Premium
Sleep disturbance is associated with longitudinal Aβ accumulation in healthy older adults
Author(s) -
Winer Joseph R.,
Mander Bryce A.,
Jagust William J.,
Walker Matthew P.
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.045646
Subject(s) - sleep (system call) , sleep disorder , medicine , longitudinal study , audiology , psychology , cardiology , cognition , neuroscience , pathology , computer science , operating system
Background Work in rodent models of Alzheimer’s disease has demonstrated that sleep disturbance leads to increased β‐amyloid (Aβ) production and decreased clearance. Impaired sleep quality has been shown to predict Aβ burden in healthy older populations cross‐sectionally. It remains untested whether objective measures of sleep quality predict the rate of Aβ accumulation over time. Method 32 healthy older adults (age 75.5 ± 4.3 at baseline, 23 female) received overnight sleep EEG recording to assess sleep architecture and slow wave activity (SWA). All subjects additionally received multiple 11 C‐PIB PET scans to assess longitudinal change in Aβ burden (Figure 1, 2.5 ± 0.8 scans over 3.6 ± 2.3 years). A global cortical 11 C‐PIB DVR (0‐90 min, cerebellar gray reference) was calculated for every PET image, and a linear mixed‐effects model was used to derive slopes of 11 C‐PIB DVR change over time for every subject. All analyses were adjusted for age at baseline and gender, and longitudinal analyses were additionally adjusted for number of PET scans per subject. Result Lower <1Hz SWA at baseline predicted not only higher cross‐sectional 11 C‐PIB DVR, but further predicted an accelerated rate of 11 C‐PIB DVR increase over time (Figure 2A , 3A&B). This was similarly true for the measure of poor sleep efficiency, predicting both baseline and rate of 11 C‐PIB DVR increase (Figures 2B , 3C&D). The significance of these associations were attenuated when baseline 11 C‐PIB DVR was included in the models, due to the strong correlation between 11 C‐PIB DVR baseline and longitudinal change (Figure 3E). Conclusion These findings demonstrate that impaired electrophysiological sleep quality (<1Hz SWA) and objective poor sleep quality are not only sensitive to an individual’s Aβ burden at the time of measurement, but also predictive of subsequent accelerated Aβ aggregation. Sleep may therefore serve as an informative, non‐invasive and scalable biomarker for assessing pathological progression of Alzheimer’s disease. If replicated in larger cohorts with high sensitivity, specificity, and accuracy, these objective sleep metrics may also provide a practical and cost‐effective measure for tracking treatment intervention efficacy.