A potential causal link between obstructive sleep apnea (OSA) and stroke is still under investigation. Evidence from a variety of studies, including studies conducted at UI Hospitals and Clinics, suggests a two-way street, according to Mark Eric Dyken, MD, associate professor of neurology and director of the Sleep Disorders Clinic, in his account of current polysomnographic data from patients with OSA.

History: Some abnormal sleep-related respiratory events have been defined as apneas and hypopneas. Apneas are characterized by a complete shutdown of airflow. Both conditions may be classified as central, obstructive, or mixed. During central events, there is no respiratory effort. Mixed phenomena start with a central component but are followed by an obstruction. Obstructions are characterized by a complete collapse of the upper airway, whereas hypopneas are associated with significantly reduced airflow when compared with normal resting respiration. During obstructive sleep apnea (OSA), respiratory efforts persist despite airway closure.

In adults, a significant apnea persists for at least 10 seconds and leads to an oxygen saturation drop of >3% (possibly followed by an arousal). A diagnosis of OSA is confirmed using polysomnography (PSG) that shows a frequency of apnea and hypopnea episodes--an apnea/hypopnea index (AHI)--greater than or equal to five events per hour.

OSA is typically associated with loud snoring, sleep-maintenance insomnia, excessive daytime sleepiness, obesity, male gender, and increasing age. Nevertheless, people who sleep alone may not be able to report snoring, and occasionally, the history obtained from a severe sleep apneic may also be limited due to sleepiness and confusion.

The literature suggests that untreated OSA may contribute to hypertension, cardiovascular disease, stroke, and death, especially in the critical care population. Patients who are seriously ill may not breathe normally for a variety of reasons; it has been documented that bilateral dysfunction of the diencephalon and pons or injury to the medullary respiratory centers has produced sleep apnea. Although stroke in certain areas of the brain may lead to OSA, studies suggest that in some cases OSA may lead to stroke.

New facts: In a random sample of 602 people between 30 and 60 years of age, the prevalence of OSA ranged between 4% and 9% for women and between 9% and 24% for men. To date, published research on the subject of a causal relationship between OSA and stroke is scarce. No large, well-designed, prospective double-blind experiments, with polysomnographic evaluations performed before and after stroke to compare treated and untreated patients with OSA, have been conducted. The available data about the relationship between stroke and OSA are drawn from nonselected, retrospectively studied populations and case reports and from selected and nonselected prospectively studied populations.

A potential relationship between stroke and OSA began to be suspected in the 1980s after some post-stroke studies using PSG revealed a high prevalence of OSA among stroke patients. In 1991, 47 selected stroke patients were found to have an overall mean AHI of 28 events per hour. Thirty-four of these individuals (72%) were diagnosed with OSA. These patients tended to be sleepier, more obese, and older than the nonapneics.

In 1988, a cohort of 198 patients with OSA were treated with tracheostomy (71) or weight loss (127) and followed for seven years. At the end of this period, only 1% in the tracheostomy group suffered a new stroke (3% died), whereas in the weight-loss group, 9% experienced a new stroke, while 17.3% died.

Practice: At UI Hospitals and Clinics, our team became interested in the association between OSA and stroke in the case of a 34-old morbidly obese man who awoke with stroke. This individual had been a snorer and was generally sleepy. Given this history, a PSG was performed, which showed severe OSA with an oxygen desaturation to 60% and an AHI of 36 events per hour. The traditional work-up suggested that this patient might have suffered a sleep-related hypertensive stroke, which was precipitated by OSA.

We then performed laboratory studies and were able to show that OSA could elicit excessive sympathetic and parasympathetic activation that would lead to respective surges in blood pressure (up to 230/130 mm Hg) and bradyarrhythmias with a complete heart block. It seemed rational to hypothesize that vacillating blood pressures with hypoperfusion and hypoxia may predispose OSA patients to cardiovascular and cerebrovascular catastrophe.

Next, our group compared the prevalence of OSA in acute stroke patients with normal gender- and age-matched control subjects. OSA was diagnosed in 71% of the stroke patients and in only 19% of the controls. We found that a higher-than-expected percentage of patients (54%) experienced their strokes during sleep (p=0.0304). In a follow-up study, we found that, of the five patients in this group who had OSA and died, four suffered their original stroke during sleep.

In general, the immediate treatment of choice for OSA is continuous positive airway pressure therapy (CPAP). In addition, we recommend implementing good dietary habits with low-level exercise (if and when possible), avoiding unnecessary central nervous system depressants, especially if taken prior to sleep, and practicing good sleep hygiene. In some cases, upper airway surgery (uvulopalatopharyngoplasty), tracheostomy, or oral devices may be considered as treatment alternatives.

A full sleep history for OSA and a PSG, when indicated, should be considered for all clinically suspect patients. Since OSA has been reported in association with a variety of significant health-related problems, it should especially be considered in individuals who present with hypertension and cardiovascular or cerebrovascular disease. The treatment of OSA in this population may be an additional way to minimize patient morbidity and mortality.