Parkinson’s can often take years to diagnose as the symptoms—slow movement, involuntary shaking and stiff muscles, to name a few—can develop so slowly over time.

The disease is currently incurable, though if diagnosed early enough the quality of life of patients can be bolstered by medicine, therapies, and in some cases surgery.

This is the part of the problem a team of scientists at Cardiff University in Wales hope to have cracked.

The experts at the university’s Neuroscience and Mental Health Innovation Institute (NMHII) and the U.K. Dementia Research Institute analyzed the accelerometry—the acceleration of motion—in 103,712 smartwatch wearers using artificial intelligence models.

By tracking the speed of motion over the course of a week, the computer programs were able to identify not only patients who had already been diagnosed with Parkinson’s but also those who were in the early stages of the disease who had not yet been diagnosed.

The researchers were also able to pinpoint when a clinical diagnosis would be made, with early onset symptoms being identified up to seven years before that point.

Dr. Kathryn Peall, clinical senior lecturer in the NMHII, explained, “Parkinson’s disease is a progressive movement disorder caused by the loss of brain cells that use dopamine. However, by the time of clinical diagnosis, approximately 50%-70% of these brain cells will have been lost. This makes early diagnosis of the disease difficult.

“We know that as Parkinson’s disease develops, there are changes to the speed of movement, so we investigated whether accelerometry could work as a prodromal marker [early signs and symptoms] for Parkinson’s disease, and ultimately allow for earlier diagnosis.”

An “unprecedented” discovery

According to research from market and consumer insights site Statista, 210 million people will be using smartwatches this year.

That figure only goes up as time goes on. By 2027 the data predicts nearly 230 million people will be making the most of wearable technologies.

Combining this consumer demand with scientific insight could be a game changer, added Dr. Cynthia Sandor of Cardiff University’s Dementia Research Institute, who said tracking accelerometry gave unique insights into Parkinson’s but not any other disorders the team examined.

The scientists said the results were so “distinct” that they could not be confused with other diseases or aging.

Sandor continued, “It suggests that accelerometry could be used to identify those at elevated risk for Parkinson’s disease on an unprecedented scale.

“In a clinical setting, continuous or semi-continuous monitoring of individuals can’t be achieved because of time, cost, accessibility, and sensitivity. But smart devices capable of collecting accelerometer data are worn daily by millions of people.”

What’s next?

The paper, which was published in the journal Nature Medicine, tempers the discovery with the need for more research.

The article says that although the team sought to “mitigate” against any biases, they had only one dataset—from the UK Biobank, an in-depth health database of more than half a million people in the country—as it was the only one large enough in terms of scale and volume of data to run the computer programs.

Bias in medical transformation is an issue the World Health Organization has issued a warning on—particularly in relation to artificial intelligence models.

In May the organization wrote, “While WHO is enthusiastic about the appropriate use of technologies, including large language models, to support health care professionals, patients, researchers, and scientists, there is concern that caution that would normally be exercised for any new technology is not being exercised consistently with LLMs.

“This includes widespread adherence to key values of transparency, inclusion, public engagement, expert supervision, and rigorous evaluation.”

The team behind the Parkinson’s paper has called for other researchers to build on their findings, writing that if the limitations to their work are addressed then wearables and health-sensor devices have the “ability to transition medicine into a digital health era,” thus improving health care itself as well as reducing costs and increasing accessibility.