2026 ECNP Congress

PS03-3006

Investigating reward-based learning across alcohol use severity

Catherine Guo — T.V. Lim, K.D. Ersche · University of Cambridge, Department of Psychiatry, Cambridge, United Kingdom

2026 ECNP Congress

Introduction

Alcohol use disorder is a major public health concern affecting millions of adults [1]. Converging neuroscientific research has demonstrated altered dopaminergic signalling and a shift toward compulsive behaviour in chronic alcohol use [2]. Dopamine plays a central role in encoding reward and updating behaviour based on feedback [3], with evidence that individuals with alcohol use disorder show impairments in reward-based learning [4]. However, this work has largely used probabilistic learning paradigms, in which outcomes are uncertain [5], leaving it unclear whether the capacity to learn from positive reinforcement is similarly impaired under deterministic conditions. Addressing this gap may clarify whether the contingencies of feedback shape learning across alcohol use severity, with implications for behavioural interventions.

Objectives

This study aimed to investigate whether deterministic positive reinforcement learning differed across alcohol use severity. We hypothesised that learning performance is impaired in individuals with alcohol use disorder, which is most impaired in alcohol-dependent individuals.

Methods

Participants (N = 92) were recruited from the general population into three groups based on clinical thresholds of the Alcohol Use Disorder Identification Test. We excluded individuals with a history of regular substance use or current use of psychoactive medication. The three groups consisted of: healthy controls with no history of affective disorders (N = 43, scores 0-6), harmful alcohol users (N= 25, scores 15-19) and alcohol-dependent individuals (N = 24, scores 20-40). All participants completed a deterministic stimulus-response task presented as a computerised game, in which they identified the correct button press from four options for several visual stimuli. Learning occurred across 10 blocks of 24 trials, with correct responses reinforced by positive feedback. Performance accuracy was compared between groups using a one-way analysis of variance with Tukey’s post-hoc test. Learning trajectories across blocks were analysed using a mixed analysis of variance with Greenhouse-Geisser correction.

Results

Accuracy improved significantly across blocks in all three groups (F(3.64, 323.71) = 206.46, p < 0.001), progressing from near-chance rates in Block 1 (M = 30.55, SD = 10.40) to high performance by Block 10 (M = 87.67, SD = 22.43). However, total accuracy did not differ significantly between the groups (F(2, 89) = 2.81, p = 0.066), and the block-by-group interaction was non-significant (F(7.27, 323.71) = 1.30, p = 0.249). Inspection of interaction plots revealed that, although learning trajectories were comparable in all three groups, the alcohol-dependent group demonstrated marginally lower performance accuracy across blocks, compared with controls and harmful alcohol users.

Conclusions

Contrary to our hypothesis, we did not find measurable impairments in learning from deterministic positive feedback in alcohol users at different levels of severity. While past evidence using probabilistic reinforcement has reported reward-learning deficits in alcohol use disorder, our preliminary findings indicate that stable contingencies may support more effective learning. It is tempting to speculate if alcohol-related learning impairments emerge more selectively, for instance under negative feedback or when established responses must be updated. Clinically, our findings suggest that deterministic reward-based learning is a potential mechanism to leverage in behavioural interventions for alcohol dependence.

References

  1. World Health Organization, 2024. Alcohol. Available from: https://www.who.int/news-room/fact-sheets/detail/alcohol. [Accessed: 30 April 2026]
  2. Everitt, B.J., Dickinson, A., Robbins, T.W., 2001. The neuropsychological basis of addictive behaviour. Brain Res. Rev. 36 (2), 129-138. https://doi.org/10.1016/S0165-0173(01)00088-1
  3. Everitt, B.J., Robbins, T.W., 2005. Neural systems of reinforcement for drug addiction: from actions to habits to compulsion. Nat. Neurosci. 8 (11), 1481-1489. https://doi.org/10.1038/nn1579
  4. Galandra, C., Basso, G., Cappa, S., Canessa, N., 2017. The alcoholic brain: neural bases of impaired reward-based decision-making in alcohol use disorders. Neurol. Sci. 39 (3), 423-435. https://doi.org/10.1007/s10072-017-3205-1
  5. Park, S.Q., Kahnt, T., Beck, A., Cohen, M.X., Dolan, R.J., Wrase, J., Heinz, A., 2010. Prefrontal cortex fails to learn from reward prediction errors in alcohol dependence. J. Neurosci. 30 (22), 7749-7753. https://doi.org/10.1523/JNEUROSCI.5587-09.2010