Back

Impaired Associative Memory, Inference, and Theta Dynamics in Postictal Psychosis of Epilepsy

Dworkin, A.; Wang, D.; Jimenez, D.; Ravenscroft, C.; Turco, F.; Johnson, C.; Chowdhury, F. A.; Pizarro, J.; Walker, M.; Balestrini, S.; Bush, D.; Vivekananda, U.

2026-01-26 neuroscience
10.64898/2026.01.23.701370 bioRxiv
Show abstract

Postictal psychosis (PIP) is a severe complication occurring in 2% of people with epilepsy (PWE) whose underlying pathophysiology remains poorly understood. Although historically considered separate from other forms of psychosis, newer evidence demonstrates a shared genetic susceptibility. People with schizophrenia are typically impaired at both associative learning and inferring connections between overlapping associations. Successful associative encoding, retrieval, and inference can each be predicted by changes in frontotemporal theta band activity, which is impaired in rodent models and people with schizophrenia. Here, we recorded high-density scalp EEG from PWE with history of PIP and well-matched control participants while they undertook a memory inference task. We found that associative memory and inference were both impaired in the PIP group, despite no difference in item recognition. Moreover, we found disrupted theta activity during memory encoding and the retrieval of inferred associations in PWE with PIP that likely originated from the medial temporal and frontal lobes. These results suggest a pattern of behavioural deficits and altered neural dynamics common to both PIP and schizophrenia. Interpreted in conjunction with previous genetic studies, they may reflect shared neural mechanisms contributing to psychopathology in both conditions and argue that PIP is a model of more general psychoses.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.