The Bidirectional Relationship Between Sleep Disorders and Stroke

Stroke continues to be among the leading causes of death and disability worldwide, with much of this burden attributable to a number of modifiable risk factors (eg, atrial fibrillation, hypertension, disorders of glucose metabolism, obesity, physical inactivity, smoking). Apart from these traditional risk factors, sleep disorders have been increasingly recognized as not only independent risk factors for stroke but also possible consequences of stroke. Moreover, the presence or treatment of sleep disorders may affect stroke recovery and functional outcomes. Therefore, understanding the complex bidirectional relationship between sleep disorders and stroke is essential to develop novel strategies in the prevention and management of stroke (see Table).

Obstructive Sleep Apnea

Obstructive sleep apnea (OSA) involves recurrent episodes of complete (apneic) or reduced (hypopneic) inspiratory airflow during sleep.¹ These respiratory episodes arise as a consequence of upper airway collapse despite ongoing respiratory effort, which is influenced by a number of anatomic and physiologic factors. The mainstay treatment for OSA is continuous positive airway pressure (CPAP).

OSA is the most common poststroke sleep disorder, with ~65% of stroke survivors having some degree of OSA, >25% of which are classified as severe.² Individuals with preexisting OSA have poorer functional outcomes, spend more time in rehabilitation, and are at increased risk of recurrent stroke and all-cause mortality.³ OSA prevalence and severity remain generally consistent over the acute to chronic stages of stroke; moreover, there does not seem to be a significant relationship between stroke topography, type, severity, or etiology and the presence of OSA.² This implies that most stroke survivors had some degree of OSA prior to their stroke, but to what degree OSA is the underlying cause or consequence of stroke in a given individual remains unclear.

Despite the strong association, screening for OSA after stroke is uncommon,⁴ which likely reflects barriers to accessing timely diagnostic testing, atypical OSA presentations among stroke survivors, and limited awareness among clinicians of the importance of OSA management in stroke recovery. Sleep evaluation frequently becomes an outpatient issue because acute stroke and stroke rehabilitation centers may not have the necessary resources or expertise to conduct sleep testing in the poststroke period. Although in-laboratory polysomnography is the gold standard method for OSA diagnosis, home sleep apnea testing (HSAT) can identify OSA accurately and may be a more convenient and cost-effective approach in stroke survivors.⁵ The ongoing Sleep for Stroke Management and Recovery (Sleep SMART) trial, which aims to determine the effect of CPAP treatment in the acute stroke or high-risk TIA period, is incorporating the use of HSAT in the inpatient setting.⁶ The American Academy of Sleep Medicine is also expected to release inpatient guidelines on stroke management, which may open the window for performing inpatient HSAT for OSA diagnosis in the immediate poststroke period.

Although traditional stroke risk factors are commonly encountered in people with OSA (eg, hypertension, obesity, diabetes), OSA by itself is associated with ~2-fold increased relative stroke risk.⁷ More severe OSA is associated with a higher stroke risk, with each 10-unit increase in apnea-hypopnea index associated with a 36% relative increase in the odds of stroke or cardiovascular death.⁸ Conversely, strong randomized controlled trial data have demonstrated that CPAP improves excessive daytime sleepiness, quality of life, and neurologic recovery after stroke.⁹ The effectiveness of OSA treatment for primary or secondary stroke prevention in people without sleepiness is difficult to demonstrate and screening for and management of OSA after stroke is underemphasized in current stroke guidelines. A major underlying confounder has been adherence to CPAP therapy in these trials, which was poor and declined over the study durations. However, an “on-treatment” reanalysis of individual data from the largest trials in this area demonstrated a robust reduction in major adverse cardiac and cerebrovascular events in stroke survivors who maintained good CPAP adherence (≥4 hours per night).¹⁰ These results require cautious interpretation because on-treatment analysis is susceptible to selection bias and people with good CPAP adherence could simply reflect overall healthier users.

The pathophysiologic mechanisms that account for stroke risk in people with OSA are likely complex and may involve acute and chronic processes related to recurrent respiratory events as well as the influence of interrelated risk factors common to both stroke and OSA. For example, acute and recurrent episodes of hypopnea and apnea lead to large negative intrathoracic pressure swings, intermittent hypoxemia and hypercapnia, and arousals from sleep. These episodes transiently result in increased sympathetic activity (which over time may be associated with refractory hypertension and arrhythmia), oxidative stress (which may precipitate inflammatory activity and endothelial dysfunction), and recurrent cerebral hemodynamic changes that lead to brain hypoxia.⁷ Moreover, the pathophysiologic changes associated with OSA may cause paradoxical embolization due to right-to-left shunt in individuals with patent foramen ovale. More chronically, these events may contribute to many of the comorbidities associated with OSA and stroke including refractory hypertension, arrhythmia, obesity, and diabetes.⁷ Cerebral small vessel disease (CSVD), which is a disorder of microscopic perforating arterioles, capillaries, and venules of the brain, may be an additional chronic pathophysiologic mechanism linking OSA and stroke. CSVD contributes to at least 25% of ischemic strokes, and OSA severity has been shown to be associated with radiographic evidence of CSVD in a dose–response manner.¹¹

Insomnia and Sleep Duration

Sleep duration and quality are influenced by a number of internal and external factors including age, social and behavioral practices, and environment (eg, light, noise, temperature). Many comorbidities are associated with changes in sleep, and stroke survivors in particular may experience poor sleep in terms of sleep efficiency, total sleep time, wake after sleep onset, and amount of slow-wave sleep.¹²

