A graph theory-based whole brain approach to assess mood state differences in adolescents and young adults with bipolar disorder

Objectives Identifying hubs of brain dysfunction in adolescents and young adults with Bipolar I Disorder (BDAYA) could provide targets for early detection, prevention, and treatment. Previous neuroimaging studies across mood states of BDAYA are scarce and often examined limited brain regions potentially prohibiting detection of other important regions. We used a data-driven whole-brain Intrinsic Connectivity Distribution (ICD) approach to investigate dysconnectivity hubs across mood states in BDAYA. Methods Functional magnetic resonance imaging whole-brain ICD data were investigated for differences across four groups: BDAYA-depressed (n = 22), BDAYA-euthymic (n = 45), BDAYA-elevated (n = 24), and healthy controls (HC, n = 111). Clusters of ICD differences were assessed for regional dysconnectivity and mood symptom relationships. Analyses were also performed for BDAYA overall (vs. HC) ICD differences persisting across mood states. Results ICD was higher in the BDAYA- depressed group than other groups in bilateral ventral/rostral/dorsal prefrontal cortex (PFC) and right lenticular nucleus (LN) (p(corrected) <0.05). In BDAYA-depressed, functional connectivity (FC) was increased between these regions with their contralateral homologues and PFC-medial temporal FC was more negative (p < 0.005). PFC-related findings correlated with depression scores (p < 0.05). The overall BDAYA group showed ICD increases in more ventral left PFC and right cerebellum, present across euthymia and acute mood states. Conclusions This ICD approach supports a PFC hub of inter- and intra-hemispheric frontotemporal dysconnectivity in BDAYA with potential trait features and disturbances of higher magnitude during depression. Hubs were also revealed in LN and cerebellum, less common foci of BD research. The hubs are potential targets for early interventions to detect, prevent, and treat BD.