Sleep and Stroke Kimberly Nicole Mims, MDa, Douglas Kirsch, MDb,* KEYWORDS Stroke Sleep Obstructive sleep apnea RLS/PLMS Parasomnias Cerebrovascular accident Central sleep apnea Insomnia Sleep duration KEY POINTS Growing evidence suggests that sleep amount and sleep disorders may impact risk for stroke; conversely, the cerebrovascular events may change sleep drive and affect breathing patterns dur- ing sleep. Treatment of sleep disorders, whether causative of stroke or caused by stroke, will likely improve sleep-related symptoms and may improve further stroke risk and long-term outcomes. Sleep apnea, both obstructive and central, is strongly associated with increased cerebrovascular events. Other sleep disorders, including insomnia, RLS/PLMS, and parasomnias may also result in increased incidence of stroke. Short and long sleep duration increase cardiovascular events by increasing sympathetic tone and low-grade inflammation. Treatment of sleep disorders reduces sleep disruption and can improve functional stroke outcome as well as decrease stroke risk. Strokes are one of the most common causes of Central sleep apnea (CSA) occurs when respira- death in the United States.1 Growing evidence tory effort is decreased or absent and is commonly suggests that sleep amount and sleep disorders associated with conditions such as heart failure. may impact risk for stroke; conversely, the cere- The most widely accepted epidemiologic data brovascular events may change sleep drive and project that 4% of men and 2% of women suffer affect breathing patterns during sleep. This article from OSA,2 although more recent data suggest describes the most up-to-date information on the that the incidence of sleep apnea in highly devel- linkage between sleep and stroke and attempts oped countries could be as high as 20% in men to demonstrate how some physicians may use and 10% in women.3 changes in sleep to limit the risk of stroke in In patients with a history of stroke or transient some patients. ischemic attack, sleep apnea incidence is signifi- cantly higher than the general population, with SLEEP APNEA AND STROKE estimates suggesting 72% for apnea-hypopnea index (AHI) >5/h and 38% for AHI >20/h.4 In a small Sleep apnea is defined by decreased airflow study evaluating sleep-disordered breathing (SDB) occurring during sleep. Two main types of sleep incidence in an inpatient stroke rehabilitation unit, apnea exist: obstructive and central. Obstructive 91% demonstrated AHI >10/h with a mean AHI of sleep apnea (OSA) is the most common type of 32/h.5 Furthermore, several prospective cohort sleep apnea and consists of complete or partial studies indicate increased risk of cardiovascular occlusion of the airway, usually accompanied by events in patients with OSA; OSA serves as an in- an associated oxygen desaturation or arousal. dependent risk factor for cardiovascular events.6–9 a Charlotte Medical Clinic, Carolinas HealthCare System, 1001 Blythe Boulevard, Suite 403, Charlotte, NC 28203, USA; b CHS Sleep Medicine, Carolinas HealthCare System, 1601 Abbey Place, Building 2, Suite 200, Char- lotte, NC 28209, USA * Corresponding author. E-mail address: [email protected] Sleep Med Clin - (2015) -–- http://dx.doi.org/10.1016/j.jsmc.2015.10.009 1556-407X/15/$ – see front matter Ó 2015 Elsevier Inc. All rights reserved. sleep.theclinics.com 2 Mims & Kirsch The American Heart Association recommends been implicated in worsening dyslipidemia, oxida- screening for OSA for stroke prevention and sug- tive stress, and endothelial dysfunction and inflam- gests treatment is reasonable, although its effec- mation. In addition, OSA is associated with a tiveness for primary prevention of stroke remains significant increase in carotid intima-media thick- unknown.10 ness and arterial stiffness evidenced as an early The 2014 recommendations by the American indication of atherosclerosis. Heart Association include stratification of antith- Intermittent hypoxia contributes to dyslipidemia rombotic therapy based on CHA2DS2-VASc score, by increasing levels of very low-density lipoprotein which does not incorporate the presence of sleep (VLDL) secretion. This increased secretion is medi- apnea. A retrospective cohort study by Yaranov ated by upregulation of stearoyl coenzyme A desa- and colleagues11 revealed that patients who had turase 1, which increases in direct proportion to atrial fibrillation and OSA developed stroke more severity of nocturnal hypoxia.17 Decreased lipo- commonly than atrial fibrillation patients without protein clearance also contributes to an increase OSA (odds ratio 3.84). in circulating VLDL. Lipoprotein lipase contributes to clearing circulating lipoproteins, and intermit- tent hypoxia inhibits its activation. Patients with