Scientists Want to Know if Fitness Trackers Can Detect Covid-19 Early

Researchers have known for years that wearables could be useful for detecting illness. Now they’re exploring whether fitness devices could help track and even contain Covid-19.

As a runner, I live by my fitness tracker data — not just to record pace and distance, but to determine how I slept the night before, measure how recovered my body is from my last workout, and see how my training is progressing. I check my data every morning when I wake up, and after every run.

In January, I noticed that my resting heart rate had, out of nowhere, jumped 10 points. I knew that meant my body was working harder than normal, as if I were sick, although I didn’t have any typical cold symptoms, like a runny nose, congestion, or sneezing — just a tightness in my chest and slight cough that could easily be attributed to running in dry air and subfreezing Colorado temps. But I was convinced the change in my data meant something was wrong with me that might derail my marathon training.

After a physical exam, my doctor ruled out anything more serious than the common cold and recommended over-the-counter meds — plus ditching my watch for a few days to prevent me from getting caught in the minutiae of my metrics.

But it turns out, I may not have been totally off base obsessing over what those metrics meant about my well-being. Researchers have seen the potential in using wearables to track physiological issues and illnesses since the original tracker, Fitbit, debuted in 2009. Now they’re hoping to use the massive amounts of continuous data recorded by wearers to track and predict cases of Covid-19.

“Every single time someone got sick with a viral infection, we could pick up their heart rate increasing well before they were symptomatic.”

Last year, in a study of 109 people using wearables, researchers uncovered — using clinical measures, like blood and urine tests, in addition to heart rate and skin temperature data from the wearables — more than 67 previously undetected issues like infections, diabetes, lymphoma, heart defects, and even precancer, according to research published in the journal Nature. Using continuous heart rate data, scientists were able to diagnose sleep apnea in one patient; in others, heart rate changes were linked to infections. One of the study authors, Michael Snyder, PhD, a professor of genetics and director of the Center for Genomics and Personalized Medicine at Stanford, even noticed that his own heart rate had increased abnormally before an eventual Lyme disease diagnosis.

The heart rate findings prompted his team to devise an algorithm which detects variations in heart rate pattern that could alert the study participants to a possible infection. And it worked: “Every single time someone got sick with a viral infection, we could pick up their heart rate increasing well before they were symptomatic,” he says.

These days, he’s applying that algorithm to crowdsourced wearable device data submitted to his lab’s Personal Health Dashboard to analyze and predict potential Covid-19 cases. Given the virus’s longer incubation time, this asymptomatic detection could be especially informative. “Preliminary research shows an increased resting heart rate about eight days before someone reports being ill with Covid-19,” he says. The algorithm works with just two weeks of baseline data, and, within the next two weeks, Snyder hopes to start notifying participants if they have any physiological signals of illness — even if they’re not yet experiencing symptoms.

Jessilyn Dunn, PhD, an assistant professor of biomedical engineering at Duke University, was part of the research team (which also included Snyder) that linked an increase in resting heart rate and skin temperature recorded by wearables to the onset of issues like Lyme disease and diabetes in a 2017 study published in the journal PLOS Biology. Now, she’s helping to spearhead a study called Covidentify to track the spread of Covid-19.

Dunn’s team will use health information like heart rate, oxygen levels, sleep schedules, and activity levels recorded by the wearables’ sensors to develop their own algorithm that could detect early symptoms of Covid-19, she explains.

“If we could track a lot of people using these devices, we would start to learn where hotspots are and start to contain it before it spreads,” says Dunn. “Having a good monitoring and containment strategy in place before the reopening of even more of the country, which will very likely bring a second wave of this, will be critical,” she says.

One of the most crucial elements of these studies, Dunn says, is collaborating with companies to get these wearable devices (most of which aren’t cheap) into underserved communities and onto essential workers. “This is a way of bringing some level of health monitoring to disadvantaged populations,” says Snyder — especially those being disproportionately affected by Covid-19.

“If we could track a lot of people using these devices, we would start to learn where hotspots are and start to contain it before it spreads.”

Other researchers are also working hard to seek answers. In March, emergency medical workers at the University of California, San Francisco Medical Center and the Zuckerberg San Francisco General Hospital were asked to start wearing Oura Rings for three months in order to collect data on Covid-19 symptoms for a study called TemPredict. Scripps Research launched their own study called DETECT to track illness emergence patterns in March as well, building on their research from earlier this year that determined Fitbit data significantly improved flu-prediction models. And in April, fitness tracker WHOOP partnered with researchers from the Cleveland Clinic and Australia’s CQUniversity to determine whether there’s a link between WHOOP users’ respiratory rates and potential Covid-19 infections.

But what should youdo with this info? If you own a health tracking device, chances are you check it regularly to see your stats. Don’t immediately freak out if you see a change in one particular metric. These new studies are still in the preliminary stages, and “a lot of these devices have not been evaluated for medical use, so I wouldn’t recommend making a medical decision based only off of data from a wearable,” says Dunn.

That said, you can use them as a screening tool — kind of like using a thermometer to take your temperature — if something feels off. And in that case, “if it’s easy for you to contact a doctor to get a better picture of what’s going on with your health, it’s probably a good idea to do it,” says Dunn.

The more knowledge you have about your body, the more informed your health decisions will be. “I’m a total believer in self empowerment,” says Snyder. “But use common sense — you might be getting ill, or you might just be stressed.” If you’re obsessively checking your stats, take a little break from your device and focus on things like sleep, good nutrition, and gentle exercise — because worrying about every little number is not going to help.