Testing the efficacy of a new intervention for high intensity drinking among emerging adults

High intensity drinking (HID; consuming 8+ standard drinks for women and 10+ for men) is most common among emerging adults (EAs), reflecting an important public health concern and NIAAA priority. HID increases risk for deleterious outcomes in the short and long-term (e.g., development of alcohol use disorders [AUD]). Experts have called for novel approaches tailored for people engaged in HID, with a key innovation proposed herein being to harness social media interactions given that EAs use these platforms frequently. Our team used iterative participatory methods to develop and pilot the first intervention for EAs engaged in HID. Given EA norms of daily social media use and the potential for social media to reach beyond clinical and school settings for intervention delivery, we used private social media 1:1 messaging with health coaches to deliver 8-weeks of motivational interviewing-based theoretically-derived interactive content. Pilot data supported intervention acceptability (e.g., 85% liked messaging) and trial methods were feasible (e.g., enrollment targets met, 92% intervention engagement, >=88% retention at 4-months). Descriptive results showed promise for alcohol outcomes at the 4-month follow-up (2-months post-intervention end) such as typical weekly drinking (Cohen’s d=.31) and alcohol use severity (AUDIT-C d=.28). Beyond alcohol, intervention content focused on stress, coping with negative affect, and promoting mental health, showing promise for depression (d=.40) and anxiety (d=.38) outcomes. Given these results, we propose a fully-powered randomized controlled trial, with an EA advisory board, to evaluate the short- and long-term efficacy of this innovative intervention in comparison to a control condition over 12 months of follow-up. Aim 1 will involve efficacy testing with the primary outcome of typical weekly drinking (reflecting overall consumption), with secondary (e.g., AUDIT-C score, depression and anxiety symptoms) and exploratory outcomes (e.g., quality of life, drug use) also evaluated. Aim 2 will employ novel machine-learning data mining techniques using participant data (e.g., demographics etc.) combined with community-level data (e.g., Census data) to identify characteristics of individuals who show positive responses, guiding future implementation of personalized medicine approaches. Moderators and mediators will also be explored. Interventions that harness social media to deliver support from health coaches in EAs’ daily lives are scalable and can be easily tailored by coaches in the moment to changing individual needs, trends and contexts of alcohol use, improving upon high-resource, static electronic interventions with limited shelf-life by allowing for a wide reach unconstrained by setting. Key innovations include ushering in a new method of health coach intervention delivery for EAs engaged in HID, holistically addressing mental health as intertwined with HID, and using machine learning analyses to evaluate predictors of response. There is potential for significant public health impact by reducing risky drinking and related consequences and preventing AUD onset in EAs.