Electroconvulsive therapy-induced motor seizure length trajectories and their association with clinical remission: a retrospective cohort study across diagnoses

Abstract

Background Electroconvulsive therapy (ECT) is a well-established treatment for severe psychiatric conditions, yet optimal process markers for predicting therapeutic response remain under investigation. Historically, motor seizure length has been considered a factor in ECT efficacy, but the relationship between seizure length and treatment response remains controversial. Methods We conducted a retrospective cohort study that included 331 inpatients who received ECT for acute psychiatric conditions between 2009 and 2015. Growth Mixture Modeling (GMM) was used to identify seizure length trajectories, while Latent Growth Curve Modeling (LGCM) examined predictors of changes. Clinical remission was assessed via the Clinical Global Impression (CGI) scale pre- and post-treatment. Results A single decreasing trajectory of seizure length over ECT sessions best fit the data. Baseline seizure length and session frequency were significantly associated with changes in seizure length. However, neither seizure length nor its reduction correlated with clinical remission. Conclusion Motor seizure length decreased across ECT sessions but was not associated with clinical remission in this study. Limitations include the single-site, retrospective design and the absence of EEG-based seizure measures, which may limit generalizability. These findings challenge the utility of seizure length as a predictive marker and underscore the need for more precise indicators to tailor ECT interventions.