Topography and associations of perivascular spaces in healthy adults: the Kashima scan study.

OBJECTIVE: We investigated whether the topography of MRI-visible perivascular spaces (PVS) is associated with markers of specific underlying small vessel disease, including cerebral microbleed (CMB) distribution, in neurologically healthy adults.

METHODS: We analyzed baseline data of an ongoing Japanese population-based cohort study. PVS were rated in the basal ganglia (BG-PVS) and centrum semiovale (CSO-PVS) on axial T2-weighted MRI using a validated rating scale (score 0-4). BG-PVS degree was classified as low (score <2) or high (score ≥2). CSO-PVS degree was classified as low (score <3) or high (score ≥3). Independent demographic, clinical, and imaging factors for high degree of BG-PVS and CSO-PVS were investigated.

RESULTS: A total of 1,575 neurologically healthy adults were included (mean age 57.1 years, SD 9.7; 47% male). In multivariable analyses, high degree of BG-PVS (n = 212, 14%) was associated with deep or infratentorial CMBs (odds ratio [OR] 2.77, 95% confidence interval [CI] 1.62-4.74), a marker of hypertensive arteriopathy; by contrast, high degree of CSO-PVS (n = 357, 23%) was associated with strictly lobar CMBs (OR 2.49, 95% CI 1.35-4.61), which share risk factors with cerebral amyloid angiopathy. Both high degree of BG-PVS and CSO-PVS were associated with hypertension (OR 2.03, 95% CI 1.46-2.82 and OR 1.39, 95% CI 1.07-1.81, respectively), lacunes (OR 3.35, 95% CI 1.92-5.86; OR 1.83 95% CI 1.08-3.08), and severe white matter hyperintensities (OR 2.17, 95% CI 1.42-3.31; OR 1.35, 95% CI 0.93-1.96), but these associations were stronger for high degree of BG-PVS.

CONCLUSIONS: In a neurologically healthy cohort, the associations of PVS differ according to their topography. PVS distribution may be useful for the early detection and classification of small vessel disease.