Parsing avenues for future fear conditioning research
Freitag, 04.06.2021:
8:30 - 10:00

Chair der Sitzung: Roland Benoit, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig
Chair der Sitzung: Tina Lonsdorf, University Medical Center Hamburg-Eppendorf
Ort: From (epi)genetics to cognition

Zusammenfassung der Sitzung

There is a long tradition of using the experimental paradigm of ‘fear conditioning’ for studying emotional memory processes in general as well as for modelling the development and treatment of anxiety- and stress-related disorders in the laboratory. This symposium takes on some of the methodological and procedural challenges from the past and showcases recent advances for the future of fear conditioning research. First, Maren Klingelhöfer-Jens will scrutinize the rank-stability of skin conductance responses and fear ratings across multiple experimental days and repeated tests of a fear conditioning paradigm. Ann-Kristin Meyer will then take stock of the emerging literature on categorical fear conditioning by reporting the results of a meta-analysis and a replication study. She will argue that this procedure has the potential to bridge the gap between research on implicit and explicit memory systems. Afterwards, Erik Müller will present behavioral, psychophysiological, and EEG evidence that merely imagined events can induce conditioned fear responses much the same way as real US. Finally, Adrian Wroblewski will highlight the value of adopting a network approach for our understanding of the neural basis of fear conditioning. In particular, he will present effective connectivity analyses that make use of dynamic causal modelling to infer directed relationships between individual brain regions. Together, the symposium will present research that outlines emerging future lines of inquiry for the research on fear conditioning.


Assessing performance adjustments in and test-retest reliability of fear conditioning

Maren Klingelhöfer-Jens1, Vincent Keyaniyan1, Manuel Kuhn2, Tina Lonsdorf1

1Institute of Systems Neuroscience, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; 2Department of Psychiatry, Harvard Medical School, and Center for Depression, Anxiety and Stress Research, McLean Hospital, Belmont, MA, USA

Fear conditioning paradigms are an important tool and used frequently to investigate fear and anxiety disorders in the laboratory. However, stability of fear responses across experimental phases and repeated measurements are investigated scarcely. The former is relevant to the common practice to control responding in later experimental phases for previous responding while the latter provides information about test-retest reliability of fear conditioning experiments. In our preregistered assessment (, we analyzed stability of fear responses (i.e., CS discrimination, responses to CS+ and CS-) in two common outcome measures (i.e., SCRs and fear ratings) collected within a longitudinal fear conditioning experiment (day 1: acquisition; day 2: extinction, reinstatement, reinstatement-test). The experiment was conducted at a baseline (T0; n=107) and a 6-months follow up (T1; n=71) measurement. At T0, stabilities of responding in SCR and fear ratings were weak to moderate between acquisition, 24h fear recall and extinction, but less stability was observed between responding in these phases and responding at the reinstatement-test. CS discrimination in SCR was generally less stable than responses to the CSs. Longitudinally, stability was weak in SCRs and weak to moderate in fear ratings across all experimental phases with higher stability of CS responses as compared to CS discrimination. The empirical basis for performance adjustments in certain experimental phases might depend on the experimental phase at hand. The poor to moderate test-retest reliabilities leave room for individual variability, but might also hamper replicability of group-based results. Further implications will be discussed.

Category conditioning put to the test: A meta-analysis and successful replication

Ann-Kristin Meyer1, Tatjana Schmidt2, Davide F. Stramaccia1, Philipp C. Paulus1, Roland G. Benoit1

1MPI for Human Cognitive and Brain Sciences; 2Philipps- Universität Marburg

Emotional events form complex memories that engage both episodic and Pavlovian memory systems. However, these systems have largely been studied separately of each other, presumably also due to methodological challenges. Episodic memory is typically tested with a large number of stimuli (e.g., lists of words), whereas in fear conditioning, these are typically confined to just a couple (e.g., CS+ and CS-). The category conditioning procedure by Dunsmoor et al. (2012) has the potential to overcome this challenge. Here, several exemplars from one semantic category (e.g. animals; CS+) are paired with an aversive stimulus, whereas exemplars from another category (e.g. objects; CS-) are not. Intriguingly, after learning, also novel exemplars of the CS+ category - that had not previously been presented - elicit a threat response. However, so far only few research groups have used the procedure and employed methods in conditioning research have been quite heterogeneous. We thus aimed to establish the replicability of this effect. First, a meta-analysis of 13 published effect sizes (n = 234) revealed a moderate difference in skin conductance between the CS+ and CS- categories. This effect remained significant after initial attempts to adjust for publication bias. Second, in a replication study (n = 30), we assessed skin conductance and heart period response. Both measures showed significant differences during fear conditioning, extinction, and reinstatement. Our results thus corroborate the extant literature and confirm the potential of this procedure. It will allow us to study how suppression affects both episodic memory and conditioned fear.

Fear conditioning with an imagined unconditioned stimulus

Erik Müller, Matthias Sperl, Christian Panitz

Differentielle Psychologie, Philipps-Universität Marburg

In classical fear conditioning, neutral conditioned stimuli that have been paired with aversive physical unconditioned stimuli eventually trigger fear responses. Here, we test whether aversive mental images systematically paired with a CS may also cause de novo fear learning in the absence of any external aversive stimulation. In the first two experiments, N=45 and N=41 participants were first trained to produce aversive, neutral, or no imagery in response to one of three different visual imagery cues. In a subsequent imagery-based differential conditioning paradigm, each of the three cues systematically co-terminated with one of three different neutral faces. Although the face that was paired with the aversive imagery cue was never paired with aversive external stimuli or threat-related instructions, participants rated it as more arousing, unpleasant, and threatening and displayed relative fear bradycardia and fear-potentiated startle. In a third experiment, two groups underwent fear conditioning with an imagined (N=24) vs. a real (N=24) electric shock while EEG and ECG were recorded. In general, the pattern of electrophysiological and cardiac fear responses was qualitatively similar for both groups, suggesting partially overlapping neurobiological mechanisms of imagery-based and classical fear conditioning. These results could be relevant for understanding the development of fear and related disorders without trauma.

Network perspective on neural activation during fear acquisition and extinction training: theoretical considerations and future directions

Adrian Wroblewski

Department of Psychiatry and Psychotherapy, University of Marburg, Germany

Fear conditioning paradigms serve as a translational model for the development and treatment of anxiety disorders. In the past years, the neural correlates of fear conditioning have been investigated intensively, but the clinical benefit remains sparse. One possible explanation could be a lack of methodological variety to extend our theoretical framework about underlying neural mechanisms of fear and extinction learning. Existing models of threat-processing and the interplay between brain regions of the fear network in humans, in which the amygdala is thought to play a central role, are mainly based on verbal theories, but an experimental validation is still missing.

Thus, I will discuss theoretical considerations about the potential of computational connectivity analyses, to parse precisely the finely balanced interactions of brain structures associated with fear and extinction learning. Specifically, I will present a planned project in which we aim to use effective connectivity analyses (Dynamic Causal Modelling; DCM) to shed light on information processing during fear acquisition and extinction training. Using a two-day fear conditioning paradigm, we plan to apply regression DCM to study whole-brain directed connectivity during fear acquisition and extinction training, and combine it with model-based classical DCM to test specific hypotheses about the interactions within the fear network. This network approach allows to experimentally validate current hypotheses about neural mechanisms of threat processing and regulation (e.g., regulative function of the ventral PFC), by means of computational modelling.

Finally, I will give an outlook on possible extensions and applications of effective connectivity for future fear conditioning research.