Plasma NfL is associated with the APOE ε4 allele, brain imaging measurements of neurodegeneration, and lower recall memory scores in cognitively unimpaired late-middle-aged and older adults
Michael Malek-Ahmadi,Yi Su,Valentina Ghisays,Ji Luo,Vivek Devadas,Yinghua Chen,Wendy Lee,Hillary Protas,Kewei Chen,Henrik Zetterberg,Kaj Blennow,Richard J Caselli,Eric M Reiman
DOI: https://doi.org/10.1186/s13195-023-01221-w
2023-04-10
Abstract:Background: Plasma neurofilament light (NfL) is an indicator of neurodegeneration and/or neuroaxonal injury in persons with Alzheimer's disease (AD) and a wide range of other neurological disorders. Here, we characterized and compared plasma NfL concentrations in cognitively unimpaired (CU) late-middle-aged and older adults with two, one, or no copies of the APOE ε4 allele, the major genetic risk factor for AD. We then assessed plasma NfL associations with brain imaging measurements of AD-related neurodegeneration (hippocampal atrophy and a hypometabolic convergence index [HCI]), brain imaging measurements of amyloid-β plaque burden, tau tangle burden and white matter hyperintensity volume (WMHV), and delayed and total recall memory scores. Methods: Plasma NfL concentrations were measured in 543 CU 69 ± 9 year-old participants in the Arizona APOE Cohort Study, including 66 APOE ε4 homozygotes (HM), 165 heterozygotes (HT), and 312 non-carriers (NC). Robust regression models were used to characterize plasma NfL associations with APOE ε4 allelic dose before and after adjustment for age, sex, and education. They were also used to characterize plasma NfL associations with MRI-based hippocampal volume and WMHV measurements, an FDG PET-based HCI, mean cortical PiB PET measurements of amyloid-β plaque burden and meta-region-of-interest (meta-ROI) flortaucipir PET measurements of tau tangle burden, and Auditory Verbal Learning Test (AVLT) Delayed and Total Recall Memory scores. Results: After the adjustments noted above, plasma NfL levels were significantly greater in APOE ε4 homozygotes and heterozygotes than non-carriers and significantly associated with smaller hippocampal volumes (r = - 0.43), greater tangle burden in the entorhinal cortex and inferior temporal lobes (r = 0.49, r = 0.52, respectively), and lower delayed (r = - 0.27), and total (r = - 0.27) recall memory scores (p < 0.001). NfL levels were not significantly associated with PET measurements of amyloid-β plaque or total tangle burden. Conclusions: Plasma NfL concentrations are associated with the APOE ε4 allele, brain imaging biomarkers of neurodegeneration, and less good recall memory in CU late-middle-aged and older adults, supporting its value as an indicator of neurodegeneration in the preclinical study of AD.
What problem does this paper attempt to address?