Electrophysiological responses to appetitive and aversive outcomes: A comparison of college drinkers and non-drinkers

Journal Pre-proof Electrophysiological responses to appetitive and aversive outcomes: A comparison of college drinkers and non-drinkers Heather E. Soder, Robert Suchting, Geoffrey F. Potts

PII:

S0304-3940(19)30652-4

DOI:

https://doi.org/10.1016/j.neulet.2019.134549

Reference:

NSL 134549

To appear in:

Neuroscience Letters

Received Date:

30 July 2019

Revised Date:

20 September 2019

Accepted Date:

9 October 2019

Please cite this article as: Soder HE, Suchting R, Potts GF, Electrophysiological responses to appetitive and aversive outcomes: A comparison of college drinkers and non-drinkers, Neuroscience Letters (2019), doi: https://doi.org/10.1016/j.neulet.2019.134549

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.

1 Electrophysiological responses to appetitive and aversive outcomes: A comparison of college drinkers and non-drinkers

Heather E. Sodera,b [email protected], Robert Suchtingb, Geoffrey F. Pottsa

a

of

University of South Florida, Department of Psychology, PCD 4118G, 4202 E. Fowler Ave, Tampa, FL 33620 b

ro

University of Texas Health Science Center at Houston, Department of Psychiatry, 1941 East Rd, BBS Building, 1st floor, Houston, TX 77054

Present address: 1941 East Rd. BBS Building, 1st floor, Houston TX 77054

re

1

-p

Corresponding author: Heather E. Soder1, University of South Florida, Department of Psychology, PCD 4118G, 4202 E. Fowler Ave, Tampa, FL 33620



Non-drinkers displayed a larger feedback-related negativity to aversive outcomes

ur na

compared to rewards 

lP

Highlights

Drinkers had a blunted feedback-related negativity to aversive outcomes compared to non-drinkers



College drinkers are less sensitive to aversive outcomes, which could contribute to

Jo

alcohol use

Abstract

The current study compared electrophysiological responses (the feedback-related negativity [FRN]) to appetitive and aversive outcomes between a group of college drinkers and non-drinkers. 50 undergraduate students completed a passive, slot machine-like task while their electroencephalographic data was recorded to extract the FRN to unexpected appetitive and aversive outcomes. In the appetitive condition, participants could expectedly or unexpectedly

2 win $1 or not win $1 and in the aversive condition participants could expectedly or unexpectedly be exposed to a loud noise burst or silence. The FRN was recorded in response to a cue indicating the outcome. Participants also reported on the number of drinks they consumed in a typical week to establish drinking status (drinker/non-drinker). Results showed that non-drinkers had a larger FRN in the aversive task compared to the appetitive task while drinkers had similar FRNs between the tasks. Drinkers had a significantly smaller aversive outcome related FRN compared to non-drinkers. Neural sensitivity to aversive outcomes might be a marker of decreased punishment sensitivity in college drinkers compared to non-drinkers, contributing to unhealthy drinking behavior.

ro

of

Keywords: feedback-related negativity; alcohol; motivation; rewards; punishments

Electrophysiological responses to appetitive and aversive outcomes: A comparison of college

-p

drinkers and non-drinkers

re

1. Introduction

Excessive alcohol use leads to negative health and social consequences [1]. Problematic

lP

alcohol use may be related to the ability to process and learn from these consequences. Identifying neuro-biological markers of sensitivity to negative outcomes can lead to better

ur na

assessment of risk and targeted intervention for alcohol related problems. Broadly, two motivational systems are involved in alcohol use: approach (seeking appetitive outcomes) and avoidance (avoiding aversive outcomes) [2,3]. A hyper-active approach system in part drives excessive alcohol use; cues related to alcohol become associated with the

Jo

rewarding effects of alcohol, leading to increased consumption [4]. A hypo-active avoidance system also related to alcohol consumption; decreased sensitivity to aversive outcomes is associated to increased alcohol-cue associations [3]. Both sensitivity to reward [5] and decreased punishment sensitivity are related to college-aged alcohol use [6]. However, alcohol is also used as a coping mechanism to deal with negative social situations, especially in college-aged drinkers

