Predicting dimensional antidepressant response to repetitive transcranial magnetic stimulation using pretreatment resting-state functional connectivity

Abstract

Repetitive transcranial magnetic stimulation is an effective treatment for depression that modulates resting-state functional connectivity (RSFC) of depression-relevant neural circuits. So far, however, few studies have investigated whether individual treatment-related symptom changes are predictable from pretreatment RSFC. Here we use machine learning to predict dimensional changes in depressive symptoms using pretreatment RSFC. We hypothesized that changes in dimensional depressive symptoms would be predicted more accurately than scale total scores. Patients with depression (n = 26) underwent pretreatment RSFC magnetic resonance imaging. Depressive symptoms were assessed with the 17-item Hamilton Depression Rating Scale (HDRS-17). Random forest regression models were trained to predict treatment-related symptom changes captured by the HDRS-17, HDRS-6 and three previously identified HDRS subscales: core mood and anhedonia (CMA), somatic disturbances and insomnia. Changes along the CMA, HDRS-17 and HDRS-6 were predicted significantly above chance, with 9%, 2% and 2% of out-of-sample outcome variance explained, respectively (all P values <0.001). CMA changes were predicted more accurately than the HDRS-17 (P < 0.05). Higher baseline global connectivity (GC) of default mode network subregions and the somatomotor network predicted poorer outcomes, while higher GC of the right dorsal attention frontoparietal control and visual networks predicted reduced CMA symptoms. HDRS-17 and HDRS-6 changes were predicted with similar GC patterns. These results suggest that RSFC spanning the default mode, somatomotor, dorsal attention, frontoparietal control and visual network subregions predict dimensional changes with significantly greater accuracy than syndromal changes after repetitive transcranial magnetic stimulation. These findings highlight the need to assess more granular clinical dimensions in therapeutic studies and echo earlier studies supporting that dimensional outcomes improve model accuracy. This study investigates the predictability of treatment-related symptom changes in depression using pretreatment resting-state functional connectivity. Machine learning models demonstrated significant accuracy in forecasting dimensional symptom changes, emphasizing the importance of assessing granular clinical dimensions for improved therapeutic outcomes.