Alcohol Hangover Impacts Learning and Reward Processing Within the Medial-Frontal Cortex

Dr. Olav Krigolsonby Dr. Olave E. Krigolson

Centre for Biomedical Research
University of Victoria
Victoria, British Columbia, Canada
V8W 2Y2. P.O. Box 1700 STN CSC
E-mail address: krigolson[at]uvic.ca

Acknowledgement

This user research article summarizes our publication “Alcohol Hangover Impacts Learning and Reward Processing Within the Medial-Frontal Cortex”. Howse AD, Hassall CD, Williams CC, Hajcak G, Krigolson OE. Psychophysiology 2018 Aug;55(8):e13081. doi: 10.1111/psyp.13081

Abstract

Alcohol intoxication impairs motor coordination, judgement, and decision-making and results in a large number of accidental deaths every year. Interestingly, research suggests that the impact of alcohol consumption continues beyond the point of intoxication and into a period of alcohol hangover. Here, we examined differences in the amplitude of reward positivity – an event-related brain potential component associated with learning and how the amplitude of this component was impacted by alcohol hangover. In line with previous findings, we found that behavioral performance was impaired during alcohol hangover. Confirming our hypotheses, we also found that hangover reduced the amplitude of the reward positivity. Importantly, our results suggest that the neural systems that underlie performance monitoring and reward-based learning are impaired during alcohol hangover.

Introduction

Alcohol intoxication results in 10,000 deaths per year in the United States (National Highway Traffic Safety Administration, 2013). In spite of the well-known risks of alcohol intoxication, more than one million people make the choice to drink and drive each year (Centers for Disease Control and Prevention, 2014). Research that potentially explains this behaviour suggests that neural decision-making systems are impaired with alcohol consumption (Bartholow et al., 2012). Further, acute alcohol use impacts other decision related cognitive processes such as planning, working memory, and risk assessment (Starkey et al., 2014). Related research by Nelson, Patrick, Collins, Lang, and Bernat (2011) found that alcohol intoxication also impaired learning related error evaluation. Specifically, Nelson and colleagues found that the amplitude of the feedback-related negativity, an event-related potential (ERP) component associated with the evaluation of performance feedback within the human medial-frontal cortex (Miltner et al., 1997), was reduced during alcohol intoxication. Extending this finding, multiple studies have reported a reduction in task performance concomitant with dampened neural responses following alcohol consumption (Ridderinkhof et al., 2012).

Importantly, the impact of high levels of alcohol consumption on behavior appears to persist beyond intoxication into hangover. Behavioral studies examining alcohol hangover report performance deficits such as slower response times and increases in response errors (Verster, 2007). For example, Kim et al. (2003) found that participants experiencing alcohol hangover exhibited deficits in cognitive function relative to controls. Motor performance also appears to be impaired with alcohol hangover; Karadayian and Cutera (2013) found that mice demonstrated an 80% reduction in motor performance during hangover. With the aforementioned evidence that error evaluation systems are impaired during alcohol intoxication and the behavioural research on alcohol hangover in mind, it stands to reason that error evaluation systems would also be impaired during alcohol hangover.

The primary goal of the present study was to assess whether alcohol hangover reduced the efficacy of neural error evaluation systems. In the present study we recorded electroencephalographic data (EEG) from two groups of participants (control, hangover) while they performed a reward gambling task. To assess the impact of alcohol hangover on outcome evaluation we focused our analysis on the reward positivity – the mirror opposite of the aforementioned feedback-related negativity (see Proudfit, 2015 for detail). Here, we predicted that the amplitude of the reward positivity would be reduced during alcohol hangover. Yoked to this reduction in performance monitoring, we also predicted that overall gambling task performance would be reduced in hangover participants relative to controls.

Methods

Participants
Fifty-eight undergraduate students took part in this study (24 male; mean age 21.5 years, range 21.0–22.0). Post experiment, participants were assigned to either a control or hangover group. Inclusion in the alcohol hangover group was determined by two criteria: (a) having consumed alcohol within 24 hours of experimental participation (but not within 10 hours of the study start time), and (b) having a score on the Modified Alcohol Hangover Severity Scale (M-AHSS) greater than zero. Participants who did not meet both of these criteria were assigned to the control group. Importantly, we found no differences in drinking behavior between hangover and control participants other than the night before participation (1.5 (0.1–2.9) versus 6.0 (3.9–8.1), t(56) 53.4, p < .001) and we found no differences between hangover and control participants in terms of sleep, smoking behaviour, and caffeine intake (all p’s > 0.05). This research was approved by the Health Sciences Research Ethics Board at Dalhousie University and followed all ethical standards prescribed in the 1964 Declaration of Helsinki.

