Paradoxical relief after seizures: a diagnostic signal distinguishing functional/dissociative from epileptic seizures

To determine whether paradoxical relief, feeling unusually better after a seizure compared to before it, is more common after functional/dissociative seizures (FDS) than epileptic seizures (ES), quantify its diagnostic accuracy, and explore its relationship with preictal symptoms.

Methods

Consecutive patients admitted to a tertiary epilepsy unit for prolonged in-patient EEG monitoring underwent a structured clinical interview on admission, before final multidisciplinary diagnostic classification. Preictal dissociative and autonomic/somatic symptom burden was assessed using items adapted from established questionnaires. Diagnostic classification incorporated clinical history, seizure semiology, video-electroencephalography findings, and collateral information. Patients with dual or indeterminate diagnoses were excluded. Associations with paradoxical relief were examined using logistic regression, followed by an exploratory mediation analysis.

Results

Of 176 patients assessed, 66 with FDS and 65 with ES were included. Paradoxical relief was reported by 46/66 patients with FDS (69.7%) and 10/65 with ES (15.4%; unadjusted odds ratio [OR] 12.65, 95% confidence interval [CI] 5.57–31.09). As a diagnostic signal for FDS, paradoxical relief had 69.7% sensitivity (95% CI 57.1–80.4), 84.6% specificity (CI 73.5–92.4), a positive likelihood ratio of 4.53 (2.51–8.19), and a negative likelihood ratio of 0.36 (0.24–0.52). FDS diagnosis remained independently associated with paradoxical relief after adjustment (OR 10.59, 95% CI 3.42–38.06). In a parallel mediation analysis, dissociative symptom burden showed a significant indirect effect, accounting for 19.5% of the association between diagnostic group and relief, whereas the indirect effect through somatic/autonomic symptom burden was not significant.

Significance

Paradoxical relief is substantially more common after FDS than ES and may provide a simple, clinically useful diagnostic signal. Its absence does not exclude FDS, and the finding requires external validation. The association with dissociative symptoms is exploratory and supports prospective investigation of whether relief reflects transient resolution of a disturbed, disembodied preictal state.