Looking from the top: enhanced top-down sensorimotor processing in somatic anxiety

Abstract

AbstractFunctional neuroimaging research on anxiety has traditionally focused on brain networks associated with the psychological aspects of anxiety. In this study, instead, we target the somatic aspects of anxiety. Motivated by the growing appreciation that top-down cortical processing plays a crucial role in perception and action, we characterize effective connectivity among hierarchically organized sensorimotor regions and its association with (trait) anxiety. We selected 164 participants from the Human Connectome Project based on psychometric measures. We used their resting-state functional MRI data and Dynamic Causal Modeling (DCM) to assess effective connectivity within and between key regions in the exteroceptive, interoceptive, and motor hierarchy. Using hierarchical modeling of between-subject effects in DCM with Parametric Empirical Bayes we first established the architecture of effective connectivity in sensorimotor networks and then tested for an association with fear somatic arousal (FSA) and fear affect (FA) scores. To estimate the out of sample effect sizes, we implemented a leave-one-out cross validation analysis. At the group level, the top-down connections to exteroceptive cortices were inhibitory in nature whereas in interoceptive and motor cortices they were excitatory. With increasing FSA scores, the pattern of top-down effective connectivity was enhanced in all three networks: an observation that corroborates well with anxiety phenomenology. The effect sizes of anxiety associated changes in effective connectivity were sufficiently large to predict whether a new participant has mild or severe somatic anxiety. Interestingly, the enhancement in top-down processing in sensorimotor cortices were associated with FSA but not FA scores, thus establishing the (relative) dissociation between somatic and cognitive dimensions of anxiety. Overall, enhanced top-down effective connectivity in sensorimotor cortices emerges as a promising and quantifiable candidate marker of trait somatic anxiety. These results pave the way for a novel and quantitative approach to the neural underpinnings of anxiety, which speak to anxiety as an embodied phenomenon and the importance of top-down cortical processing.