3 [7,8]. This unresolved complexity in the relationship between outcome processing and alcohol consumption necessitates further investigation. One way to assess the contributions of motivational systems to drinking is to measure neural responses to appetitive and aversive outcomes in controlled experimental designs. The feedback-related negativity (FRN) is an event-related potential (ERP) component that is elicited

of

by negative feedback, losing money on gambles, or unanticipated outcomes [9,10]1, occurs over the medial frontal electrodes around 250-350ms post feedback, and is largest when outcomes are

ro

unexpected [11]. The FRN is thought to reflect a signal from the midbrain to the anterior

cingulate cortex, an area that is a part of both appetitive and aversive networks [12]. The FRN

-p

was originally thought to code a dopaminergic reward prediction error [13], but recent evidence

re

suggests it might also respond to unexpected outcomes regardless of valence [14–17], consistent with its position in both approach and avoidance pathways. Commonly, ERP studies on the FRN

lP

have employed the withholding of an appetitive stimulus (i.e., loss of money), rather than the actual delivery of an aversive stimulus [18]. While withholding of an appetitive stimulus is

ur na

effective in eliciting ERPs, it may not provide a valid challenge to the aversive motivation system [19].

Despite the connection between appetitive/aversive motivation and drinking using selfreport, there are few studies linking the FRN to aversive outcomes to either alcohol use disorder

Jo

or frequency and quantity of alcohol use. The FRN to monetary loss is blunted in those with a family history of alcohol abuse [20], but the relationship between frequency of drinking and the

1

The FRN has recently been re-conceptualized, by some, as the reward positivity (RewP; Hajcak Proudfit, 2015), as there appears to be both a positive waveform to rewards and a negative waveform to non-rewards (the difference elicits a large positivity over the medial frontal electrodes). However, as our results indicate a positivity to both rewards and punishments, this terminology appears inappropriate. Note that it is always possible that the current results reflect N200/P300 contamination, as the stimuli are delivered with differing probabilities (Krigolson, 2018).

4 FRN is less clear. One study reported that an increased FRN was associated with an elevated number of drinking days in college drinkers, while another reported no difference between binge and non-binge drinkers [21,22]. Only one study has investigated neural feedback processing in those with an actual alcohol use disorder [23], but this sample had comorbid internalizing disorders. Beyond group comparisons, decreased amplitudes to winning feedback are associated

of

with increased alcohol hangover severity, suggesting alcohol has lasting effects on reward processing [24]. None of these studies administered aversive stimuli, and as such, the

ro

relationship between alcohol use and the FRN within the context of aversive stimuli remains unclear.

-p

The current study addresses two gaps in the current literature. First, most ERP studies

re

have used monetary outcomes to assess reward responsiveness (winning money) and punishment responsiveness (not winning money or losing money); i.e., receiving or not receiving an

lP

appetitive stimulus. An alternative method of assessing neural sensitivity to negative outcomes is to measure the response to the delivery of an aversive stimulus, rather than to the withholding of

ur na

an appetitive one [14]. Second, because neural response patterns observed in chronic users can indicate risk factors or consequences of repeated use [25], the current study selected a younger sample with lower consumption to assess a physiological index of a potentially geneticallymediated trait before any neuroadaptive changes to the brain [26,27].

Jo

The study had two main aims: 1) to examine the relationship between the FRN and

quantity and frequency of alcohol use in college students and 2) to compare this relationship between appetitive outcomes and aversive outcomes. Participants completed a task that either delivered or withheld appetitive outcomes ($1) and aversive outcomes (a loud noise burst). We hypothesized that drinkers would display an increased sensitivity to appetitive outcomes and/or a

5 decreased sensitivity to aversive outcomes compared with non-drinkers, consistent with the selfreport literature. Further, we expected that the neural response to aversive outcomes might be more related to drinking status, as aversive stimuli might activate underlying neural processes that are more closely related to real life alcohol-related consequences. 2. Method

of

2.1 Participants All procedures were approved by the institutional review board of the University of

ro

South Florida. Participants were English-speaking college students. Exclusionary criteria were current treatment or past hospitalization for a psychiatric disorder, current psychoactive

