You are here:
Background and Aims
Alcohol dependence is characterised by failures of choice. People drink despite severe negative consequences; other pleasures are hardly rewarding to them. A crucial unanswered question is the genesis of such choice anomalies. For instance, both insensitivity to negative consequences and inability to respond to rewards unrelated to drugs could give rise to maladaptive behaviors. The question we wish to address in this work is the role learning mechanisms per se play in the aetiology of alcohol addiction. Our approach will rest on computational characterisations of learning mechanisms that have been highly successful in teasing apart separate neurobiological contributions by structures thought to be involved in the development and maintenance of addiction. At a first level, we will combine multimodal imaging techniques with computational measurements of behavioral changes consequent on rewards and punishments. We will assess computationally and neurobiologically distinct and well circumscribed learning mechanisms. Our focus will mainly be on the acquisition of Pavlovian values for stimuli; on the effects of Pavlovian predictions on other types of choices; and on the habitization of behavior. The multimodal imaging techniques will allow both the localisation and neurochemical identification of learning mechanisms in addictive behaviors. At a second level, we acknowledge the abstract nature of reinforcement learning techniques: a century of animal literature has taught us that rewards are not just rewards and that conditioned responses are highly sensitive to the precise nature of both unconditioned and conditioned stimuli. We will thus compare learning mechanisms in scenarios involving stimuli that are either relevant or irrelevant to the addictive behavior. This involves 1) learning about cues typically predictive of drugs; and 2) learning in the presence of the drug.
Our long-term aim is to further improve treatment and prevention of alcohol addiction. Four factors will ensure maximal clinical impact: First, we propose to examine these associations at a neurobiological and behavioral level in prospective investigations of (1) a representative at-risk population of young adults and (2) in a group of alcohol-dependent patients, both in a cross-sectional manner and in a longitudinal design aiming at the prediction of the emergence and recurrence of addictive behavior. Second, based on a comprehensive cross-sectional and longitudinal neuropsychological and psychopathological characterization of subjects/patients, we will link the neurobiological findings on learning mechanisms to specific cognitive, affective and behavioral dysfunctions and learning under the influence of alcohol. Third, we will explore and test which of these dysfunctions and mechanisms are promising novel targets for interventions, resulting in the development and testing of specific pharmacological interventions for the targeted (in terms of specific individual profiles of dysfunctions) treatment of alcohol dependence in a junior group. Fourth, all our work will be flanked multivariate pattern analyses which may help to refine both patient profiles for therapeutic interventions and neurobiological and neurochemical findings. Finally, the group and work plan are carefully titrated to ensure cross-fertilization between these approaches. This will mark a major step towards individualized prevention and therapy of alcoholism.
Cooperation: Bernstein Center for Computational Neuroscience (Prof. Dr. John-Dylan Haynes), Charité Berlin (Prof. Dr. Felix Bermpohl, Dr. Ralph Buchert, Prof. Dr. Jürgen Gallinat, Prof. Dr. Andreas Heinz, Prof. Dr. Rainer Hellweg, PD Dr. Michael Rapp, Prof. Dr. Philipp Sterzer, Dr. Florian Schlagenhauf, Dr. Henrik Walter), Freie Universität Berlin (Prof. Dr. Hauke Heekeren), Technische Universität Berlin (Prof. Dr. Angela Ittel), Technische Universität Dresden (Prof. Dr. Gerhard Bühringer, Prof. Dr. Michael Smolka, Dr. Inge Mick, Dipl.-Psych. Sören Paul, Dipl.-Psych. Mirjam Petersen), Universitätsklinikum Carl Gustav Carus Dresden (Prof. Dr. Dr. Michael Bauer, Dr. Andrea Kobiella, PD Dr. Ulrich Zimmermann), University College London (Dr. Quentin Huys)