The relationship between white matter microstructure and self-perceived cognitive decline
Derek B. Archer,Elizabeth E. Moore,Ujwala Pamidimukkala,Niranjana Shashikumar,Kimberly R. Pechman,Kaj Blennow,Henrik Zetterberg,Bennett A. Landman,Timothy J. Hohman,Angela L. Jefferson,Katherine A. Gifford
DOI: https://doi.org/10.1016/j.nicl.2021.102794
2021-01-01
Abstract:Subjective cognitive decline (SCD) is a perceived cognitive change prior to objective cognitive deficits, and although it is associated with Alzheimer's disease (AD) pathology, it likely results from multiple underlying pathologies. We investigated the association of white matter microstructure to SCD as a sensitive and early marker of cognitive decline and quantified the contribution of white matter microstructure separate from amyloidosis. Vanderbilt Memory & Aging Project participants with diffusion MRI data and a 45-item measure of SCD were included [n=236, 137 cognitively unimpaired (CU), 99 with mild cognitive impairment (MCI), 73±7 years, 37% female]. A subset of participants (64 CU, 40 MCI) underwent a fasting lumbar puncture for quantification of cerebrospinal fluid (CSF) amyloid-β(CSF Aβ42), total tau (CSF t-tau), and phosphorylated tau (CSF p-tau). Diffusion MRI data was post-processed using the free-water (FW) elimination technique, which allowed quantification of extracellular (FW) and intracellular compartment (fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity) microstructure. Microstructural values were quantified within 11 cognitive-related white matter tracts, including medial temporal lobe, frontal transcallosal, and fronto-parietal tracts using a region of interest approach. General linear modeling related each tract to SCD scores adjusting for age, sex, race/ethnicity, education, Framingham Stroke Risk Profile scores, APOE ε4 carrier status, diagnosis, Geriatric Depression Scale scores, hippocampal volume, and total white matter volume. Competitive models were analyzed to determine if white matter microstructural values have a unique role in SCD scores separate from CSF Aβ42. FW-corrected radial diffusivity (RDT) was related to SCD scores in 8 tracts: cingulum bundle, inferior longitudinal fasciculus, as well as inferior frontal gyrus (IFG) pars opercularis, IFG orbitalis, IFG pars triangularis, tapetum, medial frontal gyrus, and middle frontal gyrus transcallosal tracts. While CSF Aβ42 was related to SCD scores in our cohort (Radj2=39.03%; β=-0.231; p=0.020), competitive models revealed that fornix and IFG pars triangularis transcallosal tract RDT contributed unique variance to SCD scores beyond CSF Aβ42 (Radj2=44.35% and Radj2=43.09%, respectively), with several other tract measures demonstrating nominal significance. All tracts which demonstrated nominal significance (in addition to covariates) were input into a backwards stepwise regression analysis. ILF RDT, fornix RDT, and UF FW were best associated with SCD scores (Radj2=46.69%; p=6.37x10-12). Ultimately, we found that medial temporal lobe and frontal transcallosal tract microstructure is an important driver of SCD scores independent of early amyloid deposition. Our results highlight the potential importance of abnormal white matter diffusivity as an early contributor to cognitive decline. These results also highlight the value of incorporating multiple biomarkers to help disentangle the mechanistic heterogeneity of SCD as an early stage of cognitive decline.