Functional network organization for early identification of bipolar disorder in late adolescents and young adults...

Bipolar disorder (BD) is often misdiagnosed as major depressive disorder (MDD), especially in late adolescence and young adulthood, due to delayed emergence of distinguishing symptoms. This prospective nested case-control study aimed to identify neurobiological markers for early identification of BD from MDD in this age group.

The study comprised 139 patients with bipolar depressive disorder (BDD), 148 patients with unipolar depression (UD) and 128 healthy controls (HC). During follow-up, we additionally identified 62 patients who transitioned from MDD to BD (tBD). All participants underwent resting-state functional magnetic resonance imaging (rs-fMRI) at baseline. Functional network analyses were performed based on large-scale brain networks, along with graph-theoretical analyses.

Patients with depressive episodes showed reduced FC between the sensorimotor network (SMN) and visual network (VN) and within the subcortical network (SubN). Compared to HC, all patient groups showed reduced assortativity. Sigma was elevated in BDD and tBD, with reduced maximum sparsity in BDD and increased gamma in tBD. The UD showed reduced clustering coefficient and local efficiency. Compared to UD, BDD and tBD showed higher sigma and gamma, and tBD also showed higher local efficiency. In tBD, network metrics were associated with depressive severity, anxiety, suicide risk, and aggression.

BDD and tBD showed highly modular organization in the SMN, VN and SubN, whereas UD was characterized by reduced local efficiency. These patterns may represent neural signatures with diagnostic potential for early identification of BD from depressive episodes in late adolescence and early adulthood.