(The New York Times News Service) -- It has been known for some years that people who suffer from rheumatoid arthritis are less likely to develop Alzheimer's disease.

But now, researchers at the University of South Florida think they know why, and their findings could pave the way for a treatment for the disease.

The researchers found that a protein released into the bloodstream of people with rheumatoid arthritis provided protection against Alzheimer's in mice.

Their study appears online today in the Journal of Alzheimer's Disease.

One of the next steps could be testing the protein, known as GM-CSF, on humans. Huntington Potter, the principal investigator, noted that the man-made form of GM-CSF, known as Leukine, is already approved by the FDA to treat certain cancer patients who need to generate more immune cells.

"Our study, along with the drug's track record for safety, suggests Leukine should be tested in humans as a potential treatment for Alzheimer's disease," said Potter, a professor of molecular medicine at USF Health's Byrd Alzheimer's Institute and director of the Florida Alzheimer's Disease Research Center.

People with rheumatoid arthritis a disease that leads to inflammation of the joints and surrounding tissues are less likely to develop Alzheimer's. It had been assumed that the non-steroidal anti-inflammatory drugs they were taking helped prevent the onset and progression of Alzheimer's. But recent clinical trials proved this was not the case.

So, USF researchers then looked at other explanations. They analyzed three growth factors of rheumatoid arthritis in mice that have been bred to develop Alzheimer's symptoms, and identified the protein GM-CSF as the most promising for protection against Alzheimer's. A 20-day study confirmed that the mice treated with GM-CSF performed better on tests measuring memory and learning than mice that were given a placebo. In fact, the memories in the treated mice were similar to normal-aged mice without dementia.

"We were pretty amazed that the treatment completely reversed cognitive impairment in 20 days," Tim Boyd, one of the lead authors of the study, said in a news release.

But success in mice doesn't necessarily mean the treatment will be beneficial to humans. Researchers have alleviated Alzheimer's-like symptoms in mice several times over, only to learn that the same treatment did not work in humans.

"It's essential people keep in mind that mice are not people," said Bill Thies, chief medical and scientific officer with the national Alzheimer's Association.

"Mice stay in their cages, eat their mice chow, and are of a very consistent genetic background," Thies said, while humans are much more complicated.

But at the same time, Thies said animal testing is essential.

"It gives us the possibilities we can examine in humans."

Potter says there's reason for "cautious optimism" about the latest finding. He says it's the combination of knowing that the protein GM-CSF is released during rheumatoid arthritis, and that those patients are protected against Alzheimer's; that the man-made form of the protein, Leukine, has been proven safe over nearly 20 years of use; and that the mouse study has shown benefits of using the protein.

"We expect there to be no problem setting up a clinical trial," Potter said.

Copyright 2002 The New York Times News Service. All rights reserved.