7 tesla multiscale entropy analysis reveals increased resting-state complexity in key regions for fear and anxiety in spider-fearful individuals.

Resting-state functional connectivity is limited in assessing the temporal dynamics of brain networks and, due to an insufficient signal-to-noise ratio, in detecting subtle changes at 3T MRI. Nonetheless, measures of complexity, which capitalize on temporal dynamics, have revealed alterations for some affective disorders at this field strength. Anxiety disorders have received only scant attention in this regard, despite indications of altered functional brain architecture in spider-fearful participants (SP). To address this gap, we probed resting-state complexity using 7T MRI, comparing 28 adults with SP with 45 healthy controls (HC). We computed multiscale entropy (MSE) on ten scales (1 - 0.1 Hz) for brain regions of the fear and anxiety networks in HC and SP. The MSE scales interacted with group (HC, SP) and brain area, revealing MSE increments in limbic regions in SP (versus HC). Whilst most MSE changes related to SP ranged between 1 and 0.33 Hz, the MSE of the bed nucleus of the stria terminalis (BNST), a nucleus involved in anxiety and the hormonal system, exhibited increases on all scales bar two for SP (i.e., for 0.5 - 0.125, and 0.1 Hz). MSE was also positively associated with SP severity (but not trait anxiety) in the BNST. Altogether, 7T fMRI detected elevated MSE in SP, indicating excessive intra-regional processing in brain regions key to fear and anxiety. The most pronounced effects were found in the BNST, corroborating its central role in the anxiety circuit.