RESUMEN
Background: Behavioral models suggest that strong tension-reduction alcohol-outcome expectancies (TREs) among drinkers should be associated with greater tension reduction after drinking. Yet, the few studies investigating this have found either no relationship or the opposite relationship.Objectives: We sought to explore this relationship by building upon the limitations of past studies and employing a placebo-controlled, within-subject experimental design.Methods: Sixty social drinkers (26 M, 34 F) visited the lab on two occasions spaced one week apart. Each participant was randomly assigned to receive alcoholic drinks targeting a BAC of 0.05% on one testing day and placebo drinks on the other, with the order counter-balanced. On both testing days, participants completed measures of state anxiety and fear both before drinking and following a drinking/absorption period. While completing the self-report measures, participants were anticipating an impending, mildly stressful heartbeat perception task.Results: Multilevel modeling revealed that the more strongly individuals believed that alcohol reduces tension, the less the pharmacologic properties of alcohol did so (p = .02 for the state anxiety outcome measure; p = .001 for the fear outcome measure). This was the case even with anxiety sensitivity - a known predictor of stress-response dampening - controlled for.Conclusions: These results provide further evidence for the paradoxical association of TREs and the dampening of anxiety. Additionally, the findings are consistent with the basis of expectancy challenges that aim to reframe inaccurate TREs among drinkers.