White matter tract integrity, but not amyloid burden, is related to cognition in cognitively normal older adults

Background The increased focus on preclinical Alzheimer’s disease (AD) necessitates an understanding of how general biomarkers of neuronal injury and AD‐specific biomarkers differentially contribute to the cognitive performance in cognitively normal individuals. It has been suggested that alterations in limbic white matter (WM) tracts may precede both gray matter atrophy and clinical impairment (Teipel et al., 2016). However, few studies have evaluated whether limbic tract alterations are associated with cognitive performance in individuals with normal cognition. To address this gap, this cross‐sectional, multi‐modal imaging study examined whether measures of WM tract integrity, using diffusion tensor imaging (DTI) (including fornix, hippocampal cingulum, and uncinate fasciculus), were related to cognitive performance, independent of cortical amyloid burden in cognitively normal individuals. Method Analyses included 130 cognitively normal participants from the BIOCARD Study (Table 1) who completed cognitive assessments, diffusion‐weighted MR and 11C‐PiB PET imaging during their annual visit. Cognitive performance was captured by cognitive composite scores reflecting verbal episodic memory, executive function, and visuospatial processing. WM tract integrity was measured by fractional anisotropy (FA) and radial diffusivity (RD) in the 3 above‐mentioned ROIs. PiB PET amyloid positivity was derived from cortical distribution volume ratios (cDVR) using cerebellar gray matter as the reference. Linear regressions (covarying age, sex, and education) were used to test 4 a priori cognition‐WM associations. The following relationships were examined: episodic memory and fornix and hippocampal cingulum integrity; executive function and uncinate fasciculus integrity; visuospatial processing and hippocampal cingulum integrity. Result In the linear regression models (Table 2), lower hippocampal cingulum FA and greater hippocampal cingulum RD were each associated with better episodic memory performance. Additionally, lower uncinate fasciculus RD (but not FA) was associated with better executive function performance. In contrast, amyloid positivity (i.e., cDVR > 1.06) was unrelated to any of the cognitive composite scores (all p s > .35). Conclusion Measures of WM tract integrity account for variability in cognitive performance among cognitively normal individuals, and this association is independent of amyloid burden. Additional analyses are needed to determine whether these microstructural measures reflect general, age‐related markers of neuronal injury vs. alterations related to preclinical AD.