
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
theta assessed at electrode P7 was significantly negatively correlated in
the intoxicated appointment (r = -0.534; p < .001), but not in the
sober appointment (r = -0.034; p = .834). Further matching the be-
havioral data, which had only demonstrated a significant effect of in-
toxication in the subgroup who were intoxicated during their first ap-
pointment, we further found that the intoxication-related BI effects in
RTs and parietal theta were only significantly correlated in this group (r
= -0.492; p = .020), but not in the group that was intoxicated during
their second appointment (r = -0.425: p = 0.78).
As the TF plots shown in Fig. 6 also appeared to show differences in
the alpha frequency band, we ran an add-on analysis of alpha oscilla-
tions (9.0 Hz) at electrode P7. Like for the theta frequency, the ANOVA
revealed a main effect of “intoxication status” (F(1,33) = 8.593; p =
.006; ηp2= .207), showing that alpha power was larger during the sober
appointment (126.8 μV2 ± 20.8) than during the intoxication ap-
pointment (84.6 μV2 ± 15.0). Despite the descriptive differences in
Fig. 6, we found no other significant main or interaction effects (all F ≤
3.22; p ≥ .082).
Concerning the power of theta oscillations (6 Hz) at electrode Fz, a
main effect of “intoxication status” (F(1,33) = 6.42; p = .016; ηp2 =
.163) revealed that theta power was larger during acute ethanol in-
toxication (48.7 μV2 ± 17.3) than during the sober appointment (9.7
μV2 ± 4.0). This is illustrated in Fig. 7. No other significant effects were
detected (all F ≤ 3.97; p ≥ .055).
4. Discussion
In the current study, we investigated the effects of an acute, binge-
like high-dose alcohol intoxication on cognitive (i.e. non-motor) aspects
of inhibition during cognitive flexibility. To do so, we assessed the BI
effect in a task switching paradigm in a counterbalanced within-subject
design, where each participant was tested once sober and once in-
toxicated. We furthermore assessed whether the induction or increase
of automaticity via prior practice (i.e. having performed the ~30 min
task once before) can attenuate or maybe even prevent the potential
detrimental effects of binge drinking. For this, we compared effects of
alcohol intoxication on the size of the BI effect between participants
who were able to use well-practiced and thus rather automatized S-R
associations when intoxicated (during their second appointment) and
participants who did not have comparable automatisms when in-
toxicated (due to performing the task while intoxicated during their
first appointment).
As expected, we found that that alcohol intoxication led to generally
impaired task performance in our BI task, as reflected by decreased
response accuracy and increased RTs. This is well in line with many
studies showing executive deficits during binge-like alcohol intoxica-
tion (e.g. Field et al., 2010; Stock, 2017; Stock et al., 2016b, 2014;
