Dissertation committee:
Αναστάσιος Μπονάκης, Αναπληρωτής Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Κωνσταντίνος Βουμβουράκης, Ομότιμος Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Γεώργιος Τσιβγούλης, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Γεώργιος Παρασκευάς, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Σωτήριος Γιαννόπουλος, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Ιωάννης Τζάρτος, Αναπληρωτής Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Βασιλική Ζούβελου, Αναπληρώτρια Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Summary:
Sleep-disordered breathing (SDB) and periodic limb movements during sleep (PLMS) are frequently present among patients with acute stroke. The severity of these two sleep disorders is much higher among stroke patients compared to the general population. Moreover, the detection of SDB or PLMS seems to affect the clinical outcome after stroke. In this PhD thesis, we sought to investigate the prevalence, severity and type of these two sleep disorders in consecutive acute stroke patients. Moreover, we focused on identifying risk factors that are associated with higher odds of detecting SDB or PLMS in the acute stroke setting. In addition, we tried to evaluate the association between PLMS and coexisting SDB. Finally, we investigated the potential impact of SDB or PLMS on functional outcome at three months after stroke. We prospectively studied consecutive acute stroke patients, who underwent overnight polysomnography within the first 72 h from symptom onset. Demographics, clinical and imaging characteristics were documented. Daytime sleepiness preceding the stroke, stroke severity on admission and functional outcome at three months were evaluated using the Epworth-Sleepiness Scale (ESS), National Institute of Health Stroke Scale (NIHSS) and modified Rankin Scale (mRS), respectively. Sleep stages, respiratory events and periodic limb movements (PLMs) were identified and scored using standard polysomnography criteria. We compared the baseline characteristics with the likelihood of detecting SDB or PLMS in the acute stroke setting as well as with the likelihood of excellent functional outcome at three months. For that purpose, we used univariable and multivariable logistic regression models before and after adjusting for potential confounders. 1. A total of 130 consecutive acute stroke patients were prospectively evaluated [110 with ischemic stroke and 20 with intracerebral hemorrhage, mean age 60.5 ± 10.9 years, 77% men, median NIHSS score on admission: 3 (IQR: 2–7)]. The rate of SDB detection on polysomnography recordings was 79% (95% CI: 71–86). The prevalent SDB type during the acute stroke phase was obstructive sleep apnea (81%). Three variables were independently associated with the likelihood of SDB detection in multivariable analyses adjusting for potential confounders: age (OR per 10-year-increase: 2.318, 95% CI: 1.327–4.391, p = 0.005), male sex (OR: 7.901, 95% CI: 2.349–30.855, p = 0.001) and abnormal ESS-score (OR: 6.064, 95% CI: 1.560–32.283, p = 0.017). Among patients with SDB, congestive heart failure was independently associated with the likelihood of central apnea detection (OR: 18.295, 95% CI: 4.464–19.105, p < 0.001). Among all patients, increasing NIHSS score on admission (OR: 0.817, 95% CI: 0.737-0.891, p < 0.001) and Apnea– Hypopnea Index (OR: 0.979, 95% CI: 0.962–0.996, p = 0.020) emerged as independent predictors of excellent functional outcome at 3 months (mRS-scores 0–1). 2. We prospectively assessed 126 patients with acute stroke [109 with ischemic and 17 with hemorrhagic stroke, mean age 60 ± 11 years, 68% men, median NIHSS score on admission: 3 (IQR: 2–7)]. The overall rate of PLMS in our cohort was 76%, and the rate of SDB among patients with PLMS was 83%. PLMS detection rates differed significantly (p-value: <0.001) according to SDB, with PLMS prevalence increasing with greater SDB severity. The frequency of “genuine” (independent of respiratory events) PLMS was much higher than the one of SDB-associated PLMS (50% versus 30%, respectively). SDB could independently (OR:4.869, 95% CI: 1.884–12.784, p-value: 0.001) predict the presence of PLMS in the acute stroke phase in multivariable analyses adjusting for potential confounders. Moreover, baseline stroke severity (NIHSS-score increase in per-1 point: OR: 0.819, 95% CI: 0.737–0.895, p-value < 0.001) and PLMS (OR:0.099, 95% CI: 0.009–0.482, p-value = 0.015) were significantly associated with the likelihood of excellent functional outcome (mRS-scores: 0–1) at 3 months. As shown in these two studies, the prevalence of both SDB and PLMS in patients with acute stroke is very high with the vast majority of them having severe sleep apnea or PLMS. Although, these two sleep disorders frequently coexist and interact, they are not always directly or causally associated in the acute stroke setting. The detection of SDB or PLMS is independently associated with the likelihood of poor functional outcome at three months after stroke. The common presence of mostly severe SDB or PLMS in patients with acute stroke in combination with their negative impact on post-stroke recovery point out the importance of polysomnography implementation in everyday clinical practice of acute stroke work-up and management.
Keywords:
Acute stroke, Sleep-disordered breathing, Prevalence, Risk Factors, Functional outcome
