Association between error-related negativity and binge drinking trajectories in two independent youth samples

Error processing, measured via the error-related negativity (ERN), has been associated with alcohol use disorder (AUD); however, most studies investigating this link have been cross-sectional examinations of adult samples. A key empirical question remains regarding whether the ERN represents a risk factor for AUD, a consequence of heavy drinking or the result of a simultaneously occurring third variable (e.g. anxiety). The present study compared baseline ERN amplitude across model-derived drinking frequency classes to evaluate the nature of this association.

Longitudinal multisession cohort study of two independent at-risk adolescent samples across a 2-year tracking period.

University laboratories at the University of Illinois at Chicago and The Ohio State University, United States.

Adolescents and young adults aged 16-19 years (M = 18.26, SD = 0.82, 67.9% female) with minimal alcohol exposure and established risk factors for AUD, recruited in two independent samples (n = 110; n = 164).

ERN amplitude was measured at baseline using a modified flanker task. Participants self-reported drinking behavior over the past 90 days at baseline and every 3 months up to 24 months post-baseline. Latent class mixed model analyses were performed to classify trajectories of drinking behavior and a multi-variate generalized linear model was conducted to examine the associations with class and ERN activity.

Across both samples, increases in binge drinking frequency were associated with blunted (less negative) ERN activity (Sample 1: P = 0.03, η_p ² = 0.04; Sample 2: P = 0.01, η_p ² = 0.05). Associations remained statistically significant after accounting for relevant covariates (anxiety, age, sex).

Blunted error-related negativity amplitude appears to be associated with subsequent increases in binge drinking frequency over time.