Clarifying the anxiety sensitivity and alcohol use relation: considering alcohol expectancies as moderators.

Objective: Empirical examinations of the relation between anxiety sensitivity (AS) and alcohol use have yielded inconsistent results, with some studies finding a positive association and other studies finding no association. The present study sought to clarify this relation by examining the moderational effects of alcohol expectancies (i.e., tension reduction, cognitive and behavioral impairment) on the AS-alcohol use association. Gender-specific pathways were also examined.

Method: Participants (N=158, 82 women) were alcohol nonabstaining college freshmen who completed a questionnaire battery including assessments of alcohol expectancies (four positive subscales: sociability, tension reduction, liquid courage, and sexuality; three negative subscales: cognitive and behavioral impairment, risk and aggression, and self-perception), AS, and alcohol use (past month).

Results: Regression analysis was used to test three-way AS x Gender x Alcohol Expectancy interaction terms predicting alcohol use. Only the tension reduction and cognitive and behavioral impairment subscales entered into significant (p< .05) and marginally significant (p= .05) interaction terms, respectively. The simple slopes suggested that, for men only, high AS was associated with heavy drinking but only when tension reduction expectancies were high and that high AS was associated with low levels of drinking when cognitive and behavioral impairment expectancies were high, but this was true only for women.

Conclusions: These findings provide an explanation for why AS has been inconsistently linked to alcohol use in prior research and suggest that models of alcohol use incorporating AS should consider the role of moderators.