Evaluation of polysomnography sleep parameters in patients with epilepsy

There is a well-established bidirectional link between sleep and epilepsy. Epileptic disorders can disrupt normal sleep architecture and contribute to a range of sleep disturbances. Conversely, specific features of sleep may influence seizure frequency and duration, as well as the occurrence of interictal epileptiform discharges captured on EEG. Although this interaction has long been recognized, the extent to which sleep affects seizure control and treatment response remains insufficiently clarified. The aim of this thesis is to explore how polysomnographic sleep parameters relate to seizure control and to examine whether alterations in rapid eye movement (REM) sleep may serve as a prognostic biomarker in individuals with epilepsy. The study included 20 patients with well-controlled idiopathic generalized epilepsy with generalized tonic–clonic seizures alone (13 women and 7 men; mean age 33.85 ± 10.00 years), 18 patients with poorly controlled idiopathic generalized epilepsy with generalized tonic–clonic seizures alone (9 women and 9 men; mean age 29.83 ± 11.37 years), and 17 healthy individuals who served as controls (7 women and 10 men; mean age 27.29 ± 10.21 years). Participants were recruited from the PSG–video EEG monitoring unit of the Neurology Clinic at Trakya University Faculty of Medicine between April 1, 2022, and January 1, 2025. All participants were hospitalized in the PSG–video EEG monitoring unit of the Neurology Clinic at Trakya University Faculty of Medicine between April 1, 2022 and January 1, 2025. Basic demographic and clinical characteristics, along with 24-hour polysomnographic–electroencephalographic data, were retrospectively collected for all participants. Patients who remained seizure-free during the first six months after PSG video-EEG monitoring were classified as having good seizure control, whereas those who experienced seizures were classified as having poor seizure control. These two patient groups were then compared with a healthy control group. In the six months and one year preceding polysomnography PSG video-EEG monitoring, patients with well-controlled epilepsy had significantly fewer seizures than those with poorly controlled epilepsy (p < 0.001). Although the percentage of sleep continuity was not associated with seizure control, it was significantly lower in patients with poorly controlled epilepsy than in healthy controls (p < 0.05). Although the proportion of N1 stage sleep was not associated with seizure control, it was significantly higher in the poorly controlled group compared with healthy controls (p < 0.05). REM latency was not a significant predictor of seizure control, yet it was prolonged across all epilepsy groups, with a more pronounced prolongation in patients with poorly controlled epilepsy than in healthy controls (p = 0,052). Although the total number of REM periods was not associated with seizure control, it was significantly lower in patients with poorly controlled epilepsy than in healthy controls (p < 0.05). Finally, the percentage of REM cycles was significantly lower in patients with poorly controlled epilepsy than in those with well-controlled epilepsy (p = 0,024). Our findings suggest that sleep architecture is notably altered in epilepsy, particularly among patients whose seizures are poorly controlled. In this group, both the percentage of rapid eye movement (REM) sleep and the number of REM cycles were reduced, and REM latency tended to be longer. By contrast, patients with better seizure control showed REM-related values that were largely similar to those of healthy controls, which implies that sleep patterns may return toward normal once seizures are stabilized. Taken together, these observations indicate that seizure control influences not only how often seizures occur but also the structure and physiology of sleep. In conclusion, changes in sleep architecture appear to be linked to the degree of seizure control, and several REM sleep parameters may hold promise as prognostic markers in individuals with epilepsy.