Effects of the retrieval-extinction paradigm with abstract reminders on fear memory extinction

Reactivated fear memories may enter an unstable state that could be interrupted by behavioral interventions such as the retrieval-extinction procedure. Studies of pharmacological interventions have shown that retrieval cues are not necessarily an exact reduplication of initial conditioned stimuli (CSs); they can instead be abstract word cues. However, the effectiveness of using abstract reminder cues in behavioral intervention procedures remains unknown, as do the conditions under which they are effective. The typicality of CS and the abstraction level of reminders are also likely to be important in the paradigm but have not been investigated to date. Here, in three experiments (n = 107), we manipulate the fear conditioning process, including single CS-unconditioned stimulus (US) and multiple CS-US connections, as well as the CS typicality and abstraction level of reminder cues, to explore the conditions for adopting retrieval-extinction with abstract reminders. The results of Experiments 1 and 2 show that the abstract cue is ineffective for use as a reminder in the procedure; this is the case for single exemplar category learning and perceptual discriminative learning, independent of the typicality of CS+ or the abstraction level of reminder cues. Only the outcome of Experiment 3, which contains multiple CS-US connections, shows no indication of fear return. The limitations of the study and possible interpretations of the results are discussed.