Direct Mapping of Interventions to Thought Features: A Bayesian Proof-of-Concept Study

Although uncontrollability is the core feature of perseverative thought that best accounts for its relationship to psychopathology, other features – for example, valence and content – have also been identified as potentially clinically-relevant in their own right. We describe results from a proof-of-concept study that examined the extent to which major underlying features of worry could be used to predict which of three common cognitive regulatory strategies (mindful acceptance; focused attention meditation; and thought suppression) would be helpful for regulating that worry. N = 40 adults selected for high trait worry (80% also met criteria for one or more DSM-5 anxiety-related diagnoses) generated and provided feature ratings for three idiographic thought topics. Participants then attempted to control each worry using each of the three strategies during functional magnetic resonance imaging (fMRI) in a within-subjects design (k = 468 observations). We used Bayesian multilevel modeling to test preregistered hypotheses regarding the extent to which each of five empirically-derived underlying dimensions of a worry (uncontrollability; negative valence; self-focus; apprehension; and social-memory content) could be used to predict which strategy would be most efficacious for regulating that worry. We did not find support for our preregistered hypotheses; however, in exploratory analyses, we found that mindfulness-based strategies were particularly effective compared to thought suppression for thoughts rated as higher (versus lower) in uncontrollability. Future research should test these principles in larger samples, using more diverse expressions of perseverative thought.