ALTERED STATIC AND DYNAMIC RESTING-STATE CONNECTIVITY IN YOUTH AT RISK FOR BIPOLAR DISORDER

Bipolar disorder (BD) often emerges in adolescence and early adulthood, yet early identification of high-risk individuals remains challenging due to nonspecific symptoms. Although resting-state functional connectivity (RSFC) alterations have been reported in BD and major depressive disorder, evidence in high-risk populations is limited, and static approaches may overlook clinically relevant network dynamics.
In this multisite study, static and dynamic RSFC were examined in 233 help-seeking young adults from the German Early-BipoLife cohort (mean age 24.8 ± 4.4 years, range 17–35). High risk for BD was defined using established instruments. Resting-state networks were identified using group ICA within the NeuroMark framework (GIFT). 53 intrinsic connectivity networks across seven functional domains were extracted. Static connectivity was assessed using multivariate covariance models including risk status, current MDD, and affective symptom severity, while dynamic RSFC was evaluated using sliding window analysis and k-means clustering to identify recurring connectivity states.
Dynamic RSFC revealed symptom- and risk-related alterations confined to a single state. Within this state, MDD was associated with reduced sensorimotor–visual connectivity, whereas manic symptom severity was linked to increased sensorimotor–default-mode coupling. High-risk individuals spent significantly less time in this state than low-risks. In contrast, static RSFC showed significant multivariate effects of risk status (p = 0.008) and manic symptom severity (p = 0.007), without significant univariate differences.
These findings highlight dynamic RSFC as a sensitive marker of affective vulnerability beyond static connectivity, while advanced multimodal integration approaches may further reveal clinically relevant static network patterns not captured by conventional analyses.