Trajectories of fear learning in healthy participants are able to distinguish groups that differ in individual characteristics, chronicity of fear and intrusions

Background and objectives: Studies on the development and treatment of anxiety disorders mostly focus on the comparison of predefined groups. An alternative approach is to use data-driven latent class growth analyses (LCGA) to determine differentiation between groups based on particular mechanistic factors. This study validated the use of LCGA on responses in a compact fear conditioning task and whether specific characteristics are associated with maladaptive fear learning trajectories. Methods: Healthy subjects (N = 300) completed a fear conditioning task that included uninstructed and instructed acquisition and extinction phases. Subjective fearfulness and US expectancy were used as outcome measures. Latent classes in the responses to the CS+ (coupled with a scream) and the CS- (control stimulus) were determined based on trajectories across the experimental phases. State and trait anxiety were measured during testing, and return of fear and intrusions were measured one and six weeks later. Results: Fear learning trajectories of poor extinction in responding to the CS+ and generalization of fear to the CS- were associated with higher state and trait anxiety. Individuals belonging to these trajectories reported more intrusions, fear and had higher US expectancy ratings after 1 week. Limitations: Only 56% of participants completed the six weeks follow-up measures. Conclusion: Fear learning trajectories are associated with individual characteristics, return of fear and intrusions. Next, this task will be implemented in clinical practice to assess its predictive power for the extent to which patients benefit from exposure treatments.