Association of Baseline Cerebrovascular Reactivity and Longitudinal Development of Enlarged Perivascular Spaces in the Basal Ganglia

BACKGROUND: Increasing evidence suggests that enlarged perivascular spaces (ePVS) are associated with cognitive dysfunction in aging. However, the pathogenesis of ePVS remains unknown. Here, we tested the possibility that baseline cerebrovascular dysfunction, as measured by a magnetic resonance imaging measure of cerebrovascular reactivity, contributes to the later development of ePVS. METHODS: Fifty cognitively unimpaired, older adults (31 women; age range, 60–84 years) underwent magnetic resonance imaging scanning at baseline and follow-up separated by ≈2.5 years. ePVS were counted in the basal ganglia, centrum semiovale, midbrain, and hippocampus. Cerebrovascular reactivity, an index of the vasodilatory capacity of cerebral small vessels, was assessed using carbon dioxide inhalation while acquiring blood oxygen level-dependent magnetic resonance images. RESULTS: Low baseline cerebrovascular reactivity values in the basal ganglia were associated with increased follow-up ePVS counts in the basal ganglia after controlling for age, sex, and baseline ePVS values (estimate [SE]=−3.18 [0.96]; P =0.002; [95% CI, −5.11 to −1.24]). This effect remained significant after accounting for self-reported risk factors of cerebral small vessel disease (estimate [SE]=−3.10 [1.00]; P =0.003; [CI, −5.11 to −1.09]) and neuroimaging markers of cerebral small vessel disease (estimate [SE]=−2.72 [0.99]; P =0.009; [CI, −4.71 to −0.73]). CONCLUSIONS: Our results demonstrate that low baseline cerebrovascular reactivity is a risk factor for later development of ePVS.