Treating Alzheimer's disease early can delay symptoms and improve people's quality of life, and might even change the course of the disease, but it needs an accurate test. In 2007, it looked like we had one--a paper in Nature Medicine seemed to show that an 18-strong panel of biomarkers could identify Alzheimer's disease with pretty high accuracy and pick out the cases of mild cognitive impairment (MCI) that would lead to Alzheimer's disease. Recently, however, a team of Swedish researchers looked at the same panel of biomarkers, and found that they could not reproduce the results.
In the study, published in PLoS One, the researchers used the panel to screen people with Alzheimer's disease, depression or other types of dementia, and also looked at healthy controls. While the panel could separate out the people with Alzheimer's disease from the healthy controls (though only with low accuracy), it could not tell the difference between Alzheimer's disease and other dementias.
As the researchers said in the paper, these results show just how important it is to carry out independent validation of results in explorative biomarker studies, using different sets of patients and different laboratory techniques.
Testing cerebrospinal fluid is a fairly accurate test for Alzheimer's disease, but a lumbar puncture is uncomfortable and invasive. A simple blood test for Alzheimer's is one of the holy grails of diagnostics research, but at the moment it still seems a few steps away.
- see the article in PLoS One
- check out the abstract in Nature Medicine