Interhemispheric EEG Coincidence during REM sleep correlates with mild cognitive impairment

Abstract

AbstractBackgroundThere exists a bidirectional relationship between sleep disruption and neuropathology in Alzheimer’s disease (AD). Interhemispheric EEG coincidence during sleep polysomnography (PSG) may reflect coordination of brain networks. While several EEG features of non‐REM (NREM) sleep, such as slow wave activity, are associated with proteins and measures related to AD pathogenesis, less is known about EEG features of rapid eye‐movement (REM) sleep in AD. We assessed EEG features across hemispheres during REM, NREM, and wake states, comparing participants with mild cognitive impairment to cognitively‐normal participants.MethodEEG was collected during an attended overnight polysomnogram from the Biomarkers of Alzheimer’s Disease in Sleep and EEG (BASE) cohort (n = 87, average age 70.7 ± 4.6 years), 19 of whom had Clinical Dementia Rating (CDR) of 0.5, considered Mild Cognitive Impairment (MCI) and the rest healthy controls (HC). A custom low frequency (0.5 – 8Hz) EEG oscillatory peak detector was implemented in MATLAB across six EEG leads (F3, F4, C3, C4, O1, and O2). Interhemispheric EEG coincidence between lead pairs was calculated as the number of mini‐epochs with oscillatory peaks in both leads divided by the number of mini‐epochs with a peak in either or both leads.ResultNeither sleep macro‐architecture (total time in any stage of sleep, wake after sleep onset, number of awakenings and sleep efficiency) nor spectral power density in delta or theta frequency bands differed between HC and MCI groups. Interhemispheric EEG coincidence was reduced during REM sleep in F3‐F4 (Frontal), C3‐C4 (Central) and O1‐O2 (Occipital) lead pairs in the MCI group compared to HC, all regions p<0.0001. REM F3‐F4 coincidence was negatively correlated with nighttime awakenings (r ‐0.4, p<0.0001). Performance on the Word‐Pair Task, which measures memory, was correlated with REM F3‐F4, C3‐C4, and O1‐O2 coincidence (r0.29 p0.009; r0.27 p0.01; r0.25 p0.03, respectively).ConclusionA novel neurophysiological marker of interhemispheric EEG coincidence, specific to lower frequency spectra during REM sleep, differentiates MCI from HC and correlates to memory performance. Interhemispheric EEG coincidence may reflect coordination of brain networks during REM and may be sensitive to changes in cognitive status that would otherwise go undetected by traditional analyses of sleep macro‐architecture.