Wolff et al., 2018b). As all participants conducted their practice block
of the BI task while sober (also in the intoxication session), the reduced
behavioral performance in the intoxication session is unlikely to be a
result of potential task comprehension problems. Most importantly,
however, the size of the BI effect (BI-BASE) on RTs was larger during
the intoxicated state than during the sober state, but only when parti-
cipants could not rely on automated stimulus-response mapping pro-
cesses (i.e. when they were intoxicated during their first of two ap-
pointments). This allows for the conclusion that participants, for whom
rather automated S-R mappings were not available due to not having
performed the task once before, seem to have experienced difficulties in
reactivating the currently suppressed task rules when intoxicated. In
contrast to this, participants who had the opportunity to once perform
the task for ~30 min in a sober state before performing it intoxicated,
did not show a significant increase in BI size during intoxication. In-
stead, the size of their BI effect was comparable across the sober and
intoxicated states. This apparent lack of detrimental intoxication effects
after ~30 minutes of sober task performance could be explained by the
fact that this training promotes rather automated strategies/S-R
N. Zink et al.
mappings, which generally seem to be much less vulnerable towards
the detrimental effects of alcohol than conscious top-down control (Day
et al., 2015; Garbusow et al., 2014; Stock et al., 2016a, 2017b). It
therefore seems likely that participants in the “sober first” group could
maintain good behavioral performance (i.e. a comparatively small BI
effects) during intoxication because they could rely on rather auto-
mated S-R mappings during the second appointment and thus com-
pensate the alcohol-induced impairments of attention and top-down
cognitive control. Relying on automated information like highly trained
S-R mappings may furthermore reduce the strain on the working
memory capacity of participants, which could have helped to coun-
teract the alcohol-induced BI increase and might explain the unaffected
performances of participants who performed the entire task in a sober
105
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
Fig. 6. Illustration of the time-frequency decom-
position (total evoked power) of the RIDE S-cluster at
electrode P7. Time point zero denotes the onset of
the cue stimulus on the x-axis in ms. The y-axis de-
notes the frequency band in Hz. The target stimulus
was added to the visual array at 100 ms. We quan-
tified parietal theta power at 5.5 Hz within the time
window of 180-240 ms after cue onset and found
that while intoxicated (top row), the group that was
intoxicated during their second appointment (right
column) showed significantly more theta power in
the BI condition than the group that was intoxicated
during their first appointment (left column). This
group difference was not found during the sober
appointment (bottom row), or in the BASE condition
(please see supplementary material for corre-
sponding plots). On the right side of the graph, to-
pographic maps are depicted for the quantified par-
ietal theta power. As the visual inspection of the TF
plot also seemed to show differences in the alpha
band upon visual inspection, we additionally ran an
add-on quantification of the alpha frequency band at
9 Hz in the time window from 160 to 200 ms. The
analyses revealed an alcohol-induced reduction in
alpha power, but no other effects.
state before attempting to do so while intoxicated. Further supporting
this assumption, it has been demonstrated that high-dose ethanol in-
toxication affects task switching performance only when there is a
strain on working memory that effectively reduces the working memory
capacities available for task switching (Wolff et al., 2018b).
To identify cognitive-neurophysiological sub-processes underlying
the alcohol-induced modulation of the BI effect, we analyzed both ERPs
and theta band oscillations in RIDE-decomposed data. The decom-
position of the ERP segments into an S- and a C-cluster allows to better
dissociate stimulus-related processes, which we expected to display
task- and alcohol-related modulations, from central response selection
processes, which we expected to be rather unaffected by either alcohol,
or task conditions. In addition to this, RIDE helps to deal with the
Fig. 7. Illustration of the time-frequency decom-
position (total evoked power) of the RIDE C-cluster
at electrode Fz. Time point zero denotes the onset of
the cue stimulus on the x-axis in ms. The y-axis de-
notes the frequency band in Hz. The target stimulus
was added to the visual array at 100 ms. We quan-
tified fronto-central theta power at 6 Hz within the
time window of 800-900 ms after cue onset and
found that alcohol significantly increased fronto-
central theta power, but no other significant effects.
N. Zink et al.
alcohol-increased variability in hit RTs.
In line with previous neurophysiological studies on the BI effect
(Zhang et al., 2017, 2016a, 2016b), we found condition differences (BI
vs. BASE) in the P1 and N1 ERPs of the S-Cluster, but not in the N2 or
parietal P3 quantified in the C-cluster. This further underlines the as-
sumption that the BI effect is driven by early attentional stimulus
processing and attentional selection (Luck et al., 2000; Zhang et al.,
2016c). Also matching our hypotheses that alcohol effects should be
limited to the early attentional processing stages, we found that alcohol
decreased attentional processing of the task set-relevant visual cue, as
reflected by smaller cue-P1 and cue-N1 amplitudes, which reflect de-
creased perceptual gating (Herrmann and Knight, 2001; Luck et al.,
2000) and impaired attentional selection mechanisms (Herrmann and
Knight, 2001; Luck et al., 2000), respectively. This finding is in line
with other studies showing that alcohol may decrease attentional pro-
cessing as reflected by smaller amplitudes in one or both ERPs (Stock
et al., 2017a; Wolff et al., 2018b) and furthermore impair the top-down
allocation of attention (Stock et al., 2017b), as reflected by the P1 and
N1 ERPs (Luck et al., 2000). It may therefore be assumed that decreased
attentional processing of the cue should have contributed to the overall
performance impairment observed during intoxication. This impair-
ment in the early perceptual processing of the very stimulus which
indicated the task set in effect, could have contributed to the larger
need for reactive control at the later stage of response selection and
control, as reflected by the larger mid-frontal theta power (Cavanagh
and Frank, 2014) found at 6 Hz in the C-Cluster. In accordance with
previous studies, the decreased P1 amplitude and increased frontal
theta power in the intoxicated state may reflect impaired early per-
ceptual processes and increased demand for late response control (Stock
et al., 2016b, 2017b).
