Functional recovery after ischemic stroke: Impact of different sleep health parameters

Sleep disturbances after ischaemic stroke include alterations of sleep architecture, obstructive sleep apnea, restless legs syndrome, daytime sleepiness and insomnia. Our aim was to explore their impacts on functional outcomes at month 3 after stroke, and to assess the benefit of continuous positive airway pressure in patients with severe obstructive sleep apnea. Ninety patients with supra‐tentorial ischaemic stroke underwent clinical screening for sleep disorders and polysomnography at day 15 ± 4 after stroke in a multisite study. Patients with severe obstructive apnea (apnea–hypopnea index ≥ 30 per hr) were randomized into two groups: continuous positive airway pressure‐treated and sham (1:1 ratio). Functional independence was assessed with the Barthel Index at month 3 after stroke in function of apnea–hypopnea index severity and treatment group. Secondary objectives were disability (modified Rankin score) and National Institute of Health Stroke Scale according to apnea–hypopnea index. Sixty‐one patients (71.8 years, 42.6% men) completed the study: 51 (83.6%) had obstructive apnea (21.3% severe apnea), 10 (16.7%) daytime sleepiness, 13 (24.1%) insomnia, 3 (5.7%) depression, and 20 (34.5%) restless legs syndrome. Barthel Index, modified Rankin score and Stroke Scale were similar at baseline and 3 months post‐stroke in the different obstructive sleep apnea groups. Changes at 3 months in those three scores were similar in continuous positive airway pressure versus sham‐continuous positive airway pressure patients. In patients with worse clinical outcomes at month 3, mean nocturnal oxygen saturation was lower whereas there was no association with apnea–hypopnea index. Poorer outcomes at 3 months were also associated with insomnia, restless legs syndrome, depressive symptoms, and decreased total sleep time and rapid eye movement sleep.


| INTRODUCTION
Stroke is a major public health problem.It is the second most common cause of death worldwide, and one of the leading causes of disability in adults (Johnson et al., 2019).Ischaemic stroke (IS), which accounts for 80% of strokes, is a heterogeneous condition with different underlying mechanisms and causes (Johnson et al., 2019).Sleep disturbances are frequent after IS (Bassetti et al., 2020;Hasan et al., 2021).
As they can have a direct effect on ischaemic penumbra and post-IS neuroplasticity, they may have deleterious consequences on shortterm functional recovery, disability, fatigue and quality of life (Bassetti et al., 2020;Campbell et al., 2019).
Sleep apnea syndrome (SAS) is the most frequently reported poststroke sleep disorder.It is found in approximately 70% of patients with IS, and being severe (apnea-hypopnea index [AHI] ≥ 30 per hr of sleep) in 30% of patients (Baillieul et al., 2022).Obstructive sleep apnea (OSA; Lévy et al., 2015) is the most common sub-type (Johnson & Johnson, 2010), and is associated with increased rates of stroke recurrence and mortality (Birkbak et al., 2014;Hermann & Bassetti, 2016).
The few studies on OSA impact on the post-IS functional outcomes highlighted some associations between AHI and disability, lower independence in usual activities, and cognitive dysfunction (Baillieul et al., 2022).Post-stroke OSA is a treatable condition, but it remains largely under-diagnosed especially in the acute phase after IS (Brown et al., 2019;Festic et al., 2018).The benefit of continuous positive airway pressure (CPAP) treatment after IS is still debated (Brill et al., 2018;Kim, 2019).Some studies showed an improvement of quality of life, mood, cognitive dysfunction, but others did not (Aaronson et al., 2016;Kim, 2019).In a large meta-analysis, the neurological scores were slightly better in the CPAP group, but without significant impact on functional outcomes (Brill et al., 2018).CPAP benefit for secondary stroke prevention also remains uncertain, and a recent large randomized control study did not find any reduction in cardiovascular recurrence (McEvoy et al., 2016).
Other sleep disturbances, alterations of sleep architecture, insomnia, restless legs syndrome (RLS) and excessive daytime sleepiness (EDS) also have been frequently reported after IS (Baglioni et al., 2016;Bassetti et al., 2020;Baylan et al., 2020;Hermann & Bassetti, 2016;Miano et al., 2022;Schlesinger et al., 2015).These sleep disturbances can impair quality of life, increase fatigue, anxiety and depression after IS, and can have deleterious effects on neurological and functional recovery (Boulos et al., 2017;Medeiros et al., 2011).Beyond OSA, other parameters of sleep health have been rarely studied in the context of IS.
Therefore, we carried out a combined 3-month multicentric, prospective, observational study and a nested randomized controlled trial (CPAP versus sham) in patients with severe OSA after IS.The main aim of the observational study was to evaluate functional outcomes and survival after IS according to the presence and severity of OSA.
The objective of the nest double-blind, placebo-controlled, two-arm, randomized trial was to measure CPAP impact on functional outcomes and survival at month 3 post-IS in selected patients with severe OSA.We hypothesized that OSA is associated with poorer functional outcomes after IS, and that its recovery depends on the severity of OSA according to AHI.We also hypothesized that treatment of OSA with CPAP might improve post-IS recovery in patients with severe OSA.Secondary objectives were to analyse in the whole sample the association between sleep symptoms and polysomnographic (PSG) parameters, functional outcomes and rates of death and cardiovascular events at month 3 after IS.
It was approved by the ethics committees, and was conducted in or any condition that may interfere with CPAP treatment compliance/ acceptance.All patients provided a written informed consent prior to their participation in the study.For patients who were not able to complete informed consent due to aphasia, written agreement was obtained with a reliable person, such as family.

