Clinical manifestations

Presentation and course

It may be difficult to differentiate the sleepiness and other symptoms associated with obstructive sleep apnea from such acute manifestations of cerebrovascular disease as lethargy, apathy, and neglect, particularly with strokes in specific locations, such as bilateral paramedian thalamic infarctions. The spouse of a stroke victim may describe a lack of energy, falling asleep during activities, and fatigue. Additional questioning may elicit a history of snoring (with repetitive respiratory interruptions), restless sleep, nonrestorative sleep, and weight gain prior to or following the stroke. Weakness of pharyngeal dilator muscles following a stroke may contribute to the elevated resistance in the upper airway that is common in obstructive sleep apnea. Dysphagia and dysarthria in patients with acute stroke are generally associated with sleep apnea (25). One study showed that dysphagia severity was associated with obstructive sleep apnea severity, and this association was independent of sex, modified Barthel index, and body mass index (32). However, stroke-induced dysphagia was not associated with central sleep apnea or overall sleep-disordered breathing.

Twenty-four-hour ambulatory blood pressure monitoring reveals that arterial pressure ordinarily falls during nocturnal sleep 10% to 20% below that of diurnal waking. However, 50% of patients with acute hemispheric ischemic stroke are "nondippers" (a fall of less than 10%), as are 23% of patients with lacunar infarct versus 7% of controls (53). In general, target organ damage is greater in nondippers than in dippers, even though they may have comparable 24-hour mean blood pressure levels. Interestingly, both "reverse dipping" and "extreme dipping" (greater than 20% nocturnal reduction) are independent risk factors for multiple silent lacunar infarcts in elderly Japanese patients with hypertension (49).

The incidence of periodic central sleep apnea during sleep (Cheyne-Stokes) in patients with stroke has been reported to be as high as 41% (89) (a smaller, earlier study cited 53%). Furthermore, data from the Sleep Heart Health Study reveal that the presence of central sleep apnea is associated with MRI evidence of white matter ischemic changes or frank infarcts (79), and central sleep apnea is an independent risk factor for stroke in elderly patients (68).

In a study of 1346 patients with relatively mild stroke with an NIHSS of 6 or lower, only 1.4% met the criteria for central sleep apnea (86). However, the median time from stroke recognition to portable respiratory recording was 13 days, a length of time superior to the interval of 48 to 72 hours reported by other authors (73). The timing of the stroke may be of the essence as central sleep apnea appears to be more common during the hyperacute stage of stroke (23).

Sleep disruption and nonrestorative sleep are ubiquitous following hemispheric strokes as is daytime sleepiness, even in the absence of sleep apnea. The presence of sleep continuity disturbances in the acute phase of stroke may represent a risk factor for poor outcome (98). In a rat model of middle cerebral artery ischemic stroke, sleep deprivation hindered functional recovery and impaired synaptogenesis and cell proliferation (107). Lastly, poststroke depression may affect up to one third of patients with hemispheric stroke.

Prognosis and complications

A brief sleep history should be part of the diagnostic workup for every stroke or transient ischemic attack and should include the following details:

Presence of the above symptoms may alert the clinician to the need for polysomnography, particularly if the patient is overweight or obese because obstructive sleep apnea increases morbidity and mortality in patients with stroke (74; 96). Sleepiness, fatigue, and poststroke depression are all likely to exert an unfavorable influence on stroke outcome through effects on motivation and performance (19; 46; 72). All of these conditions are often associated with reduced sleep quality, which may independently impair performance. Furthermore, the lack of restorative sleep generally leads to impaired sleep hygiene, such as excessive napping during the day, further reducing nocturnal sleep efficiency.

Obstructive sleep apnea independently contributes to cognitive dysfunction in patients with stroke through hypoxemia and sleeping discontinuity. The prospective memory test is a simple but sensitive method to detect sleep apnea-induced cognitive impairment in patients with stroke (105).

In the multicenter study Ethnic/Racial Variations of Intracerebral Hemorrhage, which recruited 3000 cases with intracranial hemorrhage and 3000 controls, 2064 (71%) cases and 1516 (52%) controls were classified as having obstructive sleep apnea by the Berlin Questionnaire (35). Cases with obstructive sleep apnea were significantly more likely to be male and have hypertension, heart disease, hyperlipidemia, and higher body mass index compared with those without obstructive sleep apnea. Obstructive sleep apnea was more common among intracerebral cases compared with controls (71% vs. 52%; odds ratio 2.28, 95% CI 2.05-2.55). The authors concluded that obstructive sleep apnea is a risk factor for intracerebral hemorrhage.