Yet still, neither the assessed ERPs, nor the alcohol-induced increase
in mid-frontal theta power or the parietal alpha power could explain the
practice effects found in the behavioral data. The posterior theta power
quantified in the S-cluster was the only neurophysiological correlate
which matched the finding that participants who lacked training (as
they did not perform the task in the sober state before) and displayed a
larger behavioral BI effect when intoxicated. As already noted in the
introduction, the theta power quantified over posterior electrodes re-
flects stimulus-related, proactive aspects of cognitive control (Cooper
et al., 2017; Freunberger et al., 2007; Gladwin and de Jong, 2005;
Sauseng et al., 2006), which is known to be related to task switching
(Freunberger et al., 2007; Gladwin and de Jong, 2005) and to reflect
top-down regulation processes in recalling the task set memory
(Freunberger et al., 2007; Gladwin and de Jong, 2005).
Opposite to the theta power effects observed at mid-frontal sites,
alcohol generally decreased posterior theta power. Given that parietal
theta is typically enhanced when a previously irrelevant visual task
must be reactivated (Freunberger et al., 2007; Gladwin and de Jong,
2005) and likely improves task switching performance via anticipatory
task-goal updating (Cooper et al., 2017), it may be deduced that smaller
posterior theta reflects alcohol-induced impairments in working
memory capacity and task set-relevant memory retrieval (Freunberger
et al., 2007). This also is in line with previous findings showing that
alcohol intoxication impairs top-down control processes (Day et al.,
2015; Garbusow et al., 2014; Stock et al., 2016a, 2017b). Further
matching this, we found that the behavioral and theta-based BI effects
were only correlated during the intoxicated appointment, but not
during the sober appointment. This effect was significant in the group
that was intoxicated during their first appointment, but only showed a
trend in the group that was intoxicated during the second appointment.
Against this background, the fact that we observed relatively larger
posterior theta power in participants with more practice/automatized
S-R mappings during the intoxicated appointment in the more chal-
lenging BI condition, suggests that the higher degree of automatization
might have left the more experienced group with more (residual) top-
down resources to invest in task set reactivation: Considering that
106
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
automaticity is supposed to reduce the strain on the residual top-down
cognitive control capacities of intoxicated participants, it might allow
for better recall of task memory/task rules, as reflected by larger pos-
terior theta power (Gladwin and de Jong, 2005). Those findings are
furthermore in line with the assumption that properties of parietal theta
oscillations ~200 ms after stimulus presentation reflect top-down reg-
ulation processes (Freunberger et al., 2007). Based thereon, we assume
that the larger theta power of the well-practiced subjects indicates that
they were able to invest more cognitive resources in overcoming the
inhibition of previous tasks, while intoxicated. This may have fa-
cilitated the recall of previously inhibited task rules and advance the
preparation for the encoding of an upcoming target stimulus in BI
condition in intoxication session. This interpretation is corroborated by
previous studies which showed that left parietal theta during pre-
paratory periods prior to actual task-relevant stimulus processing is
important for the encoding of a stimulus (Federmeier and Kutas, 2002;
Tesche and Karhu, 2000), particularly when participants engage in
controlled task processing modes (Phillips et al., 2014).
