(PS1-B27) Exploring the Impact of a Negative Affect Induction on Self-reported Alcohol Variables

Introduction: Theory suggests a primary motivation for alcohol use is reduction of negative affect (NA). Nevertheless, meta-analyses indicate that momentary NA appears orthogonal to alcohol use. Experimental work could help to clarify these theoretical and empirical discrepancies. Few studies examine how inducing NA impacts alcohol-related variables, including the reporting of substance-related experiences, without influencing actual behaviors, suggested in affect-congruent memory theory. Thus, the present work employed an experimental paradigm to explore how induced NA (or a control condition) altered reported alcohol use, expectancies, and problems.

Methods: Over 700 individuals (M_age = 36.14, 53.12% female, 71% White) recruited from Amazon’s MTurk participated. Procedures included assessing baseline affect followed by randomly assigning participants to an autobiographical emotional memory task designed to induce NA (NA induction condition) or a control condition. Follow-up measures assessed post-induction affect and alcohol variables.

Results: Initial independent samples t-tests showed significant differences between groups, such that those assigned to the NA induction reported significantly greater average alcohol intoxication (t = 2.27, p < .05, Cohen’s d = 0.18) and power and aggression expectancies (t = 2.08, p < .05, Cohen’s d = 0.16) compared to control participants. Despite these significant differences, recent work has called for separation of affect responders (defined as those who increased in NA or decreased in positive affect) compared to affect non-responders in affect induction paradigms; thus, within our experimental condition, we separated our affect responders (N = 253) from non-responders (N = 104) and re-examined differences in alcohol-reported variables. We found significant differences in tension reduction expectancies (t = 2.28, p < .05, d = 0.13), and alcohol problems based on the Rutgers Alcohol Problem Index (t = 2.22, p < .05, d = 0.26), and Alcohol Use Disorder Identification Test scores (t = 2.54, p < .01, d = 0.30), but in unanticipated directions. Those who did not experience changes in the targeted affective state (non-responders) reported greater tension reduction expectancies and problems than affect responders, contrary to hypothesized associations between heightened NA and increased problems.

Conclusions: These findings underscore recent questions about the role of NA in both the lived experience and reporting of alcohol-related variables. Future work should continue to use robust study designs to gain greater clarity into these relations. Further, the results underscore the imperative to examine results of future affect-induction paradigms based on response to the induction task. Grouping results based on assignment alone might obscure findings and lead to erroneous conclusions. Failing to respond to negative emotional memory inductions might serve as a proxy for traits related to alcohol expectancies and problems as well.