Google Research builds wearable-based health age and CKM risk framework

Google Research introduced an investigational framework that turns consumer wearable data into a "coachable" health age and simulates personalized 12-week "what-if" health trajectories. Its SensorWM world model was pretrained on 2.28 million person-weeks of de-identified Fitbit data from roughly 100,000 users and forecasts 20 physiological markers under different step, sleep and cardio routines. In offline evaluations using retrospective research cohorts, lifestyle coaches chose the counterfactual-grounded plan as best overall in 71.4% of evaluations, and physicians rated 99.2% of plans within safe exercise intensity limits. Google Research says the framework could be extended to combine wearable streams with lab biomarkers when routine blood tests are available, to detect early subclinical strain across heart, kidney and metabolic systems. The authors state the models are investigational research prototypes, not clinical diagnostic tools, and that the evaluations were offline; they describe the work as a foundation for future prospective trials of interactive coaching.

Generated from reports · updated 20 h ago

Not enough continuous observations yet to draw a trend.