Researchers at Stanford University have developed a blood test that, at least in a preliminary retrospective study, was highly accurate at determining who had Alzheimer's or was most at risk for it.
The researchers collected more than 200 stored blood samples that had been taken from people with mild cognitive impairment, Alzheimer's disease, or no cognitive impairment. They then analyzed 120 proteins in the blood samples, and found that a distinctive pattern of 18 proteins was useful in identifying the people with Alzheimer's. The proteins regulate normal functions, such as inflammation, that may go awry in Alzheimer's.
The researchers next investigated whether the proteins could be used to distinguish people with Alzheimer's, or at risk for it, in a batch of "blinded" blood samples, where they did not know the diagnosis until after making their prediction. The protein signature proved to be 89% accurate in distinguishing people with Alzheimer's from those without, and about 81% accurate in predicting which patients with mild cognitive impairment would develop Alzheimer's. This experimental blood test must be validated in other studies and by other researchers before its true predictive power is known.
Ray S, et al. "Classification and Prediction of Clinical Alzheimer's Diagnosis Based on Plasma Signaling Proteins," Nature Medicine (Oct. 14, 2007): Advance online publication.