Beyond amyloid: Contributions of neuroinflammation, cerebrovascular disease, and neurodegeneration to memory impairment in the oldest-old, the LifeAfter90 Study.
Rizvi B, Maillard P, et al. • Alzheimer's & dementia : the journal of the Alzheimer's Association • 2026
Higher plasma GFAP was associated with increased posterior cerebral artery-defined WMH volume but not global amyloid or episodic memory, while both PCA WMH and global amyloid were associated with lower medial temporal lobe cortical thickness, which in turn was associated with episodic memory in the oldest-old.
Key Findings
Results
Higher plasma GFAP concentration was associated with increased posterior cerebral artery (PCA)-defined white matter hyperintensity volume in the oldest-old.
The study cohort consisted of N=439 diverse oldest-old participants in the LifeAfter90 Study.
The association was specific to PCA-defined WMH volume, not WMHs in other vascular territories.
Plasma GFAP was used as an indicator of reactive astrocytosis and neuroinflammation.
MRI was used to quantify white matter hyperintensities by vascular territory.
Results
Plasma GFAP was not associated with global amyloid beta burden as measured by PET.
Global Aβ burden was assessed using positron emission tomography (PET).
This finding suggests neuroinflammation, as indexed by GFAP, may operate through cerebrovascular pathways rather than amyloid pathways in the oldest-old.
The cohort was a diverse sample of N=439 oldest-old individuals.
Results
Plasma GFAP was not directly associated with episodic memory performance.
Despite its associations with cerebrovascular disease markers, plasma GFAP did not show a direct association with episodic memory.
This suggests neuroinflammation's effect on memory may be mediated through downstream structural changes rather than being direct.
Episodic memory was assessed in all N=439 participants.
Results
Both PCA-defined WMH volume and global amyloid burden were independently associated with lower regional medial temporal lobe (MTL) cortical thickness.
Medial temporal lobe cortical thickness was assessed using MRI.
Both cerebrovascular disease (PCA WMH) and amyloid pathology converged on MTL atrophy as a downstream consequence.
This finding supports a model in which multiple pathological pathways contribute to neurodegeneration in the oldest-old.
Results
Lower medial temporal lobe cortical thickness was associated with worse episodic memory performance.
MTL cortical thickness served as an intermediary linking upstream pathologies (WMH and amyloid) to cognitive outcomes.
This finding implicates MTL neurodegeneration as a key mechanism through which both vascular and amyloid pathologies impair memory.
The relationship positions MTL atrophy as a downstream mediator in the pathological cascade.
Methods
The study tested a hypothesized model in which neuroinflammation acts on cerebrovascular disease and separately on amyloid pathology to promote downstream MTL atrophy and memory impairment.
N=439 diverse oldest-old participants were enrolled in the LifeAfter90 Study.
Multimodal assessment included plasma biomarkers (GFAP), MRI (WMH, cortical thickness), and PET (amyloid).
WMHs were specifically defined by vascular territory, allowing differentiation of PCA versus other territory involvement.
The study design was cross-sectional in nature based on the methods described.
Background
The authors propose that neuroinflammatory processes may play a central role in promoting mixed pathologies and cognitive impairment specifically in the oldest-old population.
The oldest-old are defined as individuals aged 90 and above in the LifeAfter90 Study.
Mixed pathology (vascular and amyloid) was implicated as relevant to this age group.
The findings are described as offering 'insights into new therapeutic and prevention strategies in the oldest-old population.'
What This Means
This research suggests that brain inflammation, measured through a blood protein called GFAP (a marker of reactive brain support cells called astrocytes), is linked to a specific type of brain damage in people aged 90 and older. Specifically, higher GFAP levels were associated with more white matter damage (small lesions in the brain's wiring) in a region supplied by the posterior cerebral artery, but were not directly linked to amyloid plaques (the protein clumps classically associated with Alzheimer's disease) or to memory problems on their own. The study involved 439 diverse participants in the LifeAfter90 Study who underwent blood tests, MRI brain scans, and PET scans.
This research suggests that in the very oldest adults, memory impairment is not driven by amyloid alone. Both the vascular brain damage associated with inflammation and amyloid plaques independently contributed to shrinkage of a brain region called the medial temporal lobe, which is critical for forming new memories. This shrinkage, in turn, was associated with worse episodic memory (the ability to remember specific events and experiences). This points to a chain of events: inflammation promotes vascular brain damage, both vascular damage and amyloid lead to brain shrinkage, and brain shrinkage leads to memory decline.
The practical importance of these findings is that they highlight multiple potential targets for treating or preventing memory loss in people over 90, not just amyloid. Reducing brain inflammation or preventing vascular brain damage could be important strategies in this age group, which has historically been underrepresented in dementia research. The study adds to growing evidence that the oldest-old may have a distinct pattern of brain disease driving cognitive decline compared to younger older adults.
Rizvi B, Maillard P, Posis A, Colbeth H, Chen N, Brickman A, et al.. (2026). Beyond amyloid: Contributions of neuroinflammation, cerebrovascular disease, and neurodegeneration to memory impairment in the oldest-old, the LifeAfter90 Study.. Alzheimer's & dementia : the journal of the Alzheimer's Association. https://doi.org/10.1002/alz.71795