Lastly, there are a few limitations to discuss: Due to concerns of the
local ethics commission, we could not include female participants in the
study. While we failed to find reports on sex effects on backward in-
hibition, further studies will be needed to assess whether the cognitive
effects of an acute binge-like intoxication are comparable across young
men and women. It would furthermore have been beneficial to include
control groups of individuals, who either perform the task twice while
sober (i.e. sober – sober), or intoxicated (intox – intox) in order to
further explore the nature and size of the observed learning/auto-
matization effect. Also, our paradigm did not have enough trials to
reliably analyze differences in the kind and effect of errors that parti-
cipants committed (i.e. the comparison of III, IIC, ICI, CII, CCI, CIC, ICC
triplets, where C = correct response and I = incorrect response) for the
two experimental conditions and intoxication states.
5. Conclusions
Alcohol generally impaired behavioral task performance in our task-
switching paradigm. This was most likely due to impaired attentional
processing of the task-relevant cue and resulted in a larger need for
reactive control at the later stage of response selection and control in
our switching task. However, we observed that prior task practice (i.e.
first performing the task while sober) attenuated the detrimental effects
of alcohol on cognitive inhibition during task switching: Without those
~30 minutes of prior sober practice, alcohol significantly impaired
cognitive flexibility, and the intoxicated participants struggled to
overcome the cognitive inhibition of a previously relevant task set, as
reflected by a larger BI effect. This phenomenon was linked to reduced
posterior theta power, which likely reflects alcohol-induced impair-
ments in working memory capacity and task set-relevant memory re-
trieval. As individuals with prior practice (i.e. who had first performed
the task sober) did not show the same alcohol-related deficit, it may be
deduced that (partial) task set automatization via S-R associations may
help to reduce the detrimental effects of alcohol on cognitive inhibition
during task switching. The most likely mechanism underlying this effect
is a shift from top-down attentional processing to more automatized S-R
mappings, which that are known to be much less affected by alcohol
than controlled processes. This likely left the well-trained individuals
with more residual cognitive control resources for overcoming the in-
hibition of previous tasks.
Author contributions
AS, NZ and CB were responsible for the study concept and design.
AS, NZ, and WC collected the data. AS, NZ and RZ analyzed the data.
All authors contributed to the manuscript and approved the final ver-
sion for publication.
N. Zink et al.
Conflicts of interest
None.
Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft
(DFG) [SFB 940 project B8].
Role of the funding source
The sponsor had no role in study design; in the collection, analysis
and interpretation of data; in the writing of the report; and in the de-
cision to submit the paper for publication.
Ethical statement
All participants gave written informed consent and were treated in
accordance with the declaration of Helsinki. The study was approved by
the local ethics committee of the TU Dresden, Germany.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.pnpbp.2018.08.034.
References
Aragues, M., Jurado, R., Quinto, R., Rubio, G., 2011. Laboratory paradigms of impulsivity
and alcohol dependence: a review. Eur. Addict. Res. 17, 64–71. https://doi.org/10.
1159/000321345.
Babor, T.F., Higgins-Biddle, J.C., Saunders, J.B., Monteiro, M.G., 2001. The Alcohol Use
Disorders Identification Test Guidelines for Use in Primary Care. Second Edition.
Balodis, I.M., Johnsrude, I.S., Olmstead, M.C., 2007. Intact preference conditioning in
acute intoxication despite deficient declarative knowledge and working memory.
Alcohol. Clin. Exp. Res. 31, 1800–1810. https://doi.org/10.1111/j.1530-0277.2007.
00482.x.
Beaton, L.E., Azma, S., Marinkovic, K., 2018. When the brain changes its mind:
Oscillatory dynamics of conflict processing and response switching in a flanker task
during alcohol challenge. PloS One 13, e0191200. https://doi.org/10.1371/journal.
pone.0191200.
Beste, C., Ness, V., Falkenstein, M., Saft, C., 2011. On the role of fronto-striatal neural
synchronization processes for response inhibition–evidence from ERP phase-syn-
chronization analyses in pre-manifest Huntington’s disease gene mutation carriers.
Neuropsychologia 49, 3484–3493. https://doi.org/10.1016/j.neuropsychologia.
2011.08.024.
Bluschke, A., Chmielewski, W.X., Mückschel, M., Roessner, V., Beste, C., 2017. Neuronal
Intra-Individual Variability Masks Response Selection Differences between ADHD
Subtypes-A Need to Change Perspectives. Front. Hum. Neurosci. 11 (329). https://
doi.org/10.3389/fnhum.2017.00329.
