Neurology. 2025 Sep 23;105(6):e214041. doi: 10.1212/WNL.0000000000214041. Epub 2025 Sep 4.
ABSTRACT
BACKGROUND AND OBJECTIVES: β-Amyloid (Aβ) likely triggers the spread of pathologic tau from the entorhinal cortex (EC) to the neocortex, but whether distinct Aβ levels exert differential influences on tau propagation beyond the EC remains unclear. We aimed to investigate the modifying effect of Aβ on the association of initial tau deposition with successive tau accumulation.
METHODS: A retrospective analysis was performed using data from 2 longitudinal observational cohort studies, the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and the Harvard Aging Brain Study (HABS), both conducted in the United States. All participants underwent a baseline study visit that included amyloid PET scans, [18F]-flortaucipir PET scans, and at least 1 follow-up tau PET scan (amyloid positivity was not required). Linear mixed-effects models were fitted to examine the modifying effect of Aβ on the relationship between baseline EC tau levels and tau accumulation over time. The outcome variables included changes in tau PET in 6 regions of interest (ROIs): the entorhinal, inferior temporal (IT), inferior parietal (IP), meta-temporal (meta-ROI), Braak III-IV, and Braak V-VI regions.
RESULTS: This study included 434 older adults from ADNI (262 [60%] cognitively unimpaired [CU] and 172 [40%] cognitively impaired participants; mean age, 73 ± 7 years; 51% female; mean follow-up duration, 3.69 ± 1.98 years) and 200 CU participants from HABS (mean age, 72 ± 9 years; 61% female; mean follow-up duration, 4.33 ± 1.75 years). In ADNI, the 3-way interaction of baseline EC tau × centiloid (CL) × time was significant across all 6 ROIs (e.g., IT region: estimate = 0.00285, 95% CIs 0.00161-0.00408; p < 0.00001), as was baseline EC tau × CL2 × time (e.g., IT region: estimate = -0.00002, 95% CIs -0.00003 to -0.00001; p = 0.00001). The negative quadratic term suggests that the potentiating association between CL and tau spread diminishes beyond a critical threshold. A similar pattern of findings was replicated in HABS.
DISCUSSION: Distinct initial Aβ levels seem to interact with EC tau deposition to affect successive tau accumulation in the rest of the brain. Interpretation should be approached with caution because causality cannot be inferred from this observational study.
PMID:40906972 | DOI:10.1212/WNL.0000000000214041
