Spatiotemporal Dissociation of Human Amygdala Response to Negative Affect
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
The amygdala is central to negative affect and a primary target of top-down regulation, yet its temporal dynamics remain poorly characterized. Most fMRI studies model stimulus-evoked amygdala activity with a canonical hemodynamic response function and collapse across subregions, potentially obscuring functional heterogeneity in time and space. We used finite impulse response modeling in two independent datasets (combined N = 358) to characterize blood oxygenation level-dependent responses during negative picture viewing and instructed cognitive reappraisal, complemented by data-driven clustering of voxelwise time courses. Across two probabilistic atlases, six of seven amygdala subregions showed prolonged activation to negative stimuli but differed in temporal profile. The centromedial amygdala showed sustained activation extending into the post-stimulus rating period, whereas laterobasal and superficial subregions peaked earlier during stimulus presentation. Reappraisal did not substantially alter response magnitude or shape, indicating that amygdala downregulation is not a reliable consequence of instructed regulation under these conditions. Data-driven analysis identified four clusters with distinct temporal profiles corresponding to laterobasal, lateral, superficial, and centromedial territories, converging with established anatomical parcellations while crossing some conventional boundaries. FIR modeling thus reveals a temporal architecture of amygdala responses to negative affective stimuli, characterized by early laterobasal and superficial responses followed by sustained centromedial activity.
Significance Statement
Most human fMRI studies model the amygdala as a single region using a canonical hemodynamic response, potentially obscuring distinct temporal dynamics during affective processing. Across two independent cohorts, flexible time-course modeling revealed a reproducible sequence: laterobasal and superficial regions responded early to negative images, whereas centromedial activity persisted after image offset into the affective-evaluation period. Although cognitive reappraisal effectively reduced self-reported negative affect, it did not reliably alter these amygdala time courses. Data-driven clustering further recovered and extended this organization, showing that temporal response profiles align with, but are not fully constrained by, conventional anatomical boundaries.