Passive Smartphone Data as a Predictor of Five-Year Mortality Risk
Researchers at the University of Illinois at Urbana-Champaign have demonstrated that simply tracking a person’s walking activity through the sensors in ordinary smartphones can forecast the likelihood of death within five years. The findings appeared in PLOS Digital Health, underscoring a shift toward using everyday digital behavior as a health signal.
In the study, data were drawn from a large cohort of 100,000 participants drawn from the UK Biobank, a long-term health study that follows volunteers over time. Participants wore their activity monitors in their pockets for a full week, providing a continuous stream of sensor information. This data captured daily walking behavior, including the six-minute walk distance and the overall pace of movement, which researchers could translate into a practical estimate of walking speed across days rather than at a single moment in time.
When combined with basic demographic information commonly collected in health assessments, the researchers were able to estimate the probability of mortality within five years with roughly 70 percent accuracy. Importantly, this level of predictive power held across different ages and both sexes, suggesting that the signal from everyday mobility patterns carries information about health risk that complements traditional risk factors.
Traditional methods for assessing walking speed rely on direct measurements taken during a clinical visit or a controlled study. Such approaches require an active involvement of a clinician or researcher and a person to come in for testing. In contrast, passive data collection uses wearable devices or smartphones already in use in daily life. This method can enable ongoing monitoring of population health at a national or regional level, offering a scalable way to detect shifts in mortality risk and intervene early where needed.
Beyond the immediate research findings, the study raises questions about how continuous mobility data can inform public health strategies. If validated in diverse settings and populations, passive walking metrics could help health systems identify communities with rising risk, tailor prevention efforts, and track the impact of interventions over time. The approach also invites consideration of data privacy, consent, and responsible data sharing as ubiquitous digital devices become more integrated into health surveillance.
Overall, the work illustrates that seemingly simple daily activities, like walking in daily routines, can reveal meaningful insights about long-term health. The use of passive smartphone sensing to gauge mortality risk represents a promising avenue for scalable health monitoring that could be adapted for use in Canada and the United States, where population health management continues to embrace digital tools for proactive care.