Methods
Subjects with panic disorder were randomly assigned to consume either a moderate dose of alcohol or a nonalcoholic placebo. All subjects were told that they were drinking alcohol to control beverage expectancies. Following the beverage administration, subjects underwent a panic challenge (35% carbon dioxide) and a series of anxiety symptom assessments.

Results
Subjects who consumed alcohol reported significantly less state anxiety both before and after the challenge. In response to the challenge, subjects who consumed alcohol experienced significantly fewer panic attacks when applying liberal panic criteria; however, this effect only approached significance when applying conservative panic criteria.

Conclusions
These findings suggest that alcohol acts acutely to reduce both panic and the anxiety surrounding panic, and they lend support to the view that drinking behavior among those with panic disorder is reinforced by this effect. We suggest that this process may contribute to the high rate at which alcohol-use disorders co-occur with panic disorder.