Yoon H.Kim H.S.Lee S.Choi T.Y.Jung S.W.Yoon H.J.Kim H.S.Yang H.J.Jeong N.Moon E.Baek J.H.Si T.M.Kallivayalil R.A.Tanra A.J.Nadoushan A.H.J.Chee K.Y.Javed A.Sim K.Pariwatcharakul P.Lin S.K.Sartorius N.Shinfuku N.Kato T.A.Kamali M.Nierenberg A.A.Park S.C.Mahidol University2026-06-212026-06-212026-08-01Asian Journal of Psychiatry Vol.122 (2026)18762018https://repository.li.mahidol.ac.th/handle/123456789/117448Objective: The investigation focused on differences in the overall network structures of depressive symptoms between patients with bipolar depression (BD) and those with unipolar depression (UD), emphasizing their unique symptom dynamics and centralities. Methods: Data from the Research on Asian Psychotropic Prescription Patterns for Antidepressants, Phase 3 (REAP-AD3), were used to estimate depressive symptom networks for 240 patients with BD and 2905 patients with UD. A Network Comparison Test (NCT) was conducted to evaluate differences in global strength, edge weights, and node centralities between the two networks. An additional NCT was performed using the same sample size in both groups. Results: Anhedonia emerged as the most central symptom in BD, while persistent sadness was the most central symptom in UD. Global strength was higher in the BD network in the full-sample NCT (p = 0.04), but not in the equal sample-size analysis (p = 0.20). However, no significant differences were identified in overall network structure invariance. Conclusions: These findings underscore distinct depressive symptom networks in BD and UD. Anhedonia and energy dysregulation were prominent in BD, whereas persistent sadness and self-rumination were more pronounced in UD. Despite the non-significance of other NCT results, the full-sample pairwise network comparison suggested that BD patients exhibit a more integrated symptom structure than UD patients, with stronger overall connectivity between symptoms, which may be linked to neurobiological distinctions such as widespread abnormalities in white matter connectivity and increased within-network connectivity in BD.PsychologyMedicineArticleSCOPUS10.1016/j.ajp.2026.1050622-s2.0-10504197133618762026