Learning Mechanisms as Predictors of Treatment Outcome in Alcohol- Dependent Patients

This project will examine 150 detoxified alcohol-dependent patients and 100 age- and gender matched controls. The main aim of this project is to assess 1) which behavioural and neuroimaging alterations (fMRI) associated with reward-based learning (see Projects 1 & 3) predict relapse within the follow-up period of 6 months, 2) how these interact with clinical and psychosocial factors which can modify the relapse risk, and 3) to provide data for genetic and imaging analyses and modelling. Furthermore, we will explore gender effects on functional imaging parameters of learning. Patients will be detoxified in an inpatient setting and followed for 6 months using the Form 90 and Time-Line Follow-Back Procedure. Clinical assessments, behavioral paradigms of learning, and brain imaging will be carried out within at least 4 half-lives after any psychotropic medication. Subjects will undergo medical management with bimonthly follow-ups and predefined in- and exclusion criteria as described previously. We will implement and apply functional imaging paradigms assessing Pavlovian-to-instrumental transfer and reversal learning as described in Projects 1 and 3. We will associate model parameters of learning with alcohol craving, intake and prospective relapse risk. Independent of these central questions, we will also assess comorbidity, psychosocial and neurobiological disease severity markers to control for specificity of findings.