Botvinick, M.M., Braver, T.S., Barch, D.M., Carter, C.S., Cohen, J.D., 2001. Conflict
monitoring and cognitive control. Psychol. Rev. 108, 624–652.
Cavanagh, J.F., Frank, M.J., 2014. Frontal theta as a mechanism for cognitive control.
Trends Cogn. Sci. 18, 414–421. https://doi.org/10.1016/j.tics.2014.04.012.
Cooper, P.S., Wong, A.S.W., McKewen, M., Michie, P.T., Karayanidis, F., 2017.
Frontoparietal theta oscillations during proactive control are associated with goal-
updating and reduced behavioral variability. Biol. Psychol. 129, 253–264. https://
doi.org/10.1016/j.biopsycho.2017.09.008.
Dajani, D.R., Uddin, L.Q., 2015. Demystifying cognitive flexibility: Implications for
clinical and developmental neuroscience. Trends Neurosci. 38, 571–578. https://doi.
org/10.1016/j.tins.2015.07.003.
Day, A.M., Kahler, C.W., Ahern, D.C., Clark, U.S., 2015. Executive Functioning in Alcohol
Use Studies: A Brief Review of Findings and Challenges in Assessment. Curr. Drug
Abuse Rev. 8, 26–40.
Depue, B.E., Ketz, N., Mollison, M.V., Nyhus, E., Banich, M.T., Curran, T., 2013. ERPs and
neural oscillations during volitional suppression of memory retrieval. J. Cogn.
Neurosci. 25, 1624–1633. https://doi.org/10.1162/jocn_a_00418.
Diamond, A., 2013. Executive functions. Annu. Rev. Psychol. 64, 135–168. https://doi.
org/10.1146/annurev-psych-113011-143750.
Dippel, G., Chmielewski, W., Mückschel, M., Beste, C., 2016. Response mode-dependent
differences in neurofunctional networks during response inhibition: an EEG-beam-
forming study. Brain Struct. Funct. 221, 4091–4101. https://doi.org/10.1007/
s00429-015-1148-y.
Everitt, B.J., Robbins, T.W., 2005. Neural systems of reinforcement for drug addiction:
from actions to habits to compulsion. Nat. Neurosci. 8, 1481–1489. https://doi.org/
10.1038/nn1579.
Federmeier, K.D., Kutas, M., 2002. Picture the difference: electrophysiological
107
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
investigations of picture processing in the two cerebral hemispheres.
Neuropsychologia 40, 730–747.
Field, M., Wiers, R.W., Christiansen, P., Fillmore, M.T., Verster, J.C., 2010. Acute alcohol
effects on inhibitory control and implicit cognition: implications for loss of control
over drinking. Alcohol. Clin. Exp. Res. 34, 1346–1352. https://doi.org/10.1111/j.
1530-0277.2010.01218.x.
Freunberger, R., Klimesch, W., Doppelmayr, M., Höller, Y., 2007. Visual P2 component is
related to theta phase-locking. Neurosci. Lett. 426, 181–186. https://doi.org/10.
1016/j.neulet.2007.08.062.
Friedman, R.S., McCarthy, D.M., Bartholow, B.D., Hicks, J.A., 2007. Interactive effects of
alcohol outcome expectancies and alcohol cues on nonconsumptive behavior. Exp.
Clin. Psychopharmacol. 15, 102–114. https://doi.org/10.1037/1064-1297.15.1.102.
Friedrich, J., Mückschel, M., Beste, C., 2018. Specific properties of the SI and SII soma-
tosensory areas and their effects on motor control: a system neurophysiological study.
Brain Struct. Funct. 223, 687–699. https://doi.org/10.1007/s00429-017-1515-y.
Garbusow, M., Sebold, M., Beck, A., Heinz, A., 2014. Too Difficult to Stop: Mechanisms
Facilitating Relapse in Alcohol Dependence. Neuropsychobiology 70, 103–110.
https://doi.org/10.1159/000362838.
