Sleep disordered breathing (SDB) and ischemic stroke (IS) are intertwined pathologies. The topography of IS may interacts with SDB (vigilance, ventilatory and upper airway control).

To evaluate the association between stroke location and SDB after a first IS.

Prospective monocentric cohort of patients aged 18–85years, hospitalized in the Intensive Care Stroke Unit for a first IS, without any past neurovascular or sleep history. Stroke location was determined on the diagnostic brain imaging (magnetic resonance imaging or computerized tomography scan). At three months post-stroke (median delay [Q1;Q3]=134days [97;227]), patients underwent a clinical examination including validated sleep-related questionnaires and a full-night polysomnography (PSG), performed according to AASM recommendations.

One hundred and forty-two patients were included. Patients were classified according to the absence or the presence of SDB [Apnea-Hypopnea Index (AHI)≥15 events/hour of sleep]. Among them 53% had a SDB with a median AHI at 28/h [21; 43] (52% obstructive, 25% coexistent, 23% central). Post-IS patients were slightly symptomatic and did not differed (for SDB and non-SDB respectively, Epworth Sleepiness Scale 8 [4; 12] vs. 6 [4; 10] and Pichot Fatigue scale 8 [3; 18] vs. 11 [4; 20]). SDB patients were older (65years old [55; 71] vs. 53 [48; 63]), and predominantly male (71%). Comorbidities were similar in both groups except for atrial fibrillation that was more frequent in the SDB group (14 (19%) vs 3 (5%), P=0.01). The macrostructure of sleep was affected by SDB (in minute, light sleep=16 [8; 31] vs 9.6 [5; 22] P=0.01 and slow wave sleep=4 [0; 9] vs 12 [5; 17] P<0.001). The cerebellar lesions were more frequent in the SDB group: 19% vs 3% P=0.003.

Those preliminary results suggest that cerebellar location might be associated with moderate to severe SDB. Moreover, post-stroke patients with IS were more prone to have atrial fibrillation.