DOI: 10.1002/alz.075860 ISSN: 1552-5260

Reduced coupling between cerebrospinal fluid flow and global brain activity is linked to tau pathology

Feng Han, JiaQie Lee, Xi Chen, Jacob Ziontz, Susan M. Landau, Suzanne L. Baker, Theresa M. Harrison, William J. Jagust
  • Psychiatry and Mental health
  • Cellular and Molecular Neuroscience
  • Geriatrics and Gerontology
  • Neurology (clinical)
  • Developmental Neuroscience
  • Health Policy
  • Epidemiology

Abstract

Background

The glymphatic system clears brain waste, including β‐amyloid (Aβ) and tau, through cerebrospinal fluid (CSF) movement. CSF inflow is coupled to global brain activity measured by global resting state fMRI (gBOLD). Weaker coupling has been proposed to reflect impaired glymphatic function and found associated with cognitive decline and Aβ accumulation in early Alzheimer’s disease (AD). Few studies have explored the role of glymphatic activity in tau pathology in the human brain, which we investigated by assessing glymphatic function with gBOLD‐CSF coupling.

Method

We identified 121 ADNI‐3 participants (72.4 ± 8.0 years; 65 females; 50 cognitively impaired, including 6 AD dementia and 44 mild cognitive impairment) who underwent both resting‐state fMRI (rsfMRI; TR = 0.607s sessions), tau PET (flortaucipir [FTP]), and amyloid PET (florbetapir or florbetaben). These participants were classified into different sub‐groups based on amyloid deposition (Aβ+: florbetapir > 1.11 SUVR; florbetaben > 1.08 SUVR) or diagnosis (cognitively impaired or unimpaired). CSF signal was extracted at the bottom rsfMRI slice, and gBOLD signal was from gray matter. Correlation analysis between gBOLD‐CSF coupling and tau load was performed on the whole cohort and the sub‐groups, respectively. Predefined regions of interest (ROIs) included Braak ROIs and a meta‐temporal ROI.

Result

CSF inflow signal was coupled with global brain fMRI. Each participant’s coupling was quantified by the gBOLD‐CSF cross‐correlation at the negative peak (Fig. 1). We found significant correlations between the coupling metric and tau load in Braak III‐IV, V‐VI, and the meta‐temporal ROI (but not in Braak I [entorhinal cortex]) across all participants and the Aβ+ subgroup. In these regions, participants with weaker gBOLD‐CSF coupling (less negative; presumably impaired glymphatic clearance) have more tau deposition (Fig. 2). Similar patterns were seen in the relatively small (N = 30) Aβ+ impaired subgroup (Fig. 3).

Conclusion

Coupling between gBOLD and CSF signals, which could be a proxy for glymphatic function, is associated with tau deposition across the neocortex, particularly in Braak V‐VI ROIs. Furthermore, this effect is particularly evident in participants with elevated amyloid or cognitive impairment. These findings suggest that gBOLD‐CSF coupling is important for tau clearance, likely through its association with the glymphatic system.

More from our Archive