Free-water metrics in medial temporal lobe white matter tract projections relate to longitudinal cognitive decline

Abstract

<jats:title>Abstract</jats:title><jats:sec><jats:title>Objective</jats:title><jats:p>Hippocampal volume is a sensitive marker of neurodegeneration and a well-established predictor of age-related cognitive impairment. Recently, free-water (FW) magnetic resonance imaging (MRI) has shown associations with pathology in Alzheimer’s disease (AD), but it is still unclear whether these metrics are associated with measures of cognitive impairment. Here, we investigate whether FW and FW-corrected fractional anisotropy (FA<jats:sub>T</jats:sub>) within medial temporal lobe white matter tracts (cingulum, fornix, uncinate fasciculus, inferior longitudinal fasciculus, and tapetum) provides meaningful contribution to cognition and cognitive decline beyond hippocampal volume.</jats:p></jats:sec><jats:sec><jats:title>Participants and Methods</jats:title><jats:p>Vanderbilt Memory &amp; Aging Project participants (n=319, 73±7 years, 59% male) with normal cognition and mild cognitive impairment (40% of cohort) underwent baseline brain MRI, including structural MRI to quantify hippocampal volume, diffusion MRI to quantify medial temporal lobe white matter tract FW and FA<jats:sub>T</jats:sub>, and longitudinal neuropsychological assessment with a mean follow-up of 3.5 years. Linear regressions were conducted to determine how hippocampal volume and white matter tract FW and FA<jats:sub>T</jats:sub> interact with baseline memory and executive function performances. Competitive model analyses determined the unique variance provided by white matter tract FW and FA<jats:sub>T</jats:sub> beyond that of hippocampal volume and other comorbidities. Linear mixed-effects models were conducted to determine how baseline hippocampal volume and white matter tract FW and FA<jats:sub>T</jats:sub> interact to explain longitudinal change in memory and executive function performances.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>FW in the inferior longitudinal fasciculus, tapetum, uncinate fasciculus, and cingulum were robustly associated with baseline memory and executive function. Further, competitive model analysis showed that tract FW contributed unique variance beyond other comorbidities and hippocampal volume for memory (ΔR<jats:sub>adj</jats:sub><jats:sup>2</jats:sup> range: 0.82-2.00%) and executive function (ΔR<jats:sub>adj</jats:sub><jats:sup>2</jats:sup> range: 0.88-1.87%). Longitudinal analyses demonstrated significant interactions of hippocampal volume and FA<jats:sub>T</jats:sub> in the inferior longitudinal fasciculus (p=0.02), tapetum (p=0.02), uncinate fasciculus (p=0.02), and cingulum (p=0.002) with decline in memory. For decline in executive function, we found significant interactions of hippocampal volume and FA<jats:sub>T</jats:sub> in inferior longitudinal fasciculus (p=0.03), tapetum (p=0.02), uncinate fasciculus (p=0.02), and fornix (p=0.02), as well as cingulum (p=0.02) and fornix (p=0.02) FW.</jats:p></jats:sec><jats:sec><jats:title>Conclusions</jats:title><jats:p>Our results highlight novel associations between FW and FA<jats:sub>T</jats:sub> measures of medial temporal lobe tract microstructure and cognitive performance such that individuals with smaller hippocampal volumes and lower tract microstructure experience greater cognitive decline. These results suggest that white matter has a unique role in cognitive decline and, therefore, could be used to provide better disease staging, allowing for more precise disease monitoring in AD.</jats:p></jats:sec>