MAPPING DEPRESSION-RELATED NETWORK DYSFUNCTION ACROSS INTRINSIC NEURAL TIMESCALES

Neural systems operate across multiple intrinsic timescales supporting affective regulation, memory, and large-scale integration. Yet most resting-state functional magnetic resonance imaging (rs-fMRI) studies of major depressive disorder (MDD) analyze this temporal structure as broadband low-frequency (0.01–0.1 Hz) signals. In this study, we investigated whether depression-related network dysfunction is frequency-specific and may be obscured in conventional broadband analyses.
Resting-state fMRI data were obtained from the Transdiagnostic Connectome Project and included 21 individuals with MDD and 21 age- and sex-matched healthy controls. Whole-brain blood-oxygen-level-dependent (BOLD) signals were decomposed into intrinsic oscillatory components using two complementary multivariate approaches: multivariate variational mode decomposition (MVMD) and variational latent mode decomposition (VLMD). Frequency-resolved functional connectivity was then estimated within canonical slow-frequency bands.
Across both decomposition methods, the strongest group differences emerged in the Slow-5 (0.010–0.027 Hz) and Slow-4 (0.027–0.073 Hz) bands, primarily involving limbic, medial temporal, thalamic, and striatal networks central to emotion regulation and reward processing. These effects were statistically significant after permutation testing and false discovery rate correction (permutation p = 0.0093, FDR-corrected p = 0.0372) and showed a moderate effect size (Hedges’ g = 0.59). In contrast, broadband connectivity analyses revealed weaker and spatially diffuse group differences.
Our findings demonstrate that depression-related dysconnectivity is selectively expressed at specific intrinsic timescales, highlighting frequency-resolved connectivity as a more sensitive framework for detecting large-scale network alterations in MDD. This approach may improve neurobiological stratification of depression and inform future translational and clinical neuroimaging studies targeting network-level dysfunction.