New blood test for Alzheimer's disease?

• Test looks for levels of 10 blood fats
• Identifying such patients could help with earlier intervention
• Test still needs to be tried in broader clinical trials

It may soon be possible to predict who will get Alzheimer's disease at least a few years before the start of symptoms, according to a paper published this week in Nature Medicine.

The paper identified levels of 10 fats seen in the blood of people who went on to develop Alzheimer's 2-3 years later. The blood test is the first to predict Alzheimer's before its characteristic memory loss.

The test was accurate about 90% of the time in distinguishing people with healthy brains from those with the fatal disease, said Howard Federoff, a professor of neurology and executive vice president for health sciences at Georgetown University Medical Center, who led the work.

A reliable test that found no trace of Alzheimer's might let people breathe easier the next time they lost their keys; a positive test years in advance would be heartbreaking, but at least would allow the person to prepare while they still have the capacity to do so, said Robert Stern, a professor at the Boston University School of Medicine who directs clinical research at the school's Alzheimer's Disease Center.

There's a chance — though still unproven — that lifestyle changes might help someone avoid or delay onset of the disease, he said. (Exercising, eating a Mediterranean diet and being socially active have all been linked to lower rates of Alzheimer's.)

Predictive tests are also hugely important for research.

In recent years, experimental drugs have repeatedly failed to show a significant effect on the course or pace of Alzheimer's. If researchers could identify likely victims years or a decade ahead of symptoms, they would gain new insights into how the disease wreaks havoc on the brain, and find drugs more likely to be effective against it, Stern said.

To develop the test, researchers took blood samples from 525 people age 70 and older and followed them over time, as some of them developed memory loss. By comparing the blood of 50 participants with memory loss with the blood of 50 who did not develop the disease, researchers were able to identify 10 fats that helped distinguish them. They then validated the test in another 40 participants, looking for those fat levels, Federoff said.

The test needs to be more widely studied before it can be used outside of clinical trials, he said. Because most of the study participants were white, he doesn't yet know whether it will be as effective in other racial and ethnic groups; or how early the test can be used.

He has also found some genetic differences in the blood that he plans to report on in a few months, he said. And he will be looking at other large data sets to see if he can find these telltale markers in people much younger than the ones in his study.

Other tests are also expected over the next few years — and there's no doubt they're needed, said Maria Carrillo, vice president of medical and scientific relations for the Alzheimer's Association, a research and advocacy group.

"This field needs better methods to detect Alzheimer's disease at its earliest time point, to allow us to intervene with potential prevention strategies," Carrillo said.