In a study of 204 patients with ischemic stroke, nocturnal hypoxemia and central apneas increased mortality, but not recurrent strokes, after ischemic stroke (45).

Obstructive sleep apnea (OSA) has significant effects on cognition impairment in patients with minor ischemic stroke, according to a study conducted in China in 94 patients where 35 had no OSA, 32 had mild OSA, and 27 had moderate-to-severe OSA (106). The authors concluded that hypoxemia may be a potential pathophysiological mechanism of obstructive sleep apnea-induced cognitive impairment.

Clinical vignette

JT was a 62-year-old male who presented with dysphasia and minimal right hemiparesis. He had a history of hypertension, noninsulin-dependent diabetes mellitus, gout, chronic alcoholism, chronic pancreatitis, and depression. He was on multiple cardiovascular medications, glipizide, and fluoxetine. On examination, his blood pressure was 110/70. He was talking incomprehensibly, and his family said that he had become more forgetful. There was a predominantly Wernicke-type dysphasia and a right visual field cut. Chest x-ray revealed moderate congestive heart failure. CT revealed old lacunae in the left hemisphere along with a new hypodensity in the left temporal-occipital area with obliteration of sulci. There were also periventricular ischemic changes. The echocardiogram revealed enlargement of the left atrium and dysfunction of the left ventricle. He was thought to have embolic stroke and was started on crystalline warfarin sodium and an angiotensin-converting enzyme inhibitor, but these medications were discontinued because of poor tolerance. He was seen in the Sleep Wake Disorder Unit three months later. In the interim, with the help of speech therapy, his dysphasia markedly improved. He complained of insomnia and fatigue since his stroke and insisted that he had no sleep complaints previously. He denied abnormal leg movements or breathing problems. At the time of the sleep study he reported that his usual sleeping hours were from 3 A.M. to 6:30 A.M. and that he was not sleepy and unable to fall asleep prior to that time. The polysomnogram revealed a bedtime of 10:47 P.M. and arise time 5:42 A.M. His sleep efficiency was 85%, with a latency to sleep of 24 minutes and 27 minutes to the first REM period. Stage 1 was 34%, delta sleep was completely absent, and REM sleep was 20%. He had 151 respiratory events, of which 47 were apneas and 104 hypopneas. They were predominantly obstructive in nature. The apnea-hypopnea index was 25.6, and the lowest oxygen saturation was 89% in REM sleep. He spent 68% of his sleep time in a supine position; respiratory events were posture-dependent. There were no periodic leg movements. The multiple sleep latency test revealed a latency of 18 minutes, and there were no sleep-onset REM periods.

Comment. This may be an example of a sleep disorder precipitated by stroke, which seems to be much less common than preexisting obstructive sleep apnea acting as a risk factor for stroke. The complaint of insomnia was only partially confirmed by the polysomnogram, but he did have significant apnea. Of note is his lack of obesity (five feet nine inches tall and 135 pounds) and the fact that there was no excessive daytime sleepiness, either subjectively or objectively. Finally, his short REM latency on the polysomnogram was related either to his previous depression (despite fluoxetine), to the presence of brain damage in the left hemisphere, or both factors acting in concert.

Biological basis

Etiology and pathogenesis

Obstructive sleep apnea may act as a risk factor for stroke because of its association with systemic hypertension and other risk factors for stroke, including atrial fibrillation. One study found that the morning surge of blood pressure was the strongest predictor of stroke during an average follow-up period of 41 months in elderly hypertensives (48); obstructive sleep apnea may contribute to that surge, and nifedipine, administered at bedtime, may reduce the surge (42). Additionally, changes in autoregulation of cerebral blood flow in severe obstructive sleep apnea may contribute to the increased morning stroke risk (97).

In 2006, the American Heart Association and American Stroke Association, together with other interested groups, issued a guideline on risk factors for stroke in which a recommendation for obtaining a history regarding symptoms of sleep-disordered breathing and referral to a sleep specialist for appropriate patients was considered Class IIb (usefulness/efficacy is less well established by evidence or opinion), Level C (lowest level of evidence – consensus opinion of experts) (37). An update of the guideline has raised the recommendation to Class I (evidence for or general agreement that the procedure is useful and effective), Level A (data derived from multiple randomized clinical trials) (38). The effectiveness of sleep apnea treatment to prevent the occurrence of stroke is still unknown according to the guideline (Class IIb, Level of Evidence C).

