Smartwatches Could Predict Parkinson's Disease

John Lister's picture

Researchers say smartwatches could predict people getting Parkinson's disease up to seven years before they developed symptoms. The next steps could be as much an ethical challenge as a technical one.

The finding came from researchers who looked at data from the UK's Biobank project. That's a database that tracks anonymized health data from half a million volunteers. The idea is that the sheer amount of data makes it easier to spot patterns and connections between different health factors and conditions.

As part of the project, just over 100,000 people agreed to wear a smart watch for a seven-day period at some point between 2013 and 2016. The watches tracked the speed of their movement. (Source: theengineer.co.uk)

Comparative Study

The researchers compared data between three groups: those who had Parkinson's at the time they wore the watch, those who have since been diagnosed with it, and those who have been diagnosed with it. They then used machine learning to try to find patterns in the data.

The resulting model, had it been used at the time the data was collected, would have predicted which people would go on to be diagnosed with Parkinson's more accurately than any risk factor. It could also have given a good forecast of how long it would be before they were diagnosed. (Source: nature.com)

The researchers note that repeating the study and getting similar results would give them more confidence, but that would require another data source (which doesn't appear to exist) or starting and waiting for a long-term study.

No Cure as of Yet

Parkinson's disease doesn't currently have a cure and is diagnosed through symptoms, while treatment is instead about managing symptoms. That means that if somebody was forecast as more likely to develop the disease, there wouldn't be anything they could do immediately.

That could create an ethical dilemma about whether they should (or would want to) know. However, the researchers suggest such people could effectively be "on standby" for clinical trials or early-stage treatments that become available in the future.

The findings also highlight the importance of data privacy. For example, smartwatch manufacturers and software developers having the right to share such data with health or life insurers could be a major ethical concern.

What's Your Opinion?

Are you surprised by the findings? Would you be happy to have your smartwatch data analyzed for medical risks? Would you want to know about your chances of developing a condition that doesn't have a cure?

Rate this article: 
Average: 4.8 (8 votes)