Dead Man Walking? Smartphone Data May Predict Mortality Risk
Nov. 3, 2022 — Maybe you’re on a leisurely neighborhood stroll or roaming the aisles of a grocery store. Chances are, your smartphone’s along for the trip, too — perhaps as a podcast player or a digital security blanket.
But what if that phone could gather data from your everyday cardio activities to predict how long you’ll live?
There may not be an app for that just yet, but researchers from the University of Illinois laid the groundwork for the possibility in a study published recently in the journal PLOS Digital Health.
“It’s well known that people [who] move more — and move more vigorously — live longer,” says Bruce Schatz, PhD, an expert in medical informatics at the University of Illinois and a co-author of the study. “We ended up trying to see what you could tell from walking motion that had some medical significance.”
Schatz and his colleagues pulled data from more than 100,000 adults ages 45-79 in the UK Biobank, a biomedical database in the United Kingdom. Participants wore wrist sensors around the clock for a week while they went about their daily routines, and researchers reviewed data from 12 consecutive, 30-second walking intervals for each study participant.
The researchers analyzed participants’ walking intensity and used it to predict their risk of death every year over a 5-year period.
Because the data was collected from 2013 to 2015, the researchers were able to check the accuracy of estimates against death records. The team’s predictions closely matched participants’ actual mortality, although the model was slightly more accurate for the earlier years than at the 5-year mark.
“It’s not giving you, personally, ‘You have 5 minutes to live,’” Schatz says. Rather, “What’s the likelihood that you’ll die in 5 years, or in 2 years?”
However, if an app capable of predicting your death date becomes available, Larry Hernandez, of San Antonio, TX, will be ready to try it. The 42-year-old is a private health insurance advisor, and such technology could be an incentive for his clients to improve their fitness, he says.
But Hernandez is also familiar with tracking his own metrics. He’s lost 60 pounds since beginning a running regimen in 2015 and continues to log a 5K daily on his Apple Watch.
If “today’s activities or yesterday’s activities actually got me another, extra year of life,” Hernandez says, “that’d be awesome.”
Stepping Toward Universal Health Care
The wrist devices worn by participants had accelerometers, which are built into even the cheapest of smartphones. These motion sensors are key to making health information accessible to the masses, Schatz says.
Smartwatches and other wearable fitness trackers are becoming increasingly popular — about 1 in 5 U.S. adults regularly wear them, according to a 2019 Pew Research Center survey — but aren’t affordable for everyone. However, 97% of Americans own a cellphone and 85% own a smartphone, according to a 2021 estimate from Pew.
The practical possibilities of using the formula created by Schatz and his colleagues are vast. A hospital system, for example, could potentially monitor most of its patients at once through their smartphones, and be alerted to changes in their walking patterns that may indicate a medical problem — all without disrupting patients’ lives.
“It’s the population screening that’s significant. It’s the catching things early when you can still do something,” Schatz says. “There’s a real opportunity here to do something for large numbers of people.”
Vienna Williams, MPH, sees an opportunity for employers. As director of the International WELL Building Institute in New York City, she helps companies from Hilton to Uber prioritize employees’ well-being.
“Wearables and sensors, they help us to really understand modifiable behavior, and that’s where we have the opportunity to intervene,” Williams says, noting the institute already uses such technology to help clients understand employee health trends. “The most important question that these things help us answer is, where do we have room to change our behavior in ways that we know help our health in the long term?”
An app that could predict likelihood of death could also help eliminate health disparities simply by being accessible to everyone with a smartphone, regardless of socioeconomic status. Even in countries with emerging economies, such as Brazil and Indonesia, a median 45% of people own a smartphone, according to a 2018 Pew Research Center survey.
“The benefits of being physically active are not disputable,” says Jan Carney, MD, associate dean for public health and health policy at the University of Vermont Larner College of Medicine in Burlington. “But the rates of physical activity among the population [are] uneven.”
The work of Schatz and his colleagues contributes to the goal of health equity, Carney says.
“Making such a simple, practical technology, you can have a lot of people in a given community know what their own activity levels are,” she says.
Future studies should be more racially and ethnically diverse, Schatz says. Although study participants reflected the U.K. population, the majority were white. Schatz’s team plans to continue its research through the National Institutes of Health’s All of Us Research Program, which aims to enroll more than 1 million people.