That doesn't stop people from trying. Doctors are looking to create them, and patients are taking them, too, even though they know that existing biomarkers — for Alzheimer's disease, for instance — have serious limitations.

Six years ago, Robin Jones of Menlo Park, Calif., found himself in a parking lot. He had no idea where his car was. That was unlike him. A 67-year-old scientist who worked on nuclear energy plants, he was good with details. "It was, you know, I can't believe this," Jones says now. "I can always remember where the car is."

At work, he had a hard time finding the right words. He'd come home and tell his wife, Anne, "Something's different."

Robin Jones made an appointment at the Stanford Center for Memory Disorders, where he went through the standard cognitive evaluation for Alzheimer's disease — memorizing lists of words and so on.

Jones did pretty well on the tests. Better, says Stanford neurologist Michael Greicius, than you'd expect from someone in the early stages of the disease.

"He didn't fit neatly into this classic presentation of Alzheimer's disease," Greicus says, "and that's one of the reasons we talked fairly early on about looking at biomarkers."

When Greicius says "biomarkers," he's referring to a very new thing in Alzheimer's disease and a very old thing in many other diseases: objective lab tests. Those are physical measures that can help doctors understand what is going on in a patient's body.

Read the full story at NPR