Reduced functional segregation of insular connectivity profiles in depression

Major depressive disorder (MDD) is characterized by imbalances in homeostatic processes as reflected in altered neural function at rest. Within the insular cortex, distinct subregions have been associated with predictive or evaluative coding of homeostatic states. However, it is still elusive if functional segregation of insular subregions is impaired in MDD.To test for loss of specificity within functional connectivity (FC) profiles of insular subregions in MDD, we measured resting-state fMRI in 433 patients with MDD and 417 healthy control participants (HCP) and compared the similarity (Pearson correlation) and density (shift function) between 6 subregional FC profiles for MDD and HCP. To evaluate the clinical utility of similarity of FC profiles, we tested whether MDD can be classified on an individual level.We found greater similarity between FC profiles of insular subregions in MDD compared to HCP which was driven by a decrease in similarity and a greater FC density within the anterior insular cortex (AIC). On the individual level, insular FC profiles significantly predicted MDD, but the accuracy scaled with symptom severity.Loss of functional segregation and a greater FC density within the AIC may indicate alterations in interoceptive signals at rest in MDD. The increase in predictive power with increasing symptom severity substantiates insular functional segregation as an index of pathophysiology and supports a model in which prolonged deviations from homeostasis might drive individuals into an energy-deficient state that characterizes depression.