Cortical thinning appears years before amyloid reaches the scan threshold in Alzheimer’s
A long-running imaging study from the University of Oslo suggests Alzheimer’s disease may leave structural fingerprints in the brain far earlier than current amyloid-PET scans can detect plaque. In healthy older adults followed with repeated MRI for nearly two decades, investigators found brain changes associated with later amyloid positivity at least seven years before plaques became visible on PET imaging. The finding shifts attention from the moment amyloid becomes measurable to the quieter biological changes that may already be underway long before that threshold is crossed.
The study was published in Nature Neuroscience and centered on a simple but clinically important question: what happens in the brain before the current gold-standard scan turns positive? By looking back through years of MRI data from cognitively well-functioning older adults, the team identified structural differences in people who would later develop detectable amyloid accumulation. The primary signal was a change in cortical thickness, a marker of gray matter structure that can reflect neurodegenerative injury, synaptic loss, or other tissue-level alterations.
This is not a trivial shift in timing. It means the disease process may be advancing long before the familiar plaque-based milestone becomes visible on PET.
Longitudinal MRI exposed a hidden pre-plaque interval
The strength of the work lies in its timeline. Participants were scanned repeatedly over nearly 20 years, allowing the investigators to determine when amyloid first became detectable and then examine prior MRI studies for earlier structural change. That design matters because a single scan can only show a snapshot. Serial imaging, by contrast, can reconstruct the sequence of disease biology with far greater precision.
James Michael Roe, who led the study while working at the Center for Lifespan Changes in Brain and Cognition at the University of Oslo, described the MRI findings as structural changes occurring many years before high levels of plaque appeared on PET. Anders Martin Fjell, who heads the center, said the unique feature of the study was the ability to examine brain structure in the years before the first scan revealed plaques in older adults who remained cognitively intact.
Those details matter for a practical reason. If the earliest signal emerges long before amyloid rises above current detection limits, then a plaque-centered diagnostic strategy may be missing an earlier phase of the disorder. For clinicians and researchers trying to identify presymptomatic disease, the gap between biological onset and scan positivity may be wider than expected.
Amyloid may not be the first visible sign of the disease process
The results also complicate a long-standing assumption in Alzheimer’s research: that amyloid accumulation is the earliest detectable event. The imaging pattern observed here suggests two biologic possibilities. One is that harmful processes tied to amyloid are already active in the brain, but remain below the threshold of current PET sensitivity. The other is more provocative: separate biological processes may be driving cortical change before amyloid plaques begin to build.
Either path leads to the same clinical tension. A scan can look negative while disease-related biology is already advancing.
Fjell described Alzheimer’s as closely linked to aging and shaped by multiple biological factors, which fits the broader reality of the disorder. Amyloid is a major pathology, but not necessarily the only one that matters at the earliest stage. Cortical thinning before plaque detection raises the possibility that neurodegeneration, vascular stress, inflammatory activity, metabolic dysfunction, or other age-related mechanisms may precede visible amyloid deposition. The study does not resolve which process comes first. It does show that the brain is changing before the current imaging threshold is reached.
What earlier detection could mean for prevention and trials
The most immediate implication is diagnostic. If structural change can be detected years earlier than amyloid positivity, MRI-based markers may help researchers identify individuals at risk before symptoms emerge and before plaque scans turn positive. That could make prevention studies easier to design, especially in older adults who still function normally but may already be entering a preclinical disease stage.
The therapeutic implications are just as important. If cortical thinning begins through mechanisms that are not fully captured by amyloid accumulation, then drugs aimed solely at plaque biology may miss a substantial part of the early disease cascade. In that scenario, future treatment strategies would need to keep pressure on other targets as well, including pathways that influence neuronal integrity and brain structure before overt plaque buildup.
The study does not prove that a non-amyloid therapy would succeed. It does, however, strengthen the argument that Alzheimer’s should not be viewed through a single molecular lens.
Imaging sensitivity may need to move upstream
Amyloid-PET remains the current standard for identifying early Alzheimer’s pathology, but this work suggests its threshold may sit later in the disease timeline than once thought. Structural MRI, especially when repeated over time, may capture a different and earlier layer of pathology. Cortical thickness changes are not specific to Alzheimer’s on their own, yet in a longitudinal design they may serve as an early warning sign that warrants closer biological follow-up.
That distinction has consequences for both research and the bedside. Trials that enroll participants only after amyloid becomes visible may already be missing a window during which disease-modifying interventions could have greater effect. Imaging strategies that look for converging signals, rather than waiting for plaques alone, may better reflect how the disease unfolds in real life.
The study also raises a basic question for the field: how long can the brain compensate while structural decline is already under way? In these cognitively healthy older adults, function was preserved even as MRI hinted at future plaque positivity. The visible damage lagged behind the biology, and the biology lagged behind the symptoms.
That sequence is exactly why earlier detection remains such a difficult target in Alzheimer’s. By the time current scans turn positive, the first chapter may already be over.