Abstract:
Background Alzheimer's disease is a degenerative disease with high incidence in the elderly, mainly characterized by impairment of memory function. Identifying and intervening in the early stage of dementia can significantly delay the progression of the disease.Objective To investigate the relationship between sleep oscillation characteristics and cognition, and discuss its potential as an electrophysiological marker for assessing cognitive impairment. Methods Wavelet transform, multiscale power spectral analysis, and phase-averaging techniques were employed to comparatively analyze sleep EEG signals collected from May 2021 to January 2023. Data were obtained from normal controls (NC), amnestic mild cognitive impairment (aMCI) patients, and Alzheimer's disease (AD) patients with varying cognitive levels at the Department of Neurology, Tianjin People's Hospital (including both outpatient and inpatient populations). Sleep spindle microstructural features were systematically investigated across different cognitive states through these multimodal analytical approaches.Results Totally 9 AD patients (mean age 70.11±6.17 years), 13 aMCI patients (67.23±5.18 years), and 15 NC subjects (67.87±7.76 years) were enrolled. In the NC, aMCI, and AD groups, with the decline of cognitive function, the sleep spindle density (F=12.92, P0.000), the sleep spindle duration (F=45.62, P< 0.001), the sleep spindle relative power (F=7.897, P<0.01), and the sleep spindle amplitude all decreased (H=12.556, P<0.01), but there was no significant correlation between sleep spindle wavelength and the degree of cognitive impairment (P>0.05). Conclusion Cognitive decline is associated with reduced sleep spindle density, relative power, amplitude, and duration, but it is not positively correlated with sleep spindle wavelength. Changes in density and duration characteristics have been significantly observed in the disease stage of aMCI, providing clues for the early diagnosis of aMCI.