Atrial fibrillation has been linked to severe, uncontrolled sleep apnea. Atrial fibrillation increases the risk of stroke by 5% to 10% per year (93). Clinical data have shown a strong relationship between sleep apnea and atrial fibrillation, and epidemiologic studies suggest that sleep apnea is a risk factor for new-onset atrial fibrillation. A large study evaluated 3542 patients without atrial fibrillation who underwent polysomnography and were followed for an average of five years (33). In patients less than 65 years old, nocturnal oxygen desaturation predicted new-onset atrial fibrillation. Sleep apnea may confer a poorer prognosis for recovery after atrial fibrillation interventions. In a study of 424 patients undergoing ablation, sleep apnea more than doubled the risk of acute intraprocedural failure (84). The effects of sleep apnea therapy on atrial fibrillation outcomes are largely unknown, and prospective randomized controlled trials are necessary to clarify this issue. In one study of 47 women and 111 men with subacute ischemic stroke admitted for neurorehabilitation, mean nocturnal desaturation was significantly associated with atrial fibrillation after adjusting for age, neck circumference, Barthel index, and high-density lipoprotein level (odds ratio = 1.19 [95% confidence interval 1.05-1.35], P = .008) (16). The authors concluded that nocturnal hypoxia due to obstructive sleep apnea is an independent predictor of atrial fibrillation in patients with subacute ischemic stroke. Another study has confirmed the high prevalence of sleep apnea in stroke-affected patients and has identified atrial fibrillation as a major source of stroke in this population, concluding that the strong correlation between age and sleep apnea drives the increased frequency of stroke related to atrial fibrillation (75).

In a study, the authors investigated whether obstructive sleep apnea is independently associated with worse cardiovascular and neurologic outcomes in patients with atrial fibrillation (29). The authors conducted a retrospective cohort study of 22,760 patients with atrial fibrillation with and without obstructive sleep apnea. They found that 4,045 (17.8%) patients had obstructive sleep apnea at baseline. Median follow-up time was 1.5 years; 1895 patients experienced major adverse cardiac events, neurologic events, or both. Patients with obstructive sleep apnea were younger, more likely male, and had increased body mass index. Those with obstructive sleep apnea had a higher prevalence of comorbidities such as diabetes, chronic obstructive pulmonary disease, and heart failure, and higher use of antithrombotic therapy. After adjustments, the presence of obstructive sleep apnea was significantly associated with major adverse cardiac and neurologic events (hazard ratio: 1.16 [95% CI: 1.03-1.31], P = .011). Obstructive sleep apnea was also an independent risk factor for stroke but not cardiovascular death, myocardial infarction, new-onset heart failure, or major bleeding. The authors concluded that among patients with atrial fibrillation, obstructive sleep apnea is an independent risk factor for major adverse cardiac and neurologic events, and more specifically, stroke.

In severe obstructive sleep apnea, oxygen saturation may fall below 85% for more than half the night and may frequently reach levels below 50%. The consequences of this degree of oxygen desaturation include early onset thickening of the carotid artery wall, even prior to the appearance of other risk factors, such as hypertension (90). In addition, accelerated carotid atherosclerosis accompanies heavy snoring independent of sleep-disordered breathing (57). Perhaps this is related to the effects of vibration on the carotid artery.

Inflammation and hypoxia are intertwined at the molecular, cellular, and clinical levels. Repeated hypoxia may damage the endothelium and trigger the release of proinflammatory factors like plasma cytokines, tumor necrosis factor-alpha, and interleukin-6 (55). Chronic intermittent hypoxia causes vascular dysfunction by increasing endothelin, augmenting neurovascular oxidative stress, decreasing vascular neuromuscular reserve, reducing vascular reactivity, and increasing susceptibility to injury (13).

Wake-up stroke refers to the presence of symptoms and signs of stroke on awakening. Wake-up stroke conjures the notion of some vascular event occurring during the night while the patient is asleep. Clinically significant sleep apnea, nocturnal hypoxemia, atrial fibrillation secondary to hypoxemia, and right-to-left shunt triggered by apnea events in patients with patent foramen ovale should be considered. In a study of 71 patients with mild to moderate stroke, the authors aimed to determine independent variables associated with wake-up stroke (43). Of the 71 patients, 26 (36.6%) had wake-up stroke. Comparing both groups, patients with wake-up stroke had a significantly higher apnea-hypopnea index (23.1 ± 19.4 vs. 12.5 ± 11.9, p = 0.016) and lower mean blood oxygen saturation (95.1 ± 1.5 vs. 95.8 ± 1.3, p = 0.046) than the patients with non-wake-up stroke. Severe sleep-disordered breathing (apnea-hypopnea index 30 or higher) was the only independent variable associated with wake-up stroke (OR 6.065, 95% CI 1.451-25.350; p = 0.014). The authors concluded that obstructive sleep apnea is an independent risk factor associated with wake-up stroke. In a case-control study conducted on 170 patients with acute stroke, those who woke up with the symptoms were labeled as wake-up stroke, and those whose stroke occurred while awake were labeled as non-wake-up stroke (66). Of 40 patients with wake-up stroke, 72.5% had underlying obstructive sleep apnea based on the Berlin Questionnaire, whereas only 45% of the 67 patients with non-wake-up stroke had underlying obstructive sleep apnea. The authors concluded that obstructive sleep apnea is an important risk factor for ischemic stroke during sleep.

