Evolution of macrosleep in the acute and subacute phase after stroke

Introduction: A restorative sleep is essential for physical and psychological health. Even though literature is highly heterogeneous, stroke patients mostly tend to have a lower sleep efficiency, reduced N2, N3 and REM sleep, and generally a more fragmented sleep, especially in the acute phase. Crucially, there might be an association between an impaired sleep architecture and impaired cognitive outcome. As such, a poor sleep efficiency, a low amount of N2 and N3 and enhanced wakefulness after sleep onset have been linked to an acute impaired outcome. However, the specific role of macrosleep variations for the long-term recovery is still unknown. Therefore, we aim at observing the evolution of sleep macrostructure at several time points after stroke and relating them to the long-term clinical outcome. We will further apply transitional sleep scoring which might allow a more profound insight into sleep stability.

Materials and methods: We have measured so far 8 adult stroke patients (6 m, 2f; age±SEM 50.7±4.0 years; range: 27–60; 5 hemispheric, 2 bithalamic, 1 cerebellar) in the acute (day range: 3–17) and 4 of those patients in the subacute phase (day range: 72–120). We applied high-density EEG during a full night of sleep, and investigated the cognitive outcome by neuropsychological examination. We relate our data also to 2 healthy not-yet age-matched controls (31±2.1 years), having spent one night at the hospital. The transitional sleep scoring method classifies the sleep recording epoch- and time-independently, thereby allowing for a detailed and continuous scoring of sleep stages and transitions throughout the night.

Results: Our very preliminary analysis shows that in the acute phase, stroke patients, as compared to controls, descriptively tend to sleep less (values expressed as means±SEM; 64.3%3.9 vs. 75.3%±6.5), to be more awake (35.7%±3.9 vs. 24.7%±6.5), to have less N3 (10.6%±1.8 vs. 23.4%±0.5) and presumably less REM sleep (12.1%±2.2 vs. 16.6%±3.8). Patients might exhibit more sleep transitions than controls (211.6±20.7 vs. 157±38.2). In the subacute phase, patients sleep shows non-significant improvements (70.2%±35.1 sleep efficiency, 19.1%±9.5 of N3, 13.9%±6.9 of REM sleep). The classical sleep scoring captured less transitions (211.6±20.7) as compared to the transitional scoring (338.4±42.4) in the acute phase, while the percentage of sleep stayed almost similar (64.3%3.9 vs. 69.8%±3.6).

Conclusions: In line with current literature, our preliminary results seem to support the evidence that stroke patients may display variations in macrosleep, especially in the acute phase, having a lower sleep efficiency, less N3 and less REM sleep. Our data also show that all-night high-density EEG recordings are feasible in this hospital setting. Furthermore, transitional scoring may provide a more profound insight into sleep stability.

Acknowledgements: This study was funded by the SNF Sinergia Grant (Project number: 160803).