Mediation of beneficial effects of an alcohol-specific inhibition training on drinking of patients with alcohol use disorder: The role of cognitive demands and inhibitory performance

Background: A cognitively demanding, alcohol-specific inhibition training (Alc-IT) might enhance treatment success in patients with severe alcohol use disorder (AUD; Stein et al., 2023). An inhibitory working mechanism for Alc-IT has been discussed, but compelling evidence supporting this hypothesis is yet lacking. The present study investigates inhibitory performance during Alc-IT and examines whether inhibitory parameters mediate drinking outcome.

Methods: Patients with AUD (N = 232) completed six sessions of either a standard or improved Alc-IT, differing in their inhibitory demands determined by Go/NoGo-ratios in a modified Go-NoGo-task, or a control training. During these training sessions, data on inhibitory performance was collected. To assess differences in inhibitory performance and its improvement, alcohol-related errors of commission and relative performance, integrating accuracy and speed, were analyzed with hierarchical linear contrast models. Mediation analyses tested whether inhibitory performance predicted drinking outcome (percent days abstinent at 3-month follow-up).

Results: Patients in improved Alc-IT started with higher errors of commission (γ_01(standard) = -2.74, p < 0.001, R² = 0.885) and a lower relative performance in the first training session compared to standard Alc-IT (γ_01(standard) = 0.51, p = 0.004). They showed a steeper increase in relative performance until the final sixth session (γ_{1(s6),(standard)} = -0.37, p = 0.024, R² = 0.882). The effect of improved Alc-IT on drinking outcome was mediated by relative performance increase (bootstrap-CI [0.15, 7.11]).

Conclusion: Higher inhibitory demands enable larger improvements across sessions. Mediation analysis supports an inhibitory working mechanism. Tailoring inhibitory demands to individual performance capacity could optimize future Alc-IT.