Back

Sleep fragmentation drives local, network-specific epileptic activity in human epilepsy

Frauscher, B.; Ho, A.; Matouskova, B.; Jaber, K.; Avigdor, T.; Minato, E.; Thomas, J.; Southwell, D. G.; Hall, J.; Klimes, P.; Peter-Derex, L.; Hannan, S.

2025-12-02 neuroscience
10.64898/2025.11.30.691386 bioRxiv
Show abstract

Although complex interactions between sleep and epilepsy have long been recognised, the directionality of the link between sleep fragmentation and epileptic activity remains unclear. We investigated causality in this relationship by experimentally manipulating sleep stability in individuals with drug-resistant focal epilepsy. Using combined stereo-electroencephalography (SEEG) and sleep recordings from 17 patients, alongside targeted auditory stimulation to induce arousals during non-rapid eye movement (NREM) sleep, we directly assessed how transient sleep disruption influences the occurrence and spatial propagation of interictal epileptiform discharges (IEDs) and examined key factors that modulate this relationship. We demonstrate that arousals acutely increase IED counts, with this effect being dependent on anatomical brain region, seizure-onset zone (SOZ) involvement, and sleep stage. Transient arousal-driven increases in IEDs were observed in neocortical regions, outside the SOZ and during NREM stage 2 sleep (N2), reflecting both regional and network-level specificity. Despite increasing IED counts, arousals did not influence IED propagation, indicating that sleep fragmentation selectively enhances local cortical excitability without engaging broader epileptic networks. Together, these findings highlight the critical role of sleep stability in shaping pathological activity, and support sleep stabilisation as a promising therapeutic strategy to reduce interictal spike burden during sleep and improve clinical outcomes.

Matching journals

The top 7 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.