| Clinical evaluation
All patients were evaluated by a stroke neurologist on admission for IS, who collected the medical history (including time of symptoms onset to identify wake-up IS), body mass index (BMI), treatments, neurological impairment severity assessed with the 11-item NIHSS (from 0, no deficit, to 40; Lyden et al., 2001), and distal upper limb motor function (DMF: 0 = normal-full extension maintained for 5 s; 1 = some movement-any change from full extension within 5 s; and 2 = no movement-no strength at all; Egelko et al., 2021).NIHSS and DMF scores were repeated at day 15 ± 4 (preferably at D15, range D11-D19), and at month 3 after IS.The stroke aetiology after extensive evaluation and the acute phase treatment were recorded.

The level of function and independence in daily live activities
were assessed with the 10-item Barthel Index (BI) at day 15 ± 4 and month 3 post-IS to measure the consequences of the motor and cognitive disorders (total score ranging from 0 to 100 = completely autonomous; Mahoney & Barthel, 1965;Sulter & Steen, 1999).
Functional disability was also assessed with the mRS at day 15 ± 4 and month 3 (0: asymptomatic; 1: some symptoms but no significant disability; 2: slight disability, unable to carry out all previous activities but able to look after own affairs without assistance; 3: moderate disability, requires some help but able to walk without assistance; 4: moderately severe disability, unable to walk without assistance; 5: severe disability, bedridden, requiring constant nursing care).Significant disability was defined by a mRS score > 2 (New & Buchbinder, 2006).
Sleep disturbances were assessed at day 15 ± 4 after IS with a face-to-face consultation, self-reported questionnaires, and one in-laboratory PSG.RLS was assessed in all patients during a faceto-face clinical interview, according to standard criteria (Allen et al., 2003).A second clinical sleep assessment was performed at month 3.

| Self-report questionnaires
Several self-administered questionnaires were administered at day 15 ± 4 (patients were asked to rate their condition currently, thus since and not before the stroke) and at month 3 post-IS: Epworth  et al., 1996;Beck et al., 1998).Patients also completed the Berlin F I G U R E 1 Study flowchart.AHI, apnea-hypopnea index; CPAP, continuous positive airway pressure; PSG, polysomnography; SAS, sleep apnea syndrome.Performed with SAS version 9.4 questionnaire for OSA screening (Netzer et al., 1999) 2003).In case of severe aphasia, questionnaires were completed with the help of a family person or the main caregivers.

| Randomization and controlled trial
Patients with severe OSA (i.e.AHI ≥ 30 per hr, obstructive AHI ≥ 50% total AHI) were randomized 1:1 to CPAP treatment (with an auto-CPAP device; Respironics ® , Philips) or sham treatment (CPAP machine Respironics ® , Philips, modified to include a hidden leak in the connector between the mask and the CPAP tube and a restrictor in the ventilator to restrict the flow, which prevents the machine from shutting down because the machine detects a leak, this system generates non-therapeutic pressure levels at the mask; Farré et al., 1999;Rodway et al., 2010).The randomization sequence was computer-generated, using random blocks in an order unknown by the investigators.The list was established by a statistician and was only accessible to the personnel in charge of randomization.The study was conducted in a double-blind manner, i.e. both subjects and physicians were blinded to treatment allocation.The interventional trial lasted 3 months, after which effective CPAP treatment was offered to all randomized patients.

