Distinct predictors of short- versus long-term depression outcomes following electroconvulsive therapy

Abstract

Patients and clinicians considering electroconvulsive therapy (ECT) for treatment-resistant depression are faced with limited information about the likely long-term outcomes, and the individual characteristics that predict those outcomes. We aimed to identify sociodemographic and clinical predictors of acute ECT response and subsequent long-term depression severity. This prospective longitudinal study followed adult patients at a single academic ECT center. Among 114 participants, 105 completed an index ECT series and 70 were classified as acute ECT responders. Over a 2-year follow-up period, 82 subjects provided data on depression severity (Patient Health Questionnaire; PHQ-9). Better acute ECT response was predicted by less medication resistance, shorter index episode, and psychotic features (p < 0.05). PHQ-9 scores during the two-year follow-up period improved from baseline at all time points (p < 0.000001) but individual scores varied widely. Lower long-term PHQ-9 scores were predicted by better acute therapeutic response to ECT (p = 0.004) but not by ECT adverse effects (p > 0.05). Married status and greater baseline clinician-rated severity were not associated with acute ECT response but those variables did predict lower PHQ-9 scores longitudinally (p < 0.001), independent of other baseline features, initial ECT response, or intensity of ongoing treatment. These findings confirm previously identified predictors of short-term ECT response and demonstrate that distinct individual characteristics predict long-term depression outcomes. An individual's social context appears to strongly influence long-term but not short-term outcomes, suggesting a potential target for post-ECT therapeutic interventions.

Keywords: Electroconvulsive therapy; Long-term; Predictors; Social support; Treatment-resistant depression.