Individual differences predictive of drinking to manage anxiety among non-problem drinkers with panic disorder

Background: Alcohol problems co-occur with anxiety disorders at a rate that far exceeds chance ('comorbidity'). One view suggests that risk for developing a comorbid alcohol use disorder is increased when alcohol is used routinely to cope with anxiety symptoms ('self-medication'). Indicating that this view is overly broad, however, the literature suggests that only a subgroup of anxiety-disordered individuals tend to drink to manage their symptoms. Therefore, we set out to identify psychological characteristics that might mark those for whom drinking to cope with anxiety is most likely. Our survey of the literature identified several possibilities, including anxiety-related personality traits (anxiety sensitivity, self-consciousness and Trait Anxiety); higher-order personality dimensions (Harm Avoidance, Reward Dependence, and Novelty Seeking); and, finally, alcohol outcome expectancies (specifically, those related to tension-reducing effects from alcohol). Methods: In a sample of nonproblem drinkers with panic disorder, we regressed predictor variables on several alcohol use measures, including drinking aimed at the control of anxiety symptoms ('self-medication'). Results: Although each variable related positively to a self-medicating style of drinking, expectancies for tension reduction from alcohol use accounted for about twice as much variance as did the other predictors. With simultaneous entry in a step-wise regression analysis, only tension-reduction alcohol outcome expectancies and the Harm Avoidance personality dimension were retained as significant predictors. Conclusions: Findings suggest that development of a self-medicating style of drinking among individuals with panic disorder is predicated, in part, on specific psychological characteristics of the individual. Alcohol outcome expectancies emerged as the single most important predictor of drinking behavior in this anxiety- disordered sample.