-p

medications, and hearing impairment or implanted hearing device that could block the ear canal.

re

A total of 58 participants consented and participated in the study. 50 of those participants were included in the final analysis. Two were excluded for excessive artifacts and 6 were excluded

lP

because they did not rate the white noise burst as unpleasant. Drinking status was determined by self-reported number of drinks in a typical week within the past year using the Daily Drinking

ur na

Questionnaire [8]. Number of drinks per week was calculated by summing the number of drinks reported for each day in a typical week. Those who reported zero drinks per week constituted the “non-drinkers” group (n = 32), while participants who reported greater than zero drinks per week constituted the “drinkers” group (n = 18).

Jo

2.2 Measures

2.2.1 Sample characteristics. Demographic variables (age, gender, year in college,

mother’s highest level of education, father’s highest level of education, and childhood household income) were all included to be considered as covariates. Participants also completed the Beck’s

6 Depression Inventory (BDI), State-Trait Anxiety Inventory (STAI), and Barratt Impulsiveness Scale (BIS-11) to assess current depression, anxiety, and impulsivity, respectively2 (Table 1). 2.2.2 Experimental tasks. Participants completed two tasks: the reward prediction task (RPT) and the punishment prediction task (PPT) to assess neural sensitivity to appetitive and aversive outcomes, respectively [11,17]. In short, two stimuli are presented sequentially followed

of

by an outcome screen. The first stimulus predicts the second stimulus with 80% accuracy (to establish a prediction) and the second stimulus informs predicts the outcome with 100%

ro

accuracy. 20% of the time participants receive unexpected outcomes, which are known to

maximally engage the reward system [28]. The stimuli are lemons and gold bars in the RPT and

-p

lightning bolts and clouds in the PPT. Trials are not inter-mixed and are presented in a blocked

re

format, alternating between RPT and PPT two times for a total of 4 blocks. Each block consisted of 240 trials. Gold bars in the second position return $1 with a cash noise, while lightning bolts

lP

return a loud (~98 dB) burst of white noise delivered during the outcome screen (see Supplement for information on the auditory device and example stimuli). Lemons return $0, while clouds

ur na

return silence. ERPs collected in response to the second stimulus can inform about the neural response to expectation violation to delivered or withheld rewards and punishments. Conditions in the RPT are: unpredicted rewards, unpredicted withheld rewards, predicted rewards, and predicted withheld rewards. Conditions in the PPT are: unpredicted punishments, unpredicted

Jo

withheld punishments, predicted punishments, and predicted withheld punishments. 2.4 Procedure

2

The BDI and STAI were added after the start of the study and thus, 8 participants were missing these scales. Any analysis including these questionnaires reflect this different sample size (non-drinkers = 28, drinkers = 14). Missing data was identical across drinking groups.

7 Eligible participants were invited to an in-person session. After consenting, participants were fitted for the EEG net and earphones and completed the two EEG tasks. The tasks were administered in 2 blocks each, counterbalanced (ABAB or BABA) across participants. A computerized survey including self-report data was also administered in a counterbalanced manner either before or after the EEG tasks. At the end of the experiment, participants were paid

of

a portion of their winnings from the RPT (~$10) and received extra credit. 2.5 Data Analysis

ro

2.5.1 EEG Acquisition/Analysis. EEG data was acquired with a 128-channel Geodesic Sensor Net (EGI, Eugene) at a sampling rate of 250 Hz, rereferenced to the vertex with 0.1-1 Hz

-p

analogue filtering and a 20 Hz low-pass digital filter offline. The EEG data were segmented 200

re

ms before and 800 ms after the onset of the second stimulus and sorted by condition. We screened for artifacts (eye blinks, movement, and bad electrodes) using EGI’s Netstation artifact

lP

detection tool. The mean number of trials left after artifact rejection per condition is presented in Supplemental Table 1. Cleaned segments were averaged within conditions, baseline corrected

ur na

(200 ms pre-stimulus period), and rereferenced to the average reference. Difference waves were created for each individual average to define the FRN [13]. As the neural response to rewards and punishments is largest when outcomes are unexpected, the following difference waves were created per group: unpredicted withheld rewards minus unpredicted rewards (from the RPT) and

