Background: Cerebral small vessel disease (SVD) contributes to stroke and dementia, yet venous involvement in its pathophysiology remains unclear. Resting-state fMRI captures spontaneous low-frequency oscillations (sLFOs), reflecting cerebral perfusion dynamics primarily within the venous system. This study applied sLFO lag-time mapping to examine venous perfusion in SVD. Methods: We analyzed 572 community-dwelling adults (age 62.2±8.5 years, 61.2% women) from the ILAS cohort with high-resolution 3T MRI imaging. Participants were stratified into controls and SVD subtypes based on MRI markers: Type 1 (pure WMH), Type 2 (WMH + lacune), Type 3 (mixed CMBs), and Type 4 (strictly lobar CMBs). Types 1 and 2 were combined as ischemic SVD. Three perfusion indices were derived: mean positive lag time, mean negative lag time, and voxel-wise lag-time distribution. Group differences were assessed using voxel-wise ANCOVA with post-hoc analyses and hierarchical clustering, controlling for age, sex, hypertension, and motion parameters. Results: SVD Type 1+2 showed significantly prolonged mean positive lag times compared to controls, indicating delayed global superficial venous drainage. Voxel-wise analyses revealed region-specific alterations: in the superficial venous system, prolonged lag was found in the occipital cortex in Type 1+2 and in the occipital pole and medial frontal cortex in Type 4, while frontal motor regions showed shortened lag in all SVD types. In the deep venous system, only Type 4 showed prolonged lag, in voxels located in the cerebellum, superior frontal gyrus, and parietal lobule, while Types 1+2 and 3 exhibited regional lag shortening in temporo-occipital and peri-hippocampal areas. Type 4 exhibited the most prolonged venous lag in both the superficial and deep venous systems. Hierarchical clustering further distinguished Type 4 from other subtypes in both systems, suggesting a distinct venous perfusion dysfunction profile potentially associated with CAA. Conclusion: This study demonstrates that rs-fMRI sLFO lag-time mapping can sensitively detect early venous perfusion alterations in covert SVD. Ischemic SVD showed delayed global superficial drainage, while a presumed CAA-related subtype exhibited spatially and temporally distinct perfusion features in both venous systems. These findings highlight venous dysfunction as an underrecognized feature of SVD and support lag-time mapping as a non-invasive biomarker for early detection and subtype differentiation.
Chung et al. (Thu,) studied this question.