A new test may help researchers understand why a toxin builds up in the brains of Alzheimer's patients.
Alzheimer's causes progressive loss of memory
Amyloid beta protein accumulates in the brain in Alzheimer's disease but whether the body produces too much or cannot break it down is unclear.
But by labelling the protein with a carbon isotope, doctors can measure the rate of turnover, a report in Nature Medicine suggests.
Experts said the test could help improve diagnosis and treatment.
Doctors are already able to measure amounts of amyloid beta protein - or abeta - in the cerebrospinal fluid but that doesn't indicate why the build-up is occurring.
By working out whether the body is producing too much or is unable to break it down, researchers develop drugs too accurately target the right process.
Furthermore, the test may also prove useful in the diagnosis of Alzheimer's prior to the onset of clinical symptoms.
The team at the Washington School of Medicine gave eight healthy volunteers an intravenous infusion which contained an amino acid - Leucine - that had carbon molecules with one extra neutron.
They then took samples of cerebral spinal fluid - the fluid that surrounds the brain - over a period of 36 hours.
The body uses amino acids to form proteins so the researchers were able to measure how the carbon isotope was taken up in the production of amyloid beta protein and then how long it took to break the labelled proteins down.
They found the the production rate of the amyloid protein was 7.6% per hour and the clearance rate was 8.2% per hour - much faster than anyone had predicted.
"Abeta has the second-fastest production rate of any protein whose production rate has been measured so far," said lead author Dr Randall Bateman, assistant professor of neurology at Washington University Medical School.
"In a time span of about six or seven hours, you make half the amyloid beta found in your central nervous system."
There are Alzheimer's drugs in development which either decrease or increase the clearance of the protein.
"This new test could let us directly monitor patients in clinical trials to see if the drug is really doing what we want it to," said Dr Batemam.
"If further study confirms the validity of our test, it could be very valuable for determining which drugs go forward in clinical trials and at what doses."
He added: "We hope to study whether we can develop ways to identify potential Alzheimer's patients on the basis of a metabolic imbalance between Abeta synthesis and clearance rates."
Rebecca Wood, Chief Executive, Alzheimer's Research Trust, said: "The results so far suggest that both the production and clearance of amyloid from the brain is much faster than has previously been supposed.
"It is very interesting and, if validated, the method could provide new avenues for scientists looking for better early diagnostic tests and improved ways of targeting and monitoring the effects of new drugs."
But she added: "The research used measurements of cerebral spinal fluid that surrounds the brain, but we need to assess what extent they reflect what it happening inside the brain itself.
"More work is needed before we can say this is a valid and reliable method, but this research could bring us a step closer towards finding an answer to Alzheimer's disease."
Dr Susanne Sorensen, Head of Research at the Alzheimer's Society said: "This test could be of real value in research into the processes that leads to Alzheimer's disease and in understanding what the drugs do.
"However, it is not, in its present format, useful as a way to diagnose Alzheimer's routinely."