Can Big Data Help Prevent Alzheimer’s Disease?

A wholly digital, 400,000-person study will gather health data via a smartphone app to identify risk factors for dementia

3 min read

Animated gif illustration of a memory test on the Savonix Alzheimer’s Disease Discovery (ASSIST) Study app
Gif: Savonix

During the Framingham Heart Study, a long-term research study initiated in 1948 that collected health data from thousands of people, researchers discovered that high cholesterol and elevated blood pressure increase one’s risk of heart disease. Thanks to that insight, at-risk individuals can reduce their chances of developing the condition by taking drugs to lower cholesterol and blood pressure.

Could the same be done for Alzheimer’s disease, a notoriously opaque and complex progressive brain disease?

This week, San Francisco-based Savonix, a digital health company, and Boston University School of Public Health announced the start of the Alzheimer’s Disease Discovery (ASSIST) Study, a virtual study that will gather health and lifestyle data from 400,000 people in the hopes of identifying lifestyle risk factors for Alzheimer’s.  

“There is ample reason to believe that lifestyle factors matter quite a bit for Alzheimer’s and other dementias, and there hasn’t been a big study that has shown that,” says study leader Sandro Galea, dean of the Boston University (BU) School of Public Health. While individuals have no control over their age or genetic risk factors for Alzheimer’s, lifestyle factors—such as sleep, nutrition, and exercise—are modifiable and therefore could point to clear interventions, he adds.

Inspired by the 2017 Apple Heart Study, a virtual study in which researchers gathered and analyzed heart rate data recorded by Apple Watches and iPhones from more than 400,000 participants, Mylea Charvat, CEO and founder of Savonix, decided a similar large-scale virtual study ought to be done for the brain.

“All we’ve done to date are really small, isolated studies,” says Charvat. “We need large, population studies that include sensitive, cognitive data. And you can’t do that with a $600-per-hour neuropsychologist going out and testing everybody.” So Savonix and BU decided to try it with an app.

Like the Apple Heart Study, the ASSIST study will gather purely digital data—there are no doctor’s visits, no brain scans. Instead, participants enroll online by signing a consent form and completing a short health history questionnaire. The team hopes to recruit 400,000 individuals from a diverse range of ages, genders, and ethnicities via digital ads and outreach efforts.

Once enrolled, participants are asked to download the Savonix Mobile app. In the app, users take a series of cognitive tests in areas such as memory, learning, and attention, based on established, validated neurocognitive tasks. The Savonix test is designed and run by neuropsychologists and tests multiple brain domains, says Charvat, a Stanford-trained neuropsychologist. The test is not FDA-cleared and is not considered a device by the FDA. Currently, the company sells the test to businesses such as pharmaceutical companies and clinicians, and has no plans to sell or market it directly to individuals, says Charvat.

In addition to taking the Savonix test, participants using iPhones are also asked to share Health App data, including exercise, nutrition, and sleep information gathered from phones and wearables. That’s a lot of sensitive health data, and there are privacy reasons to be concerned about sharing details about one’s health with an app. Yet the study organizers point out that unlike traditional health apps, this is a research study, so data handling is supervised by two Institutional Review Boards, and all user information will be encrypted and fragmented among secure servers. The company will not sell user’s data, says Charvat.

Anyone who is age 22 or older can participate, and the first stage of the study takes about 45 minutes to complete. Participants will be contacted again in 18 months for a follow-up quiz, which takes about 15 minutes—putting the total participation time at 1 hour.

After the health and lifestyle data is gathered, it will be analyzed to try to determine how risk factors overlap or cluster to produce cognitive changes. Galea suspects there won’t be one single risk factor for Alzheimer’s, but rather a constellation of factors that increase one’s risk of developing the disease. Understanding that constellation could lead to drugs or treatments to prevent the disease, similar to how cholesterol-lowering statins are reducing rates of heart disease.

For now, the ASSIST study is funded for a three-year period, but Charvat and Galea hope it will evolve into a long-term study, like the Framingham Heart Study. “There is a lot of opportunity for this to become a study that is longer and larger,” says Galea. “Absolutely.”

The Conversation (0)