Abstract
AIMS: Major depressive disorder (MDD) is a common psychiatric disorder whose causes and manifestations are diverse and numerous. To facilitate targeted therapeutic interventions, we characterized the abnormalities in effective connectivity within the cognitive-affective (CCN-AN) circuits to identify predictive biomarkers of TMS efficacy based on a large multicenter dataset and an independent dataset from patients receiving TMS. METHODS: Both functional and effective connectivity (FC, EC) were analyzed. As there was only one significant connection observed in FC, classification based on the differences in EC was performed using REST-meta-MDD. Furthermore, correlations between these abnormal connectivity and depression severity, as well as depression and suicidality alleviation, were calculated to determine their predictive implications for TMS efficacy using an independent dataset. RESULTS: Overall increased connectivity from the AN to the CCN and decreased connectivity from the CCN to the AN in MDD were observed using EC. These disruptions drove the classification accuracy up to 79.1%. Furthermore, the connection from the right inferior parietal lobule (IPL. R) to the right amygdala (AMYG.R) was negatively correlated with depression scores. Notably, the IPL connectivity to the anterior cingulate cortex (ACC) and the AMYG.R were closely correlated with depression and suicidal ideation alleviation following TMS treatment. CONCLUSIONS: These findings suggest that MDD is characterized by disruptions in both top-down and bottom-up emotion regulation systems. Notably, the key abnormal connectivities, particularly those from the IPL to ACC and AMYG, could predict the efficacy of TMS treatment. This insight refines MDD diagnosis and paves the way for more precise targeted interventions in the future.