Tasks and Investigated Components in Social Cognition Research Among Adults With Alcohol Use Disorder: A Critical Scoping Review

Public Health Significance Statement This scoping review stresses problems in the assessment procedures and scope of social cognition (the perception and interpretation of social information) research in alcohol use disorder (AUD). The tasks often do not allow for clear interpretations regarding the social cognitive ability under investigation and/or do not adequately reflect everyday situations. Moreover, important social cognition components such as the ability to identify and take into account social contexts and rules, or the tendency to provide negative causal explanations for ambiguous social events, have been largely neglected. Addressing these issues will help maximize the impact of social cognition research on the understanding and treatment of severe AUD. Objective: Social cognition research in alcohol use disorder (AUD) has accumulated over the past decades and has implications for understanding the interpersonal problems reported in this population and for improving clinical outcomes. However, recent criticism of classically used social cognition tasks calls for an evaluation of social cognition assessments in AUD. Moreover, available literature reviews focus on a restricted subset of social cognition components, leaving the outcomes and significance of studies assessing the remaining components unknown. Hence, to qualify and broaden our understanding of the available evidence and identify research perspectives, we systematically charted and critically appraised the tasks used and social cognition components investigated in AUD. Method: We searched databases for studies comparing patients with AUD and healthy controls on behavioral social cognition assessments. We extracted the number of times specific social cognition components were investigated and the tasks assessing them. Results: Of the 74 included records, 58 investigated emotion recognition, 14 investigated theory of mind (ToM), three investigated social perception/knowledge, and two investigated attributional biases. Most emotion recognition tasks required complex categorization, and presented unimodal static and context-free emotional stimuli among verbal labels. ToM was mostly assessed with the reading the mind in the eyes and faux-pas tests. Conclusions: Emotion recognition and ToM have been extensively investigated yet most tasks are multidetermined, lack ecological validity, or fail to assess the targeted ability. Conversely, social perception/knowledge and attributional biases, despite their clear relevance to AUD, are insufficiently studied. We propose concrete ways to address these issues.