In a study of 335 patients (mean age 64 years) with wake-up stroke or transient ischemic attack, 202 (60%) had at least one long obstructive sleep apnea (greater than 20 seconds) and 116 (35%) a right-to-left shunt; 69 (21%) had both (21). There were significantly more wake-up strokes and transient ischemic attacks in subjects with right-to-left shunt plus long obstructive sleep apnea than those without this association (27/69 vs. 70/266; OR 1.91, 95% CI 1.08 to 3.38; p=0.03). The authors concluded that the combination of long obstructive sleep apnea and right-to-left shunt could be a potentially treatable risk factor for cerebrovascular ischemic events. The authors hypothesized that long obstructive sleep apnea can lead to right-to-left shunting and propitiate paradoxical embolism.

To investigate the relationship between nocturnal atrial fibrillation and wake-up stroke, Riccio and colleagues prospectively assessed every patient with acute ischemic stroke or transient ischemic attack admitted to hospital over a 3-year period (78). They studied 356 patients, 274 (77.0%) with a diagnosis of acute ischemic stroke and 82 (23.0%) with transient ischemic attack. A total of 41 (11.5%) events occurred during night sleep. Newly diagnosed atrial fibrillation was detected in 27 patients of 272 without known atrial fibrillation (9.9%). The authors found an independent association between newly diagnosed atrial fibrillation and wake-up ischemic stroke and transient ischemic attack (odds ratio 3.6, 95% confidence interval 1.2-7.7, p = 0.019) and concluded that the odds of detecting newly diagnosed atrial fibrillation were 3-fold higher among wake-up cerebrovascular events than among non-wake-up events.

In a study of 837 patients with home sleep monitoring wake-up, stroke was not associated with sleep apnea (85).

Severe premorbid obstructive sleep apnea aggravates ischemic stroke as noted in laboratory experiments by Cananzi and colleagues. The authors developed an air-bag device that enabled them to apply an upper airway obstruction in mice over three days simulating obstructive sleep apnea (12). Cerebral ischemia was induced by performing middle cerebral artery occlusion surgery the day after completion of rounds of obstructive sleep apnea. The occlusion lasted 60 minutes. The authors found that control mice developed 10.5 ± 5.1% injury, whereas obstructive sleep apnea mice showed 26.2 ± 5.46% infarct, a significant increase. Although the animal model developed by Cananzi and colleagues does not replicate precisely human obstructive sleep apnea, it has sufficient parallelisms to strongly suggest that severe premorbid obstructive sleep apnea significantly aggravates the impact of ischemia on the brain. Prestroke injury to the vascular bed damage may occur through a variety of pathophysiologic mechanisms including endothelial damage, sympathetic activation, and acute hypertension.

On the other hand, animal models also show that brains exposed to intermittent brief, mild ischemia reveal a measure of protection against ischemic stroke, a phenomenon known as ischemic preconditioning. The neuroprotective effect in ischemia preconditioning is mediated by an attenuation of the mechanisms of injury, along with the activation of defense mechanisms, and enhancement of endogenous repair processes (36). Even remote ischemia exerted in a limb may have a neuroprotective effect (71). Questions have been asked about whether mild intermittent hypoxia in obstructive sleep apnea has a neuroprotective effect and where is the dividing line between aggravation of ischemic injury following severe obstructive sleep apnea and neuroprotection induced by ischemia preconditioning (27).

In a case series, 18 consecutive male patients with a median age of 41 (35-50) years received care for acute ischemic stroke (95). All of these patients had laboratory-confirmed asymptomatic COVID-19 infection based on a positive SARS-CoV-2 serological test result. The median time from a positive serological test result to acute ischemic stroke was 54.5 (0-130) days, and the median National Institutes of Health Stroke Scale score was 5 (1-25). Ten patients (56%) presented with a large vessel occlusion. Three patients (17%) had a possible cardiac source of embolus. The authors concluded that the risk for acute ischemic stroke is higher in adults 50 years of age or younger during the convalescent period of a COVID-19 infection without respiratory symptoms. Obstructive sleep apnea is a risk factor for COVID-19 infection, and this mechanism of stroke should be considered in case of infection (60).

