Associations of peak width of skeletonized mean diffusivity with cardiovascular disease risk and cognitive decline in clinically normal older adults: Neuroimaging / New imaging methods. (7th December 2020)

• Record Type: Journal Article
• Authors: Scott, Matthew R.; Schultz, Aaron P.; Buckley, Rachel F.; Chou, Hsiang‐Chin Lori; Hampton, Olivia L; Hanseeuw, Bernard J; Jacobs, Heidi I.L.; Properzi, Michael J.; Rabin, Jennifer S.; Johnson, Keith A.; Sperling, Reisa A.; Chhatwal, Jasmeer P.
• Is Part Of: Alzheimer's & dementia. Volume 16(2020)Supplement 4
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-12-07
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.043812
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Cardiovascular disease (CVD) is associated with cognitive decline, both alone and synergistically with Aβ burden, and can have a stronger relationship with cognitive trajectories than many standard imaging biomarkers. However, the effects of CVD and Aβ on cognitive decline have not been examined alongside diffusion markers of small vessel disease (SVD). With data from the Harvard Aging Brain Study (HABS), we examined the influence of CVD risk and peak width of skeletonized mean diffusivity (PSMD; a DTI measure of SVD) on cognitive trajectories in clinically normal participants, while accounting for Aβ burden. Method: 264 clinically normal participants from HABS (mean [SD]: age=73.2[7.1] years; follow‐up range=6.3[1.8] years) underwent at least four longitudinal assessments of the Preclinical Alzheimer's Cognitive Composite (PACC96), as well as PiB‐PET imaging and DTI at baseline. PSMD was calculated from skeletonized mean diffusivity maps using FSL. CVD risk was quantified using the Framingham Heart Study‐CVD risk score. We used linear mixed‐effect models to examine the influence of baseline PSMD and CVD on longitudinal PACC, while controlling for, or interacting with, PiB DVR, correcting for baseline age, years of education, and biological sex. Result: PSMD and CVD were significantly correlated ( r =0.38, p< 1.0x10 ‐9 ), though this weakened when correcting for age ( r =0.15, p =0.01), while PSMD and PiB showed no association ( r =0.09, p =0.13). Higher PSMD was associated with faster cognitive decline ( t (1808)=‐2.40, p =0.02) and this effect remained ( t (1806)=‐2.26, p =0.02) after correcting for PiB. The effect of PSMD was marginally significant ( t (1804)=‐1.93, p =0.05) when correcting for both CVD ( t (1804)=‐4.44, p <1.0x10 ‐5 ) and PiB ( t (1804)=‐7.79, p <1.0x10 ‐13 ). No significant interactive effect of PSMD and PiB on cognitive decline was observed. Conclusion: CVD risk and PSMD were significantly correlated, though this was largely driven by age, while PiB and PSMD were not. PSMD was significantly associated with faster cognitive decline. After adjusting for both CVD and PiB however, the effect of PSMD on cognitive decline was attenuated. It is possible that CVD risk may capture variance related to cognitive decline beyond that of PSMD. However, future research is needed to understand which domains of the CVD risk score are driving these findings.