Appropriate regulation of cerebral blood flow (CBF) is critical for optimal neurological functioning. Alterations in CBF, as measured by perfusion MRI, occur across a wide range of diseases including Alzheimer’s disease (AD). The emergence of blood-based biomarkers has allowed for non-invasive assessment of AD-related pathophysiology, axonal injury, and astrocytic dysfunction—all factors that impact CBF. Dysregulation of cerebral perfusion co-occurs with or may even precede abnormality of other AD pathophysiologic markers (e.g., tau and β-amyloid), but links between cerebral perfusion and biofluid markers of AD and neuronal/astroglial health are poorly understood. In a cohort of older adults, we examined associations between plasma AD biomarkers and CBF in limbic regions where pathological proteins are known to accumulate early in disease processes.

122 older adults (mean age=75.4 [σ=7.2]; 81% cognitively normal [CN]; 19% mild cognitive impairment [MCI]; 52% women; 90% white; 31% APOE-ε4 carriers) were included from the UCSF Brain Aging Network for Cognitive Health (BRANCH) who completed a pseudo-continuous arterial spin labeling (pCASL)-MRI scan and blood draw within one year of their MRI. Partial volume corrected gray matter CBF values from pCASL-MRI were quantified in a composite medial temporal lobe (MTL) region comprising the entorhinal cortex, hippocampus, and parahippocampal gyrus, as well as in each subregion individually. Plasma samples were assayed for phosphorylated-tau (pTau)181, Aβ42/40 ratio, glial fibrillary acidic protein (GFAP), and neurofilament light (NfL) chain (Quanterix Simoa). Multiple linear regression models examined regional CBF as a function of plasma biomarkers, covarying for age, sex, and APOE-ε4 genotype. Associations were studied across the entire cohort, followed by regression models testing the moderating effect of diagnosis (CN vs. MCI) on associations between CBF and plasma biomarkers. Finally, we conducted sensitivity analyses in the CN cohort to test robustness of relationships in cognitively unimpaired individuals.

In the entire cohort, higher plasma GFAP related to lower MTL CBF (β=-0.22, p=.027), with subregional analyses revealing a specific relationship with entorhinal CBF (β=-0.28, p=.006) but not hippocampal (β=-0.17, p=.089) or parahippocampal (β=-0.17, p=.094) CBF. Clinical severity did not moderate observed relationships between plasma GFAP and MTL (β=-0.07, p=.837) or entorhinal (β=-0.26, p=.467) CBF. Sensitivity analyses in CN participants revealed a similar negative association, with higher plasma GFAP relating to lower entorhinal perfusion (β=-0.28, p=.016) and approaching significance for the entire MTL (β=-0.22, p=.052), but not with hippocampal (β=-0.16, p=.165) or parahippocampal (β=-0.18, p=.114) CBF. The remaining plasma biomarkers (pTau181, Aβ42/40 ratio, NfL) did not relate to perfusion across the entire cohort or in CN participants.

In older adults with mild or no cognitive symptoms, we found specific links between entorhinal perfusion and astrocyte function but not with markers of axonal injury, tau, or β-amyloid. CBF regulation occurs largely at the level of astrocytic endfeet, and astrocytic integrity as measured by plasma GFAP may be a key marker associated with subtle early cerebrovascular alterations in regions vulnerable to AD. Future studies combining multimodal non-invasive techniques to simultaneously characterize cerebrovascular and proteinopathic changes in aging may assist with earlier detection of neurodegenerative disease.