Epidemiology

Most surveys have reported a 75% incidence of obstructive sleep apnea in patients with stroke. In a 10-year follow-up of subjects enrolled in a large, community-based epidemiologic survey, the relative risk for stroke was significantly greater in those individuals who reported excessive daytime sleepiness, a symptom that may be associated with sleep apnea syndrome (31). More importantly, the ongoing multicenter Sleep Heart Health Study found that even mild degrees of sleep-disordered breathing posed a modest but significant risk factor for stroke with an odds ratio of 1.58 (87). Finally, a number of longitudinal studies have shown that obstructive sleep apnea is a risk factor for stroke. Two studies reported that even after adjustment, the presence of sleep apnea was associated with a greater odds ratio (1.97) for stroke and death compared with a control group after an average follow-up interval of 3.3 years (102) and after 10 years (hazard ratio 1.76) (83). The MESA study (Multiethnic Study of Atherosclerosis) reported a higher incidence of cardiovascular events, including stroke, in patients with a diagnosis of obstructive sleep apnea (103).

The association of stroke with sleep apnea may change with time. Parra and colleagues described polysomnographic findings during the first three days after a first-ever stroke and compared them with the findings three months later (73). The apnea-hypopnea index declined from 71.4 to 61.6 at the later time-point, primarily due to a drop in central events (including Cheyne-Strokes respiration), whereas the number of obstructive events remained stable. This suggests that obstructive apneas usually predate an acute stroke, whereas central apneas may arise from the stroke itself.

Stroke may be more prevalent in women with sleep apnea. In one large study, 71,779 female nurses 40 through 65 years of age without previously diagnosed vascular disease were followed for eight years, and frequency of snoring was assessed using mailed questionnaires; 398 women suffered a stroke (44). The age-adjusted relative risk of stroke was 1.60 (95% CI 1.21 to 2.12) for occasional snorers and 1.88 (95% CI 1.29 to 2.74) for habitual snorers. After further adjustments, the association between snoring and vascular disease remained positive at 1.33 (95% CI 1.06 to 1.67) for regular snorers. Although the study did not identify patients with sleep apnea, the assumption was made that habitual snoring was a marker of obstructive sleep apnea and likely the background risk factor for vascular disease.

In another study, the authors provided evidence that stroke has a higher incidence in Chinese women 35 years old or younger, with a sleep apnea/hypopnea index of 5/hour or higher (15). The authors used a universal insurance claims database and identified a large cohort of patients with sleep apnea using polysomnography. They identified 29,961 patients with sleep apnea and compared the sex- and age-specific stroke risk with a control group. The sleep apnea cohort had a higher stroke incidence in women with an adjusted hazard ratio between males and females of 1.21 (95% CI 1.01 to 1.24; p < 0.05) and 1.44 (95% CI 1.20 to 1.72; p < 0.05), respectively. Stratified by age, the effects of sleep apnea on stroke risk in women decreased with advancing age (adjusted HR 4.90, 95% CI 1.93 to 12.4 for subgroup aged 20 to 35 years; adjusted HR 1.64, 95% CI 1.01 to 2.65 for subgroup aged 36 to 50 years; adjusted HR 1.38, 95% CI 1.01 to 1.89 for subgroup aged 51 to 65 years). The authors concluded that young Chinese women with sleep apnea are at higher risk of stroke and hypothesized that structural and hormonal differences in young women might have increased the risk of stroke.

In a study of 394 old males aged 70 to 100 years, the authors concluded that severe obstructive sleep apnea/hypopnea (AHI [apnea/hypopnea index]=30/hour or higher) increases the risk of ischemic stroke in an elderly male non-institutionalized population, independently of known confounding factors (69).

In a cross-sectional analysis of 1475 individuals, the authors found that subjects with an apnea-hypopnea index of 20/hour or greater had increased odds for stroke (odds ratio 4.33; 95% CI 1.32 to 14.24; p = 0.02) compared with those without sleep-disordered breathing (AHI < 5) after adjustment for known confounding factors (02). They concluded that there is a strong association between moderate to severe sleep-disordered breathing and prevalent stroke, independent of confounding factors.

In a large population-based study, among 842 subjects (median age 65 years) the median respiratory event index (REI) score was 14, and 63% had sleep-disordered breathing (09). REI was associated with recurrent ischemic stroke (hazard ratio 1.02; 95% CI 1.01-1.03), but not with mortality alone (hazard ratio 1.00; [95% CI 0.99-1.02). The authors concluded that sleep-disordered breathing may represent an important modifiable risk factor for poor stroke outcomes.

