Neural connectivity biotypes: Predictors of clinical outcomes and improvement patterns of iTBS treatment in adolescents and young adults with depression

Abstract

<p>Background The heterogeneity of depression limits the treatment outcomes of intermittent theta burst stimulation (iTBS) and hinders the identification of predictive factors. This study investigated functional network connectivity and predictors of iTBS treatment outcomes in adolescents and young adults with depression. Aim This study aimed to identify default mode network (DMN)-based connectivity patterns associated with varying iTBS treatment outcomes in depression. Methods Data from a randomised controlled trial of iTBS in depression (n=82) were analysed using a data-driven approach to classify homogeneous subgroups based on the DMN. Connectivity subgroups were compared on depressive symptoms and cognitive function at pretreatment and post-treatment. Furthermore, the predictive significance of baseline inflammatory cytokines on post-treatment outcomes was evaluated. Results Two distinct subgroups were identified. Subgroup 1 exhibited high heterogeneity and greater centrality in the posterior cingulate cortex and retrosplenial cortex, while subgroup 2 showed more homogeneous connectivity patterns and greater centrality in the temporoparietal junction and posterior inferior parietal lobule. No main effect for subgroup, treatment or subgroup×treatment interaction was revealed in the improvement of depressive symptoms. A significant subgroup×treatment interaction related to symbol coding improvement was detected (F=5.22, p=0.026). Within subgroup 1, the active group showed significantly greater improvement in symbol coding compared with the sham group (t=2.30, p=0.028), while baseline levels of interleukin-6 and C-reactive protein emerged as significant indicators for predicting improvements in symbolic coding (R 2 =0.35, RMSE (root-mean-square error)=5.72, p=0.013). Subgroup 2 showed no significant findings in terms of cognitive improvement or inflammatory cytokines predictions. Conclusions Data-driven network analyses offer valuable insights into iTBS treatment outcomes in depression, providing clues for predicting cognitive improvements from an inflammatory perspective.</p>