Depressive disorder is one of the most common psychiatric illnesses. It is associated with high mortality and disability rates. Effective treatment of depression is critical to reducing its detrimental effects. In approximately one third of patients with refractory depression, symptoms cannot be relieved by regular antidepressants(Fava, 2003). Electroconvulsive therapy (ECT) has been documented to reverse refractory depression with high efficacy(Bahji et al., 2019; Ross et al., 2018). Although some previous investigations disclosed structural and functional changes in frontal, temporal, and parietal cortices in depression patients after ECT(Dukart et al., 2014; Joshi et al., 2016; Perrin et al., 2012). However, despite its remarkable efficacy, the neural substrate of ECT remains elusive.

As a cardinal feature of refractory depression, depressive rumination is characterized as repetitive, uncontrollable, self-referencing thoughts about negative mood states. Ruminative thinking is associated with negative information and reduced problem-solving ability(Kaiser et al., 2018). Previous experimental studies have implicated rumination in the onset, maintenance, and adverse outcomes of depression(Disner et al., 2011; Enns and Cox, 2005).

Therefore, the development of effective treatments may be supported by assessment of rumination in depressed individuals. Measured by the self-reported Ruminative Response Scale (RRS), rumination was defined based on three distinct factors: depression-related, brooding, and reflective rumination. Brooding rumination is conceptualized as “a passive comparison of one's current situation with some unachieved standard,” whereas reflective rumination is “a purposeful turning inwards to engage in cognitive problem-solving to alleviate one's depressive symptoms”(Treynor, 2003). Brooding rumination examines current distress from a self-critical evaluative viewpoint. Reflective rumination examines current distress from a more neutral and distanced perspective(Kross, 2012). Rumination is known to be associated with activity of the default mode network (DMN), which includes the precuneus and the medial frontal cortex(Hamilton et al., 2015). A recent brain connectome study reported rumination network dysfunction in depressive populations(Zhang et al., 2020). In addition, Watkins and Nolen-Hoeksema have proposed that reward-related brain activity is involved in the ruminative response(Watkins and Nolen-Hoeksema, 2014).

Resting-state functional magnetic resonance imaging (fMRI) is a non-invasive and convenient technique to reveal intrinsic brain activity. Degree centrality (DC) is a voxel-wise data-driven graph-theoretical approach which can identify brain networks and calculate the number of functional connections(Buckner et al., 2009; Lohmann et al., 2010). DC can directly reflect the functional connectivity (FC) capability between different voxels. Higher DC values are considered to indicate stronger functional connections with other voxels and conveyal of more information. Currently, considerable studies have used DC analysis to reveal the pathogenesis of psychiatric diseases, including depression. A recent study reported increased DC values in the prefrontal cortex and improved identified areas after antidepressant treatment(Zhang et al., 2021). Previous research related to DC has reported that the angular gyrus (AG) is implicated in the antidepressive effect of ECT(Wei et al., 2018). A neuroimaging study has confirmed that abnormal FC between the precuneus and the orbitofrontal cortex is associated with ruminative activity(Cheng et al., 2018). However, few studies have explored the neural substrate of ECT in relieving rumination.

In the present study, we analyzed the efficacy of ECT in ameliorating rumination and we investigated the underlying mechanism from a neural imaging perspective. We primarily made a longitudinal comparison of RRS scores in depressive patients. Subsequently, we contrasted the DC in depression patients before and after ECT with that of healthy controls (HC) over corresponding time intervals, aiming to identify central brain changes linked to ECT. Further, we examined the relationship between DC in identified areas and RRS scores. Based on previous evidence, we speculate that the beneficial effects of ECT on reward-related networks and the DMN are associated with rumination levels.