Brain Scan Finds Alzheimer’s Risk In Adults With Normal Memory

Matching Memory Scores Hid Different Alzheimer’s Risks, Brain Scans Show

A new way of analyzing brain scans may help identify healthy older adults at greater risk of Alzheimer’s disease before standard memory tests reveal a problem. In a study published in Science Advances, two groups of older adults, sorted by their scans, scored about the same on memory tests at the start. One group then declined faster, built up more of an Alzheimer’s-linked protein, and more often moved toward Alzheimer’s or mild cognitive impairment, an early stage of memory and thinking trouble.

Researchers built the tool by comparing two extremes: SuperAgers, adults 75 and older whose memory scores beat those of middle-aged adults, and patients already diagnosed with Alzheimer’s. The resulting brain pattern sorted SuperAgers from patients correctly about 80% of the time in the original group and 71% of the time in a separate group of 143 people.

Alzheimer’s is hard to catch early. Damage can build quietly before symptoms appear, and memory loss is often the first obvious sign. Spotting vulnerable brains sooner could give doctors time to monitor patients, and perhaps someday step in. A scan result might eventually work like a cholesterol reading, a heads-up that prompts a closer look.

Brain Scan Pattern Separates SuperAgers From Alzheimer’s Patients

Most studies of exceptional memory in old age compare SuperAgers with typical older adults. That group is a mixed bunch, some aging steadily and some already in a symptom-free early stage of Alzheimer’s, so using it as the yardstick can blur the differences that matter. Anchoring the analysis to the two extremes was meant to make those differences easier to see.

Scans came from 290 people, 180 SuperAgers and 110 age-matched Alzheimer’s patients, drawn from two existing research projects. The MRI scans tracked which of 100 brain regions rise and fall in activity together while a person lies still. A computer model learned which connections best separated the two groups.

This pattern spanned regions that handle planning, memory, and emotional control. Patients showed stronger connections involving the insula, a fold deep in the brain, and the front tips of the brain’s side lobes. SuperAgers showed stronger connections across regions at the front, top, and back of the brain. The pattern also tracked brain age, an estimate of how old a brain looks based on its connections rather than the calendar. Patients’ brains tended to look older than SuperAgers’ did. In a side analysis, cognitively typical older adults landed between the two extremes, pointing to a sliding scale of cognitive aging rather than two separate boxes.

Healthy Adults With Matching Memory Scores Diverged in Alzheimer’s Risk

A harder test used people with no memory problems. The researchers scored scans from 818 cognitively healthy adults in the same two projects, plus an independent set of 497 from a third project, on how closely each matched the SuperAger or the Alzheimer’s pattern. About one in four in each set landed in the Alzheimer’s-like camp: 211 of 818 and 127 of 497.

Over follow-up visits, the two camps split. Alzheimer’s-like adults declined faster on overall dementia ratings and on tests of planning and problem-solving. They also accumulated amyloid-beta, the sticky protein buildup tied to Alzheimer’s, more quickly, despite starting with similar levels. Tau, a second Alzheimer’s-linked protein, did not differ.

Alzheimer’s-like adults were also more likely to progress to mild cognitive impairment or Alzheimer’s, in both sets of participants. Adding the scan score improved progression forecasts beyond demographics, amyloid and tau brain imaging, and APOE, a well-known Alzheimer’s risk gene. One caution: Alzheimer’s-like adults were older on average, the one starting difference seen in both sets. A follow-up analysis that separated out starting age still showed the gap in the independent set.

Two Gene Regions Emerge as Unconfirmed Leads

To explore possible genetic roots, the team ran an early genetic scan in 506 participants from that third project, 151 Alzheimer’s-like and 355 SuperAger-like. Three genetic variants near two genes, CLYBL and FRMD6, stood out. CLYBL has been linked to metabolism and inflammation, and FRMD6 to how brain cells grow and connect. None met the strict statistical standard normally required in genetics, and no second group confirmed them. The authors call them candidates for future study.

Brain Scan Alzheimer’s Risk Test Is Not Ready for Doctors’ Offices

None of this amounts to a diagnostic test. The study found differences between groups, so it cannot say what will happen to any one person. The authors flagged other gaps: the participants need to be more ethnically diverse, the pattern relied on a single type of brain measurement, and the Alzheimer’s-like pattern was tied to amyloid buildup, so it may not reflect healthy-brain resilience alone. Larger studies that follow people forward are still needed.

Still, the core discovery holds up: two people can look identical on a memory test and yet carry very different brain patterns linked to very different risks of future decline. Memory tests alone miss that gap, so reading the brain itself looks like the more promising route to earlier prevention, once a scan proves its worth in individual patients.

---

Disclaimer: This article summarizes a single research study for general information only. It is not medical advice and cannot diagnose, treat, or predict Alzheimer’s disease in any individual. Anyone with concerns about memory or brain health should speak with a qualified health care professional.

---