Neuroimaging signatures of aging and Alzheimer's disease display distinct associations with plasma biomarkers, and their combination with plasma biomarkers may improve the detection and prediction of cognitive impairment.
Key Findings
Results
SPARE-AD displayed the strongest association with plasma phosphorylated tau 217 among all plasma biomarkers examined.
SPARE-AD is a machine learning-derived neuroimaging signature specifically indexing Alzheimer's disease-related brain changes
Phosphorylated tau 217 (p-tau217) is a plasma biomarker associated with Alzheimer's disease pathology
Associations were assessed using partial correlations across 1297 participants without dementia
The finding suggests SPARE-AD captures AD-specific pathological processes also reflected in p-tau217
Results
SPARE-BA correlated most strongly with plasma neurofilament light (NfL) among all plasma biomarkers examined.
SPARE-BA (Spatial Pattern of Atrophy for Recognition-Brain Age) indexes brain aging patterns
Neurofilament light (NfL) is a plasma biomarker associated with neurodegeneration and general brain injury
The association was assessed using partial correlations in the same sample of 1297 participants without dementia
This finding indicates that SPARE-BA and SPARE-AD capture distinct biological processes reflected in different plasma biomarkers
Results
The combination of neuroimaging signatures with plasma biomarkers achieved the highest diagnostic discriminability compared to either modality alone.
Logistic regression models were used to evaluate diagnostic accuracy
Both SPARE-BA and SPARE-AD were derived from T1-weighted MRI images
The combined approach outperformed neuroimaging signatures alone and plasma biomarkers alone in diagnostic discrimination
The sample included 1297 participants without dementia drawn from a diverse cohort
Results
The combination of neuroimaging signatures with plasma biomarkers showed the strongest associations with clinical progression.
Cox proportional hazard models were used to evaluate clinical progression
Clinical progression was assessed in participants without dementia at baseline
Both individual neuroimaging signatures and individual plasma biomarkers were less predictive of progression than their combination
This suggests complementary information is captured by imaging and fluid biomarkers
Methods
SPARE-BA and SPARE-AD neuroimaging signatures were calculated from a diverse sample of 1297 participants without dementia.
Both signatures were derived using T1-weighted MRI images
The study explicitly aimed to derive signatures in a diverse sample
Participants were free of dementia at the time of neuroimaging
The study included participants from multiple cohorts including those with representation from diverse backgrounds, including from the Maestre lab which recruits from Latino populations
Results
Neuroimaging signatures of aging and Alzheimer's disease display distinct, rather than overlapping, associations with plasma biomarkers.
SPARE-AD was most strongly linked to p-tau217, a marker of AD-specific tau pathology
SPARE-BA was most strongly linked to NfL, a marker of general neurodegeneration and aging-related neuronal injury
The divergent biomarker associations suggest the two signatures index biologically distinct processes
This distinction supports the clinical utility of measuring both signatures separately
What This Means
This research suggests that two different types of brain scan measurements — one that tracks how much the brain has aged beyond what is expected (SPARE-BA) and one that detects patterns of brain change associated with Alzheimer's disease (SPARE-AD) — are linked to different proteins found in blood. Specifically, the Alzheimer's-related brain pattern was most closely tied to a blood protein called phosphorylated tau 217, which is associated with Alzheimer's pathology, while the brain aging pattern was most closely tied to neurofilament light, a protein that rises with general brain cell damage and aging. These findings were derived from over 1,200 people who did not have dementia at the time of their brain scans.
Importantly, this research suggests that combining these brain scan measurements with blood biomarker tests works better than using either approach alone for identifying people who may have early Alzheimer's disease or who are at risk of cognitive decline. When both types of information were used together, they more accurately identified people with cognitive impairment and better predicted who would experience worsening cognition over time.
This matters because early and accurate detection of Alzheimer's disease and related brain aging is a major challenge in medicine. This study suggests that pairing relatively accessible blood tests with specialized brain scan analyses could provide more complete information about an individual's brain health than either approach alone, potentially enabling earlier and more precise identification of people who may benefit from intervention or monitoring.
Gonzales M, Ho N, Wang D, Rashid T, Brandigampala S, Dong M, et al.. (2026). Integration of neuroimaging signatures of aging and Alzheimer's with fluid biomarkers.. Alzheimer's & dementia : the journal of the Alzheimer's Association. https://doi.org/10.1002/alz.71889