Regional growth rates of white matter hyperintensities are associated with beta-amyloid burden

There is increasing evidence for an association between white matter hyperintensities (WMH) and brain beta-amyloid deposition. How WMH are longitudinally associated with brain beta-amyloid burden requires further investigation, particularly with respect to co-existent vascular risk factors and differences across white matter regions. We measured WMH on MRI and vascular risk factors in a combined neuroimaging data set of cognitively normal and individuals with dementia comprised of the ADNI, AIBL and OASIS3 studies, which includes harmonized centiloid estimates of beta-amyloid burden from PET imaging. WMH were measured using the TrUE-Net algorithm. Vascular risk factors were extracted from provided clinical data and used to calculate individual revised Framingham Stroke Risk Profile (FSRP) scores. Linear mixed effects modelling was used to determine the relationship between the growth rate of WMH and baseline beta-amyloid burden, controlling for age, sex, APOE4 status, and vascular risk factors. 1243 participants [49 % female, mean age 71.7 y (SD 7.6 y)] had at least 3 brain MRIs. Linear mixed models demonstrate robust independent cross-sectional relationships between WMH and baseline beta-amyloid burden (beta coefficient=0.27, p < 0.001), age (beta coefficient=0.04, p < 0.001) and vascular risk factors (beta coefficient=0.25, p < 0.001). Growth rates of WMH increased with baseline beta-amyloid burden (slope=0.021, p < 0.001) and decreased with anti-hypertensive medications (slope=-0.019, p = 0.002), above and beyond age, APOE4 status, and other vascular risk factors. The longitudinal association for beta-amyloid burden persisted in a similar analysis for parietal WM. Our study suggests that in Alzheimer disease research cohorts, WMH progression is associated with age and beta-amyloid burden, particularly in parietal white matter, and slowed by anti-hypertensive treatment.