Associative Learning and Regional White Matter Deficits in Mild Cognitive Impairment

Background: While diagnostic criteria for Alzheimer's disease (AD) include neuroimaging biomarkers, there remains no definitive biomarker of mild cognitive impairment (MCI). MCI is a risk factor for AD that may be amenable to early intervention. Early decline in white matter (WM) integrity identified by diffusion tensor imaging (DTI) is a predictor of future progression of neurodegeneration. Objective: Identify regionally specific WM differences between individuals with MCI and those with age-associated memory impairment (AAMI) and relationships with specific memory decrements. Methods: DTI and neuropsychological data were acquired from 38 participants (23 MCI and 15 AAMI). A region of interest approach was used to evaluate regional differences between groups and correlative relationships with performance on memory tasks. Results: Fornix WM had higher mean (MD), radial (DR), and axial (DA) diffusivity in MCI participants relative to AAMI. Temporal stem (TS) WM had higher MD and DR in MCI than in AAMI. In MCI, TS MD and DR varied, while fornix MD and DR was uniformly high, and in AAMI, TS MD and DR were uniformly low and fornix MD and DR varied. In MCI, TS MD and DA were inversely associated with associative learning but not list learning. Conclusions: In addition to supporting prior evidence implicating the fornix in early AD pathology, these data implicate a profile of neurodegeneration associated with early MCI. Further, they suggest that associative learning tasks are more sensitive to early neurodegeneration and may be useful in identifying individuals at risk for AD.