Jo

unpredicted withheld punishments minus unpredicted punishments (from the PPT). This allowed the comparison of unexpected absence (getting nothing) and unexpected presence (getting the outcome) between the two types of outcomes (appetitive or aversive). For conditional waveform plots, see Supplementary Figures 3 and 4. In line with suggestions from the literature [29] and after visual inspection of the peak latency and scalp topographies (see Supplemental Figure 5),

8 the mean amplitude was extracted from 245-295 ms over a montage of medial frontal electrodes (see Figure 1). 2.5.2 Statistical strategy. Potential confounding variables were tested following recommendations in the literature [30,31]. Groups were tested for differences across theoretically important demographic and affective variables (e.g., age, gender, education, depression, anxiety,

of

and impulsivity) using independent samples t-tests or chi-square tests where appropriate; none of these variables differed between groups and therefore were not included as covariates in

ro

statistical modeling (Table 1).

Linear mixed effect (LME) modeling was used to examine the FRN difference waves on

-p

unexpected trials as a function of task (aversive vs. appetitive), drinking group (non-drinkers vs.

re

drinkers), and the interaction between task and group, with a random effect for participant id to account for correlated observations within participant. Planed simple effects models explored

lP

magnitude and direction of task differences within each drinking group as well as drinking group differences within each task. Normality of residuals was assessed graphically (e.g., histogram of

ur na

residuals, qq plot), finding only mild deviation from normality, potentially due to one high value of ERP amplitude. Sensitivity to this potential outlier (z = +3.4) was assessed via removal and reanalysis of the statistical model. This sensitivity analysis found no deviation from the inferences resulting from the models with the value in the model. As such, the value was retained

Jo

as a theoretically possible case that preserved the observed distribution. Follow-up analyses tested a continuous measure of drinking (typical number of drinks per week) to further explicate the relationship between task and drinking. Analyses were performed using packages nlme v. 3.1-139 [32] and lme4 v. 1.1-21 [33] in the R statistical computing environment v. 3.6.0 [34]. 3. Results

9 Drinking groups did not differ on any major demographic or affective variables, including impulsivity, depression, and anxiety (Table 1). On average, drinkers reported about 6 alcoholic drinks in a typical week (ranging from 1-14 drinks). These drinks were consumed primarily over the weekend, on average across 2 days. The primary linear mixed effects (LME) model of ERP amplitude found a significant

of

interaction between task and drinking group (p = 0.010; 16.8% marginal R2 for the fixed effects in the model [35]. Assumptions related to collinearity (i.e., VIF), linearity (i.e., evidence for

ro

nonlinear effects), and heteroscedasticity were satisfied. Table 2 provides details for each model parameter and Figure 2 provides a graphical depiction of this interaction. Within drinking group,

-p

non-drinkers experienced significantly lower ERP amplitude in the appetitive task than the

re

aversive task (b = -1.82, p = 0.006; marginal R2 = 10.1%), while drinkers showed no difference between tasks (b = 0.55, p = 0.242; marginal R2 = 2.6%). Within stimuli, the aversive task

lP

demonstrated lower ERP amplitude in drinkers compared to non-drinkers (b = -2.84, p < 0.001; marginal R2 = 24.5%), while the appetitive task was not different across drinking groups (b = -

ur na

0.46, p = 0.520; marginal R2 = 0.8%). Model assumptions for these follow-up tests of simple effects were similarly satisfied as noted above. Follow up analyses testing continuous drinking (typical number of drinks per week) in place of dichotomous non-drinkers versus drinkers found the same overall pattern, although the interaction was no longer statistically significant (p =

Jo

0.080). Detail regarding this analysis is provided in the Supplement. 4. Discussion

The present study found that non-drinkers were generally more sensitive to aversive

outcomes, as compared to appetitive outcomes. In contrast, drinkers were equally sensitive to both appetitive and aversive outcomes. Further, in comparison to non-drinkers, the drinkers had a

