Smartphones with fitness trackers are capable of predicting a person’s risk of dying over the next five years, according to a recent study. The scientist at the University of Illinois of Urbana-Champaign revealed that this technology may unlock more opportunities for better and easier health screenings.
According to the researchers, measuring walking intensity is possible. However, the ability of a wearable device, like a smartwatch, worn for 24 hours to measure overall activity is not. The researchers were curious about whether motion sensors embedded in smartphones may be enough to provide correct prediction data.
The study published in PLOS Digital Health looked at data from over 100,000 participants in the UK Biobank who wore activity monitors on their wrists with motion sensors for 7days. Bruce Schatz and his team used motion sensors to collect data on short stretches of motion intensity.
Importantly, this is the first study on the dataset involving the largest national cohort. According to scientists, short strides of motion intensity data obtained from these fitness trackers are equivalent to data that can be gotten by a smartphone in an individual’s pocket.
The research team was able to recognize the demographic characteristics coupled with predictive models of death risk using only six minutes a day of steady walking collected by the sensor. With the help of the passively collected data on walking speed, the researchers were able to predict accurately the risk of death over a five-year time frame—not dependent on sex and age. The predictive model used only walking to stimulate smartphone monitors.
While previous studies have looked at similar walking metrics to examine mortality risk. This study, however, focused on the quality of an individual’s walk rather than the number of walking movements. In addition to that, their findings in previous work showed sensors in carried smartphones could digitally model physical distance and oxygen saturation during a six minutes walk test.
For the results, the team found that passive measures with motion sensors can achieve a 5-year death risk estimate of about 70 percent accuracy similar to wearable devices in active measures of walking speed and walking pace.
Their findings add to the growing body of research looking at several ways to predict death risks. “Our scalable methods offer a feasible pathway towards national screening for health risk”, the authors write in the study.
Above all, the study authors aim to conduct more research on a diverse population to further explain health risks in detail and ensure the predictive model is more accurate across a wide variety of participants. Likewise, describe potential health risks without interfering with people’s day-to-day activities.