Impact of sleep disturbance on longitudinal cognitive performance in patients with transient ischemic attack or mild stroke

Abstract

Introduction: Sleep disturbances including obstructive sleep apnea (OSA) and poor sleep quality are common after stroke, while its association with cognitive changes following transient ischemic attack (TIA) or mild stroke remains unclear. We aim to determine whether sleep duration, OSA parameters, or nocturnal hypoxemia is associated with a greater cognitive decline after stroke. Methods: We prospectively followed-up patients with acute TIA/mild stroke [National Institute Health Stroke Scale (NIHSS) < 7] who underwent baseline sleep questionnaire [Pittsburgh Sleep Quality Index (PSQI)], and serial cognitive assessments [Montreal Cognitive Assessment (MoCA) 5-min, Stroop Test] at baseline and one-year. We also evaluated apnea-hypopnea index (AHI) and nocturnal hypoxemia by Home Sleep Apnea Test (HSAT) at one-year. Primary outcome was one-year change in MoCA 5-min score. Results: One hundred and five patients with TIA/mild stroke (mean age 63 years, 65 % male) were included. Baseline short sleep (< 6 hour/night) and AHI ≥ 20/hour at one-year were independently associated with a decline in the MoCA 5-min total score after covariates adjustment [short sleep: β = −2.36 95 % confidence interval (CI) (−4.13, −0.59), p = 0.009; AHI ≥ 20/hour: β = −1.79 (−3.26, −0.32), p = 0.017; remained significant after multiple comparisons correction]. A lower mean MinSpO2 was associated with a decline in executive function [Stroop interference index: β = 0.29 (0.04, 0.53), p = 0.021], but not with MoCA 5-min score at one-year. Moderation analysis indicated AHI ≥ 20/hour was associated with a pronounced decline in executive function only in men. Conclusions: Short sleep after stroke onset, AHI ≥ 20/hour and nocturnal hypoxemia at one-year contributed to an impaired cognitive trajectory at one-year following stroke in patients with TIA/mild stroke.