There is growing evidence that small vessel disease and leukoaraiosis are worse in subjects with sleep apnea disorder. White matter disease in the form of leukoaraiosis is associated with sleep-disordered breathing in patients with acute stroke (41), and there is a greater risk of silent strokes in high-risk individuals who have nocturnal oxygen desaturation (30). A study found that obstructive sleep apnea with an AHI of more than 15/hour is a risk factor for cerebral white matter changes in middle-aged and older patients (OR 2.08; 95% CI 1.05 to 4.13) (50). Also, it has been shown that obstructive sleep apnea with an AHI of more than 15 is a risk factor for silent cerebral infarction in patients older than 65 (OR 2.44; 95% CI1.03 to 5.80) (20). In another study, moderate-to-severe obstructive sleep apnea was associated with multiple indicators of cerebral small vessel disease, including white matter hyperintensities, cerebral microbleeds, and enhanced perivascular spaces as seen in MRI (92). It has also been suggested that moderate-to-severe obstructive sleep apnea can be one of the independent predictors of cerebral microbleeds, which are considered a surrogate marker of overt stroke (51). White matter hyperintensities represent a combination of microinfarcts, gliosis, and edema.

In a study of dynamic susceptibility with contrast and dynamic contrast-enhanced MRI in 27 patients, to determine cerebral blood flow and blood brain barrier permeability, expressed as leakage rate and volume, the authors found a decrease in cerebral blood flow and an increase in leakage in the perilesional zones close to white matter hyperintensities (104; 100). The authors concluded that blood brain barrier impairment and hypoperfusion in relation to white matter hyperintensities suggest deterioration of a physiologic regulatory mechanism.

The extensive white matter lesions partially disconnect the thalamus and other basal ganglia from preferentially frontal cortical regions. In fact, the frontal cortex in patients with obstructive sleep apnea suffering hypoxia may be abnormally thin (61). A thin cortex might be the consequence of transneuronal degeneration as a result of disconnection at the periventricular white matter level, although direct cortical damage by hypoxia cannot be ruled out. Obstructive sleep apnea may be a risk factor for subcortical ischemic vascular dementia, or Binswanger disease, a form of vascular dementia in the elderly (80). Binswanger disease is manifested by executive dysfunction, gait disorder, and urinary incontinence. White matter hyperintensities progression has been associated with lower extremity dysfunction during aging and verbal memory deficits (67; 94). Executive dysfunction and slow-gait speed progression have been associated with sleep apnea, and there is some evidence that CPAP treatment improves gait control in severe obstructive sleep apnea (77; 56; 03).

These observations support the hypothesis that brain regions with poor hemodynamic reserve are preferentially affected in sleep apnea and coincide with reports of permanent alteration of auditory event-related potentials in these patients. In fact, obstructive sleep apnea may be a risk factor for subcortical ischemic vascular dementia (28), a notion supported by the observation that older women (mean age 82.3 years) with obstructive sleep apnea more than 15/hour were more likely to develop mild cognitive impairment or dementia (adjusted odds ratio [AOR] 1.85; 95% CI 1.11 to 3.08) (101; 81). In a home polygraphic study of 2636 men (median age 76.0 ± 5.3 years), it was found that nocturnal hypoxemia was associated with cognitive mental decline as measured by the Modified Mini-Mental (MMM) scores (04). For each five points of increment in desaturation index there was an annualized decline in MMM of 0.36 points (P=0.01. In a study of patients with minimal cognitive impairment or with Alzheimer disease, the results showed that in patients with sleep apnea, mental decline started at an earlier age (MC, group 1: 72.63 vs. 83.67; MC, group 2: 72.15 vs. 83.45; MC, group 3: 77.40 vs. 89.89; p < 0.01) (AC, group 3: 83.46 vs. 88.13; p < 0.05) (70). These studies further suggest that early therapeutic intervention in sleep apnea is desirable.

