Circadian Rhythms and Seizure Prediction in Epilepsy

Epilepsy, affecting over 50 million people worldwide, is characterised by unpredictable seizures that disrupt daily life and pose significant health risks. Research over the past decade has revealed that seizure incidence is not entirely random but follows endogenous biological cycles, chiefly circadian (∼24 h) and multidien (multi-day) rhythms. These rhythms reflect oscillations in neural excitability linked to molecular clocks, hormonal fluctuations and systemic physiological processes. By decoding the phase and amplitude of such cycles, clinicians and researchers aim to forecast periods of heightened seizure susceptibility, paving the way for tailored interventions, adaptive medication schedules and closed-loop neurostimulation. Advances in dynamical systems theory have further formalised seizure onset as a critical transition, marked by subtle warning signals in brain activity. Coupled with long-term electrophysiological recordings—via intracranial or subscalp devices—and non-invasive wearables, these insights are transforming epilepsy management from reactive treatment to proactive risk mitigation.

Research from Nature Portfolio

Recent studies have applied dynamical systems theory to hippocampal seizure onset, demonstrating that mathematical bifurcations predict critical transitions in both animal models and human intracranial EEG. Active probing of neural circuits, through controlled perturbations, uncovers latent signatures of excitability and delivers superior forecasting accuracy compared with passive recording alone. Complementary work has validated critical slowing down—an increase in signal variance and autocorrelation—as a reliable biomarker of impending seizure events, detectable across hours to days in long-term iEEG. Foundational investigations have also detailed how interictal epileptiform activity oscillates with both circadian and stable multidien periods (commonly 20–30 days), with seizures clustering on the rising phase of these rhythms. Integrating circadian and multidien phase information yields a composite biomarker that stratifies seizure risk with a large effect size, establishing a quantitative framework for personalised seizure forecasting.

Circadian Rhythms and Seizure Prediction in Epilepsy publication trend

The graph below shows the total number of articles in circadian rhythms and seizure prediction in epilepsy across all publications each year (not limited to Nature Index journals).

Technical terms

Circadian rhythm: An endogenous roughly 24-hour cycle in behavioural and physiological processes.

Multidien rhythm: Biological oscillation spanning multiple days, synchronised with cycles of neural excitability.

Interictal epileptiform activity (IEA): Abnormal electrical discharges observed between seizures, reflecting brain irritability.

Critical slowing down: Dynamical systems indicator marked by increased variance and autocorrelation preceding a state transition.

Phase-locking: Alignment of seizure events to a specific phase of an underlying biological rhythm.

Area under the receiver operating characteristic curve (AUC): Quantitative measure of a model’s discrimination ability between high- and low-risk states.

References

1. The critical dynamics of hippocampal seizures. Nature Communications (2024).
2. Critical slowing down as a biomarker for seizure susceptibility. Nature Communications (2020).
3. Multi-day rhythms modulate seizure risk in epilepsy. Nature Communications (2018).
4. Forecasting seizure likelihood from cycles of self-reported events and heart rate: a prospective pilot study. EBioMedicine (2023).
5. Resting-state background features demonstrate multidien cycles in long-term EEG device recordings. Brain Stimulation (2023).
6. A new era in electroencephalographic monitoring? Subscalp devices for ultra–long‐term recordings. Epilepsia (2020).