The All of Us Research Program's wearables dataset

Area of Science:

• Digital health technologies
• Wearable sensors
• Real-world data in research

Background:

• Digital health technologies (DHTs) offer continuous health monitoring and disease detection capabilities.
• The All of Us Research Program aims to collect comprehensive health data from diverse populations.
• Characterizing large DHT datasets is crucial for advancing medical research methodologies.

Purpose of the Study:

• To characterize one of the largest and most demographically diverse digital health technology datasets.
• To describe the Fitbit data collected within the All of Us Research Program.
• To highlight the potential of this dataset for studying relationships between digital health metrics and clinical outcomes.

Main Methods:

• Utilized Fitbit data from over 59,000 participants in the All of Us Research Program.
• Collected data spanning 14 years, including over 39 million step and 31 million sleep observations.
• Integrated Fitbit data with electronic health records, physical measurements, genomics, and survey data for 46% of participants.

Main Results:

• The dataset is one of the largest and most demographically rich DHT datasets available.
• It encompasses extensive step and sleep data, providing granular insights into participant activity and rest.
• Significant overlap exists between DHT data and other comprehensive health information, enabling multi-modal analyses.

Conclusions:

• This large, demographically diverse DHT dataset is a valuable resource for medical research.
• The multi-modal data linkage facilitates novel studies on digital health metrics and clinical outcomes.
• The findings support the advancement of DHT methodologies through large-scale, real-world data utilization.