Multiple sleep–wake disturbances after stroke predict an increased risk of cardio‐cerebrovascular events or death: A prospective cohort study

Abstract

AbstractBackground and purposeContradictory evidence on the impact of single sleep‐wake‐disturbances (SWD), such as sleep‐disorderd breating (SDB) or insomnia, in patients with stroke, on the risk of subsequent cardio‐ and cerebrovascular events (CCE) and death, exists. Very recent studies in the general population suggest that the presence of multiple SWD increases cardio‐cerebrovascular risk. Hence, the aim of this study was to asssess whether a novel score capturing the burden of multiple SWD, a so called "sleep burden index", is predictive for subsequent CCE including death in a prospectively followed cohort of stroke patients.MethodsPatients with acute ischemic stroke or transient ischemic attack (TIA) were prospectively recruited. Four SWD were analyzed: (i) SDB with respirography; (ii) insomnia (defined using the insomnia severity index [ISI]); (iii) restless legs syndrome (RLS; defined using the International RLS Study Group rating scale); and (iv) self‐estimated sleep duration at 1 and 3 months. A “sleep burden index”, calculated using the mean of z‐transformed values from assessments of these four SWD, was created. The occurrence of CCE was recorded over a mean ± standard deviation (SD) follow‐up of 3.2 ± 0.3 years.ResultsWe assessed 437 patients (87% ischemic stroke, 13% TIA, 64% males) with a mean ± SD age of 65.1 ± 13.0 years. SDB (respiratory event index ≥ 5/h) was present in 66.2% of these patients. Insomnia (ISI ≥ 10), RLS and extreme sleep duration affected 26.2%, 6.4% and 13.7% of the patients 3 months post‐stroke. Seventy out of the 437 patients (16%) had at least one CCE during the follow‐up. The sleep burden index was associated with a higher risk for subsequent CCE, including death (odds ratio 1.80 per index unit, 95% confidence interval 1.19–2.72; p = 0.0056).ConclusionThe presence of multiple SWDs constitutes a risk for subsequent CCE (including death) within the first 3 years following stroke. Larger systematic studies should assess the utility of the sleep burden index for patients' risk stratification in clinical practice.