My Healthcare News & Research — January 9, 2026

An Alzheimer's Blood Test Reaches the Clinic — and Meets the Real World

For decades, proving that someone's memory loss was actually Alzheimer's disease — and not depression, a thyroid problem, a vitamin B12 deficiency, or one of the other dementias — meant an expensive PET brain scan or a spinal tap. Both are accurate. Both are also slow, costly, and off-putting enough that many people never get a firm answer at all. A blood draw that could do much of that work has been the field's long-promised shortcut. As of the start of 2026, that shortcut is real: two blood tests are FDA-cleared and moving into everyday practice. This edition looks at what they actually measure, what the studies really showed, and a late-2025 stumble that explains why the fine print matters as much as the headline.

What Was Actually Cleared

There are now two distinct tests, and they do different jobs. The first, Lumipulse (made by Fujirebio Diagnostics), was cleared by the FDA in May 2025 as the first blood test to aid in diagnosing Alzheimer's. It measures the ratio of two proteins in plasma: p-tau217 (a form of tau that rises early in Alzheimer's) and beta-amyloid 1-42 (a fragment tied to the amyloid plaques that build up in the disease). The ratio between them tracks whether amyloid plaques are likely present in the brain. It is cleared for adults 55 and older who already have signs or symptoms of cognitive decline — not for healthy people curious about their risk.

The second test, Roche's Elecsys p-tau181, was cleared in October 2025 and is the first blood test approved for use in the primary-care setting. Its role is narrower and, in a way, more modest: it is designed to help a family doctor rule out Alzheimer's-related amyloid pathology in a patient with memory complaints, so that people who almost certainly do not have the disease can be spared a referral and a scan. It was developed with Eli Lilly and cleared for the same 55-and-older, symptomatic population.

What the Studies Really Showed

The validation numbers are genuinely good. In the data behind Lumipulse's clearance, roughly 92% of people with a positive result were confirmed to have amyloid plaques on a follow-up PET scan or spinal fluid test, and about 97% of people with a negative result were confirmed to be plaque-free. Roche's p-tau181 test was cleared on the strength of a 312-participant study in which it correctly ruled out amyloid pathology about 98% of the time when it read negative — which is exactly what you want from a rule-out tool.

In plain terms: a negative result on either test is strong reassurance, and a positive Lumipulse result usually means amyloid really is there. That is a real advance over guessing from a memory questionnaire alone. It is worth being clear about what these numbers are, though. They come from carefully run research cohorts, where blood is drawn, spun, and frozen under tight protocols. The messier question is how the tests behave in an ordinary clinic — and that is where 2026 opens with a cautionary note.

The Real-World Gap

At a major Alzheimer's research conference in December 2025, a Mayo Clinic team reported what happened when they ran the Lumipulse test on their own memory-clinic patients using the FDA-cleared cutoffs. The test still caught essentially everyone who truly had amyloid — but it also flagged a large share of people who did not. Among 75 patients confirmed amyloid-negative by PET or spinal fluid, 30 (40%) tested falsely positive. The share of positive results that were truly correct dropped from the pivotal study's 92% down to about 75%, and overall accuracy in that cohort fell to roughly 77%.

The likely culprit turned out to be practical, not conceptual. Fujirebio said it had identified a manufacturing issue with recent commercial lots of the test and placed them on a voluntary quality hold while it released corrected kits; this was a company action, not a formal FDA recall. Researchers also debated whether part of the problem was the chosen cutoff thresholds, or the well-known fragility of beta-amyloid 1-42, which can stick to tube walls and shift with how long a sample sits before it is processed. Whatever the exact mix of causes, the lesson is the useful one: a test that shines in a research lab can drift in the real world, and a single bad batch or a rushed blood draw can turn a good result into a misleading one.

What It Means for You

Alzheimer's now affects roughly 7 million Americans, a number expected to nearly double by 2060, and surveys suggest about four in five people would want to know early if they were developing it. For anyone facing that fear, a $100-ish blood draw that is faster, cheaper, and far less daunting than a PET scan or spinal tap is genuinely welcome news. But a few honest caveats should travel with it:

The practical takeaway: if you or someone you love is having real memory trouble, these blood tests are a reasonable, low-burden first step to bring up with a doctor — a way to move faster toward an answer. Treat a negative result as strong reassurance, treat a positive result as a reason for careful follow-up rather than a verdict, and don't chase the test out of anxiety if there are no symptoms. The technology has arrived. The judgment around it still has to come from a clinician who knows the person, not from a number on a lab slip. For related background, see our pages on Alzheimer's disease, vascular dementia, and lab tests.

Sources