Category: Addictive Behaviors

PS1- #A13 - Discrepancy in Implicit and Explicit Drinking Identity Is Associated With Current but Not Future Risk of Alcohol Use Disorder

Friday, Nov 17
8:30 AM – 9:30 AM
Location: Indigo Ballroom CDGH, Level 2, Indigo Level

Keywords: Addictive Behaviors | Alcohol | Cognitive Biases / Distortions

Drinking identity, or the extent to which one views one’s self as a drinker, is a promising cognitive risk factor for predicting alcohol use and misuse (Lindgren et al., 2016). Drinking identity has been evaluated with implicit and explicit measures – measures that tap into cognitive processes that are more reflexive, automatic, and impulsive (implicit) or more reflective, controlled, introspective (explicit). Research to date has established the unique contribution of implicit and explicit identity in predicting drinking behaviors (Lindgren et al., 2013, 2016). We know of no studies that have assessed the discrepancy between implicit and explicit drinking identity or evaluated how that discrepancy relates to drinking behaviors. Related work in depression (Creemers et al., 2012, 2013) has shown that the discrepancy between implicit and explicit self-esteem is associated with greater depressive symptoms, suicidal ideation, and loneliness. This study’s goal was to assess the discrepancy in implicit and explicit drinking identity and test whether it predicted current and future risk of alcohol use disorder [AUD]. A sample of 506 undergraduates (57% women) completed implicit and explicit drinking identity measures at baseline. Risk of AUD was assessed at baseline, three-month, and 12-month follow-up via the Alcohol Use Disorder Identification Test (Babor et al., 2001). Drinking identity variables were z-transformed, and two discrepancy variables were created following Creemers et al. (2013): a magnitude variable that indicates the size of the implicit-explicit discrepancy and a direction variable that indicates whether implicit identity is stronger than explicit identity. Zero-inflated negative binomial regression models tested how discrepancy magnitude, direction, and their two-way interaction predicted baseline AUDIT scores and 3-month and 12-month changes in AUDIT scores. Sex was controlled for in analyses. Preliminary results indicated significant cross-over interactions when predicting baseline risk of AUD. There was a positive relationship between discrepancy magnitude and AUD risk when explicit identity was stronger than implicit identity, and there was a negative relationship between discrepancy magnitude and AUD risk when implicit identity was stronger than explicit identity. None of the discrepancy variables significantly predicted changes in AUD risk at 3- or 12-month follow up. Results suggest that implicit-explicit discrepancy is associated with current AUD risk but is not associated with changes in future risk. Thus, discrepancy in implicit and explicit drinking identity may be a marker of current vulnerability but does not appear to be a mechanism for future risk.

Kristen P. Lindgren

University of Washington

Kirsten P. Peterson

Research Coordinator
University of Washington

Cecilia C. Olin

Research Coordinator
University of Washington
Seattle, Washington

Bethany A. Teachman

University of Virginia
Charlottesville, Virginia

Scott Baldwin

Associate Professor
Brigham Young University
Provo, Utah