TREATMENT Combined Neuroimaging, Neurocognitive and Psychiatric Factors to Predict Alcohol Consumption Following Treatment for Alcohol Dependence

Abstract — Aims: Resumption of hazardous drinking after treatment is common in alcohol use disorders (AUD). This study exam-ined the ability of multimodality magnetic resonance, neurocognitive, psychiatric and demographic, to predict alcohol consumption after treatment for AUD. Methods: Seventy treatment-seeking participants completed 1.5T magnetic resonance studies, yielding regional gray matter (GM) and white matter (WM) surrogate markers of neuronal integrity (N-acetylaspartate: NAA) and cell mem-brane turnover/synthesis (choline: Cho), assessment of major psychiatric disorders and comprehensive neurocognitive assessment after ∼1 month of abstinence. Participants were followed up 6–12 months after treatment and classified as Abstainers (no alcohol consump-tion; n = 26) and Resumers (any alcohol consumption; n = 44). Abstainers and Resumers were contrasted on various outcome measures, and those that significantly differed between groups were entered as factors in a logistical regression model to predict drinking status at follow-up. Results: The following variables were independent predictors of resumption of drinking: temporal GM NAA, frontal WM NAA, frontal GM Cho, processing speed and comorbid unipolar mood disorder. With each standard deviation unit decrease in temporal GM NAA, frontal WM NAA, frontal GM Cho and processing speed, the odds of resumption of drinking were increased 3.1