Learning in Young Adults as Predictor for the Development of Alcohol Use Disorders

Hazardous alcohol use and alcohol use disorders (AUDs) are extremely prevalent in industrialized countries, affecting about 6 million individuals in Germany alone. The onset of most cases occurs during adolescence and early adulthood. Therefore, targeted prevention would be desirable especially in young people who are at high risk to develop AUDs. Since our knowledge about predisposing factors is limited, this project aims to identify mechanisms underlying liability for dysfunctional alcohol consumption (i.e. hazardous alcohol use, alcohol abuse and alcohol dependence). Based on the hypothesis that addiction is a disorder due to aberrant learning, the researchers expect that inter-individual differences in learning behavior should be associated with liability for as well as resiliency against AUD. To test the hypotheses, the researchers will characterize 200 men at age 18, and then prospectively assess alcohol consumption and development of AUDs over a period of three years. At baseline, the researchers will study three clusters of predictive variables: (i) individual learning parameters, estimated by computational modeling of behavioral performance in learning tasks such as Pavlovian-to-instrumental transfer, probabilistic reversal learning, and habitization-devaluation; (ii) individual neural correlates of learning, assessed by functional brain imaging during learning; and (iii) already established risk factors such as family history of alcoholism and impulsivity. The specific aim is to test a set of related hypotheses. The researchers assume that high risk for AUD at baseline (cross-sectional design), increase of alcohol consumption after 3 years and incidence or progression of AUD during follow-up (prospective data) will be associated with decreased reward sensitivity, decreased punishment sensitivity, increased Pavlovian approach behavior ('sign tracking'), increased 'go' effect of conditioned appetitive stimuli, increased habitization, increased activation of striatal and prefrontal brain regions by the Pavlovian-to-instrumental transfer process, decreased correlation between striatal brain activity and prediction error during reversal learning.