The scan everyone trusts turns out to be years behind

Amyloid-PET has functioned as the gold standard for spotting early Alzheimer's disease for years, the scan doctors and researchers turn to when they want to know if the sticky plaques that define the disease have started forming in someone's brain. A new study from the University of Oslo, published in Nature Neuroscience, suggests that gold standard has been missing a substantial head start. Researchers found measurable structural changes in the brain's cortex at least seven years before amyloid plaques become detectable on a PET scan at all.

That's a genuinely uncomfortable finding for a field that has organized much of its early-detection and drug-timing strategy around amyloid-PET results. If the scan itself only catches the disease seven-plus years after something has already started going wrong structurally, then a lot of what researchers have assumed about the "earliest" stage of Alzheimer's may need revising.

How you catch something before you know when to look for it

The methodological trick behind this finding is almost as interesting as the result itself. You can't design a study to look for signs of Alzheimer's seven years before diagnosis if you don't know in advance which of your healthy volunteers will eventually develop the disease. The Oslo team solved that problem with patience: they followed cognitively healthy older adults with repeated MRI brain scans over nearly two decades, a genuinely rare kind of longitudinal dataset in neuroscience, where funding cycles rarely stretch that far.

That long follow-up let researchers work backward. Once enough time had passed to see which participants eventually crossed the threshold for detectable amyloid on PET scans, and which didn't, the team went back through each person's decade of prior MRI scans looking for a divergence point, some structural signal that showed up earlier in the people who would later develop plaques than in those who wouldn't. They found one: measurable cortical thickness changes, differences in the physical thickness of the brain's outer layer, showed up at least seven years before amyloid became detectable on PET.

What the lead researchers are actually claiming, and what they're not

James Michael Roe, a postdoctoral researcher at the Center for Lifespan Changes in Brain and Cognition at the University of Oslo when the study was conducted, was careful about how to interpret the finding. "We found that structural changes in the brain occur many years before high levels of plaque are seen on PET scans, which is the brain scan currently used to identify the earliest signs of Alzheimer's disease," he said. He described it as "the earliest signal detected on brain scans to date," useful specifically for tracking disease progression before symptoms emerge.

Crucially, Roe and his co-author, Professor Anders Martin Fjell, stopped short of claiming they've identified the true root cause of Alzheimer's. Two competing explanations remain open. One possibility is that harmful biological processes, either driving amyloid accumulation or resulting from it, are already active well before plaques reach detectable levels. The other is that some entirely separate process is causing these early structural changes independently of amyloid buildup altogether. Fjell was direct about what's at stake in resolving that distinction: "If the latter is true, it suggests it is important to continue developing drugs that target processes other than amyloid plaque accumulation. But we need more research on this."

Why the amyloid hypothesis itself is under quiet reconsideration

That caveat matters more than it might first appear, because it touches one of the most consequential and contested theories in Alzheimer's research: the amyloid hypothesis, the idea that plaque buildup is the primary driver of the disease rather than simply an early marker of it. Most current FDA-approved Alzheimer's drugs, including antibody therapies designed to clear amyloid from the brain, are built directly on that hypothesis. If cortical thinning genuinely precedes amyloid accumulation by seven-plus years, one interpretation is that whatever triggers that early structural change might be the more fundamental driver of disease, with amyloid arriving as a downstream consequence rather than the original cause.

That interpretation, if it holds up under further research, would have real implications for treatment timing. Current amyloid-targeting therapies are typically administered once amyloid becomes detectable via PET scan or blood biomarkers, precisely the point this study suggests may already be years too late to intervene at the disease's actual point of origin. It's a similar logic to what's driving recent efforts to identify blood-based biomarkers that predict which patients will actually respond to amyloid-clearing drugs like lecanemab, since treatment timing and patient selection have both become central bottlenecks in translating anti-amyloid science into real clinical benefit.

A finding that fits a broader pattern in aging brain research

This isn't the only recent discovery suggesting the aging brain undergoes significant, previously underappreciated restructuring well before any diagnosable disease shows up. Other research has found that the brain's own resident immune cells get substantially replaced starting around age 50, a decades-long remodeling process happening quietly in cognitively normal adults. Findings like these, taken together, are steadily pushing the field's working definition of "early" Alzheimer's further and further back from the point of noticeable symptoms, and further back even from the point of detectable amyloid, toward a much longer preclinical window than researchers assumed even five years ago.

What this changes, and what it doesn't, right now

For patients and families, this study doesn't change anything about today's diagnostic pathway or treatment options; amyloid-PET and blood biomarker tests remain the tools clinicians actually have available, and this new cortical-thickness signal isn't yet a validated, clinically deployable test. What it does change is the research roadmap. If cortical thinning really does mark an earlier, more fundamental stage of Alzheimer's pathology, the next scientific priority becomes figuring out what's driving that structural change in the first place, and whether intervening at that stage, years before today's earliest detectable signal, could meaningfully alter the disease's course. That's a substantially harder problem than refining an existing PET scan. But identifying that there's a seven-year window nobody knew to look inside is, on its own, the kind of finding that reorganizes where a field decides to look next.