10 significantly smaller neural response to aversive outcomes, suggesting they were less sensitive to this specific type of negative outcome. This reduced sensitivity to aversive outcomes was related to increased alcohol use in this college-aged sample, potentially by an inability to learn from the negative consequences associated with drinking. These results can be understood within the context of the decision-making literature; people are generally more risk-averse than risk-seeking [36]. Losses tend to induce a larger “feel

of

worse” affective response than wins induce a “feel good” response, consistent with the larger

ro

FRN difference in the PPT compared to the RPT. The drinkers did not demonstrate this ‘risk averse’ FRN response pattern, suggesting that typical loss aversion in decision-making may be

-p

blunted in this college sample. Those who “feel loss” equally as much as they “feel wins” may

re

struggle to turn down immediate rewards, even when those choices might lead to future negative consequences. On the other hand, those who “feel loss” more strongly may have an easier time

lP

turning down immediate reward in the service of longer-term beneficial outcomes. The present results have important implications for future studies on reward and

ur na

punishment processing. The FRN to winning money was not comparable to the FRN to avoiding a loud noise burst and the FRN to winning nothing was not comparable to receiving the loud noise burst. It was revealed only in the aversive domain that drinkers are less sensitive to negative outcomes, emphasizing the need to consider different types of wins and losses. Previous

Jo

studies on the FRN and drinking have only looked at monetary outcomes [20,22,23] and thus, may have missed important information about other types of negative outcomes. Specifically, aversive outcomes might be more related to the real-life consequences of drinking and may be more related to drinking. These effects may not be limited to alcohol use; future studies could test if this pattern is consistent across other drugs of abuse.

11 One result that was inconsistent with our hypotheses was that drinkers did not have a larger neural response to appetitive stimuli compared to non-drinkers. Further, one might suspect that impulsivity or depression would be higher in the drinking group, but both were unrelated to drinking status in our sample. These inconsistent findings might be a reflection of the sample (younger individuals without chronic alcohol use) or of the stimuli (natural reward rather than

of

alcohol cues). An important future direction includes comparing alcohol cues with natural reward cues in a similar sample to test if the differences in the neural response to these cues is

ro

related to drinking status. It is also possible that more refined constructs related to reward sensitivity are important in drinking behaviors. A meta-analysis revealed that positive or

-p

negative urgency, rather than reward sensitivity, were most strongly related to problematic

re

alcohol use [37]. Therefore, the urge to act impulsively when motivated by positive or negative mood states might be more relevant for predicting drinking patterns compared to raw neural

lP

response to unexpected appetitive or aversive events employed in this design. While the present study was the first to compare the FRN in response to appetitive and

ur na

aversive outcomes in college students, it had limitations. First, it is unknown if the participants drank alcohol the night before participation. Although alcohol hangover can have an effect on FRN and performance [24], the majority of our participants reported primarily drinking over the weekend. Next, the stimuli used in the experiment were created to resemble a slot-machine.

Jo

Experience with gambling was not accounted for and could have affected the FRN amplitudes. Further, the participants were not evaluated for the presence of an alcohol use disorder and were not asked about alcohol-related problems, so it is unclear if our drinking group experienced negative outcomes related to drinking. However, as the mean age was under 21, the threat of negative consequences is constant. Future studies should look to compare the FRN between

12 college students with and without an alcohol use disorder. Finally, the task was a passive prediction task, not an active task that required learning or decision-making. Future studies could determine if these neural responses are actually related to learning from consequences and if this learning is related to drinking status. In conclusion, the brain processes appetitive and aversive feedback differently among

of

college drinkers and non-drinkers. Non-drinkers have an increased sensitivity to aversive outcomes, whereas drinkers do not show this pattern. Insensitivity to aversive outcomes might

ro

lead to increased drinking by an inability to learn from negative consequences or an inability to evaluate future outcomes during decision-making. Future studies should look to clarify the

-p

mechanism by which insensitivity to punishment is associated with drinking and should also

re

consider these different types of rewards and punishments to clarify which contributes more to

lP

alcohol use.

