Spatiotemporal asymmetries on brain energy landscape uncover system entrapment related to depression severity

Abstract We investigate the spatiotemporal organization of resting-state brain activity in individuals with and without major depressive disorder (MDD), identifying stable and recurring whole-brain functional co-activation patterns that serve as attractor-like configurations. A particularly prominent brain state, marked by suppressed default mode and frontoparietal networks and heightened salience system engagement, occurring more frequently and with shorter dwell times in MDD and correlating with greater anhedonia severity. Transition dynamics further reveal that MDD participants exhibit reduced transitions between visual-attentional and limbic-default mode systems, which is associated with higher overall depression symptoms, suggestive of affective and cognitive rigidity. By evaluating the dynamic properties in relation to white matter architecture, we find that MDD individuals preferentially transition along energetically costly trajectories, particularly from salience-reactive to introspective states, despite the presence of structurally facilitated alternatives, implicating inefficient structure-function coupling. Finally, characterizing the energetic asymmetries of entry and exit transitions uncovers the landscape on which the brain travels between source- and sink-like attractors, with MDD dynamics disproportionately becoming trapped between a local maximum and a deep basin. These results offer a mechanistic account of how depression may emerge from maladaptive state trajectories within an energetically imbalanced neural landscape.