| Outcomes
The primary outcome was functional independence (assessed with the BI 3 months after IS) in four groups of patients defined in function of OSA presence and severity (absent, mild, moderate or severe on the basis of the AHI: < 5, [5-15], [15-30] and ≥ 30 per hour of sleep, respectively).In severe OSA, BI at month 3 was compared between the two treatment groups (CPAP versus sham-CPAP).
Secondary outcomes were disability (mRS), neurological recovery (NIHSS score), and death and cardiovascular event rates at month 3 in the four groups of patients defined by the AHI.These outcomes and BI changes were also compared in patients with no/mild OSA and in patients with moderate/severe OSA.Clinical and PSG parameters were compared between patients with good and poor functional outcomes.

| Statistical analysis
Patients' characteristics were described using numbers and percentages for categorical variables, and medians (25th percentile; 75th percentile) for continuous variables, because their distributions were mostly skewed according to the Shapiro-Wilk test.Non-parametric statistical tests were used due to the small sample size.The Chi-square or Fisher's exact tests were used to compare categorical variables Sex, men had RLS versus only 27% of non-EDS patients, p = 0.02).According to the Berlin questionnaire, 24 patients (39.3%) were at high risk of OSAS, among whom 15 (65.2%) had moderate to severe OSAS, and nine (36.0%) had AHI < 15 per hr ( p = 0.04; Table 2).
In patients with EDS, severe OSA was more frequent (50% had AHI ≥ 30 per hr versus 16% of non-EDS patients, p = 0.03).the two groups with AHI ≥ 15 per hr and < 15 per hr (Figure 2; Table 2).Similar results were obtained when the percentages of relative change were compared (Table 2).
No difference was found in BI, NIHSS and mRS changes between baseline and month 3 in patients with AHI ≥ 15 per hr and < 15 per hr, and after excluding patients with severe OSA (AHI ≥ 30 per hr; n = 10) included in the interventional trial (data not shown).
Comparison of baseline characteristics in patients with favourable outcome (mRS ≤ 2 or BI ≥ 95) and with significant disability (mRS > 2 or BI < 95) at month 3 showed that the mean oxygen saturation during sleep was lower and the percentage of patients with ≥ 3% ODI ≥ 15 per hr was higher in the second group, without any difference in AHI (Table 3).Moreover, moderate to severe insomnia and higher fatigue scores and depressive symptoms were more frequent in the significant disability group.Baseline RLS was associated only with mRS > 2 but not BI < 95 (Table 3).Lower REM sleep percentage was strongly associated with poorer functional outcomes (BI and mRS) at month 3. Total sleep time (TST) and sleep efficiency were lower only in patients with BI < 95 (Table 3).Results remained unchanged when patients with severe OSA in the randomized controlled trial were excluded (data not shown).
At month 3, 18.7% of patients had EDS, 17.8% moderate to severe insomnia, and 16.3% depressive symptoms.Chalder's fatigue scale and BDI-II scores at baseline and at month 3 were higher in patients with poorer outcome (mRS ≤ 2 or BI ≥ 95).Fatigue symptom improvement was associated with good functional outcome (Table 4).
Conversely, no significant difference was observed between relative changes in the ISI, BDI-II and ESS scores and functional recovery (BI and mRS).

| Nested randomized controlled trial
Twelve of the 13 patients with severe OSA (one refused CPAP treatment) were randomized in the CPAP group (n = 7) and in the sham-CPAP group (n = 5).One patient in the CPAP group was secondarily excluded due to the occurrence of severe cognitive problems that did not allow a correct assessment (e.g.filling the self-report questionnaires and assessment of the IS-related disability; Figure 1).Among the CPAP group (n = 6), the median age was 73.01 years [71.20;73.84]and five (83.3%) were men.Among the sham-CPAP group (n = 5), the median age was 71.08 years [60.49;77.74]and one (20.0%)was male.No differences were observed between the two groups in terms of age, sex and BMI.
Two patients in the CPAP group underwent an invasive procedure for IS (i.e.thrombectomy, thrombolysis or both) versus three patients in the sham-CPAP group.Baseline BI tended to be higher in the CPAP group than in the sham-CPAP group (80 [25; 90] versus 45 [10;75]; not significant).BI percentage relative changes between 3 months and 2 weeks were not different between treatment groups (25% [11.1; 80.0] in the CPAP group versus 33.3% [17.7; 111.1] in the sham-CPAP group).

