Alzheimer’s is characterized by a toxic buildup of amyloid and tau proteins in the brain. This buildup destroys neurons. Several blood tests that can diagnose the disease already exist, but they aren't sensitive enough to predict the disease's onset.

A new study, published in the journal  and conducted by Howard Federoff and colleagues at Georgetown University, evaluated 525 people aged 70 and over for five years. At the onset of the study, the group did not show any signs of mental impairment. Each year of the study, the researchers conducted a detailed cognitive examination and took blood samples from all the participants.

During this time, 28 people developed Alzheimer's or mild cognitive impairment, which is considered to be the earliest noticeable sign of dementia, including Alzheimer's disease.

Related News: Vitamin E May Slow the Progression of Alzheimer's Disease »

An analysis of the participants' blood highlighted 10 metabolites that were depleted in people who had mild cognitive impairment and who went on to develop Alzheimer's, compared with those who didn't. In subsequent trials, the researchers showed that measuring these chemicals could predict who would develop Alzheimer's within the next three years, with up to 96 percent accuracy.

Commenting on the fact that the 10 metabolites play a key role in supporting cell membranes, maintaining neurons, or sustaining energy processes, Mark Mapstone, one of the researchers at the University of Rochester Medical Center in New York, said, "We think the decrease in these chemicals reflects the breakdown of neural populations in the brain."

Learn More: What Are the Symptoms of Alzheimer's Disease? »

Fast and Inexpensive

Once the test is confirmed in a larger study group, it may offer an inexpensive and fast way of predicting Alzheimer's. What’s more, the test may even be able to predict the disease well in advance of onset. According to Mapstone, these metabolic changes in the brain might occur up to 20 years before symptoms appear.

The research team is planning to look back at other dementia studies in which blood has been taken over decades, to see whether the chemical changes can be detected that early.

Related News: Hearing Loss Leads to Brain Tissue Loss in Older Adults »

Will People Want to Know?

Since there are currently no treatments available, it’s anyone’s guess whether people will want to take a blood test to learn that they are likely to develop Alzheimer’s. Gladstone believes that if people in their early 40s find out, they could make a habit of eating the right foods, avoiding head trauma, and participating in more exercise in order to slow the disease's onset.

Also, people who know that they are at risk may benefit by getting their affairs in order, planning for future care, and helping loved ones prepare.

More Research Needed

Commenting on the new blood test, Maria Carrillo, Ph.D., the Alzheimer’s Association's vice president of medical and scientific relations, said, “The Alzheimer’s disease field needs methods to detect and diagnose Alzheimer’s at its earliest time point in order to allow earlier intervention with new treatments and future potential prevention strategies." (Carrillo was not involved in this study.)

Carrillo added that current biological markers for early disease, including levels of abnormal proteins in cerebrospinal fluid, structural and functional MRI of the brain, and brain PET amyloid imaging, are limited by their inability to specifically diagnose (“rule in”) the disease, or because they are invasive and can be expensive. “Blood-based biomarkers would be a great and useful option—more accessible, less invasive, easier to gather and less expensive to process. Several are under development for preclinical Alzheimer’s disease. More research investment in this area is urgently needed,” she said.

See How It Works: A Map of the Human Brain »

Emphasizing that the research results, while intriguing, are preliminary, Carrillo said, “We are in the early days of discovery in this area of Alzheimer's research. Much more research is needed in this exciting and potentially useful area—on this panel, on the others that are in development, and on uncovering and verifying additional blood-based biomarkers for Alzheimer's."