Wenfeng ZHEN, Han WU, Shuo LIN, Dandi ZHU, Siyuan LIAN, Li REN, Yilang TANG, Yanping REN, Xin MA
Alpha Psychiatry - 2026;27(5):51154-0
Background: Late-life depression (LLD) is commonly complicated by comorbid anxiety. However, whether the severity of baseline anxiety is associated with differential symptomatic and cognitive outcomes following electroconvulsive therapy (ECT) in LLD remains unclear, particularly with respect to cognitive safety in high-anxiety older patients. Methods: This prospective self-controlled pre-post study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. Ninety-three patients aged >=60 years with LLD received an acute course of ECT and were stratified into mild-to-moderate anxiety (Hamilton Anxiety Rating Scale [HAMA] score 14-28) and severe anxiety (HAMA score >=29) groups according to established HAMA severity cutoffs. Depressive symptoms (24-item Hamilton Depression Rating Scale [HAMD-24]), anxiety symptoms (HAMA-14), and cognitive performance (Montreal Cognitive Assessment [MoCA] and Repeatable Battery for the Assessment of Neuropsychological Status [RBANS]) were assessed at baseline and after completion of ECT treatment. Linear mixed-effects models (LMMs) with participant-level random intercepts were used to examine differential treatment trajectories while adjusting for demographic characteristics and baseline symptom severity. Hierarchical regression analyses were performed to determine whether improvement in anxiety symptoms (DeltaHAMA) predicted antidepressant response (DeltaHAMD) and cognitive change (DeltaMoCA). Results: Both groups demonstrated significant reductions in depressive and anxiety symptoms following ECT. Baseline HAMA and HAMD-24 scores were not significantly correlated (r = 0.12, p = 0.24), suggesting that anxiety and depressive symptom severity represented distinct constructs. After multivariable adjustment, the severe anxiety group showed an antidepressant response comparable to that of the mild-to-moderate anxiety group (HAMD-24 total score interaction, p = 0.155), but exhibited significantly greater improvements in HAMA total and subdomain scores (all p < 0.01), as well as in the HAMD anxiety/somatization and retardation factors (both p < 0.05). Global cognitive performance remained stable across groups, with no significant differences observed in MoCA or RBANS total scores (both p > 0.60). Improvement in the RBANS attention domain was the only cognitive outcome that differed significantly between groups, favoring the severe anxiety group (p = 0.004). Greater improvement in anxiety symptoms predicted a stronger antidepressant response (beta = 0.31, p < 0.001); however, this association did not differ between anxiety severity groups (p = 0.316) and was not related to cognitive change (p = 0.443). Conclusions: In patients with LLD, more severe baseline anxiety does not predict a greater overall antidepressant response to ECT but is associated with preferential improvement in anxiety-related and psychomotor symptom dimensions. ECT does not adversely affect global cognitive functioning regardless of baseline anxiety severity. These findings suggest that anxiety severity may help inform expectations regarding domain-specific symptomatic improvement and support the cognitive safety of ECT across the spectrum of anxiety severity in older adults.