| Cardiovascular events and deaths
At month 3 of follow-up, one death due to a non-cardiovascular cause in the moderate OSA group and one cardiovascular event in the mild OSA group were reported.

| DISCUSSION
In this study, no association was found between severity of OSA according to AHI and functional independence at 3 months post-IS, although the mean nocturnal oxygen saturation during sleep was lower in patients with poorer functional outcome.A decrease in TST and in REM sleep, insomnia, RLS, and the presence of depressive and fatigue symptoms were also associated with poorer functional outcome.
Obstructive sleep apnea was frequent (83.6%) in our cohort, and was moderate to severe in 44.3% of patients, in agreement with the literature (Seiler et al., 2019).In our study, patients with OSA were not different from the other patients, but tended to be older and more frequently men.The risk factors for stroke are the same as those for OSA (i.e.older age, male and higher BMI), which could explain the relative homogeneity of our population; however, with a fairly small sample.No specific stroke characteristic was associated with OSA.
The use of the Berlin questionnaire in this condition is not a sensitive screening tool, confirming previous findings (Srijithesh et al., 2011).
OSA can play a negative role after stroke, but different underlying mechanisms may exist.In animal models, previous exposure to apnea is associated with a larger ischaemic lesion (Cananzi et al., 2020).In humans, post-IS OSA has been associated with poorer outcome (Lisabeth et al., 2019), worse quality of life and more cognitive impairment (Ott et al., 2020).A potential association between AHI and significant disability at month 3 was previously suggested; however, AHI was assessed in the first 7 days following IS, patients with CSA were not excluded, and most patients had mild-to-moderate IS (Ott et al., 2020).In our study, no association was found between AHI and functional outcomes, but abnormalities in oximetric parameters were associated with poor functional outcome.Although clinically relevant, the time spent below 90% of saturation was lower in our study than in other studies (Ferdinand & Roffe, 2016;Ott et al., 2020), results potentially related to key demographic characteristics of our population (e.g.19% obese and 25% smokers).Intermittent hypoxaemia could be potentially harmful for the suffering brain tissue in the ischaemic penumbra area in the acute phase, resulting in a larger ischaemic lesion.Hypoxic burden is now widely acknowledged as a better prognosis factor than AHI for cardiovascular outcomes (Trzepizur et al., 2022).SAS could also interfere with neuroplasticity mechanisms via sleep fragmentation, reduction in sleep efficiency, SWS and REM sleep, oxidative stress, or indirectly via the associated mood and quality of life disturbances (Baillieul et al., 2022;Lisabeth et al., 2019).Besides, related hypoxaemia may potentiate these mechanisms and add specific ones (endothelial dysfunction, sympathetic overactivity).
Few studies showed improvement in quality of life, mood, cognitive dysfunction after post-IS OSA treatment; however, the overall impact of OSA management on functional recovery remains uncertain (Aaronson et al., 2016;Kim, 2019).In addition, these studies had several limitations.Using AHI as the only diagnostic tool for OSA is insufficient because it may not be the best metric to reflect OSA severity, and because the negative impact of OSA on disability may be more related to oxygen desaturation.Moreover, the heterogeneity of the included patients, type of sleep studies, central or obstructive origin of sleep-disordered breathing, poor adherence to treatment, time taken to diagnose SAS and to start treatment, and different OSA management could have masked the beneficial impact of the treatment.Also, acceptance of CPAP following acute IS was often limited.In our study, CPAP efficacy could not be properly assessed due to the small number of randomized patients, and data on CPAP compliance were unfortunately not available for all patients.More interventional studies on OSAS management poststroke are needed to determine when (i.e.time after IS), how and which patient may benefit the most from this treatment, especially regarding its impact in reducing the risk of worse functional outcomes.Also, compliance with CPAP can be improved by the patient's entourage, and the involvement of the main caregivers during the follow-up of patients at home and in consultation.
A shorter TST and lower sleep efficiency were frequently reported in our population, and insomnia symptoms were reported by > 20% of patients.Previous studies highlighted frequent sleep complaints and impaired PSG parameters in post-stroke patients (Baglioni et al., 2016;Pace et al., 2018).Moreover, some studies reported an association between insomnia and cardiovascular and metabolic comorbidities, stroke severity and poorer functional outcomes (Matas et al., 2022;Miano et al., 2022;Pace et al., 2018;Terzoudi et al., 2009).Sleep duration prior to stroke may also be associated with post-stroke mortality, following an U-shaped curve, with higher risk for shorter or longer sleep time (Wang et al., 2022).
