Postictal psychosis in temporal lobe epilepsy: a case-control study

To investigate the prevalence of postictal psychosis (PP) in patients with temporal lobe epilepsy (TLE) and to estimate the predictive value of various variables for the development of PP.

By retrospectively reviewing the charts of all patients evaluated with video-electroencephalogram (EEG)-monitoring at our unit between January 1995 and February 2012, we identified 684 patients with TLE, of which 48 patients had a history of PP. Patients with TLE and PP were compared with 200 controls (patients with TLE without a psychotic history) on demographic, clinical, EEG and magnetic resonance imaging (MRI) variables.

The prevalence of PP in our TLE sample was 7.0%. Aggressive behaviour during PP was present in 22.9% of the sample. Univariate analysis revealed that PP was significantly associated with early age at epilepsy onset (P = 0.007), longer duration of epilepsy (P = 0.002), presence of ictal fear (P = 0.005), impaired intellectual function (P = 0.045), and bilateral ictal and interictal epileptiform activity (both P < 0.0001). Using logistic regression analysis, ictal fear [odds ratio (OR) 2.88; P = 0.015] and bilateral interictal EEG activity (OR 6.40; P < 0.0001) were predictive of PP development. No association of PP with MRI pathology or epilepsy-relevant aetiological factors was found.

PP is a frequent and potentially dangerous complication within the course of TLE. Bilateral or widespread functional central nervous system disturbances rather than distinct structural brain alterations or certain predisposing aetiologies of epilepsy appear to be a risk factor for the development of PP. Ictal fear may be a predictive clinical variable of PP in TLE.