Altered brain activity in unipolar depression unveiled using connectomics

Abstract

Over 20 years of neuroimaging experiments into aberrant task-based brain activity in unipolar depression have failed to reliably delineate a convergent set of anatomical regions. Here we examined whether study-derived coordinates might delineate a dysfunctional brain network in unipolar depression rather than isolated neuroanatomical foci, utilizing data from 57 studies with 99 individual neuroimaging task-based experiments, testing either emotional or cognitive processing (n = 1,058). We further assessed clinical relevance by computing optimal network-based personalized targets in 26 individuals who previously received transcranial magnetic stimulation for unipolar depression. Although coordinates were neuroanatomically heterogeneous, they localized to highly robust distributed brain networks. Importantly, these networks closely recapitulated clinically meaningful and independently derived models of depression circuitry, quantified by spatial correlation (P < 0.00002). Therapeutic outcome of transcranial magnetic stimulation was dependent on how effectively this circuit was targeted (P = 0.018). These findings indicate that neuroimaging findings in depression, which previously appeared irreconcilable, localize to highly robust and clinically meaningful distributed brain networks. Cash et al. examined whether coordinates derived from neuroimaging studies might delineate a dysfunctional brain network in unipolar depression rather than isolated neuroanatomical foci. The authors found highly robust distributed brain networks that closely recapitulate clinically meaningful models of depression circuitry.