Night-to-night variability in sleep and amyloid beta burden in normal aging.

Abstract

Alzheimer's disease is associated with sleep disturbances and accumulation of cerebral amyloid beta. The objective was to examine whether actigraphy-detected sleep parameters might be biomarkers for early amyloid burden. Participants underwent a week of actigraphy and an amyloid positron emission tomography (PET) scan. Sleep duration and continuity disruption (sleep fragmentation and nocturnal awakenings) were extracted and compared between amyloid-positive and amyloid-negative participants. Then multiple linear regressions were used between mean or night-to-night intra-individual variability (standard deviation) of sleep parameters and brain amyloid burden in a voxel-wise analysis. Eighty-six subjects were included (80.3±5.4 years; 48.8% of women). Amyloid-positive participants had a higher variability of sleep fragmentation compared to amyloid-negative participants. This parameter was associated with a higher amyloid burden in the frontal and parietal regions, and in the precuneus, in the whole sample. This study highlights the relevance of using variability in sleep continuity as a potential biomarker of early amyloid pathogenesis.