Can smartphones predict mortality risk?
Passive smartphone monitoring of people’s walking activity can be used to build population-level models of health risk and mortality, a recent study finds Digital Health PLOS study. Read our summary of findings from Bruce Schatz of the University of Illinois at Urbana-Champaign, USA, and colleagues below or skip to complete article in Digital Health PLOS.
Previous studies have used fitness measures, including walk tests and self-reported walking pace, to predict individual mortality risk. These measures focus on quality rather than quantity of movement; measuring an individual’s walking speed has become common practice in some clinical settings, for example. The rise of passive monitoring of smartphone activity opens up the possibility of population-level analyzes using similar metrics.
Study design and results
The researchers studied 100,000 participants from the UK Biobank National Cohort who wore activity monitors with motion sensors for 1 week. Although the wrist sensor is worn differently than how smartphone sensors are worn, their motion sensors can both be used to extract walking intensity information from short bursts of walking – one version daily walk test.
The team was able to successfully validate predictive mortality risk models using just 6 minutes per day of regular walking collected by the sensor, combined with traditional demographic characteristics. The equivalent of walking speed calculated from these passively collected data was a predictor of 5-year mortality independent of age and gender (combined C index 0.72). Predictive models only used walking intensity to simulate smartphone screens.
“Our results show that passive measurements with motion sensors can achieve similar accuracy to active measurements of walking speed and walking rhythm,” the authors state. “Our scalable methods offer a feasible route to national health risk screening.”
Schatz adds, “I spent a decade using inexpensive phones for clinical models of health status. These have now been tested on the largest national cohort to predict population-wide life expectancy.