The association between resting‐state functional magnetic resonance imaging and aortic pulse‐wave velocity in healthy adults. Issue 8 (7th February 2020)

• Record Type: Journal Article
• Authors: Hussein, Ahmad; Matthews, Jacob L.; Syme, Catriona; Macgowan, Christopher; MacIntosh, Bradley J.; Shirzadi, Zahra; Pausova, Zdenka; Paus, Tomáš; Chen, J. Jean
• Is Part Of: Human brain mapping. Volume 41:Issue 8(2020)
• Page Start: 2121
• Page End: 2135
• Publication Date: 2020-02-07
• Subjects: aging -- aortic stiffness -- arterial stiffening -- blood pressure -- blood‐oxygenation level dependent signal (BOLD) -- BOLD signal amplitude -- functional connectivity -- functional MRI (fMRI) -- phase‐contrast imaging -- pulse wave velocity -- resting‐state fMRI; Brain mapping -- Periodicals; 611.81
• DOI: 10.1002/hbm.24934
• Languages: English
• ISSNs: 1065-9471

Abstract

Abstract: Resting‐state functional magnetic resonance imaging (rs‐fMRI) is frequently used to study brain function; but, it is unclear whether BOLD‐signal fluctuation amplitude and functional connectivity are associated with vascular factors, and how vascular‐health factors are reflected in rs‐fMRI metrics in the healthy population. As arterial stiffening is a known age‐related cardiovascular risk factor, we investigated the associations between aortic stiffening (as measured using pulse‐wave velocity [PWV]) and rs‐fMRI metrics. We used cardiac MRI to measure aortic PWV (an established indicator of whole‐body vascular stiffness), as well as dual‐echo pseudo‐continuous arterial‐spin labeling to measure BOLD and CBF dynamics simultaneously in a group of generally healthy adults. We found that: (1) higher aortic PWV is associated with lower variance in the resting‐state BOLD signal; (2) higher PWV is also associated with lower BOLD‐based resting‐state functional connectivity; (3) regions showing lower connectivity do not fully overlap with those showing lower BOLD variance with higher PWV; (4) CBF signal variance is a significant mediator of the above findings, only when averaged across regions‐of‐interest. Furthermore, we found no significant association between BOLD signal variance and systolic blood pressure, which is also a known predictor of vascular stiffness. Age‐related vascular stiffness, as measured by PWV, provides a unique scenario to demonstrate the extent of vascular bias in rs‐fMRI signal fluctuations and functional connectivity. These findings suggest that a substantial portion of age‐related rs‐fMRI differences may be driven by vascular effects rather than directly by brain function.