RLS was common in our sample, and its frequency was higher than in previous studies possibly due to the older age of our population (Manconi et al., 2018).Furthermore, RLS was more common in patients with poor neurological outcomes, consistent with literature (Medeiros et al., 2011).We can hypothesize that patients with more severe neurological deficit are bedridden longer, which is a risk factor for RLS.On the other hand, the PLM index was not associated with functional outcomes.The association between RLS and worse clinical outcome after IS may be explained by the disruption in sleep continuity and architecture, overactive sympathetic activity, and low-grade chronic brain inflammation (Manconi et al., 2021).These findings may have direct implications for the management of poststroke patients; however, the beneficial impact of RLS and insomnia treatment on stroke recovery is still unknown (Duss et al., 2018).
Disturbances in post-stroke sleep architecture could be explained by brain lesions or oedema caused by the cerebral infarct (alteration of cortico-thalamo-cortical loops), by the secretion of proinflammatory cytokines in the acute phase (Lambertsen et al., 2012), or by environmental factors, anxiety and post-stroke stress.In our study, baseline sleep parameters were worse (lower TST, sleep efficiency and REM sleep) in patients with significant disability at month 3 post-IS, in line with a previous study reporting decreased REM sleep after stroke with an association with poor outcome (Pace et al., 2018).
As brain plasticity processes take place mainly during SWS and REM sleep (Li et al., 2017;Tononi & Cirelli, 2006), a decrease in these sleep stages can have deleterious effects.Sleep disturbances could also cause fatigue, EDS and mood deterioration.Indeed, in patients with poorer functional outcomes, depressive symptoms, fatigue and insomnia, but not EDS, were more frequent.However, the use of ESS to assess EDS may have underestimated this problem in these elderly patients 15 days post-stroke.The daily life situations evoked by the scale are sometimes artificial under these conditions, so must then be assumed and therefore lack precision.Objective assessment of sleepiness via mean sleep latency tests, although costly and time-consuming, could be considered in the future.Altogether, this highlights the importance of screening and diagnosing sleep disorders in the context of recent IS, and also of determining whether their management may contribute to reduce post-IS functional disabilities.
Our study has several limitations.First, due to low recruitment, the study was stopped prematurely (early termination for futility), the multivariate analyses taking age into account for example were not feasible, and a possible confounding bias could not be eliminated.
These recruitment difficulties are partly linked to the strict patient selection criteria, particularly frequent symptom improvement at day 14 ± 5 after thrombolysis/thrombectomy.However, it would have been also difficult to transfer unstable or more clinically severe patients to the sleep laboratory.Recruitment difficulties were also due to the patients' reluctance to undergo PSG after a recent stroke and to participate in a randomized controlled trial where they could have been in the sham group.Moreover, 15 patients were excluded because of severe central SAS.This high frequency may be linked to the early PSG recording after IS.High levels of breathing instability were already reported following IS, with a decrease in SAS severity between the acute and the chronic IS phases (Baillieul et al., 2022).
The diagnosis, management and consequences of persistent central SAS after stroke were out of the study scope and should be investigated in future studies.Our study was not designed to assess the rate of recurrence or death, because follow-up was limited to 3 months, thus explaining the low number of vascular reoccurrences or deaths.
accordance with the principles of good clinical practice and local regulations.The observational study began in September 2011.The randomized controlled trial (CPAP versus sham) started in March 2012 in patients with severe OSA after stroke.2.1 | Patients Consecutive patients, aged between 18 and 85 years, hospitalized in stroke units for a supra-tentorial IS confirmed by magnetic resonance imaging (MRI) or computer tomography (CT) imaging data, and nonminor neurological deficit or relevant deficit at inclusion and persisting on day 7 after IS were included.Clinically relevant deficit was defined by a National Institute of Health Stroke Scale (NIHSS) score ≥ 4, isolated aphasia (NIHSS item 9 score ≥ 1), extinction/inattention (NIHSS item 11 score = 2) or deficit in distal motor function (DMF score ≥ 1).Exclusion criteria were: haemorrhagic stroke, infra-tentorial IS, already known OSA treated or not using CPAP, severe central sleep apnea (CSA) diagnosed by PSG (AHI ≥ 30 per hr of sleep, including > 50% of the total number of apneas and/or hypopneas of central origin), severe arterial hypertension or unstable cardiovascular disease (e.g.unbalanced heart failure, recent myocardial infarction, poorly tolerated atrial fibrillation or severe pulmonary embolism), severe EDS, pre-stroke disability (modified Rankin score [mRS] > 2 before IS), misuse of illicit substances or alcohol, Sleepiness Scale (ESS; score ≥ 11 indicates EDS; Johns, 1991); Insomnia Severity Index (ISI; score ≥ 15 indicates moderate to severe insomnia; Bastien, 2001); Chalder's Fatigue Scale (Chalder et al., 1993); and Beck Depression Inventory version II (BDI-II; score ≥ 20: moderate to severe depressive symptoms; Beck T A B L E 1 Baseline sociodemographic and clinical characteristics in the whole sample (N = 61)

