MONTREAL – Smart watches could detect Parkinson’s disease several years before the first symptoms appear by analyzing the user’s movements, an experiment conducted in the United Kingdom shows.
Researchers from the University of Cardiff in Wales used artificial intelligence to analyze data from around 105,000 smartwatch users.
By measuring the speed of their movements during a single week between 2013 and 2016, they were able to predict which users would eventually develop Parkinson’s disease. They were also able to distinguish Parkinson’s from other diseases that might affect the user’s movements.
However, the authors of the study concede that further experiments will be necessary to verify the validity of their results.
Parkinson’s disease begins to take root in the patient’s brain several years before the onset of symptoms and is well advanced when we finally realize something is wrong.
It could therefore be beneficial to detect it as soon as possible to limit the damage… but not necessarily today, warns Professor Louis-Éric Trudeau, a specialist at the University of Montreal.
“Currently, there is no therapy that prevents neuron death,” he said. So even if we tell a patient that in seven years’ time they have a good chance of developing Parkinson’s disease, that won’t open the door for us to treatments that might fix it.
He therefore wonders how important it is to tell a patient what to expect in a few years’ time, while at the same time admitting that there is nothing that can be done at the moment. The information could be very interesting for the insurers, but maybe a little less for the patient, he adds.
Mr. Trudeau compares the situation to websites that offer a full analysis of our genetic profile to produce a report that informs us of our risk of suffering from such a disease or health condition.
Whether he knows it or not should be entirely up to the patient, he said. There are people who just don’t want to know. And among those who want to know will be those ill-equipped to deal with the response and in need of support.
In his opinion, the use of smartwatches as a screening tool for Parkinson’s disease would be more relevant in a research context.
“As part of a clinical trial, we could use such an approach to characterize our participants and already plan to better track disease progression,” he explained.
“And among our controls, who don’t usually get the disease, there may be some who develop it. This is all information that is really important for clinical research.”
The results of this study were published in the journal Nature Medicine.