Real-time flu prediction using Fitbit data

Heart rate and sleep data from Fitbits could be used to track influenza in real time, according to new research. Scientists used de-identified sensor data from nearly 50,000 Fitbit users in the five states with the most users. Previous research has found that people's heart rates tend to spike ahead of infectious illness. In this study, researchers found that weeks when users’ resting heart was above average and sleep levels did not dip below normal positively correlated with the CDC’s state-level estimates predicting flu activity. Adding in the Fitbit data to the CDC’s prediction method increased its accuracy. Although the research was only conducted in five states and the prediction model needs to be validated, researchers suggest the Fitbit data could help with better public health models for disease.