Pathology of Obstructive Sleep Apnea and OSA who used positive pressure therapy as treat- Stroke ment had increased lipoprotein lipase activity.18 Although the specific causal mechanism linking sleep However, several other studies contradict a rela- apnea to increased strokerisk hasyet tobeidentified, tionship between OSA and dyslipidemia, and addi- several direct and indirect relationships contributing tional studies have been suggested to further to atherosclerosis are known. Atherosclerosis, tradi- investigate the relationship.19 tionally viewed as solely a disease of lipid storage, is Intermittent hypoxia resulting from OSA is highly now thought to be multifactorial, with several pro- associated with oxidative stress. Oxygen free rad- cesses contributing to plaque development. Factors icals lead to lipid peroxidation, which are acquired contributing to atherosclerotic development include more easily by macrophages; this causes macro- hypertension, metabolic syndrome (diabetes, dysli- phage foaming and provides a substrate for the pidemia), and smoking. Inflammatory mediators of progression of the atherosclerotic plaque.17 atherosclerosis include markers of systemic inflam- Although most studies verify an increase in lipid mation (eg, interleukin [IL]-6, C-reactive protein peroxidation and oxidized low-density lipoprotein [CRP], intracellular adhesion molecules [ICAMs]), in patients with OSA, the lack of benefit seen fibrinogen, and lipoprotein (a). with antioxidant therapy raises the question of OSA causes repetitive episodes of decreased whether oxidative stress is a result of vascular oxygenation mimicking asphyxia and results in inflammation instead of atherosclerosis20 (Fig. 1). negative intrathoracic pressure. Increased arousals OSA is also correlated with an increase in in- from sleep occur in response to decreased oxygen flammatory mediators and cytokines thought to levels and increased circulating carbon dioxide contribute to endothelial dysfunction. A direct pro- levels. Arousals during sleep increase sympathetic portional relationship is observed with elevation of activation, resulting in brief increases in blood pres- inflammatory markers in patients with increased sure. Patients with OSA demonstrate increased AHI, resulting in increased serum levels of markers, incidence of refractory hypertension, perhaps as a including CRP and IL-6. Several studies indicated result of the changed nocturnal blood pressure.12 an increase in CRP was independently associated OSA may also cause insulin resistance resulting in with OSA and nocturnal hypoxemia, although con- diabetes mellitus type 2,13 thought to be secondary tradictory studies found increased CRP to be more to an increase in circulating cortisol. Leptin, a hor- independently associated with body mass index mone released by adipocytes in response to food, (BMI) than OSA severity. IL-6, responsible for CRP is decreased in patients with OSA, lowering their production by the liver, also increases in patients metabolic rate, decreasing the sensation of full- diagnosed with OSA compared with those without, ness, and contributing to metabolic syndrome and although contradictory studies exist linking this increased weight gain.14 mediator to BMI as well.21 Intracellular adhesion The predominant abnormality of OSA stems from molecules, which facilitate leukocyte adhesion to intermittent hypoxia occurring during apnea and vascular endothelium, increase in OSA patients hypopnea events. In several mice models, intermit- compared with controls, and they increase in direct tent hypoxia resulted in increased formation of fatty proportion to nocturnal hypoxemia.22 OSA and streaks in the aortic arch and acceleration in devel- nocturnal hypoxemia severity also increase tumor opment of disease in those genetically prone to necrosis factor-a with additional influence by age atherosclerosis.15,16 Intermittent hypoxia has and BMI.23 Although data are limited on IL-8, Sleep and Stroke 3 Oxidave stress Endothelial dysfuncon and inflammaon Inhibion of Insulin lipoprotein Dyslipidemia Diabetes Atherosclerosis resistance lipase Hypertension Increased sympathec acvaon Fig. 1. Intermittent hypoxia during apneic events increases atherosclerosis via multiple mechanisms. selectins, and vascular cell adhesion molecules, that is, large-vessel versus Small-vessel strokes. data suggest these markers are significantly However, no studies have yet demonstrated a sig- elevated in patients with OSA, again associated nificant difference in type of ischemic stroke seen with influence by age and BMI21 (Fig. 2). In har- in patients diagnosed with sleep-disordered vested endothelial cells in patients with OSA, the breathing (SDB) compared with those without. nuclear factor, NF-kB,