White matter microstructure and cerebrospinal fluid biomarkers of Alzheimer’s disease in middle‐aged cognitively unimpaired participants (the ALFA study)

Background Previous studies comparing white matter (WM) integrity and biomarkers of Alzheimer’s disease (AD) at early stages have yielded inconsistent results and findings have remained inconclusive. In this work, we used diffusion‐weighted imaging (DWI) and cerebrospinal fluid (CSF) biomarkers to s...

Full description

Saved in:
Bibliographic Details
Published inAlzheimer's & dementia Vol. 16
Main Authors Operto, Greg, Milà‐Alomà, Marta, Suárez‐Calvet, Marc, Falcon, Carles, Kollmorgen, Gwen, Eichenlaub, Udo, Zetterberg, Henrik, Blennow, Kaj, Molinuevo, Jose Luis, Gispert, Juan Domingo
Format Journal Article
LanguageEnglish
Published 01.12.2020
Online AccessGet full text

Cover

Loading…
More Information
Summary:Background Previous studies comparing white matter (WM) integrity and biomarkers of Alzheimer’s disease (AD) at early stages have yielded inconsistent results and findings have remained inconclusive. In this work, we used diffusion‐weighted imaging (DWI) and cerebrospinal fluid (CSF) biomarkers to study the relationship between neuropathological burden and neural injury in the WM in a middle‐aged cognitively unimpaired population: the ALFA+ study. Methods N=318 cognitively unimpaired individuals from the ALFA+ study, aged from 45 to 74 (mean: 60.79∓4.66) underwent DWI and lumbar puncture. Biomarkers were measured with Elecsys® and NeuroToolKit robust prototype assays (Roche Diagnostics). Using tract‐based spatial statistics, we assessed their association with DWI metrics using individuals models including CSF markers of core AD pathology (A𝛃42, A𝛃42/40 ratio, p‐tau), neuronal injury (neurofilament light [NfL], t‐tau), synaptic dysfunction (neurogranin [NGR], 𝛂Synuclein), and inflammation (sTREM2, YKL40, GFAP, IL6, S100b). Age, sex and APOE e4 carriership were included as covariates. Supplementary analyses were performed including levels of AD core biomarkers as additional covariates and stratifying the sample into A/T groups. Results DWI metrics showed significant associations with A𝛃42, YKL40, IL6, NFL, NGR, pTau, tTau, S100, 𝛂Synuclein; none with GFAP, sTREM2, nor A𝛃42/40 ratio (Figures 1 and 2). Positive associations in fractional anisotropy (FA) (and negative with diffusivity) were found with A𝛃42, YKL40, NFL, NGR, S100, p‐tau and t‐tau. Negative association in FA (positive in diffusivity) was found for IL6 only. p‐tau showed the only association to be significant in both directions with FA, positive in cingular/callosal and negative in corticospinal tracts. Post‐hoc analyses on p‐tau showed significant interaction with A/T stage (Figure 3) with positive association with FA in A‐T‐ and negative in A+T+ subjects (inversely with diffusivity). Conclusion WM microstructural properties correlate in extensive regions with CSF AD core biomarkers, inflammation and axonal degeneration markers. Such associations illustrate the interplay of multiple pathways in early stages of Alzheimer’s continuum such as deleterious effects of amyloid deposition or the involvement of glial activation/inflammatory processes, resulting in observed reduced diffusivity. Furthermore, interaction with A/T stage suggest the presence of non‐monotonic dynamics occurring along the Alzheimer’s continuum.
ISSN:1552-5260
1552-5279
DOI:10.1002/alz.043027