Your smartphone could be used to calculate your mortality risk.

A study of over 100,000 adults discovered that motion sensor data from just 6 minutes of walking accurately predicted five-year mortality risk as well as other leading methods.

Data collected by your smartphone while out walking may be sufficient to estimate your mortality risk for the next five years.  Shpakdm / Shutterstock
Data collected by your smartphone while out walking may be sufficient to estimate your mortality risk for the next five years.  Shpakdm / Shutterstock

Data from just 6 minutes of walking, collected by motion sensors in smartphones, may be sufficient to predict someone's risk of death in the next five years.

Previous studies estimated mortality risk by measuring daily physical activity levels with wearable motion sensors in devices such as fitness watches. Despite their growing popularity, smart watches and fitness trackers are still primarily worn by a wealthy minority.

Most people have smartphones with similar sensors, but calculating mortality risk from activity data is difficult because people don't carry their phones around all day, according to Bruce Schatz of the University of Illinois Urbana-Champaign.

Schatz and his colleagues examined data from 100,655 participants in the UK Biobank study, which has been collecting information on the health of middle-aged and senior adults in the UK for more than 15 years, to find an alternative predictor that is measurable with smartphones. Participants in that study wore motion sensors on their wrists for one week. Approximately 2% of the participants died over the next five years.

The researchers ran motion sensor and death data from about one-tenth of the participants through a machine learning model, which created an algorithm that estimated five-year mortality risk based on acceleration during a 6-minute walk.

"There's a very characteristic pattern for many diseases, specifically heart or lung diseases, where people slow down when they're out of breath and speed up again in short doses," Schatz says.

The model was then tested using data from other participants, and its c-index score - a metric commonly used in biostatistics to assess accuracy - was 0.72, which is comparable to other metrics for estimating life expectancies, such as daily physical activity or health risk questionnaires.

"This predictor is as effective as or more effective than traditional risk factors," says Ciprian Crainiceanu of Johns Hopkins University in Maryland.

While this study used wrist-worn motion sensors, smartphones can also measure acceleration during short walks, according to Schatz, who is currently planning a larger smartphone study. "If people carry phones around, you could do weekly or daily predictions, which you can't get any other way," he says.

PLoS Digital Health, DOI: 10.1371/journal.pdig.0000045


Font Size
lines height