Author:

Abstract:

English Summary Beyond probability: How fear is shaped by intensity and specificity of danger. Adaptations of the Pavlovian fear conditioning procedure in humans. Kim Haesen Promotors: Prof. Dr. Bram Vervliet & Prof. Dr. Frank Baeyens Copromotor: Prof. Dr. Tom Beckers Centre for The Psychology of Learning and Experimental Psychopathology Fear can be understood as the emotional correlate of the expectation of danger. Three components of this expectation shape the level of fear: Where is the danger? How likely is the danger? How bad is the danger? Thus, fear intensity is a function of the estimated specificity of the situation in which the danger is expected to occur (S), the estimated probability that the danger will occur in this situation (P) and the estimated intensity or aversiveness of the expected danger (I). Understanding how fear works requires a detailed understanding of these three components and their interaction, which we have called the SPI-framework of fear. In this dissertation, we have started this investigation by focusing on the specificity and intensity factor. Previous research on Pavlovian fear conditioning has focused mostly on the probability-factor, by investigating how breaking the contingency between the situation (the conditional stimulus, CS) and the aversive event (unconditional stimulus, US) leads to extinction of the fear elicited by the CS. This is viewed as a laboratory model for exposure-based psychotherapy for anxiety disorders and is therefore used as a screening model for novel behavioral or pharmacological adjuncts to exposure treatments. In our first research line, we have compared the standard CS-exposure (extinction) intervention (aimed at fear reduction through lowering the US-probability) with a US-exposure intervention (aimed at fear reduction through lowering the expected US intensity). We found that US-exposures outperformed standard CS-exposures: the achieved fear reduction generalized across contexts in the former, but not the latter. An additional study refuted the possibility that US-exposures lower fear by influencing the estimated CS—US probability, hence supporting the notion that fear reductions resulted from a change in expected US intensity. In the second research line, we developed a novel procedure to investigate fast-specific fear learning, by borrowing the one-trial overshadowing procedure from basic learning research in animals and applying it to human fear conditioning. We found in healthy individuals that, even after as little as a single conditioning trial, fear is constricted to the exact stimulus configuration that is paired with the aversive US, with subsets of the configuration activating only lower levels of fear. Anxiety patients failed to show this fast specific fear-learning pattern. Clinically, these findings suggest that fear extinction treatments may be optimized by including interventions aimed at US devaluation (Research Line 1) and that pathological overgeneralization of fear may be related to deficits in associative/configural processing of fearful situations (Research Line 2). This showcases how the SPI-framework is fruitful for expanding and innovating the domain of Pavlovian fear conditioning and its implications for clinical practice.