Abstract
INTRODUCTION
Sleep disturbances are linked to Alzheimer’s disease (AD) and related dementias (ADRD) pathology and cognitive decline, but how sleep architecture, biomarkers, and cognition are organized within a conditional dependence framework remains unclear.
METHODS
We studied 98 adults spanning normal cognition (NC; n = 21), mild cognitive impairment (MCI; n = 34), and clinically diagnosed AD dementia (n = 43) using overnight polysomnography, plasma phosphorylated tau181 (p-tau181), neurofilament light chain (NfL), brain-derived neurotrophic factor (BDNF), amyloid-β positron emission tomography (PET), and multidomain cognitive testing. A regularized Gaussian graphical model was used to characterize conditional associations and identify cross-domain connectivity among sleep, biomarkers, and cognition.
RESULTS
The 17-variable network contained 26 edges, with significantly denser within-domain than cross-domain connectivity (40.5% vs 9.6%; p < 0.001). Direct sleep-cognition connections were sparse. Rapid eye movement (REM) sleep latency had the highest bridge strength (0.387), directly connecting the sleep domain with NfL, amyloid-β PET, and BDNF. Plasma p-tau181 was the only biomarker directly connected to global cognition. Exploratory mediation analyses were consistent with indirect associations involving p-tau181 in three of five cognitive domains.
DISCUSSION
Sleep architecture and cognition were conditionally linked primarily through biomarker nodes. REM sleep latency and plasma p-tau181 may represent key cross-domain connectors linking sleep architecture, neurodegeneration, and cognition across the AD continuum.


If you do not see content above, kindly GO TO SOURCE.
Not all publishers encode content in a way that enables republishing at Neuro.vip.

This post is Copyright: | November 18, 2026
Neuro-Dementia