Age Ageing. 2025 Nov 28;54(12):afaf353. doi: 10.1093/ageing/afaf353.
ABSTRACT
BACKGROUND: Although cerebrovascular reactivity (CVR) correlates with cognitive performance in neurodegenerative conditions, the age-related spatial patterns of CVR alterations and their relationships with grey matter (GM) atrophy and cognition are underexplored.
METHODS: In this cross-sectional study, 301 cognitively unimpaired participants (181 younger, 18-34 years; 120 older: 60-89 years) underwent multi-echo resting-state functional magnetic resonance imaging (fMRI) for CVR measurement. Voxel-wise t-tests compared regional CVR between age groups, with significant clusters defined as regions of interest (ROIs). Mediation analyses examined regional GM atrophy as a mediator of the ageing-CVR relationships within ROIs. Multivariable linear regression and restricted cubic spline analyses evaluated the association between ROI-CVR and cognition in older adults.
RESULTS: Compared with younger adults, older adults showed lower CVR primarily in the temporal, basal ganglia, cingulate, brainstem and cerebellum regions, while higher CVR in the frontal, parietal, occipitotemporal, thalamus and caudate regions. Regional GM atrophy partially mediated age-related CVR increases in the right frontal pole (P = .004) and fusiform/lingual gyrus (P = .001), as well as age-related CVR reduction in bilateral brainstem/cerebellum vermis 45 (P < .001). The proportions mediated were 55.9%, 56.6% and 79.2%, respectively. Among older adults, six ROI-CVRs were associated with executive function, exhibiting linear or nonlinear relationships.
CONCLUSIONS: Resting-state CVR demonstrated regionally heterogeneous age-related decreases or increases, partly mediated by GM atrophy. In older adults, CVR in age-sensitive regions was selectively associated with executive function through linear and nonlinear patterns. Cerebrovascular ageing may involve region-specific vascular adaptations and macrostructural-microvascular (GM-CVR) interactions. Region- and range-dependent CVR could serve as a biomarker for executive function changes.
PMID:41370625 | DOI:10.1093/ageing/afaf353