Longitudinal cognitive and depressive symptom trajectories during ECT: the role of patient vs. treatment factors

Background

Electroconvulsive therapy (ECT) is currently one of the most effective treatments for severe depression. While short-term cognitive side effects and clinical outcomes following acute ECT have been widely studied[1], less is known about the factors that influence their trajectories over time. In particular, it remains unclear whether changes in cognition and depressive symptoms are primarily driven by treatment-related parameters or by patient-related characteristics.

Objectives

This study aimed to investigate factors influencing longitudinal trajectories of both cognitive performance and depressive symptoms during ECT.

Methods

We examined longitudinal changes in cognitive performance and depressive symptoms in patients with major depressive disorder (N = 72; mean age 53 years; 47% female) undergoing ECT. Patients received ~12 index treatment followed by maintenance treatment. Bifrontal ECT was administered under general anaesthesia (Remifentanil, Succinylcholine, Propofol/Etomidate). Assessments were conducted at five time points (baseline, mid-index treatment, post-index ECT, 2- and 6-month follow-up). Cognitive domains included global cognition (MoCA), verbal memory (RAVLT), attention and processing speed (DSST), working memory (n-back), cognitive flexibility (IE/ED), and subjective cognition (PDQ-5). Depressive symptoms were assessed using MADRS and BDI-II. Responder status (≥47% MADRS or ≥50% BDI reduction) was examined as a moderator. Multilevel models tested the influence of covariates (age, sex, education, seizure length, stimulation charge, number of treatments, and time since last ECT session in days) on cognitive and clinical trajectories.

Results

Significant time effects were observed across cognitive domains (all p < .001), with the exception of cognitive flexibility (ED-Cost), which did not change over time (p = .60). Hippocampal-associated functions showed transient declines during the acute phase with recovery to baseline (e.g., verbal memory: F(4,142.7) = 17.64, p < .001, η² ≈ 0.33), whereas prefrontal functions remained stable during the acute phase and improved significantly above baseline during maintenance treatment (e.g., working memory: F(4,141.3) = 6.91, p < .001, η² ≈ 0.16). Higher age predicted poorer cognitive performance (p ≤ .002), while higher education predicted better performance (p < .001) in all domains. Responder status and treatment parameters showed no effects on objective cognition (all p > .05). Subjective cognition complaints showed a significant time × responder interaction (F(4,148.0) = 2.67, p = .035, η² ≈ 0.07).

Depressive symptoms decreased significantly over time (BDI-II: F(4,155.6) = 19.75, p < .001, η² ≈ 0.34; MADRS: F(4,150.7) = 40.33, p < .001, η² ≈ 0.52). Age was associated with slightly lower symptom severity (β = −0.144, p = .043), while a higher number of acute treatments predicted higher symptom levels (β = 1.23, p = .045); all other treatment parameters were not significant.

Conclusions

Overall, these findings indicate that cognitive trajectories during ECT are primarily determined by patient characteristics (age and education), whereas depressive symptoms trajectories only showed contributions from age and number of acute treatments, and no significant influence of treatment-related parameters.

References

1. Landry M, Moreno A, Patry S, Potvin S, Lemasson M. Current Practices of Electroconvulsive Therapy in Mental Disorders: A Systematic Review and Meta-Analysis of Short and Long-Term Cognitive Effects. J ECT. 2021 Jun 1;37(2):119-127. doi: 10.1097/YCT.0000000000000723. PMID: 33009218.