Gladwin, T.E., de Jong, R., 2005. Bursts of occipital theta and alpha amplitude preceding
alternation and repetition trials in a task-switching experiment. Biol. Psychol. 68,
309–329. https://doi.org/10.1016/j.biopsycho.2004.06.004.
Herrmann, C.S., Knight, R.T., 2001. Mechanisms of human attention: event-related po-
tentials and oscillations. Neurosci. Biobehav. Rev. 25, 465–476.
Keil, A., Debener, S., Gratton, G., Junghöfer, M., Kappenman, E.S., Luck, S.J., Luu, P.,
Miller, G.A., Yee, C.M., 2014. Committee report: publication guidelines and re-
commendations for studies using electroencephalography and magnetoencephalo-
graphy. Psychophysiology 51, 1–21. https://doi.org/10.1111/psyp.12147.
Knight, J.R., Wechsler, H., Kuo, M., Seibring, M., Weitzman, E.R., Schuckit, M.A., 2002.
Alcohol abuse and dependence among U.S. college students. J. Stud. Alcohol 63,
263–270.
Koch, I., Gade, M., Philipp, A.M., 2004. Inhibition of Response Mode in Task Switching.
Exp. Psychol. 51, 52–58. https://doi.org/10.1027/1617-3169.51.1.52.
Larson, M.J., Clayson, P.E., Clawson, A., 2014. Making sense of all the conflict: a theo-
retical review and critique of conflict-related ERPs. Int. J. Psychophysiol. Off. J. Int.
Organ. Psychophysiol. 93, 283–297. https://doi.org/10.1016/j.ijpsycho.2014.06.
007.
Lister, R.G., Gorenstein, C., Fisher-Flowers, D., Weingartner, H.J., Eckardt, M.J., 1991.
Dissociation of the acute effects of alcohol on implicit and explicit memory processes.
Neuropsychologia 29, 1205–1212.
Luck, S.J., Woodman, G.F., Vogel, E.K., 2000. Event-related potential studies of attention.
Trends Cogn. Sci. 4, 432–440. https://doi.org/10.1016/S1364-6613(00)01545-X.
Mayr, U., Keele, S.W., 2000. Changing internal constraints on action: the role of backward
inhibition. J. Exp. Psychol. Gen. 129, 4–26.
Montgomery, C., Fisk, J.E., Murphy, P.N., Ryland, I., Hilton, J., 2012. The effects of heavy
social drinking on executive function: a systematic review and meta-analytic study of
existing literature and new empirical findings. Hum. Psychopharmacol. Clin. Exp. 27,
187–199. https://doi.org/10.1002/hup.1268.
Mückschel, M., Stock, A.-K., Beste, C., 2014. Psychophysiological mechanisms of inter-
individual differences in goal activation modes during action cascading. Cereb.
Cortex N. Y. N 1991 (24), 2120–2129. https://doi.org/10.1093/cercor/bht066.
Mückschel, M., Dippel, G., Beste, C., 2017. Distinguishing stimulus and response codes in
theta oscillations in prefrontal areas during inhibitory control of automated re-
sponses. Hum. Brain Mapp. 38, 5681–5690. https://doi.org/10.1002/hbm.23757.
Nunez, P.L., Pilgreen, K.L., 1991. The spline-Laplacian in clinical neurophysiology: a
method to improve EEG spatial resolution. J. Clin. Neurophysiol. Off. Publ. Am.
Electroencephalogr. Soc. 8, 397–413.
Ouyang, G., Herzmann, G., Zhou, C., Sommer, W., 2011. Residue iteration decomposition
(RIDE): A new method to separate ERP components on the basis of latency variability
in single trials. Psychophysiology 48, 1631–1647. https://doi.org/10.1111/j.1469-
8986.2011.01269.x.
Ouyang, G., Sommer, W., Zhou, C., 2015a. Updating and validating a new framework for
restoring and analyzing latency-variable ERP components from single trials with
residue iteration decomposition (RIDE). Psychophysiology 52, 839–856. https://doi.
org/10.1111/psyp.12411.
