Impaired connectivity and topology of functional networks in recurrent major depressive disorder with and without suicidal ideation

Suicidal ideation is a severe manifestation of major depressive disorder (MDD), particularly among patients with recurrent episodes. Investigating functional network impairments in recurrent MDD patients with suicidal ideation (MDDSI) may aid in clinical diagnosis. This study aimed to examine connectivity and topological alterations in MDDSI and evaluate their potential as diagnostic imaging markers. Participants included 94 MDDSI, 53 recurrent MDD patients without suicidal ideation (MDDNSI), and 84 healthy control (HC). We used network-based statistics and graph theory analyses to identify aberrant connectivity and topology. Classification performance of features showing significant between-group differences was subsequently tested at the individual level. MDDSI exhibited more widespread dysconnectivity among networks than MDDNSI when compared with HC, including decreased connectivity between default mode network (DMN) and visual network (VN), between DMN and sensorimotor network (SMN), and between SMN and dorsal attention network (DAN). MDDSI and MDDNSI had shared dysconnectivity between VN and SMN, and within SMN. Additionally, MDDSI showed reduced connectivity between VN and DAN, and between DMN and SMN relative to MDDNSI, which were negatively correlated with SI severity. Although MDDSI displayed nodal topological alterations in occipital lobe relative to MDDNSI and HC, this difference did not reach statistical significance. The individual-level classification combining connectivity and topological metrics outperformed classifiers using either metric type alone, achieving an accuracy of 0.74–0.78 in identifying MDDSI patients. Our findings reveal impaired connectivity of functional networks in MDDSI. A combination of connectivity and topological metrics may hold potential as imaging markers to aid clinical diagnosis. Not applicable.