Working memory in context: the role of alcohol distractors in low to heavy alcohol drinkers

Background: While widely known that both motivational and cognitive control processes are important in addiction, most studies have focused on these processes independently. Deficits in working memory have long been associated with the development and maintenance of alcohol use disorder, yet findings on a behavioural level have been mixed. This may be context-dependent; working memory performance may be reduced only in the presence of alcohol-related cues or contexts. This study aimed to investigate whether adding an alcohol-related context to a working memory task affects performance i.e., whether working memory performance would be reduced when presented with an alcohol-related context relative to a neutral context. And whether these alcohol-context effects would be associated with severity (AUDIT) and quantity (TLFB) of alcohol use.
Methods: We developed an online N-back flanker task, building on the standard letter N-back task, by bilaterally flanking letters with either alcohol-related or neutral flankers. Here we aimed to capture the extent to which alcohol flankers would interfere with cognitive control related processes (i.e., working memory) compared to neutral flankers. Four working memory loads were included in the task, these being the 0-back, 1-back, 2-back and 3-back. A total of 257 low to high drinking participants (age range of 16-60) completed the task as part of a larger-scale online study. A linear mixed effects model approach with maximum likelihood estimation and random intercepts was used to assess whether working memory load, flanker type, alcohol use severity (AUDIT), quantity (TLFB) or their interaction affected task performance measured as accuracy (percentage correct responses). Akaike’s information criterion was used for model selection.
Results: There was both an effect of working memory load (0-back, 1-back, 2-back, 3-back) and flanker (alcohol, neutral) on accuracy. As working memory load increased, accuracy reduced. Accuracy was also lower when presented with alcohol-related flankers relative to neutral-related flankers. An interaction effect was observed between working memory load and flanker type; reduced accuracy was found when presented with alcohol flankers and a higher working memory load relative to neutral flankers and a lower working memory load. No main or interaction effects were found with alcohol use severity (AUDIT) or quantity (TLFB).
Conclusions: Our findings suggest that an alcohol-related context impacts working memory performance, particularly when cognitive demand is high. However, working memory performance was not associated with severity or quantity of alcohol use.