Enhancing Generalization of Virtual Reality Exposure for Public Speaking Anxiety Via Imagery-Based Extinction Memory Updating: A Randomized Controlled Trial

Exposure-based interventions are highly effective in the treatment of anxiety disorders; however, fear reduction achieved during exposure is often limited to the specific situations and stimuli encountered during treatment. As a result, fear may re-emerge when individuals face novel situations, a phenomenon commonly referred to as return of fear. This limited generalization is assumed to reflect the context- and stimulus-specific nature of extinction learning. While increasing variability across exposure contexts can enhance generalization, this approach typically requires longer treatment duration and increased therapeutic burden. Therefore, strategies that promote extinction generalization without increasing exposure time are of high clinical relevance.

Experimental laboratory research has demonstrated that extinction generalization can be enhanced by updating the extinction memory through mental imagery. Specifically, retrieving a recently acquired extinction memory and integrating novel, non-experienced stimuli or situations through imagination has been shown to reduce fear responses to both imagined and previously unencountered stimuli. This process is thought to increase the accessibility and robustness of inhibitory extinction memories by integrating new information into the existing memory network. The present study aims to translate this experimentally established mechanism into a clinically relevant intervention.

In this randomized controlled trial, individuals with elevated public speaking anxiety complete a standardized, single-session virtual reality (VR) speech exposure consisting of repeated short oral presentations across different VR contexts. Following exposure, participants are randomized to one of two imagery-based post-exposure procedures. In the control condition, participants engage in guided mental rehearsal of the previously experienced VR exposure situations. In the experimental condition, participants additionally update the extinction memory by vividly imagining having the same successful exposure experience in two novel public speaking situations: (1) giving a speech in front of a small real audience and (2) giving a speech in a VR environment with a substantially larger audience than experienced during exposure. These imagined situations are not directly experienced during the intervention but are intended to be integrated into the existing extinction memory through guided imagery.

Generalization of exposure effects is assessed 7-14 days after the intervention using a multi-context behavioral approach test (BAT). The BAT includes oral presentations in (a) situations that were directly experienced during VR exposure, (b) situations that were previously imagined during the extinction memory updating procedure, and (c) novel situations that were neither experienced nor imagined before. This design allows for the differentiation between responses to previously experienced, imagined, and entirely new public speaking contexts, thereby providing a stringent test of extinction generalization.

Primary outcomes include public speaking anxiety at follow-up as well as subjective fear ratings during each behavioral approach test, with a focus on differences between study groups across the different BAT contexts. Secondary outcomes include anticipatory anxiety prior to each BAT and behavioral performance indexed by speech duration. Baseline measures of anhedonia and positive emotionality are assessed to examine individual differences in treatment response.