None of the authors have potential conflicts of interest to be disclosed.

ur na

This work was supported in part by the National Science Foundation [1428999]. The funding

Jo

source was not involved in any aspect of the research design, analysis, or manuscript preparation.

13 References SAMHSA, 2014 National Survey on Drug Use and Health (NSDUH), 2014.

[2]

C.S. Carver, T.L. White, Behavioral inhibition, behavioral activation, and affective responses impending reward and punishment: The BIS/BAS Scales, J. Pers. Soc. Psychol. 67 (1994) 319–333.

[3]

R.M. O’Connor, C.R. Colder, Influence of Alcohol Use Experience and Motivational Drive on College Students’ Alcohol-Related Cognition, Alcohol. Clin. Exp. Res. 33 (2009) 1430–1439. doi:10.1111/j.1530-0277.2009.00973.x.

[4]

N. Kambouropoulos, P.K. Staiger, The influence of sensitivity to reward on reactivity to alcohol-related cues, Addiction. 96 (2001) 1175–1185. doi:10.1046/j.13600443.2001.968117510.x.

[5]

I.H.A. Franken, P. Muris, BIS/BAS personality characteristics and college students’ substance use, Pers. Individ. Dif. 40 (2006) 1497–1503. doi:10.1016/j.paid.2005.12.005.

[6]

S.H. Shin, H.G. Hong, S.-M. Jeon, Personality and alcohol use: the role of impulsivity., Addict. Behav. 37 (2012) 102–7. doi:10.1016/j.addbeh.2011.09.006.

[7]

C.L. Park, S. Armeli, H. Tennen, The daily stress and coping process and alcohol use among college students., J. Stud. Alcohol. 65 (2004) 126–135. doi:10.15288/jsa.2004.65.126.

[8]

G.A. Marlatt, R.L. Collins, G.A. Parks, Social determinants of alcohol consumption: the effects of social interaction and model status on the self-administration of alcohol, J. Consult. Clin. Psychol. 53 (1985) 189–200.

[9]

W.H.R. Miltner, C.H. Braun, M.G.H. Coles, Event-related brain potentials following incorrect feedback in a time-estimation task: Evidence for a “generic” neural system for error detection, J. Cogn. Neurosci. 9 (1997) 788–798. doi:10.1162/jocn.1997.9.6.788.

ur na

lP

re

-p

ro

of

[1]

[10] W. Gehring, A. Willoughby, The medial frontal cortex and the rapid processing of monetary gains and losses, Science (80-. ). 295 (2002). http://science.sciencemag.org/. [11] G.F. Potts, L.E. Martin, P. Burton, P.R. Montague, When things are better or worse than expected: The medial frontal cortex and the allocation of processing resources, J. Cogn. Neurosci. 18 (2006) 1112–1119. doi:10.1162/jocn.2006.18.7.1112.

Jo

[12] S. Leknes, I. Tracey, A common neurobiology for pain and pleasure, Nat. Rev. Neurosci. 9 (2008) 314–320. doi:10.1038/nrn2333. [13] O.E. Krigolson, Event-related brain potentials and the study of reward processing: Methodological considerations, Int. J. Psychophysiol. 132 (2018) 175–183. doi:10.1016/J.IJPSYCHO.2017.11.007. [14] D. Talmi, R. Atkinson, W. El-Deredy, The feedback-related negativity signals salience prediction errors, not reward prediction errors, J. Neurosci. (2013). doi:10.1523/JNEUROSCI.5695-12.2013. [15] T.U. Hauser, R. Iannaccone, P. Stämpfli, R. Drechsler, D. Brandeis, S. Walitza, S. Brem,

14 The feedback-related negativity (FRN) revisited: new insights into the localization, meaning and network organization., Neuroimage. 84 (2014) 159–68. doi:10.1016/j.neuroimage.2013.08.028. [16]

E.J. Hird, W. El-Deredy, A. Jones, D. Talmi, Temporal dissociation of salience and prediction error responses to appetitive and aversive taste, Psychophysiology. 55 (2018). doi:10.1111/psyp.12976.

