DECODING THE HYPE: Blood Test to Predict Alzheimer’s Disease?

Posted by The Evidence Blog on March 18, 2014

by Jill Shuman, Manager of the Hayes News Service and Registered Dietitian

No doubt you’ve heard the news: “A new blood test can predict whether you will develop Alzheimer’s disease.” As a relative of someone with Alzheimer’s disease, you’re likely relieved that there’s now “A blood test with astonishing accuracy to predict Alzheimer’s disease!” And as a clinician, you’re probably excited to learn that “A new blood test is a game changer in predicting who will develop Alzheimer’s disease.”

With the population living longer due to other medical interventions, it is estimated that 1 in 9 people over 65 years of age has Alzheimer’s disease (AD). The costs of caring for people with AD and other dementias will soar from an estimated $203 billion in 2014 to a projected $1.2 trillion per year by 2050. Is it no wonder, then, that such exuberant headlines create hope and hype for a disease that has been described as “a slow descent into madness?”

Recently published in the journal Nature Medicine, researchers described a set of blood biomarkers for AD that might make it possible to identify those at risk before they develop symptoms. Without having any predetermined biomarkers in mind, the research team recruited 525 adults who were at least 70 years of age; 46 of the recruits were diagnosed with AD when they signed up for the study. Another 28 went on to develop a type of cognitive impairment that often progresses to AD.

Researchers chose 53 of these 74 people and took blood samples to compare with 53 others in the study who had no cognitive deficits. This analysis allowed them to compare differences in certain blood components between the two groups that might predict the development of AD. The team eventually identified significantly higher or lower levels of 10 phospholipids in the blood samples of those with the cognitive failure. It is believed that these 10 lipids indicate the breakdown of neural cell membranes as AD accelerates.

To test their hypothesis further, the researchers then went back to 41 recruits—21 who had developed early AD over the course of the study and 20 who had not. Blinded as to which of the 41 people was in which group, the researchers were able to use the blood samples to correctly sort the recruits with more than 90% accuracy.

The idea of predicting AD isn’t new. It’s already possible to detect signs of the disease long before symptoms such as memory loss begin to appear. But the tests require either a spinal tap—which is painful—or an magnetic resonance imaging (MRI) scan, which is time consuming and expensive. So a blood test to predict AD would be an enormous breakthrough for both diagnosis and treatment.

So what’s the real story? This may be the real thing eventually—but not quite yet. The study was small. Most of the study recruits were white and they were all at least 70 years of age, so it’s not clear whether the results are generalizable to people of other races or age groups. The test still needs to be validated by other laboratories and in larger, more heterogeneous groups of people. And at this point, there is no commercial product available to test for the 10 biomarkers of interest, although a patent has been filed related to a technology of this type. Other data have suggested that biomarkers are often not as prognostic as suggested by early research results.

One important issue has to do with accuracy of the test among those that test positive. In the study, 5% of the patients converted from normal cognition to mild impairment or full-blown AD. If that percentage represents a real-world screening population, then the test would have a positive predictive value of just 35%—meaning that nearly 2 of 3 of the positive results would actually be false. That seems unacceptably high, given the anxiety and life decisions that a positive test is likely to generate.

While overly exuberant about their data, the researchers, themselves, consider their results a single—albeit important—step along a long road. “This is a major step toward the commercialization of a preclinical disease biomarker test that could be useful for large-scale screening to identify at-risk individuals.” They also intend to design a clinical trial that will use the markers to identify people at high risk for AD and to test a therapeutic agent that might delay or prevent the emergence of the disease.

Topics: Hayes Blog

Sign up to receive updates from our blog

Our latest articles

New Call-to-action