Experimental Task
Participants completed a simple two choice gambling task programmed in MATLAB Version 7.14 using the Psychophysics Toolbox extension. On each trial, participants gambled by choosing between two differently colored squares using a USB gamepad. Unbeknownst to participants, one of the colored squares won 60% of the time when selected whereas the other won 10% of the time with the color and winning square being randomly selected at the start of each block of trials. After the participant selected one of the squares a fixation cross appeared for 400 to 600 ms which was then followed by a feedback stimulus – the word “win” or “loss” – for 1000 ms indicating the outcome of the gamble. Participants completed 25 blocks of 20 gambling trials and were told how many trials they had “won” after each block.

Behavioural and EEG Data Collection
The task program recorded response times and gamble selection for each experimental trial. EEG data were recorded from 64 electrode locations at 500 Hz with an actiCHamp amplifier and PyCorder software. All electrode impedances were kept below 20 kΩ throughout data collection. Following completion of the experimental task, participants were interviewed using the timeline follow-back method (Sobell & Sobell, 1992) in order to assess recent alcohol and substance use behavior and also completed the M-AHSS (Penning et al., 2013).

Data Analysis
For our behavioural data we computed the mean response time and mean performance for each participant. EEG data were processed offline with BrainVision Analyzer 2.1 software. First, excessively noisy or faulty electrodes were removed, the EEG data were rereferenced to an average mastoid reference and were filtered with band-pass (0.1 Hz to 30 Hz) and notch (60 Hz) filters. Independent component analysis (ICA) was then used to identify and remove ocular artifacts (Delorme & Makeig, 2004) and any channels that were removed initially were interpolated. Epochs were constructed from 200 ms before to 600 ms after the onset of the feedback stimuli. All segments were then baseline corrected using the 200 ms epoch immediately preceding stimulus onset. Next, segments that had gradients of greater than 10 μV/ms and/or a 100 μV absolute within-segment difference were removed. Following artifact rejection, segments were averaged to create win and loss ERP waveforms for each participant. Difference waveforms were computed by subtracting the average loss from the average win waveform. Finally, grand-averaged waveforms were computed for the win, loss, and difference waveforms for both the hangover and no hangover groups. The reward positivity for each participant was quantified as the peak positive deflection of the difference waveform 200 to 400 ms post feedback onset at electrode FCz. A full summary of our EEG pre-processing pipeline can be found at https://www.krigolsonlab.com/data-analysis.html.

Results

Overall, hangover participants demonstrated greater variability in performance than controls (control: 17.8%, hangover: 20.5%; t(56) = 2.32, p = .025). Further, mean performance was reduced in the hangover group in comparison to the control group (control: 61%, hangover: 58%; t(56) = 4.29, p = .0007, Cohen’s d = 1.12, see Figure 1) during the initial learning phase of each block of trials.

Figure 1: Performance across collapsed trials averaged over all blocks and measured in percent of better selections out of all selections. Participants were informed that the target square may switch on some trials. Time of target switch is indicated by a dashed line.

Figure 1: Performance across collapsed trials averaged over all blocks and measured in percent of better selections out of all selections. Participants were informed that the target square may switch on some trials. Time of target switch is indicated by a dashed line.

We also found that the amplitude of the reward positivity was reduced for hangover participants relative to control participants (2.07 μV versus 4.66 μV, t(56) = 4.26, p = .0008, d = 1.11, see Figure 2).

Figure 2: Grand-averaged ERP waveforms for wins and losses across all trials and blocks for the (a) control group, and the (b) hangover group. Latency of the component is consistent with previous reports of the reward positivity. (c) Difference waves for each group (control, hangover) calculated as win - loss.

Figure 2: Grand-averaged ERP waveforms for wins and losses across all trials and blocks for the control group (left), and the hangover group (center). Latency of the component is consistent with previous reports of the reward positivity. Right: Difference waves for each group (control, hangover) calculated as win – loss.

Within the hangover group we also found that hangover severity was negatively correlated with the magnitude of the reward positivity (Pearson’s r = -.61, p = .0003, see Figure 3).

Figure 3: Peak component magnitudes plotted against hangover severity in the hangover group, Pearson’s r = 0.61.

Figure 3: Peak component magnitudes plotted against hangover severity in the hangover group, Pearson’s r = 0.61.

Discussion

In the present experiment we hypothesized reduced task performance during alcohol hangover. Furthermore, we predicted that the amplitude of the reward positivity would be reduced for participants experiencing alcohol hangover. In line with previous behavioral studies (Verster, 2007), we found that task performance was indeed reduced for hangover participants relative to controls. Consistent with our primary hypothesis, we also observed a reduction in the amplitude of the reward positivity in hung-over participants suggesting that an error evaluation system within the medial-frontal cortex (c.f., Holroyd & Coles, 2002) is affected by hangover. Our data are consistent with the notion that the impact of higher volumes of alcohol consumption on brain activity continues beyond alcohol intoxication into hangover. Importantly, other than the number of drinks consumed the night prior to the study and the score on the M-AHSS, we found no differences between participants with and without hangovers in terms of hours of sleep, other substance use, and general drinking behavior.