2. 4
| PolysomnographyOne PSG session was recorded at the Sleep Unit of each hospital, between 23:00 hours and 07:00 hours, with standard methods (electroencephalogram [EEG] channels, submental and both tibialis anterior muscles electromyogram, electro-oculogram, nasal airflow, thermistor sensor, chest and abdominal straps, electrocardiogram, body position, oximetry and video-recording) at day 15 ± 4. The EEG was performed after placement of the electrodes (C3/A2-C4/A1-Cz/O1) according to the international 10/20 placement system.PSG recordings in the sleep units were monitored by nurses.All PSG recordings were manually scored, according to the American Academy of Sleep Medicine (AASM) 2007 international criteria(Iber et al., 2007), to extract the sleep stages: stages 1, 2, 3 (i.e.slow-wave sleep [SWS]) and rapid eye movement (REM) sleep, arousals from sleep, periodic legs movements during sleep (PLM;Ferri et al., 2016) and respiratory events (AHI).A PLM index ≥ 15 per hr was considered pathological, as defined by the International Classification of Sleep Disease (ICSD), third edition.An apnea was defined by a decrease of nasal respiratory flow ≥ 90% for at least 10 s.A hypopnea was defined by a nasal respiratory flow decrease ≥ 50% of the pre-event baseline for ≥ 10 s or ≥ 30% for ≥ 10 s, followed by ≥ 3% arterial oxygen desaturation or an arousal.Apnea was classified as obstructive or central according to the presence or absence of respiratory effort.

