Memory Digest

Two New Alzheimer's Tests Made News This Week. Here's How to Tell Them Apart

Sunday, October 18, 2026

If you've been reading health news this week, you may have seen two very different stories about testing for Alzheimer's. One came from a hospital memory clinic and a medical journal. The other came from a startup launch and lit up X. Both are about catching the disease earlier, but they sit at very different points on the road from "interesting idea" to "test your doctor can trust."

Story one: a blood test, studied the careful way. Researchers at Hospital del Mar in Barcelona ran a randomized trial with 220 people who came to their memory clinic with thinking or memory complaints. Everyone got the usual workup. Half also learned their blood test results (a protein called p-tau217, plus a marker of nerve damage called NfL) at about three months; the other half waited until nine months. At three months, doctors felt very confident about the cause of symptoms for 50% of the early group, compared with under 5% of the group still waiting. People who got results early were more likely to start Alzheimer's treatment sooner, and those with reassuring results were more often sent back to their regular doctor instead of facing more rounds of testing. Just as important, getting the results early did not raise anxiety or depression, and stress actually went down. The study was published October 5 in JAMA Neurology. (Source: Earlier Disclosure of Blood-Based Biomarkers, JAMA Neurology.) It was one clinic in Spain, and it didn't measure whether people did better over the long run, but it is the kind of evidence that helps blood tests move into everyday care. Our article on blood tests for Alzheimer's explains how these tests work and who they're meant for.

Story two: an eye-tracking headset, still early. A Y Combinator startup called Saccade launched a 10-to-15-minute test that uses a virtual-reality headset to record fast eye movements and give a risk score for Alzheimer's. Its launch post says it is already testing people in their 30s and older in San Francisco clinics. (Source: Saccade launch page, Y Combinator.) The science behind the idea is real: a review of 71 studies found that certain eye-movement tasks consistently told apart people with Alzheimer's or mild cognitive impairment from healthy people. But the same review noted that most studies didn't confirm who actually had Alzheimer's biology, which makes it hard to know how well these tests work in the real world. (Source: Eye tracking in Alzheimer's: a systematic review.) The test is not FDA-cleared, and a company profile lists an FDA submission target of 2027. Several neuroscientists on X pushed back hard on the launch claims. That doesn't mean the idea is worthless. It means it hasn't been proven yet.

The takeaway. "Earlier" is only helpful if the test is accurate, if a clinician can explain what the result means for you, and if there's a sensible next step. That's why most experts still recommend these tests for people who already have symptoms and are being evaluated by a doctor, not as a screening gadget for healthy 35-year-olds. If you or someone you love is noticing changes, our guide to getting a diagnosis is a good place to start.

Have a restful Sunday, and we'll see you tomorrow.