A prospective examination of TAC features as predictors of consequences and alcohol use disorders

PROJECT SUMMARY Young-adult alcohol misuse is a public health problem causing thousands of assaults, injuries, and deaths each year. Both college and non-college adults frequently report high drink counts (e.g.,10+ drinks) and many alcohol-related consequences. Approximately 13% meet criteria for past-year alcohol use disorder (AUD) and problems associated with AUDs frequently continue into midlife. Although measurement of drinking is necessary for assessing these risks, concerns have been raised about exclusive reliance on self-reported drink counts, especially on heavier drinking nights. Our pilot work supports this concern, showing that over 12 weekend days: 1) 70% of the sample reported blackouts (55% reporting multiple blackouts), and 2) the rate of blackout was 1 in every 3 drinking days. These and associated impairments may affect the accuracy of self-reported drinking and lead to underestimation of the links between alcohol consumption and outcomes (consequences and AUD). To address this concern, we have used objective watchband-like transdermal alcohol concentration (TAC) biosensors in our studies. TAC assesses alcohol use in near real-time and records the manner in which alcohol is being consumed through continuous measurement of alcohol intoxication. TAC features from the rising portion of the curve for each drinking day – rise rate, peak, and rise duration – capture dynamic aspects of drinking behavior that standard drink counts do not. Our pilot data provide evidence that TAC features predict consequences and are associated with baseline AUD. These findings are promising, but they are limited to small samples over short time spans (1-4 weeks). These short time spans limit generalizability of TAC effect size estimates and have prevented prospective tests of TAC features as predictors of outcomes that develop over long time spans (i.e., AUD). The proposed research addresses these gaps using a large-scale measurement burst design consisting of eight 14-day bursts across two years among 500 young adults (ages 19-22) who frequently engage in heavy episodic drinking. Aim 1 will examine TAC features as predictors of alcohol-related consequences, relative to EMA self-reports. Aim 2 will assess AUD using DSM criteria at baseline, 12, and 24 months and focus on examining the contribution of TAC features, relative to EMA self-reports, in predicting AUD. The scientific premise of our research is fully supported by the literature and our past studies. Together, these warrant further study of how well TAC features can predict: 1) event level alcohol-related consequences, and 2) AUDs. Our approach uses a rigorous longitudinal and multimodal measurement burst design to address these significant gaps in the field.