Prevention

The treatment of patients with severe obstructive sleep apnea with CPAP during an average 10.1-year period reduced fatal and nonfatal events, including stroke, compared with untreated patients with sleep apnea (62). Another study demonstrated the efficacy of CPAP treatment for the prevention of vascular events (ie, myocardial infarction, stroke, acute coronary syndrome, and vascular death) in patients with mild to moderate obstructive sleep apnea (AHI 30.9 +/- 21.8) versus untreated controls (AHI 15.3+/-13.0). Over a 6-year follow-up period, treated patients had a 28.5% risk reduction compared with the control group, independent of other possible confounding risk factors (10). Another small study found that administration of auto-titrating CPAP to patients with transient ischemic attack and obstructive sleep apnea nonsignificantly reduced the incidence of vascular events during a 90-day period following the transient ischemic attack compared with a non-treated control group (2% [one of 45 patients] vs. 12% [three of 25]) (05). Timely diagnosis and treatment of obstructive sleep apnea may also alleviate the worsened functional outcome of patients with stroke who have sleep apnea (19; 46; 82). Another important modifiable risk factor may be insufficient or excessive sleep. The relative risk for ischemic stroke in postmenopausal women followed for 7.5 years was 1.22 in those women who reported sleeping less than or equal to six hours per night and who had no cardiovascular disease at baseline (18). In the same group of women, longer sleep (more than or equal to nine hours per night) was also an independent risk factor for stroke. Similar findings in the general population were reported by Grandner and colleagues (39). Analysis of data from the very large Nurses’ Health Study has also revealed a slightly enhanced stroke risk (hazard ratio 1.04) from cumulative lifetime rotating night-shifts, suggesting circadian rhythm disruption along with sleep deprivation as causative factors (08). Finally, a meta-analysis has confirmed the deleterious effect of short sleep on ischemic stroke incidence in 15 studies comprising 474,684 individuals (relative risk 1.15); in this meta-analysis, long sleep was also significant (RR 1.65) (14).

In a study of Medicare beneficiaries, 5757 met inclusion criteria as follows: aged 65 years or older, newly diagnosed with obstructive sleep apnea, had initiated CPAP therapy, and were followed over 25 months (99). Of these, 407 (7%) patients experienced stroke. CPAP adherence was associated with a reduced risk of stroke (hazard ratio 0.98; 95% confidence interval 0.96, 0.99) over 25 months, indicating a 2% reduction in the risk of stroke for each month of CPAP adherence. Stratification showed that the protective effect remained significant for the 12- and 6-month outcome models, but not the 3-month outcome model. The authors concluded that CPAP adherence was associated with a significantly reduced risk of stroke.

Differential diagnosis

Vigilance and attention span can be affected in patients with acute stroke by conditions other than excessive sleepiness. Lesions of the right parietal lobe or deep white matter may induce neglect of the opposite side and may present as lethargy. Frontal lobe damage can be associated with apathy or, in more severe cases, with abulia in which patients show no motor or cognitive initiative. Cerebral edema during the first several weeks after a large ischemic lesion or hemorrhage impairs attention span and alertness. Psychomotor retardation can be a feature of poststroke depression, which can be seen in up to 30% of patients with stroke. Metabolic and toxic conditions need to be ruled out along with other causes of daytime sleepiness, such as narcolepsy and periodic leg movements of sleep.

Finally, congestive heart failure must be ruled out as a cause of periodic respiration during sleep.

Diagnostic workup

Polysomnography, the most valuable diagnostic test, should be performed in any stroke patient with loud snoring, restless sleep with or without involuntary leg movements, or excessive daytime sleepiness, particularly if the patient is obese. Ten or more respiratory events (apneas and hypopneas) per hour of sleep is a commonly used threshold for the diagnosis of obstructive sleep apnea, but 20 or more respiratory events per hour of sleep may be more significant in predicting morbidity and mortality. A decision regarding management should not just take into account the number of respiratory events, however, but should also be based on the severity of nocturnal hypoxemia, cardiac arrhythmias, and sleepiness. Sleepiness can be quantified objectively by the multiple sleep latency test and subjectively by the Epworth Sleepiness Scale.

Polysomnography may also reveal the presence of periodic movements of sleep, the clinical significance of which is based on association with arousals and complaints of insomnia or hypersomnolence. Indeed, in patients with renal failure, the presence of periodic limb movements of sleep is itself a risk factor for stroke (59).

Management

Obstructive sleep apnea adversely affects the length of hospitalization and functional capacity after ischemic strokes (19; 46). In patients with acute stroke, altered vasomotor reactivity may be aggravated by sleep apnea (01). The authors reported the occurrence of intracranial blood flow steal in response to changing vasodilatory stimuli like carbon dioxide elevations in patients with sleep apnea and stroke. This phenomenon has been termed the “reversed Robin Hood syndrome.” It might play a pivotal role in clinical deterioration after an acute stroke and has led to the notion that noninvasive ventilatory correction in select patients with acute stroke might have a beneficial effect on sleep apnea and brain perfusion.

