More than one in five people in the United States have a smartwatch or fitness tracker, like a Fitbit device. Beyond helping individuals monitor their own heart rate, sleep, and physical activity, these devices offer tremendous potential to advance our collective understanding of health through research.
Until recent years, data from these devices have been difficult for most researchers to obtain. But that’s changing, thanks in part to All of Us. Program participants have the opportunity to contribute Fitbit data along with other kinds of information for research. The program strips out personal identifiers and adds the data to the Researcher Workbench for use by registered researchers. Today, the platform houses Fitbit data from more than 15,600 participants, making it the largest, broadly accessible set of digital health technology data to date–and the dataset is still growing. Already, these data are revealing new insights about health and disease, with more findings on the horizon.
Wearables Data Offer a Holistic View of Health
All of Us collects participant data from multiple sources, such as survey responses, electronic health records, and blood samples. These sources provide snapshots of information that tell us about a person’s health at specific moments in time. Through the Workbench, researchers can combine these data points with Fitbit’s continuous stream of data for a more holistic view of health. For example, researchers can use Fitbit data to examine the relationships between physical activity, psychiatric diagnoses, and genomic data.
“Combining Fitbit data with other data types provided by the program offers rich and nuanced insights, allowing us to explore and answer different kinds of research questions,” said Hiral Master, P.T., Ph.D., M.P.H., a senior scientific project manager for the All of Us Data and Resource Center (DRC) at Vanderbilt University Medical Center.
The types of data available and the size of the dataset make All of Us unique. “This data resource combines rich biomedical data with physical activity monitoring that is more extensive and detailed than any previous dataset. This will allow researchers to start asking questions about activity and health that couldn't have been asked before," explained Amy Price, Ph.D., a neuroscientist in the Division of Medical and Scientific Research at All of Us.
Promising Research With Fitbit Data
In the two and a half years since the Fitbit data became available, hundreds of researchers have already started using it. As of July 2023, about a quarter of the 5,771 projects on the Workbench used Fitbit data. In addition, Fitbit data from All of Us participants have been used in several peer-reviewed publications, including in major scientific journals, such as Epidemiology, Nature Medicine, and JAMA Network Open.
In one widely publicized article, Master and her colleagues found that walking more than 8,200 steps a day was associated with reduced risk of chronic diseases, including obesity, diabetes, high blood pressure, sleep apnea, gastroesophageal reflux disease, and major depressive disorder.
“Walking more was associated with a lower risk of diabetes and cardiovascular disease. This was consistent with findings from earlier studies. We were also able to explore some previously unidentified activity-disease associations. Specifically, we found that daily step counts were associated with a lower risk of depression, gastroesophageal reflux disease, or sleep apnea, which was novel,” Master said. “We were thus able to find evidence of the impact of daily steps on a wide range of conditions.”
More studies could be published in the coming years to further our knowledge about other conditions. Price speculates that Fitbit data could be used, among other things, to understand how diseases like Alzheimer's progress. “In Alzheimer’s and frontotemporal dementia, doctors receive reports that their patients don’t want to go out as much, even in early stages. These are usually self-reports provided by a caregiver,” Price said. “With Fitbit, we have actual data about participants’ daily activity patterns, so we can start to examine the relationship between physical activity and neurological diagnoses.”
Growing the Fitbit Dataset
The wearable data currently available in the Workbench come from individuals who already owned a Fitbit device. This group of participants is about 80% White, and only 6% are Hispanic or Latino. The lack of diversity of this sample limits how easily findings can be generalized to other groups. To expand the number and diversity of participants with Fitbit data, the program has created the WEAR study, which provides free Fitbit devices to invited participants from underrepresented communities.
Future data releases will include Fitbit data from even more participants, as well as additional Fitbit data types. Data from additional types of wearable devices may also be available in the future.
Getting Started With Fitbit Data
For now, Master encourages researchers to explore the trove of existing Fitbit data in the Workbench. They can learn from other researchers by exploring projects underway in the Research Projects Directory and team up with colleagues at their home institution or other institutions on the cloud-based platform.
For those ready to start using the data, she notes that the DRC offers a number of resources on the publicly accessible User Support Hub. “Our goal at the DRC is to facilitate and empower researchers. I know there are a lot of smart people out there. We want to make sure everyone can contribute to accelerating health discoveries,” Master said.
As of April 2024, new types of Fitbit device data, such as the type of device, approximate battery level, and more, are now available in the Workbench. Visit the User Support Hub to learn more.
This article appears in the August 2023 issue of All of Us Research Roundup. Subscribe to receive future issues of the bimonthly researcher newsletter.
View the full August edition of the All of Us Research Roundup here.
All of Us is a registered service mark of the U.S. Department of Health & Human Services (HHS).
