Characterizing White Matter Microstructure In Asymptomatic Older Adults At Elevated Risk For Alzheimer's Disease

Abstract

Growing evidence suggests that Alzheimer’s disease (AD) is associated with axonal tract alterations. These white matter (WM) changes may emerge very early, prior to clinical symptom onset and may precede cortical grey matter changes. However, reliably characterizing these WM alterations in vivo and differentiating them from normal age-related change has been challenging. To address this challenge, the overarching goal of this dissertation was to examine differences in WM microstructure attributable to known AD-risk factors: age, genetics, and the presence of AD-related pathology. Advanced diffusion-imaging methods were used to characterize WM microstructure in a large sample of older adults at elevated familial risk for AD who remained clinically asymptomatic (n=146). Additionally, participants underwent genetic testing, lumbar punctures, and positron emission tomography (PET) scanning to derive AD-risk biomarkers. In Study 1, I implemented a multivariate, data-driven statistical technique, Partial Least Squares (PLS), to identify covariance patterns between whole-brain, voxelwise white matter microstructure and AD-risk factors. Neurite Orientation Dispersion and Density Imaging (NODDI) data were collected to derive three WM microstructure indices: neurite density (NDI), orientation dispersion (ODI), and isotropic volume fraction (ISOVF). Each of these measures was associated with age, APOE4 genotype, and amyloid-beta and tau pathology biomarkers. Older age was associated with all three NODDI WM indices. NDI was uniquely sensitive to AD-risk indexed by AD pathology biomarkers. Study 2 extended these analyses to examine WM microstructural associations with cognition (episodic memory, processing speed, and executive control) in the same preclinical AD cohort using a whole-brain exploratory approach. WM microstructure, indexed by NODDI, was associated with episodic memory and executive control. However, most associations did not remain when accounting for age-related variance, suggesting that WM-cognition associations may not be specific to AD-risk factors. This dissertation represents one of the first, and among the most comprehensive investigations into WM microstructure in the context of multiple AD-risk factors, occurring before clinical syndrome onset. These findings demonstrate WM microstructural alterations are among the earliest neural changes to accompany AD-related pathology, providing a window into the impact of AD on brain structure, and informing novel opportunities for surveillance and intervention at the earliest disease stages.