Summary: A Swedish study questioned the accuracy of the proposed “Dubois criteria” for diagnosing Alzheimer’s. Professor Dubois says the criteria are very specific, and so will miss some cases, but will result in few false positive diagnoses.
Researchers are still working to validate the biomarkers included in the proposed criteria. If you are considering requesting one of these tests before researchers can confirm its accuracy, there are some practical considerations you may want to discuss with your doctor.
I wrote last month about a Swedish study questioning the accuracy of the proposed “Dubois criteria” for diagnosing Alzheimer’s. These criteria incorporate biomarkers such as results of imaging scans and spinal fluid tests.
Commenting on the Swedish study, Bruno Dubois, leader of the group of experts who developed the criteria, says other research has shown “a high accuracy of the new proposed criteria.” Professor Dubois is Director of the Memory and Alzheimer’s Institute at the Salpêtrière Hospital in Paris.
Accuracy is a complicated concept when it comes to medical tests, though, and few tests are 100 percent accurate. A test can be very sensitive – it will detect all people with a disease, but may also result in false positive results for some people. A test can also be very specific – it will miss some people with a disease, but will result in fewer false positive results. There’s often a trade-off between sensitivity and specificity.
“The paper you are referring to showed in fact a low sensitivity, which is not surprising given the high specificity of the criteria,” says Professor Dubois. “There is no doubt that these criteria have a very high specificity given the fact that we add reference to biomarkers.” Low sensitivity and high specificity (along with the different definitions of Alzheimer’s discussed in the previous post) would help explain why so few people diagnosed with Alzheimer’s under the current criteria met the biomarker-enhanced Dubois criteria in the Swedish study.
Should you request these new tests?
Even before scientists can validate these biomarkers, patients and families are asking their doctors to order these tests. Is this a good idea?
The panels convened by the Alzheimer’s Association and U.S. National Institute on Aging have proposed the use of similar biomarkers in research, but have called for “extensive testing” before they are widely used in doctors’ offices.
Professor Dubois himself says there should be limits to widespread use of the Dubois criteria and biomarkers for diagnosing Alzheimer’s. “These criteria apply in clinical settings, as long as clinicians use a good memory test and ask for an MRI that allows for an assessment of hippocampal volume,” he says. In his opinion, tests involving other biomarkers such as proteins in spinal fluid or scans to measure the amount of beta amyloid in the brain are not yet ready for use outside of research.
If you’re thinking about requesting one of these newer tests, there are some practical considerations you might discuss with your doctor:
- Will the test make my diagnosis more certain?
- Will my treatment change? What will happen if my test results indicate I have Alzheimer’s or am at increased risk?
- What will happen if my test results indicate I do NOT have Alzheimer’s, but I still have problems with memory and thinking?
- Will test results affect my insurance or employment opportunities?
- Who will pay for the test?
- What does the test involve? Are there any risks I should know about?
In his comments on my previous post, Richard Taylor pointed out that announcements about Alzheimer’s often end with “buyer beware” statements because of the uncertainties involved. These biomarker tests are valuable for research and if validated, should help with diagnosis. But if you’re thinking about asking your doctor to order one of these tests now, buyer beware!