Arterial Spin Labeling Hemodynamic Indices Relate to Cognitive and Small Vessel Disease Trajectories Over 11 Years

BACKGROUND: Hemodynamic insufficiency may contribute to the development of neuroimaging signs of cerebral small vessel disease (SVD). The arterial spin labeling (ASL) spatial coefficient of variation (ASL-sCoV) is a proxy marker of arterial transit time and may better capture SVD-related hemodynamic disturbances compared with ASL-derived cerebral blood flow. We investigated ASL-derived hemodynamic indices in relation to cognition and neuroimaging markers of SVD over an 11-year follow-up period. METHODS: Participants without dementia or stroke were enrolled in the Vanderbilt Memory and Aging Project, a longitudinal observational cohort study in Nashville, TN. Participants underwent serial multimodal 3T brain magnetic resonance imaging to quantify SVD burden and neuropsychological assessment from 2012 to 2024 (4.9±3.5 years mean follow-up). Pseudo-continuous ASL assessed cerebral blood flow and ASL-sCoV in total gray matter. Baseline ASL-sCoV and cerebral blood flow were individually related to the cross-sectional burden and longitudinal trajectory of each SVD neuroimaging marker (white matter hyperintensities, enlarged perivascular spaces, cerebral microbleeds, and lacunar infarcts) and cognitive performances using linear and linear mixed-effects regression models. Models were adjusted for demographics, cognitive status, Framingham Stroke Risk Profile (minus age), apolipoprotein E -ε4 status, intracranial volume (for SVD outcomes), and follow-up time (for longitudinal models). RESULTS: Among participants (n=667, 68±9 years, 18% mild cognitive impairment, 51% female), higher ASL-sCoV was cross-sectionally associated with SVD markers, including white matter hyperintensities and enlarged perivascular spaces ( p FDR -values<0.006). Higher baseline ASL-sCoV was associated with a faster longitudinal increase in the burden of white matter hyperintensities ( p FDR =0.03) and accelerated decline in executive function, information processing, language, and visuospatial performances ( P <0.05). Gray matter cerebral blood flow was not associated with SVD burden or cognition cross-sectionally ( p FDR -values>0.05) or longitudinally ( p FDR -values>0.06). CONCLUSIONS: ASL-sCoV may serve as an important hemodynamic measure for identifying perfusion deficits tied to SVD burden and progression. Readily derived from existing ASL data sets, ASL-sCoV could provide substantial insight into the longitudinal clinical consequences of SVD.