Abstract
STUDY OBJECTIVES: The objective of this study was to determine which respiratory and architectural sleep parameters are related to cognitive function and cognitive status (mild cognitive impairment [MCI] versus normal cognitive aging [NCA]) in community-dwelling individuals with hypertension. Additionally, it aimed to determine whether the results changed in the presence or absence of vascular brain lesions (silent brain infarcts and extensive white matter hyperintensities [WMHs]). METHODS: In a cohort of individuals with hypertension and without previous stroke or dementia, we conducted in-hospital polysomnography including electroencephalography, electro-oculography, electromyography, and magnetic resonance imaging to assess silent brain infarcts and WMHs. Cognitive testing was carried out with a screening test (Dementia Rating Scale version 2 [DRS-2]) and a complete cognitive visit. RESULTS: This study included 158 participants with a median age of 65.0 years; 32.3% were females, and the median apnea-hypopnea index was 22.3 events/h. MCI was diagnosed in 24 study participants, and the rest had NCA. Regarding respiratory parameters, total DRS-2 scores (β; 95% CI) 0.121; 0.026, 0.215 were positively associated with mean O₂ saturation, whereas total (-0.022; -0.036, -0.009), executive function (-0.016; -0.026, -0.006) and memory (-0.017; -0.029, -0.004) DRS-2 scores were all negatively associated with the percent of time with oxygen saturation < 90% after correcting for education, vascular risk factors, and magnetic resonance imaging lesions. Regarding sleep architecture, Attention DRS-2 scores (0.0153; 0.001, 0.306) were independently associated with total sleep time. Similar results were obtained in the absence of silent brain infarcts or WMHs in the stratified analysis. None of the sleep parameters were associated with cognitive status. CONCLUSIONS: Low oxygen saturation contributes to cognitive performance, and this effect appears even in the absence of vascular brain lesions in individuals with hypertension.