It is not difficult to see how impairments in error evaluation systems pose potential danger. Given the proposed role of the medial-frontal system in motor control (Krigolson & Holroyd, 2007) and our results here, it is quite plausible that reduced error evaluation capability due to alcohol hangover is the mechanism behind the increase of accidents reported in this state. Indeed, the hangover state has been associated with increased accidents and motor impairment (Verster, 2007) and the current study suggests a potential mechanism governing the association between hangover and these accidents. Insofar as the current results suggest that hangover is associated with attenuated neural response to reward, these findings also have broader implication for alcohol misuse. Specifically, blunted neural response to reward has also been associated with negative affect (Bartholow et al., 2012) and depression (Foti et al., 2014) and as such alcohol hangover may also exaggerate these conditions.

References

  • Bartholow, B. D., Henry, E. A., Lust, S. A., Saults, J. S., & Wood, P. K. (2012). Alcohol effects on performance monitoring and adjustment: Affect modulation and impairment of evaluative cognitive control. Journal of Abnormal Psychology, 121(1), 173–186.
  • Centers for Disease Control and Prevention. (2014). Department of Health and Human Services, Atlanta, Georgia. Retrieved from https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812169
  • Delorme, A., & Makeig, S. (2004). EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of Neuroscience Methods, 134 (1), 9–21.
  • Foti, D., Carlson, J. M., Sauder, C. L., & Proudfit, G. H. (2014). Reward dysfunction in major depression: Multimodal neuroimaging evidence for refining the melancholic phenotype. NeuroImage, 1(101), 50–58.
  • Holroyd, C. B., & Coles, M. G. H. (2002). The neural basis of human error processing: Reinforcement learning, dopamine, and the error-related negativity. Psychological Review, 109, 679–709.
  • Karadayian, A. G., & Cutrera, R. A. (2013). Alcohol hangover: Type and time-extension of motor function impairments. Behavioural Brain Research, 24, 165–173.
  • Kim, D., Yoon, S., Lee, H., Choi, B., & Go, H. (2003). The effects of alcohol hangover on cognitive functions in healthy subjects. International Journal of Neuroscience, 113, 581–594.
  • Krigolson, O. E., & Holroyd, C. B. (2007b). Hierarchical error processing: Different errors, different systems. Brain Research, 1155, 70–80.
  • Miltner, W. H. R., Braun, C. H., & Coles, M. G. H. (1997). Eventrelated brain potentials following incorrect feedback in a time estimation task: Evidence for a “generic” neural system for error detection. Journal Cognitive Neuroscience, 9, 788–798.
  • National Highway Traffic Safety Administration. (2013). Department of Transportation, U. S. Government, Washington, DC. Retrieved from https://www.cdc.gov/motorvehiclesafety/impaired_driving/ impaired-drv_factsheet.html
  • Nelson, L. D., Patrick, C. J., Collins, P., Lang, A. R., & Bernat, E. M. (2011). Alcohol impairs brain reactivity to explicit loss feedback. Psychopharmacology, 218, 419–428.
  • Penning, R., McKinney, A., Bus, L. D., Olivier, B., Slot, K., & Verster, J. C. (2013). Measurement of alcohol hangover severity: Development of the Alcohol Hangover Severity Scale (AHSS). Psychopharmacology, 225, 803–810.
  • Proudfit, G. H. (2015). The reward positivity: From basic research on reward to a biomarker for depression. Psychophysiology, 52, 449– 459.
  • Ridderinkhof, R., de Vlugt, Y., Bramlage, A., Spaan, M., Elton, M., Snel, J., & Band, G. H. (2002). Alcohol consumption impairs detection of performance errors in mediofrontal cortex. Science, 298, 2209–2211.
  • Sobell, L. C., & Sobell, M. B. (1992). Timeline followback: A technique for assessing self-reported alcohol consumption. In R. Z. Litten & J. Allen (Eds.), Measuring alcohol consumption: Psychosocial and biological methods (pp. 41–72). New York, NY: Humana Press
  • Starkey, N. J., & Charlton, S. G. (2014). The effects of moderate alcohol concentrations on driving and cognitive performance during ascending and descending blood alcohol concentrations. Human Psychopharmacology, 29(4), 370–383.
  • Verster, J. C., (2007). Alcohol hangover effects on driving and flying. International Journal of Disability and Human Development, 6, 361–367.
©Brain Products GmbH 2018