Thalamic Volumes and Functional Networks Linked With Self-Regulation Dysfunction in Major Depressive Disorder

Abstract

Aims: Self-regulation (SR) dysfunction is a crucial risk factor for major depressive disorder (MDD). However, neural substrates of SR linking MDD remain unclear.

Methods: Sixty-eight healthy controls and 75 MDD patients were recruited to complete regulatory orientation assessments with the Regulatory Focus Questionnaire (RFQ) and Regulatory Mode Questionnaire (RMQ). Nodal intra and inter-network functional connectivity (FC) was defined as FC sum within networks of 46 thalamic subnuclei (TS) or 88 AAL brain regions, and between the two networks separately. Group-level volumetric and functional difference were compared by two sample t-tests. Pearson's correlation analysis and mediation analysis were utilized to investigate the relationship among imaging parameters and the two behaviors. Canonical correlation analysis (CCA) was conducted to explore the inter-network FC mode of TS related to behavioral subscales. Network-based Statistics with machine learning combining powerful brain imaging features was applied to predict individual behavioral subscales.

Results: MDD patients showed no group-level volumetric difference in 46 TS but represented significant correlation of TS volume and nodal FC with behavioral subscales. Specially, inter-network FC of the orbital part of the right superior frontal gyrus and the left supplementary motor area mediated the correlation between RFQ/RMQ subscales and depressive severity. Furthermore, CCA identified how the two behaviors are linked via the inter-network FC mode of TS. More crucially, thalamic functional subnetworks could predict RFQ/RMQ subscales and psychomotor retardation for MDD individuals.

Conclusion: These findings provided neurological evidence for SR affecting depressive severity in the MDD patients and proposed potential biomarkers to identify the SR-based risk phenotype of MDD individuals.

Keywords: canonical correlation analysis; major depression disorder; network‐based statistic‐predict; self‐regulation; thalamic subnuclei.