According to researchers at the Washington University School of Medicine in St. Louis, they’ve come up with an algorithm that can predict Alzheimer’s disease and its progress.

The researchers use brain scans to determine your Alzheimer’s risk and, if the disease has already begun, how much time you have left until your memory and sense of self start to disappear.

If you’re not familiar with algorithms, they’re processes and formulas computers use in their programming to solve problems or crunch data so that the program yields practical information.

For example, when you browse the Internet, algorithms often decide what ads you get to see online. It does this by taking your data – the kinds of websites you like to look at – and translating it into what kinds of advertising is most likely to appeal to you.

For this latest study, the researchers first compiled data from brain scans that measured levels of amyloid beta in older people’s brains. Amyloid is a protein that often accumulates early in the progress of Alzheimer’s disease – often before memory loss or other brain problems begin.

The scientists then tracked the brain health of these people to show how the growing amount of amyloid in brain scans paralleled any memory issues that evolved.1 This resulted in the algorithm that, they say, can predict the onset of Alzheimer’s disease.

But does it really work?

Looking Inside the Brain 

The study that gave rise to this new algorithm involved an analysis of brain scans taken of 236 people who were part of Alzheimer’s studies at Washington University. The people’s average age at the start of the research was 67 years old. All of the participants underwent two or more brain scans that, on average, were 4 1/2 years apart.

The scans used were PET scans – positron emission tomography scans. This type of imaging uses radioactive dyes to take a picture of the brain and can reveal the presence of amyloid protein.

As part of the investigation, the researchers included data from more than a thousand clinical assessments of 180 study subjects which included information about their cognitive health.

According to researcher Suzanne Schindler, M.D., PhD, the study showed that there is a “tipping point” in the buildup of amyloid in the brain that signals a relatively rapid further accumulation of amyloid plaque and full-blown Alzheimer’s disease. And after you reach that tipping point, you can predict how the disease will progress.

“You may hit the tipping point when you’re 50; it may happen when you’re 80; it may never happen,” Dr. Schindler says. “But once you pass the tipping point, you’re going to accumulate high levels of amyloid that are likely to cause dementia. If we know how much amyloid someone has right now, we can calculate how long ago they hit the tipping point and estimate how much longer it will be until they are likely to develop symptoms.”

Is this the Fool Proof Test We’ve Been Waiting For? 

I have to tell you that what Dr. Schindler says gives me pause. First, she notes that accumulating high levels of amyloid is “likely” to cause dementia. That word right there indicates that this algorithm is not completely reliable – not yet anyway. Over and over again it’s been shown that a lot of people carry around amyloid beta in their head and never develop Alzheimer’s disease. Whether that’s because they haven’t hit the “tipping point” yet is a possibility.

In an online discussion of this study, Murat Bilgel, PhD, a staff scientist with the National Institute on Aging, points out that “It is important to remember that not every individual who has elevated amyloid levels develops cognitive impairment or dementia.

“Further analyses are necessary to provide insight into who among the amyloid accumulators might develop future cognitive impairment.” 2

Another issue is the relatively small size of the study. The main part of the study only included 236 people. That’s not a huge sample size. So the same results might not show up in a larger group of study subjects.

Still, this research provides another potential step towards developing ways to predict Alzheimer’s disease and potentially treat it before serious – and intractable – brain injury occurs. And that is welcome news.


  1. https://pubmed.ncbi.nlm.nih.gov/34504028/ 
  2. https://www.alzforum.org/news/research-news/can-single-amyloid-pet-scan-predict-time-
    symptom-onset