a
For the data available at both evaluation times.b % of relative change was defined as 100 Â (3 months À2 weeks)/2 weeks.betweengroups (i.e.AHI ≥ 15 per hr versus < 15, mRS > 2 versus ≤ 2, BI ≥ 95 or < 95) and Mann-Whitney U-tests for continuous variables.Given the lack of a validated BI threshold, a receiver-operating characteristic (ROC) analysis was performed to determine the BI cut-off point that best differentiated patients with mRS ≤ 2 (low disability) and > 2 (significant disability)(25, 26).The optimal BI threshold of 95 was defined by the highest Youden Index [(specificity + sensitivity) À 1].Scores at month 3 and changes in disability and clinical parameters between day 15 ± 4 and month 3 post-IS, and percentages of relative change (defined as 100 Â (month 3day 15 ± 4)/ day 15 ± 4) were compared between groups using the Mann-Whitney U-test and among groups using the Kruskall-Wallis test.Withingroup differences between day 15 ± 4 and month 3 post-IS were compared with the Wilcoxon signed-rank test.To avoid a confounding factor related to the treatment effect, primary and secondary outcomes were also analysed after excluding patients with severe OSA included in the interventional study.Statistical significance was set at p < 0.05.Statistical analyses were performed with SAS version 9.4.3 | RESULTSNinetypatients hospitalized in the stroke units of the participating centres were screened for inclusion in this study between 2011 and 2019 (n = 60 from Montpellier, n = 17 from Grenoble, n = 12 from Nimes, n = 1 from Béziers; Figure 1).Seven patients were excluded because they refused PSG (n = 5) or were clinically unstable (n = 2).Then, 22 patients were excluded after PSG due to severe central SAS (n = 15), technical problems (n = 3), substance misuse (n = 3), or no comprehensive clinical sleep assessment (n = 1).Therefore, 61 patients (42.6% of men, median age of 71.8 years [60.6;76.3]) with baseline (i.e.day 15 ± 4 post-IS) comprehensive assessment were included.Diagnosis of supratentorial IS was made in all cases by MRI, but four patients with CT-scan due to an MRI contraindication.The median length of hospital stay after IS was 13.5 (10;18) days.Most patients were then transferred to a neurological rehabilitation centre, while 12 patients went directly home.During the follow-up, one patient died, one refused the CPAP treatment, six were lost to follow-up, two withdrew their consent, and one developed severe cognitive impairment.Finally, 50 (84%) patients had a complete evaluation at baseline and at month 3.3.1 | Baseline evaluation (n = 61 patients)The median NIHSS score at hospital admission for IS was 4 (range 2-9), and was ≥ 4 in 75% of patients.IS aetiology was mainly cardioembolic (27.1%), atheromatous (22.5%) and unknown (47%; Table1).Initial acute stroke treatment was mechanical thrombectomy in eight (14.8%)patients, intravenous thrombolysis in nine (16.7%), associated thrombectomy and thrombolysis in three (5.6%), and other medical treatment (antiplatelets, anticoagulants) in 34 patients.At baseline (day 15 ± 4), moderate to severe insomnia was reported by 13 (24.1%)patients, moderate to severe depressive symptoms by three (5.7%), and EDS by 10 (16.7%) patients.RLS was found in 20 (34.5%) patients, and was severe or very severe in 45%.EDS was more frequent in patients with RLS (70% of patients with EDS F I G U R E 2 Changes in: (a) the National Institute of Health Stroke Scale (NIHSS) score; (b) the modified Rankin scale (mRS); and (c) the Barthel Index (BI) between day 15 ± 4 and month 3 of follow-up after ischaemic stroke (IS).Performed with SAS version 9.4 T A B L E 3 Baseline demographic, clinical and PSG characteristics in patients with and without disability at month 3 (BI and mRS) 3.1.1| Primary outcome (n = 50 patients) At month 3 after IS, the median BI increased from 55.0 [25.0; 85.0] to 95.0 [55.0; 100.0].However, BI at month 3 did not differ between the four groups of patients (AHI (< 5, [5-15], [15-30] and ≥ 30 per hr)), or Focusing on BI changes to assess functional recovery limited the result interpretation because of the lack of a reliable and validated threshold for this score; however, a BI cut-off was determined by ROC analysis to differentiate patients with different mRS scores.Most patients were transferred to a neurological rehabilitation centre following their hospitalization in the stroke unit, but data on the duration and type of rehabilitation in re-education centres were unfortunately not collected in our study.A possible centre effect cannot be excluded because most patients were included at Montpellier Hospital.Finally, only patients with supra-tentorial IS were included, although the impact of sleep disorders and OSA on the functional outcomes in patients with haemorrhagic or infra-tentorial strokes remains unknown.This study has several strengths.It was a prospective, multicentric study in which the sample was selected according to stringent criteria.Patients were systematically evaluated by a vascular neurologist and a sleep specialist, and validated questionnaires were used.Unlike most studies using ventilatory polygraphy alone, our objective sleep assessment was performed by in-laboratory PSG.The interventional study was a randomized, double-blind, placebo-controlled trial nested in a prospective cohort, an uncommon design in other studies.In conclusion, poorer functional outcomes at 3 months after a moderate to severe IS were associated with lower mean oxygen saturation, but not with AHI.Other parameters such as RLS, low TST and REM sleep, insomnia, depressive symptoms and fatigue were more frequent in patients with less favourable outcome.More prospective studies are needed to better understand the bidirectional relationship between the different sleep disorders and stroke, their co-existence, severity, and the effect of their management on recovery after stroke in the framework of precision medicine.