Early cerebrospinal fluid proteomic signatures preceding amyloid PET positivity predict Alzheimer’s disease progression

Cerebrospinal fluid (CSF) amyloid-beta 42 (Aβ42) declines decades before Alzheimer’s disease (AD) symptoms, yet the proteomic correlates of this early decline, and their potential alterations prior to the onset of measurable Aβ42 abnormality, remain poorly characterized, limiting early risk stratification efforts. We analyzed over 6,000 CSF proteins in 201 non-demented, amyloid positron emission tomography (PET)-negative participants from the Alzheimer’s Disease Neuroimaging Initiative at baseline. Linear mixed-effects (LME) models identified proteins associated with longitudinal Aβ42 rate of change, followed by pseudo-temporal trajectory modeling to characterize the dynamic patterns of these candidate proteins before CSF Aβ42 abnormality. Nested cross-validation with protein selection within each training fold was used to evaluate prediction of future amyloid PET positivity. A fixed 10-protein candidate panel was subsequently derived, and its associations with disease progression and AD-related endophenotypes were assessed using survival analysis and LME models. Findings were evaluated in one internal and two external cohorts. A total of 235 proteins were significantly associated with longitudinal Aβ42 rate of change, the majority showing negative associations. Pseudo-temporal clustering identified four trajectory patterns. The presynaptic vesicle machinery cluster showed significant alterations relative to the stable reference group years before Aβ42 abnormality. In nested cross-validation, adding proteomic features selected within each training fold to demographics and baseline amyloid measures improved discrimination of PET conversion in both logistic regression (AUC = 0.845) and elastic-net Cox models (C-index = 0.835). Separately, a higher score from the fixed 10-protein panel was associated with accelerated progression from cognitively normal to mild cognitive impairment due to AD or AD (HR = 2.40, 95% CI 1.34–4.30, P = 0.003), longitudinal cognitive decline, and CSF p-tau accumulation. These findings were supported by the internal independent subset and external cohorts. In an amyloid PET-negative population, we identified a CSF proteomic signature associated with longitudinal Aβ42 trajectory, with exploratory evidence that proteins involved in presynaptic vesicle machinery were already altered before measurable Aβ42 abnormality. Integrating early CSF proteomic signatures with amyloid measures and demographics may help identify individuals at elevated risk before amyloid PET positivity, offering a potential molecular window for timely preventive intervention.

Authors

Institutions

Publication Details

Journal
Alzheimer s Research & Therapy
Published
2026-11-20
DOI
https://doi.org/10.1186/s13195-026-02258-3
Primary Topic
Dementia and Cognitive Impairment Research
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Early cerebrospinal fluid proteomic signatures preceding amyloid PET positivity predict Alzheimer’s disease progression

Dandan Sheng, Yumei Liang, Siqi Xie, for the Alzheimer’s Disease Neuroimaging Initiative et al.
3 links
Alzheimer s Research & Therapy
Dementia and Cognitive Impairment Research
article

Early cerebrospinal fluid proteomic signatures preceding amyloid PET positivity predict Alzheimer’s disease progression

Dandan Sheng, Yumei Liang, Siqi Xie, for the Alzheimer’s Disease Neuroimaging Initiative, Lan Ding, Jianping Jia, Ting Yang, Zhibo Wang
article en

Abstract

Cerebrospinal fluid (CSF) amyloid-beta 42 (Aβ42) declines decades before Alzheimer’s disease (AD) symptoms, yet the proteomic correlates of this early decline, and their potential alterations prior to the onset of measurable Aβ42 abnormality, remain poorly characterized, limiting early risk stratification efforts. We analyzed over 6,000 CSF proteins in 201 non-demented, amyloid positron emission tomography (PET)-negative participants from the Alzheimer’s Disease Neuroimaging Initiative at baseline. Linear mixed-effects (LME) models identified proteins associated with longitudinal Aβ42 rate of change, followed by pseudo-temporal trajectory modeling to characterize the dynamic patterns of these candidate proteins before CSF Aβ42 abnormality. Nested cross-validation with protein selection within each training fold was used to evaluate prediction of future amyloid PET positivity. A fixed 10-protein candidate panel was subsequently derived, and its associations with disease progression and AD-related endophenotypes were assessed using survival analysis and LME models. Findings were evaluated in one internal and two external cohorts. A total of 235 proteins were significantly associated with longitudinal Aβ42 rate of change, the majority showing negative associations. Pseudo-temporal clustering identified four trajectory patterns. The presynaptic vesicle machinery cluster showed significant alterations relative to the stable reference group years before Aβ42 abnormality. In nested cross-validation, adding proteomic features selected within each training fold to demographics and baseline amyloid measures improved discrimination of PET conversion in both logistic regression (AUC = 0.845) and elastic-net Cox models (C-index = 0.835). Separately, a higher score from the fixed 10-protein panel was associated with accelerated progression from cognitively normal to mild cognitive impairment due to AD or AD (HR = 2.40, 95% CI 1.34–4.30, P = 0.003), longitudinal cognitive decline, and CSF p-tau accumulation. These findings were supported by the internal independent subset and external cohorts. In an amyloid PET-negative population, we identified a CSF proteomic signature associated with longitudinal Aβ42 trajectory, with exploratory evidence that proteins involved in presynaptic vesicle machinery were already altered before measurable Aβ42 abnormality. Integrating early CSF proteomic signatures with amyloid measures and demographics may help identify individuals at elevated risk before amyloid PET positivity, offering a potential molecular window for timely preventive intervention.

Alzheimer s Research & Therapy
Capital Medical University (CN), National Clinical Research Center for Digestive Diseases (CN), National Clinical Research (US)
Changping Laboratory, U.S. Department of Defense, Foundation for the National Institutes of Health, Michael J. Fox Foundation for Parkinson's Research, Eli Lilly and Company, Pfizer, Sanofi, Biogen, University of Southern California, Teva Pharmaceutical Industries, Celgene, Meso Scale Diagnostics, BioClinica, Novartis Pharmaceuticals Corporation, Northern California Institute for Research and Education, Jazz Pharmaceuticals, Neurocrine Biosciences, Verily Life Sciences, Aligning Science Across Parkinson’s, Voyager Therapeutics, Weston Family Foundation, Eisai, Beijing Municipal Science and Technology Commission, Servier, H. Lundbeck A/S, Sun Pharma, IXICO, National Institutes of Health, Genentech, Johnson and Johnson, Allergan, Avid Radiopharmaceuticals, Canadian Institutes of Health Research, National Institute on Aging, National Institute of Biomedical Imaging and Bioengineering
Good health and well-being
Dementia and Cognitive Impairment Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon