WASHINGTON (AP) -- Scientists for the first time have peered into people's brains to directly measure the ebb and flow of a substance notorious for its role in Alzheimer's disease.

The delicate research was performed not with Alzheimer's patients but with people suffering severe brain injuries -- because a brain injury increases the risk of developing dementia later in life.

The goal is to learn why, so that doctors one day might be able to lower that risk.

But with this first-step study, a team of scientists from Missouri and Italy got a surprise.

Too much of that Alzheimer's-related protein, called beta-amyloid, is thought to be harmful. So the team had expected beta-amyloid levels to spike right after the injury and then drop as patients recovered.

Instead, beta-amyloid levels increased as patients improved and dropped if they got worse, lead researcher Dr. David Brody, a neurologist at Washington University in St. Louis, reported Friday in the journal Science.

What's going on? Beta-amyloid seems to be a marker of increased brain activity as patients improved.

If so, what started as a study of Alzheimer's risk might have implications for how the brain-injured are tracked in intensive-care units -- although that will take much more research to prove.

"Our study is just the beginning," Brody said. "Amyloid-beta measurements in the brain may turn out to be a good indicator of how well the cells are communicating with each other.'

Beta-amyloid is best known as the sticky goo that makes up the hallmark plaques inside the brains of Alzheimer's victims. But it doesn't start out as gunk. Soluble forms are found in the fluid that bathes the brain, although scientists don't understand its purpose, or just what happens to trigger formation of those bad plaques.

Nor do they understand how brain trauma so often leads to later Alzheimer's. One possibility is that extra beta-amyloid speeds whatever dementia-forming process might be lurking among brain cells. Another theory is that the injury decreases a person's "cognitive reserve," so that symptoms merely become apparent sooner.

Brody thought brain-injured patients could offer a precious opportunity to start testing that first possibility. These patients were undergoing brain surgery anyway. What if surgeons could insert an extra tiny tube at the same time that would allow hour-to-hour sampling of brain fluid, to measure beta-amyloid?

It's a technique called intracerebral microdialysis, and colleagues at Washington University already had performed it in mice -- linking increased synaptic activity, or communication between brain cells, to increased beta-amyloid.

Brody teamed with Dr. Sandra Magnoni of the Ospedale Maggiore Policlinico, a major trauma center in Milan that has experience with the technique. They asked the families of patients suffering brain injuries from car crashes, falls or hemorrhages from burst blood vessels if they'd agree to the experiment. Eighteen said yes.

Roughly 24 hours after the initial injury, the catheter was placed in patients' brains, where it stayed for three to seven days. ICU doctors and nurses otherwise provided routine care, tracking patients' neurologic changes with a standard tool called the Glasgow Coma Score.

Brody tracked the beta-amyloid levels and found they mirrored that coma score, with a direct relationship to each patient's neurological status.

That means the findings agreed with the previous research on mice. That's important, as so much Alzheimer's research must be performed in animals, said Dr. Ramona Hicks, a specialist in traumatic brain injury at the National Institutes of Health, which helped fund the work.

Brody cautions that beta-amyloid might have spiked in everyone right at the time of injury, something his study started too late to measure.

Also, it only measured total beta-amyloid levels, not a strung-together form that's thought to be toxic to cells, something Brody hopes to try next.

While the work raises more questions than it answers, it brings researchers a valuable new tool for studying both Alzheimer's risk and just what happens during brain-injury recovery.

"It sort of sets a platform for future studies," said NIH's Hicks.

Copyright 2008 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.