Lower cortical thickness is associated with reduced gray matter cerebral blood flow across the AD continuum: Neuroimaging / Optimal neuroimaging measures for early detection

Background Cortical thinning is an early marker of Alzheimer’s disease (AD) neurodegenerative changes. We examined relationships of cortical thickness and cerebral blood flow (CBF) to assess whether structural brain changes in AD‐related regions are associated with functional changes in blood flow dynamics. Previous reports of CBF indicate reduced blood flow progressing along the AD continuum from normal cognition (NC) to mild cognitive impairment (MCI) to AD. We hypothesized that reduced cortical thickness in AD‐specific brain regions would relate to reduced global gray matter (GM) CBF. Method Participants enrolled in the Wake Forest Alzheimer’s Disease Research Center (ADRC) Clinical Core cohort (N=373, mean age=70.5 years) underwent neuropsychological assessment, and T1 and pcASL MRI to assess brain structure and CBF. Participants were adjudicated as cognitively unimpaired (NC, N=212) or impaired (MCI or AD, N=161) using NIA‐Alzheimer’s Association criteria. T1 MRI were processed using FreeSurfer v5.3; an “AD‐signature” meta‐ROI cortical thickness value was calculated, which was thresholded to classify participants as neurodegeneration‐negative (N‐, above threshold) or ‐positive (N+, below threshold). PcASL MRI were processed to yield CBF images that were normalized to MNI space (Fig.1). A 2‐tissue partial volume correction was applied to yield voxelwise GM CBF. Mean GM CBF across a set of cortical AAL ROIs was calculated for each participant. A MANCOVA model adjusted for age, sex and education measured differences in GM CBF between N‐ and N+ groups. We additionally assessed cognitive status as a moderator of these associations. Result We found a significant difference in GM CBF between N‐ and N+ groups (p=0.010). Participants with lower meta‐ROI cortical thickness had reduced GM CBF compared to participants above the threshold (Fig.2). The interaction of cognitive status and N+/‐ group revealed lower CBF for impaired participants (Fig.3, unadjusted p<0.001), which remained significant after adjustment (p=0.014). Conclusion In our ADRC cohort, neurodegenerative changes within brain regions vulnerable to AD pathology were associated with global GM CBF reductions. Cognitively unimpaired N+ participants exhibited similar levels of GM CBF as impaired N‐ participants. Future studies will assess whether unimpaired N+ participants with reduced CBF have a higher risk of conversion to MCI or AD.