Sleep-Disordered Breathing Is Associated With Recurrent Ischemic Stroke

Brown, Devin L. MD; Shafie-Khorassani, Fatema MPH; Kim, Sehee PhD; Chervin, Ronald D. MD, MS; Case, Erin BA; Morgenstern, Lewis B. MD; Yadollahi, Azadeh PhD; Tower, Susan MD; Lisabeth, Lynda D. PhD


Stroke. 2019;50(3):571-576. 

In This Article

Abstract and Introduction


Background and Purpose: Limited data are available about the relationship between sleep-disordered breathing (SDB) and recurrent stroke and mortality, especially from population-based studies, large samples, or ethnically diverse populations.

Methods: In the BASIC project (Brain Attack Surveillance in Corpus Christ), we identified patients with ischemic stroke (2010–2015). Subjects were offered screening for SDB with the ApneaLink Plus device, from which a respiratory event index (REI) score ≥10 defined SDB. Demographics and baseline characteristics were determined from chart review and interview. Recurrent ischemic stroke was identified through active and passive surveillance. Cause-specific proportional hazards models were used to assess the association between REI (modeled linearly) and ischemic stroke recurrence (as the event of interest), and all-cause poststroke mortality, adjusted for multiple potential confounders.

Results: Among 842 subjects, the median age was 65 (interquartile range, 57–76), 47% were female, and 58% were Mexican American. The median REI score was 14 (interquartile range, 6–26); 63% had SDB. SDB was associated with male sex, Mexican American ethnicity, being insured, nonsmoking status, diabetes mellitus, hypertension, lower educational attainment, and higher body mass index. Among Mexican American and non-Hispanic whites, 85 (11%) ischemic recurrent strokes and 104 (13%) deaths occurred, with a median follow-up time of 591 days. In fully adjusted models, REI was associated with recurrent ischemic stroke (hazard ratio, 1.02 [hazard ratio for one-unit higher REI score, 95% CI, 1.01–1.03]), but not with mortality alone (hazard ratio, 1.00 [95% CI, 0.99–1.02]).

Conclusions: Results from this large population-based study show that SDB is associated with recurrent ischemic stroke, but not mortality. SDB may therefore represent an important modifiable risk factor for poor stroke outcomes.


Sleep-disordered breathing (SDB) is a highly prevalent condition in poststroke patients, with the majority affected by obstructive, rather than central, sleep apnea.[1] SDB, a condition that also appears to be prevalent before stroke,[2] is an established risk factor for incident stroke.[3] Multiple prospective studies have shown that SDB is associated with poststroke mortality. As examples, a respiratory event index (REI) score ≥30 was associated with an increased risk of death in 161 Spanish patients after first stroke/transient ischemic attack in an adjusted analysis.[4] Another study showed that an REI score >10 was associated with death 6-month poststroke, after adjustment for confounders in 114 UK patients.[5] In a study of 132 Swedish patients in rehabilitation, an obstructive apnea index ≥15 was associated with risk of death after adjustment for multiple confounders.[6] Nevertheless, studies on the relationship between mortality and SDB in stroke patients have not included diverse populations, larger samples, or population-based data.

Multiple meta-analyses based on prospective cohort studies now support the association between SDB and incident stroke.[3,7,8] Yet, only scant prospective data have been published to support the relationship between SDB and ischemic stroke recurrence, and none has taken into account competing risks. One study from Poland found a higher unadjusted risk of recurrent stroke/transient ischemic attack at 2 years in 91 first-ever ischemic stroke/transient ischemic attack patients with an REI score >5 as compared with those with an REI score ≤5.[9] Approximately 160 000 recurrent ischemic strokes occur each year and these events are even more costly than incident strokes.[10] This underscores the need for the identification of novel, modifiable risk factors for recurrent stroke, such as SDB.

Mexican Americans (MAs), the largest subgroup of Hispanics, who are the most populous minority group in the United States, are at a higher risk of recurrent stroke[11] than non-Hispanic whites (NHWs).[12] Furthermore, we have previously published that after ischemic stroke, SDB is more prevalent and severe in MAs than NHWs.[13] The association between SDB and stroke outcomes in MAs is unknown and whether SDB has a differential association with outcomes after stroke by ethnicity is also unknown. We therefore sought, among subjects with ischemic stroke who enrolled in a bi-ethnic population-based study to: (1) investigate the association between SDB and recurrent ischemic stroke; (2) investigate the association between SDB and all-cause poststroke mortality; and (3) determine whether ethnicity modifies the effect of SDB on these outcome measures, or whether the strength of the ethnicity-outcome association is confounded by SDB. In the context of these questions, marginal probabilities for cause-specific events were assessed to correct results for competing risks.