Sleep may be a modifiable factor associated with dementia prevention and optimal brain health. Yet, the neurobiological pathways underlying the protective effects of sleep for the human aging brain are not well established. The glymphatic system relies on circulating cerebrospinal fluid (CSF) through perivascular spaces to clear toxic proteins (e.g., beta-amyloid) from the brain parenchyma. Prior work in animals and humans has demonstrated that the dynamic movement of CSF is heightened during sleep. However, questions remain regarding the relationship of CSF flow dynamics with objectively measured sleep behavior and cognition. In the present study, we examined the role of CSF flow dynamics within the glymphatic system on sleep efficiency, sleep duration, sleep variability, and cognition.

Sixty-five healthy older adults were recruited for this study as a part of a larger parent clinical trial (mean age: 66.71+6.74 years, 63% women, mean education: 17.03+2.45 years, mean MoCA score: 26.37+2.46). CSF flow dynamics were examined using phase-contrast MRI scans and included the following measures: subarachnoid space (SS) C2-C3 flush peak (SS flush peak), SS stroke volume, SS diameter, ratio of cerebral aqueduct-to-SS stroke volume (CA-SS ratio). Sleep behavior was objectively measured via a wrist-worn GENEActiv tri-axis accelerometer worn for approximately 30 days. Behaviors of interest included sleep efficiency, defined as the percentage of time asleep within a given sleep period (i.e., time from sleep onset to wake up), and sleep duration, defined as the total time asleep within a given sleep period. Variability in sleep efficiency and duration were characterized using an intra-individual coefficient of variation (standard deviation/mean X 100). Cognitive measures included the Montreal Cognitive Assessment (MoCA) (general cognition) and the visual mnemonic similarity test (MST) (fine memory discrimination). Participants were selected for available data including MRI, sleep behavior, cognitive measures, and demographics of interest. Partial correlation models were conducted to analyze the relationships between CSF flow dynamics, sleep behavior, and cognition. All models were adjusted for age at MRI scan. Models with cognitive measures were further adjusted for years of education.

Sleep behavior. Higher SS flush peak was associated with lower variability in both sleep efficiency (r=-.29, p=.02) and sleep duration (r=-.28, p=.03) (i.e., more consistent sleep). Greater SS stroke volume was trend-level associated with better sleep efficiency (r=.25, p=.054). Bigger SS diameter was trend-level associated with lower variability in sleep duration (r=-.25, p=.056). No significant relationships were observed between CA-SS ratio and sleep behavior.

Cognition. Higher SS flush peak was associated with better performance on the MoCA (global cognition) (r=.27, p=.05). Higher CA-SS ratio was associated with better performance on the MST (fine memory discrimination) (r=.30, p=.03). No significant relationships were observed between SS stroke volume or SS diameter and cognition.

These results suggest that better CSF flow dynamics in older adults are associated with more consistent day-to-day sleep efficiency and sleep duration and better global cognition and fine memory discrimination. This lends support to the idea that CSF flow in support of glymphatic clearance may be one mechanism through which sleep supports age-associated brain health.