Insomnia involves persistent difficulty in initiating or maintaining sleep despite adequate opportunity, which results in dissatisfaction with sleep and daytime impairment of function.¹ It is the most prevalent sleep disorder in the general population and is frequently encountered after stroke, with a pooled prevalence of 36% in the chronic phase.² Notably, whereas the diagnostic criteria for insomnia largely rely upon self-reported sleep symptoms and variables (eg, sleep onset latency and sleep duration), the addition of objectively measured reduced sleep time (<6 hours as recorded by polysomnography) confers a higher risk of comorbidities. For example, insomnia with objectively reduced total sleep time was associated with a 2-fold increased risk of hypertension and a 2.5-fold increased risk of cardiovascular disease including stroke compared with people with normal sleep duration without insomnia.¹³ By contrast, people with insomnia but without objectively reduced sleep time did not have a significantly increased risk of hypertension or cardiovascular disease. This suggests that insomnia with objectively determined sleep disturbance is a more severe phenotype than insomnia alone.

Similarly, prolonged sleep duration (>8 to 9 hours in most adults, depending on age) also affects all-cause mortality risk and confers an increased risk of stroke in a dose-dependent manner.¹⁴ For example, a consistently long sleep duration was associated with a 70% increased risk of stroke incidence compared with those who slept between 7 and 7.9 hours.¹⁴ Thus, there appears to be a U-shaped relationship between sleep duration and stroke risk, with higher risk linked with both long and short sleep durations with increasing risk as one moves to further extremes.

The mechanisms that underlie the associations among sleep duration, insomnia, and stroke risk are unclear but likely involve a combination of direct and indirect processes. For example, chronic insomnia and short sleep duration are associated with hypertension, cardiovascular disease, metabolic abnormalities, and mood disorders, which are all risk factors for stroke.¹⁵ Both short and long sleep durations are associated with arteriosclerosis and blood pressure variability, which could imply proinflammatory, endocrine, or dysautonomic mechanisms.¹⁶ On the other hand, altered sleep duration may simply reflect overall poorer health.

Restless Legs Syndrome and Periodic Limb Movements of Sleep

Restless legs syndrome (RLS) is a common chronic sensorimotor disorder characterized by a strong urge to move the limbs when at rest, associated with nocturnal predominance and cessation of the urge once the limbs are moved.¹ The overall prevalence of RLS is 5% to 10% of the general population, but is more common among stroke survivors in the acute and chronic periods (>10%).² Subcortical infarcts involving the pyramidal structures, basal ganglia, and brainstem are associated with poststroke RLS, which coincides with cases of non–stroke-related RLS resulting from dysfunction in these structures.¹⁷

RLS has been associated with a number of cerebrovascular risk factors including hypertension, hypercholesterolemia, diabetes, obesity, and atherosclerotic disease;¹⁸ however, its direct association with stroke is less certain. Two meta-analyses reported either limited evidence for an association between RLS and stroke¹⁹ or absence of such an association.²⁰ In both cases, studies were limited by the method of ascertainment of RLS diagnosis, multiple confounders, and the overall quality of the studies including methodologic rigor, sample size, and control of bias. Conversely, results of a large-scale prospective investigation of insurance claims demonstrated that pharmacologic treatment of RLS was associated with a reduction in cardiovascular risk including stroke.²¹ As such, part of the benefit of RLS treatment could arise indirectly as a result of the improvement of insomnia (see previous section).

Periodic limb movements of sleep (PLMS) involve spontaneous repetitive flexion movements of the limbs during sleep.¹ Although common with aging and especially comorbid with RLS, PLMS are frequently seen among stroke survivors with an estimated prevalence of ≥70%.²² PLMS have also been linked to a number of other diseases including OSA and stroke.¹⁹ An increased frequency of PLMS is associated with greater cerebral small vessel disease burden²³ and is a risk factor for cerebrovascular disease including stroke.²⁴ Given the high prevalence of both PLMS and OSA after stroke, it is notable that a synergistic effect between PLMS and OSA has not been observed.²⁴

Several mechanisms have been proposed to link RLS and PLMS with stroke risk aside from their propensity to contribute to sleep fragmentation and their association with vascular risk factors. These include systemic inflammation, oxidative stress, hypothalamic-pituitary-adrenal system overactivation, and sympathovagal imbalance leading to transient increases in heart rate and blood pressure, particularly when associated with arousals from sleep.²⁵ For example, the generation of PLMS may arise from within the thoracolumbar segments of the spinal cord with resultant repetitive coactivation of preganglionic sympathetic fibers. These periods of increased sympathetic outflow may cause transient cyclic hemodynamic changes that ultimately result in vascular remodeling, shear stress, or platelet activation, which are factors linked with hypertension and stroke.²⁶

Conclusion

The relationship between sleep disorders and stroke is complex and bidirectional (see Figure). Sleep disorders and stroke share comorbid and interrelated risk factors that may influence functional outcomes and stroke recovery. There is growing evidence to suggest that treatment of OSA may reduce stroke recurrence, although this remains controversial. Conversely, studies that have examined the effect of treating nonrespiratory sleep disorders in reducing stroke risk and poststroke outcomes remain limited. Future research should be focused on screening methods to determine poststroke sleep disorders as well as the effect of treatment on stroke risk and poststroke outcomes.