Tracing the Dynamics and Content of Ruminative Thought using the FAST Paradigm

In recent years, it has been emphasized the importance of studying cognitive processes in mental health using paradigms that allow to track the dynamics of the stream of thought. In this study, we investigated how rumination, depression, and anxiety symptoms relate to the content and dynamics of self-generated thought using the Free Association Semantic Task (FAST). Ninety-one participants completed the FAST task online, generating chains of concepts from seed words and rating each for valence and self-reference. Self-report measures of rumination, depression and anxiety symptoms were collected. We found that higher levels of these symptoms predicted the generation of more negatively valenced concepts and greater self-reference for negative content. Markov chain analysis revealed that individuals with higher rumination and symptomatology were more likely to transition from positive to negative states and to persist in negative states. Using natural language processing, we found that participants with elevated rumination, depression and anxiety symptoms produced concepts that were semantically closer to core beliefs characteristic of these disorders, as proposed by cognitive theories. These findings support the view that rumination is a rigid and sticky form of cognition, constrained to negative affective states, and highlight the utility of combining computational approaches with free association paradigms in the study of psychopathology.