A decade-long comparison of depressive symptom networks in Asian patients with depressive disorders: Findings from REAP studies in 2023 and 2013

J Affect Disord. 2026 Mar 13:121616. doi: 10.1016/j.jad.2026.121616. Online ahead of print.

ABSTRACT

BACKGROUND: Network analysis provides a framework for examining interrelationships among symptoms beyond traditional categorical diagnostic models.

OBJECTIVE: Data were analyzed from two decade-apart, large-scale, multi-country surveys in Asia. Network structures of depressive symptoms among Asian patients with depressive disorders in 2023 and 2013 were compared.

METHODS: Ten depressive symptom domains, rated by clinicians in accordance with National Institute for Health and Care Excellence (NICE) guidelines, were modeled using the graphical Least Absolute Shrinkage and Selection Operator. Network structures were compared using the Network Comparison Test and evaluated with respect to expected influence centrality, stability, and clustering. Subgroup network structure comparisons were additionally conducted according to gender and geographic region.

RESULTS: The Network Comparison Test demonstrated significant differences in both network structure and global strength between 2013 and 2023. In 2023, among 2298 patients, depressive symptom networks comprised 38 connections, with the strongest associations observed between low self-confidence and guilt or self-blame and between disturbed sleep and poor or increased appetite; low self-confidence emerged as the most central symptom. In contrast, in 2013, among 1303 patients, 36 connections were identified, with persistent sadness and disturbed sleep forming the strongest association and persistent sadness representing the most central symptom. Substantial subgroup differences in network structure were also observed according to gender and region.

CONCLUSIONS: These differences may reflect contextual influences, including the COVID-19 pandemic and related sociocultural factors. The findings may further identify self-critical symptoms as promising targets for clinical intervention.

PMID:41833617 | DOI:10.1016/j.jad.2026.121616