U.S. researchers say they are developing a blood test that can help to detect Alzheimer's disease two to six years before the onset of the disease.

The test identifies changes in a handful of blood plasma proteins that cells use to send messages to one another.

The research team at the Stanford University School of Medicine discovered a connection between shifts in the cells' "dialogue" and the changes in the brain that mark Alzheimer's.

By "listening" to the way different proteins communicate, the researchers say they can measure whether something is going wrong in the cells. Their "chatter" takes on a different tone, explains Tony Wyss-Coray, associate professor of neurology and neurological sciences and senior author of the study.

They found that the blood test could indicate who would develop Alzheimer's with 90 per cent accuracy.

The study will appear in the Oct. 15 advance online edition of Nature Medicine.

Currently, the only way to clinically diagnose Alzheimer's is by ruling out other causes of cognitive declines, such as stroke and tumours. What often remains is Alzheimer's, the most common cause of dementia.

Even the clinical diagnosis is imperfect, and the only definitive diagnosis is by brain autopsy after a person has died.

The researchers came to their findings after obtaining 259 blood samples from individuals who had symptoms ranging from nothing abnormal to mild cognitive impairment to advanced Alzheimer's.

Starting with 120 communication proteins, the team developed an analysis procedure to recognize if there was a pattern seen in Alzheimer's that could be compared with that of people without the condition.

They discovered that as few as 18 proteins were sufficient to identify an Alzheimer's-specific pattern.

Among blood samples from 92 individuals who ranged from no symptoms to full dementia, the protein analysis matched the clinical diagnosis 90 per cent of the time.

In their study, the team determined that the 18 proteins that indicate Alzheimer's are also involved in the production of new blood cells, immune processes and apoptosis, the process of programmed cell death when a cell is no longer needed.

"Our hypothesis is that there is something wrong with the production of certain blood cells, which may be needed to clear that stuff that accumulates in the brain in Alzheimer's disease," said Wyss-Coray.

"That makes a lot of sense, and it is very exciting to think of immune cells and molecules interacting with the brain."

The authors emphasize their findings need to be confirmed in further clinical trials.