The association of subjective sleep characteristics and plasma biomarkers of Alzheimer's disease pathology in older cognitively unimpaired adults with higher amyloid-β burden

We aimed to investigate the association of subjective sleep characteristics and plasma Alzheimer's disease (AD) biomarkers in older cognitively unimpaired adults with higher amyloid-beta (A beta) burden. Unimpaired cognition was determined by education-adjusted performance for the Mini-Mental State Examination and exclusion of dementia and mild cognitive impairment via standardized neuropsychological tests. We used Pittsburgh Sleep Quality Index (PSQI) to assess subjective sleep quality. The participants also underwent examination of plasma AD biomarkers and F-18-florbetapir PET scan. Correlation and multiple linear regression analyses were used to investigate the association between subjective sleep characteristics and AD biomarkers. A total of 335 participants were included and 114 were A beta-PET positive. Multivariable regression analysis showed sleep duration > 8 h and sleep disturbance were associated with A beta deposition in total participants. Two multiple linear regression models were applied and the results revealed in participants with A beta-PET (+), falling asleep at >= 22:00 to <= 23:00 was associated with higher levels of A beta 42 and A beta 42/40. Other associations with higher A beta 42/40 and standard uptake value ratio contained sleep efficiency value, sleep efficiency >= 75%, no/mild daytime dysfunction and PSQI score <= 5. Higher p-Tau-181 level was associated with sleep latency > 30 min in A beta-PET (+) group and moderate/severe sleep disturbance in A beta-PET (-) group. Our data suggests sleep duration <= 8 h and no/mild sleep disturbance may be related to less A beta burden. In participants with A beta deposition, falling asleep at 22:00 to 23:00, higher sleep efficiency (at least >= 75%), no/mild daytime dysfunction, sleep latency <= 30 min, and good sleep quality may help improve AD pathology.