The efficacy of positive airway pressure therapy in stroke outcomes remains controversial. Continuous positive airway pressure reduced the incidence of recurrent strokes during a 5-year follow-up period in patients with sleep apnea compared with patients who did not tolerate the treatment (64). One study reported a reduction of deaths from cardiovascular events (cardiac and cerebrovascular) in patients with severe sleep apnea treated with CPAP over an average period of 10.1 years (62), and another study described a 28.5% reduction in vascular morbidity and mortality even in cases of less severe obstructive sleep apnea (apnea-hypopnea index 30.9 +/- 21.8) compared with nontreated patients (average follow-up six years) (10).

One study showed that therapy with CPAP plus usual care, as compared with usual care alone, did not prevent cardiovascular events in patients with moderate-to-severe obstructive sleep apnea and established cardiovascular disease (65). However, the study results were clinically uncertain because patients received on average 3.7 hours of CPAP applications per night, which is considered to be a subtherapeutic duration of therapy. Another study evaluated the efficacy of CPAP therapy for stroke prevention in patients with obstructive sleep apnea (54). In a systematic review of randomized controlled trials, two independent reviewers explored different databases and evaluated the risk of bias. Subgroup analysis and meta-regression revealed no effect on proposed outcomes. The authors concluded that although there was no evidence that CPAP therapy improves stroke outcomes, concerns regarding risk of bias, CPAP adherence, and the population included in each clinical trial might have reduced the strength of the findings to support the benefit in all patients. The authors finalized by stating that future research exploring these relevant outcomes is needed.

In another study, 679 patients with stroke were screened to assess the effect of CPAP treatment on prevention of new vascular events among patients with stroke and obstructive sleep apnea (40). On follow-up at 3, 6, and 12 months from randomization, significantly better stroke outcomes were found. Recurrence of vascular events showed nonsignificant favorable outcomes, likely because the follow-up time was insufficient. In a systematic review and meta-analysis of 10 randomized controlled trials examining the effectiveness of CPAP in stroke patients with sleep-disordered breathing, the results showed an overall neurofunctional improvement with CPAP (SMD 0.5406, 95% CI 0.0263-1.0548) in the combined analysis of NIH Stroke Scale and Canadian Neurological Scale (07). The overall data indicated that CPAP might be beneficial for neurologic recovery and should be considered in patients with clinically significant sleep apnea who tolerate PAP devices.

Patients with cerebrovascular disease and hypertension have a high prevalence of obstructive sleep apnea. The use of portable polysomnography and autotitrating CPAP in the patients' homes improved both the diagnosis and the treatment for sleep apnea compared with usual care but did not lower blood pressure (06). In a study of 126 subjects screened for arterial hypertension and obstructive sleep apnea, repeated cognitive measurements were performed (52). The authors concluded that arterial hypertension and obstructive sleep apnea appear to be risk factors in common neurocognitive disorder, and cognitive decline may be mitigated by treatment of these risk factors.

Results in the study by Osorio and colleagues showed that treatment with CPAP slowed onset and progression of mental decline (CMC, group 1: 72.63 vs. 82.10; CMC group 2: 72.11 vs. 82.10; p < 0.01) (70). In a subsequent study, a different group of authors investigated the impact of CPAP use on cognition in patients with obstructive sleep apnea and cognitive impairment due to neurodegenerative or vascular disease (22). Duration of CPAP therapy and good CPAP adherence for a duration of 2 to 12 months was associated with a 2.3-point higher follow-up MoCA score (p < 0.001) and a 1.2-point higher follow-up MMSE score (p = 0.01). The authors concluded that use of CPAP therapy is associated with improved cognitive outcomes in patients with obstructive sleep apnea and cognitive impairment due to a neurodegenerative or vascular etiology.

For patients who do not tolerate continuous positive airway pressure, alternative therapies include various modes of surgery or dental devices. Weight reduction for obese patients and treatment for alcoholism should be encouraged. Besides CPAP, intravenous theophylline and oxygen inhalation are the most effective therapies available for stroke patients with periodic breathing.

Subcortical lesions can induce periodic leg movements of sleep (58). If periodic leg movements of sleep are a significant problem, ropinirole .25 to .75 mg or pramipexole up to 0.75 mg prior to sleep is recommended.

Comprehensive swallowing intervention had therapeutic effects on obstructive sleep apnea and dysphagia after stroke, a mechanism that the authors related to enhancing oropharyngeal muscle strength and changing upper airway structure (76).

Special considerations

Anesthesia

Airway difficulties in patients with sleep apnea have been frequently reported in the anesthesiology literature. Careful monitoring and early provision of an airway are necessary and prophylactic tracheostomy should be considered in particularly serious cases.