[17] H.E. Soder, G.F. Potts, Medial frontal cortex response to unexpected motivationally salient outcomes, Int. J. Psychophysiol. 132 (2018) 268–276. doi:10.1016/j.ijpsycho.2017.11.003.

of

[18] G.F. Potts, Impact of reward and punishment motivation on behavior monitoring as indexed by the error-related negativity, Int. J. Psychophysiol. 81 (2011) 324–331. doi:10.1016/J.IJPSYCHO.2011.07.020.

ro

[19] C.L. Barberini, S.E. Morrison, A. Saez, B. Lau, C.D. Salzman, Complexity and competition in appetitive and aversive neural circuits, Front. Neurosci. 6 (2012) 170. doi:10.3389/fnins.2012.00170.

-p

[20] G. Fein, M. Chang, Smaller feedback ERN amplitudes during the BART are associated with a greater family history density of alcohol problems in treatment-naïve alcoholics, Drug Alcohol Depend. 92 (2008) 141–148. doi:10.1016/j.drugalcdep.2007.07.017.

lP

re

[21] I.H.A. Franken, I. Van Den Berg, J.W. Van Strien, Individual differences in alcohol drinking frequency are associated with electrophysiological responses to unexpected nonrewards, Alcohol. Clin. Exp. Res. 34 (2010) 702–707. doi:10.1111/j.15300277.2009.01139.x.

ur na

[22] S. Lannoy, F. D’Hondt, V. Dormal, J. Billieux, P. Maurage, Electrophysiological correlates of performance monitoring in binge drinking: Impaired error-related but preserved feedback processing, Clin. Neurophysiol. 128 (2017) 2110–2121. doi:10.1016/j.clinph.2017.08.005. [23] H. Hixson, K.L. Burkhouse, S.M. Gorka, H. Klumpp, A preliminary examination of the relation between neural sensitivity to reward and history of alcohol use disorder among adults with internalizing psychopathologies, Neurosci. Lett. 690 (2019) 17–22. doi:10.1016/j.neulet.2018.10.003.

Jo

[24] A.D. Howse, C.D. Hassall, C.C. Williams, G. Hajcak, O.E. Krigolson, Alcohol hangover impacts learning and reward processing within the medial-frontal cortex, Psychophysiology. 55 (2018). doi:10.1111/psyp.13081. [25] J.E. Grant, S.R. Chamberlain, Impulsive action and impulsive choice across substance and behavioral addictions: Cause or consequence?, Addict. Behav. 39 (2014) 1632–1639. doi:10.1016/j.addbeh.2014.04.022. [26]

D.M. Olvet, G. Hajcak, The error-related negativity (ERN) and psychopathology: Toward an endophenotype, Clin. Psychol. Rev. 28 (2008) 1343–1354. doi:10.1016/j.cpr.2008.07.003.

[27] H.E. Soder, T.A. Webber, M.A. Bornovalova, J.Y. Park, G.F. Potts, A test of dopamine

15 hyper- and hyposensitivity in alcohol use, Addict. Behav. 90 (2019) 395–401. doi:10.1016/j.addbeh.2018.12.002. [28] W. Schultz, P. Dayan, P.R. Montague, A Neural Substrate of Prediction and Reward, Science (80-. ). 275 (1997) 1593–1599. http://science.sciencemag.org/. [29] T.D. Sambrook, J. Goslin, A Neural reward prediction error revealed by a meta-analysis of ERPs using great grand averages, Psychol. Bull. 141 (2015) 213–235. doi:10.1037/bul0000006.

of

[30] S.J. Pocock, S.E. Assmann, L.E. Enos, L.E. Kasten, Subgroup analysis, covariate adjustment and baseline comparisons in clinical trial reporting: current practice and problems, Stat. Med. Stat. Med. 21 (2002) 2917–2930. doi:10.1002/sim.1296.

ro

[31] S.F. Assmann, S.J. Pocock, L.E. Enos, L.E. Kasten, Subgroup analysis and other (mis)uses of baseline data in clinical trials, Lancet. 355 (2000) 1064–1069. doi:10.1016/S0140-6736(00)02039-0.

