The association between white matter hyperintensities and amyloid and tau deposition

• Published in: NeuroImage clinical 2023-01, Vol.38, p.103383-103383, Article 103383
• Main Authors: Alban, Sierra L., Lynch, Kirsten M., Ringman, John M., Toga, Arthur W., Chui, Helena C., Sepehrband, Farshid, Choupan, Jeiran
• Format: Article
• Language: English
• Subjects: Alzheimer Disease - pathology; Amyloid; Amyloid beta-Peptides - metabolism; Amyloidogenic Proteins; Cognitive Dysfunction - pathology; Female; Humans; Male; Regular; tau Proteins - metabolism; White Matter - pathology
• ISSN: 2213-1582; 2213-1582
• DOI: 10.1016/j.nicl.2023.103383

•Significant relation between β-amyloid and white matter hyperintensities was found.•Age and sex differences were present across various cardiovascular risk factors.•Cardiovascular risk factors were not predictive of white matter hyperintensities.•White matter hyperintensities were not correlated with Tau deposition.•White matter hyperintensities and β-amyloid colocalize in the neocortex. White matter hyperintensities (WMHs) frequently occur in Alzheimer’s Disease (AD) and have a contribution from ischemia, though their relationship with β-amyloid and cardiovascular risk factors (CVRFs) is not completely understood. We used AT classification to categorize individuals based on their β-amyloid and tau pathologies, then assessed the effects of β-amyloid and tau on WMH volume and number. We then determined regions in which β-amyloid and WMH accumulation were related. Last, we analyzed the effects of various CVRFs on WMHs. As secondary analyses, we observed effects of age and sex differences, atrophy, cognitive scores, and APOE genotype. PET, MRI, FLAIR, demographic, and cardiovascular health data was collected from the Alzheimer’s Disease Neuroimaging Initiative (ADNI-3) (N = 287, 48 % male). Participants were categorized as A + and T + if their Florbetapir SUVR and Flortaucipir SUVR were above 0.79 and 1.25, respectively. WMHs were mapped on MRI using a deep convolutional neural network (Sepehrband et al., 2020). CVRF scores were based on history of hypertension, systolic and diastolic blood pressure, pulse rate, respiration rate, BMI, and a cumulative score with 6 being the maximum score. Regression models and Pearson correlations were used to test associations and correlations between variables, respectively, with age, sex, years of education, and scanner manufacturer as covariates of no interest. WMH volume percent was significantly associated with global β-amyloid (r = 0.28, p