Diagnosing dementia

A key in issue the fight against Alzheimer’s disease is accurate identification of the disease. In many diseases, like chickenpox or COVID, there are unambiguous clinical signs or reasonably accurate diagnostic tests. But the main sign of Alzheimer’s, memory loss, can also be part of normal aging. Other conditions and dementias can also interfere with memory function, making clear identification tricky. While there is still much work to be done, recent advances in Alzheimer’s disease testing reflect improved test performance and a reduction in the invasiveness of the tests.

Standard practice for Alzheimer’s disease diagnosis involves:

1) Asking the patient and the caregiver for a medical history, including duration of symptoms, past medical problems, current medication use, diet and lifestyle, ability to carry out activities of daily living and changes in personality or behavior.

2) Administering a psychiatric evaluation for depression or other mental health conditions.

3) Ordering blood, urine and other standard lab tests to identify other possible causes of the problem.

4) Performing a brain scan to support an Alzheimer’s diagnosis and rule out other possible causes (https://www.nia.nih.gov/health/how-alzheimers-diseasediagnosed).

The focus is on ruling out other contributors to memory loss rather than positively identifying Alzheimer’s disease.

Several new tests look promising for the sensitive and specific identification of Alzheimer’s. The terms “sensitive” and “specific” in research reflect how well a test works. “Sensitivity” is used to describe how well a test identifies people who have a disease. At the same time, specificity refers to how well a test identifies people who do not have a disease. It is also important, both clinically and in research, to ensure that sick people don’t get a negative test result (false negative), and that well people don’t get a positive result (false positive).

The most data on disease identification comes from tests of the proteins that make up plaques (Beta Amyloid) and tangles (Tau), which are hallmark signs of the disease. These proteins are identified using positron emission tomography (PET), a brain imaging technology in which substances containing radioactive isotopes are introduced into the body, allowing localization of physiological processes.

A study of Beta Amyloid PET showed a low false negative but a high false positive rate. Longitudinal data correctly classified 89% of participants who progressed to AD but only 58% of participants who did not progress to AD (Amivid; Martinez et al., 2017). However, more recent studies using blood samples showed promising results using a less invasive technique. Plasma B-amyloid showed specificity/sensitivity of 91% and 71%, respectively, with an overall diagnostic accuracy of 86%. The biomarker-positive group was 7.9 times more likely to develop clinical AD (Nabers et al., 2018).

Tau testing, specifically the optimization of the radioactive isotope, is in an early stage of development. A recent Tau study identified mild cognitive impairment with a sensitivity of 38% and specificity of 95% (Leuzy et al., 2020). A positive Tau test was strongly indicative of Alzheimer’s disease, but a negative test did not exclude the diagnosis.

Another classic sign of Alzheimer’s disease is brain atrophy, or shrinkage, due to cell death. Atrophy is measured using magnetic resonance imaging (MRI). This medical imaging technique uses a magnetic field and computergenerated radio waves to create detailed images of the body’s tissues. A new study used artificial intelligence modeling methods to identify early Alzheimer’s in simple brain scans. The model identified a handful of brain size and shape parameters from hundreds of possibilities that were optimal for discrimination. They reported 98% accuracy even when the sample included other diseases such as Parkinson’s and frontotemporal dementia (Inglese et al., 2022).

In addition to tests for classic signs of AD, there are tests of new emerging targets in various stages of development. Cellular energy production, specifically the function of mitochondria, the powerhouse of the cell, has been targeted using PET and blood samples. Neuroinflammation, or activation of the brain’s innate immune system following injury or disease, such as accumulation of Beta Amyloid, is also a new target. Both are in the initial stages of development.

At LSU Health Shreveport, our team is working on a test based on the cardiovascular contribution to Alzheimer’s disease. There is growing evidence that high cholesterol, high blood pressure and diabetes increase the risk of the disease, particularly in the preliminary stages. We are studying a blood test for a molecule called hydrogen sulfide, which is important for signaling in the cardiovascular and nervous systems. We have shown that hydrogen sulfide levels are increased when cognitive function decreases. Hydrogen sulfide was associated with cognitive performance and an M R I - based measure of vascular disease in the brain (Disbrow et al., 2021).

Our blood test showed a sensitivity of 80% and a specificity of 98%.

Even with these advances, none of these tests are 100% sensitive and specific. We are advancing rapidly, but our ability to measure our targets is not perfect. We may be including more than a single disease in our current definition of “Alzheimer’s disease,” so our quest for better identification will likely improve the precision of our description of the disease and the tests we are using.

To learn more about the Center for Brain Health, visit www.lsuhs.edu/cbh.

Elizabeth Disbrow, Ph.D., professor, Department of Neurology, and director, Center for Brain Health at LSU Health Shreveport.