-p

[32] J. Pinheiro, D. Bates, S. DebRoy, D. Sarkar, R.C. Team, _nlme: Linear and nonlinear mixed effects models_.R package version 3.1-139, (2003). https://cran.rproject.org/package=nlme.

re

[33] D. Bates, M. Machler, B. Bolker, S. Walker, Fitting linear mixed-effects models using lme4, (2014). http://dx.doi.org/10.18637/jss.v067.i01. [34] R.C. Team, A language and environment for statistical computing, (2019). https://www.rproject.org/.

lP

[35] S. Nakagawa, H. Schielzeth, A general and simple method for obtaining R2 from generalized linear mixed-effects models, Methods Ecol. Evol. 4 (2013) 133–142. doi:10.1111/j.2041-210x.2012.00261.x.

ur na

[36] D. Kahneman, A. Tversky, Prospect Theory: An Analysis of Decision under Risk, 1979. http://www.albacharia.ma/xmlui/bitstream/handle/123456789/31987/kahnmtversky.pdf?se quence=1.

Jo

[37] K. Stautz, A. Cooper, Impulsivity-related personality traits and adolescent alcohol use: A meta-analytic review, Clin. Psychol. Rev. 33 (2013) 574–592. doi:10.1016/J.CPR.2013.03.003.

16

Jo

ur na

lP

re

-p

ro

of

Figure 1. FRN-difference by drinking group. Difference waves for PPT reflect unpredicted withheld punishment minus unpredicted punishment and difference waves for RPT reflect unpredicted withheld rewards minus unpredicted rewards. PPT = Punishment Prediction Task; RPT = Reward Prediction Task. A. Difference waves for the Drinkers. B. Difference waves for the Non-Drinkers. ERPs are averaged over the montage and time window is indicated by dotted lines.

17

Jo

ur na

lP

re

-p

ro

of

Figure 2. Drinking group by task interaction on FRN-difference. Difference wave FRN amplitude to unexpected outcomes as a function of drinking group (non-drinkers/drinkers) and task (reward prediction task/punishment prediction task).

18 Table 1. Participant characteristics and variable means

Non-Drinkers (n=32)

Drinkers (n=18)

Significance

M/% 20.88 78% F 2.34 3.56 2.13 2.38 45.07 34.71 57.88

SD 5.563 1.125 1.883 1.212 1.289 7.076 10.687 8.911

M/% 21.06 78% F 2.39 3.83 2.50 2.28 47.43 37.57 57.66

SD 2.817 1.195 2.007 1.339 1.018 8.491 14.130 9.864

M -3.323 -1.500

SD 2.701 2.783

M -0.486 -1.037

SD 1.781 1.640

ro

PPT Difference Wave RPT Difference Wave Drinking Variables

P .898 .977 .895 .636 .317 .784 .347 .468 .939 P <.001 .520

-p

Age Gender Year in College Youth Household Income Mother’s Education Father’s Education Depression (BDI) Anxiety (STAI) Impulsivity (BIS-11) ERP Amplitudes

of

Sample Characteristics

ur na

lP

re

M SD M SD P Drinks per week 0 0 5.78 3.64 <.001 Days per week 0 0 2.11 0.76 <.001 Note. P value for sample characteristics/drinking variables reflects either independent samples t-test or chi-square test where appropriate. P value for difference waves represent simple effects models described in the text. See supplement to interpret income and education variables. F = female; PPT = punishment prediction task; RPT = reward prediction task; BDI = Beck’s Depression Inventory; STAI = the trait scale on the State Trait Anxiety Inventory; BIS-11 = Barratt Impulsiveness Scale.

b

SE

df

t

p

Jo

Table 2. Fixed effects from primary linear mixed effect model Predictor

3.32

0.43

91

7.74

< 0.001

Task – Appetitive (vs. Aversive)

-1.82

0.53

48

-3.418

0.001

Drinking Group – Drinkers (vs. Non-Drinkers)

-2.84

0.72

91

-3.964

< 0.001

Drinking Group x Task

2.37

0.89

48

2.671

0.010

Intercept