Blood Test for Alzheimer's Risk Identifies Disease Decades Before Symptoms

Alzheimer's disease (AD) begins silently — amyloid plaques and tau tangles accumulate in the brain for 15–20 years before cognitive symptoms emerge. Identifying this preclinical stage has historically required expensive PET imaging or invasive lumbar punctures, limiting large-scale screening. This study, published in JAMA Neurology, evaluated whether plasma phosphorylated tau 217 (p-tau217) — a blood-based biomarker — could accurately identify preclinical AD across diverse, cognitively unimpaired populations.

The study pooled data from 1,767 cognitively normal participants across eight international cohorts: BioFINDER-1, BioFINDER-2, ADNI, ALFA+, PREVENT-AD, TRIAD, Wisconsin ADRC, and the Amsterdam Dementia Cohort. Participants had plasma p-tau217 measured using the ALZpath Simoa assay, alongside amyloid PET, tau PET, and/or CSF biomarkers. The primary outcome was the ability of plasma p-tau217 to detect amyloid positivity (A+), defined by PET or CSF, in cognitively normal individuals.

Plasma p-tau217 demonstrated high diagnostic accuracy for identifying amyloid-positive individuals, with area under the curve (AUC) values ranging from 0.83 to 0.93 across cohorts (pooled AUC ~0.88). Using a two-threshold approach — a low cutoff to rule out AD pathology and a high cutoff to rule it in — approximately 60–70% of participants could be classified with high confidence, leaving only 30–40% in an intermediate gray zone requiring confirmatory testing. Sensitivity exceeded 85% and specificity exceeded 80% in most cohorts at optimized thresholds.

The study also examined p-tau217's relationship with tau PET burden. Among amyloid-positive individuals, higher plasma p-tau217 correlated strongly with greater tau PET signal (r = 0.65–0.75), particularly in the entorhinal cortex and temporal regions — areas affected earliest in AD. This suggests p-tau217 not only detects amyloid pathology but also tracks the downstream tau spreading that drives neurodegeneration. Importantly, p-tau217 levels were significantly elevated even in A+ individuals with no tau PET signal, indicating it rises early in the amyloid cascade.

Age, sex, and APOE ε4 status influenced p-tau217 levels, with APOE ε4 carriers showing higher values independent of amyloid status — a finding with implications for threshold calibration in clinical use. The authors propose that plasma p-tau217 could serve as a first-line screening tool in prevention trials and memory clinics, triaging who needs confirmatory PET or CSF testing and dramatically reducing the cost and burden of preclinical AD detection. They estimate that using this two-threshold strategy could reduce the need for confirmatory imaging by roughly 60–65% without meaningfully sacrificing diagnostic accuracy.

The study's main caveat is that all cohorts were predominantly white and highly educated, limiting generalizability. Additionally, plasma p-tau217 thresholds varied across cohorts and assay platforms, underscoring the need for standardization before widespread clinical deployment. Nonetheless, this is among the largest and most methodologically rigorous validations of blood-based preclinical AD detection to date, representing a significant step toward accessible, scalable Alzheimer's screening.