Ouyang, G., Sommer, W., Zhou, C., 2015b. A toolbox for residue iteration decomposition
(RIDE)–A method for the decomposition, reconstruction, and single trial analysis of
event related potentials. J. Neurosci. Methods 250, 7–21. https://doi.org/10.1016/j.
jneumeth.2014.10.009.
Perrin, F., Pernier, J., Bertrand, O., Echallier, J.F., 1989. Spherical splines for scalp po-
tential and current density mapping. Electroencephalogr. Clin. Neurophysiol. 72,
184–187.
Phillips, J.M., Vinck, M., Everling, S., Womelsdorf, T., 2014. A long-range fronto-parietal
5- to 10-Hz network predicts “top-down” controlled guidance in a task-switch para-
digm. Cereb. Cortex N. Y. N 1991 (24), 1996–2008. https://doi.org/10.1093/cercor/
bht050.
Polich, J., 2007. Updating P300: an integrative theory of P3a and P3b. Clin.
Neurophysiol. Off. J. Int. Fed. Clin. Neurophysiol. 118, 2128–2148. https://doi.org/
10.1016/j.clinph.2007.04.019.
Rohrbaugh, J.W., Stapleton, J.M., Parasuraman, R., Zubovic, E.A., Frowein, H.W., Varner,
J.L., Adinoff, B., Lane, E.A., Eckardt, M.J., Linnoila, M., 1987. Dose-related effects of
ethanol on visual sustained attention and event-related potentials. Alcohol Fayettev.
N 4, 293–300.
Sabah, K., Dolk, T., Meiran, N., Dreisbach, G., 2018. When less is more: costs and benefits
of varied vs. fixed content and structure in short-term task switching training.
Psychol. Res. https://doi.org/10.1007/s00426-018-1006-7.
Sauseng, P., Klimesch, W., Freunberger, R., Pecherstorfer, T., Hanslmayr, S., Doppelmayr,
N. Zink et al.
M., 2006. Relevance of EEG alpha and theta oscillations during task switching. Exp.
Brain Res. 170, 295–301. https://doi.org/10.1007/s00221-005-0211-y.
Sinai, M., Goffaux, P., Phillips, N.A., 2007. Cue- versus response-locked processes in
backward inhibition: Evidence from ERPs. Psychophysiology 44, 596–609. https://
doi.org/10.1111/j.1469-8986.2007.00527.x.
Smith, J.L., Mattick, R.P., Jamadar, S.D., Iredale, J.M., 2014. Deficits in behavioural in-
hibition in substance abuse and addiction: a meta-analysis. Drug Alcohol Depend.
145, 1–33. https://doi.org/10.1016/j.drugalcdep.2014.08.009.
Stalnaker, T.A., Takahashi, Y., Roesch, M.R., Schoenbaum, G., 2009. Neural substrates of
cognitive inflexibility after chronic cocaine exposure. Neuropharmacology 56 (Suppl.
1), 63–72. https://doi.org/10.1016/j.neuropharm.2008.07.019.
Stock, A.-K., 2017. Barking up the wrong tree: why and how we may need to revise
alcohol addiction therapy. Front. Psychol. 8 (884). https://doi.org/10.3389/fpsyg.
2017.00884.
Stock, A.-K., Beste, C., 2014. Binge drinking and the differential influence of ethanol on
cognitive control subprocesses: a novel field of neurotoxicology. Arch. Toxicol. 88,
9–10. https://doi.org/10.1007/s00204-013-1172-8.
Stock, A.-K., Blaszkewicz, M., Beste, C., 2014. Effects of binge drinking on action cas-
cading processes: an EEG study. Arch. Toxicol. 88, 475–488. https://doi.org/10.
1007/s00204-013-1109-2.
Stock, A.-K., Riegler, L., Chmielewski, W.X., Beste, C., 2016a. Paradox effects of binge
drinking on response inhibition processes depending on mental workload. Arch.
Toxicol. 90, 1429–1436. https://doi.org/10.1007/s00204-015-1565-y.
Stock, A.-K., Schulz, T., Lenhardt, M., Blaszkewicz, M., Beste, C., 2016b. High-dose al-
cohol intoxication differentially modulates cognitive subprocesses involved in re-
sponse inhibition. Addict. Biol. 21, 136–145. https://doi.org/10.1111/adb.12170.
Stock, A.-K., Hoffmann, S., Beste, C., 2017a. Effects of binge drinking and hangover on
response selection sub-processes-a study using EEG and drift diffusion modeling.
Addict. Biol. 22, 1355–1365. https://doi.org/10.1111/adb.12412.
Stock, A.-K., Wolff, N., Beste, C., 2017b. Opposite effects of binge drinking on consciously
vs. subliminally induced cognitive conflicts. NeuroImage 162, 117–126. https://doi.
org/10.1016/j.neuroimage.2017.08.066.
Tallon-Baudry, C., Bertrand, O., Delpuech, C., Pernier, J., 1997. Oscillatory γ-band (30–70
Hz) activity induced by a visual search task in humans. J. Neurosci. 17, 722–734.
Tenke, C.E., Kayser, J., 2012. Generator localization by current source density (CSD):
implications of volume conduction and field closure at intracranial and scalp
108
Progress in Neuropsychopharmacology & Biological Psychiatry 89 (2019) 97–108
resolutions. Clin. Neurophysiol. Off. J. Int. Fed. Clin. Neurophysiol. 123, 2328–2345.
https://doi.org/10.1016/j.clinph.2012.06.005.
Tesche, C.D., Karhu, J., 2000. Theta oscillations index human hippocampal activation
during a working memory task. Proc. Natl. Acad. Sci. U. S. A. 97, 919–924.
Verleger, R., Jaśkowski, P., Wascher, E., 2005. Evidence for an integrative role of P3b in
linking reaction to perception. J. Psychophysiol. 19, 165–181.
Verleger, R., Metzner, M.F., Ouyang, G., Śmigasiewicz, K., Zhou, C., 2014. Testing the
stimulus-to-response bridging function of the oddball-P3 by delayed response signals
and residue iteration decomposition (RIDE). NeuroImage 100, 271–280. https://doi.
org/10.1016/j.neuroimage.2014.06.036.
Watson, P.E., Watson, I.D., Batt, R.D., 1980. Total body water volumes for adult males
and females estimated from simple anthropometric measurements. Am. J. Clin. Nutr.
33, 27–39.
Widmark, E.M.P., 1932. Die theoretischen Grundlagen und die praktische
Verwendbarkeit der gerichtlich-medizinischen Alkoholbestimmung. Urban und
Schwarzenberg, Berlin.
Wolff, N., Mückschel, M., Beste, C., 2017. Neural mechanisms and functional neuroana-
tomical networks during memory and cue-based task switching as revealed by residue
iteration decomposition (RIDE) based source localization. Brain Struct. Funct. 222,
3819–3831. https://doi.org/10.1007/s00429-017-1437-8.
Wolff, N., Giller, F., Buse, J., Roessner, V., Beste, C., 2018a. When repetitive mental sets
increase cognitive flexibility in adolescent obsessive-compulsive disorder. J. Child
Psychol. Psychiatry. https://doi.org/10.1111/jcpp.12901.
Wolff, N., Gussek, P., Stock, A.-K., Beste, C., 2018b. Effects of high-dose ethanol in-
toxication and hangover on cognitive flexibility. Addict. Biol. 23, 503–514. https://
doi.org/10.1111/adb.12470.
Zhang, R., Stock, A.-K., Beste, C., 2016a. The neurophysiological basis of reward effects
on backward inhibition processes. NeuroImage 142, 163–171. https://doi.org/10.
1016/j.neuroimage.2016.05.080.
Zhang, R., Stock, A.-K., Fischer, R., Beste, C., 2016b. The system neurophysiological basis
of backward inhibition. Brain Struct. Funct. 221, 4575–4587. https://doi.org/10.
1007/s00429-016-1186-0.
Zhang, R., Stock, A.-K., Rzepus, A., Beste, C., 2017. Self-Regulatory Capacities Are
Depleted in a Domain-Specific Manner. Front. Syst. Neurosci. 11. https://doi.org/10.
3389/fnsys.2017.00070.