iHub offers an elective rotation for residents, who exemplify the next generation of innovation in our teaching hospital. They are on the front lines of patient care, are digitally savvy, and come to us with diverse backgrounds and interests. Their perspective is invaluable to our work, and our Resident Perspective blog series is a place for them to share their thoughts on digital health.
In the last few years, fitness trackers have permeated our society, with nearly everyone you pass wearing an Apple Watch, Fitbit or some other fitness tracker.
As a physician, I often recommend that patients increase their exercise and improve their diet, while occasionally suggesting the use of fitness trackers to help them monitor their activity. But keeping in mind my training in evidence-based medicine, I have questions.
How accurate are they? Is there evidence to support their use beyond just lifestyle and personal style?
Upon digging deeper into the subject, I found that accuracy is variable from device to device, and largely unknown due to the lack of formal studies. A comparison of devices done by CNET in 2015 showed Apple Watch, when calibrated, is the most accurate for distance traveled with 0.3% deviation. However, Apple Watch, when not calibrated, had up to 10% deviation, one of the highest in the group. Even with the same device, there is still a wide margin of error.
A formal study published in the Journal of Personalized Medicine in 2017 compared seven different devices for measuring heart rate and energy expenditure for walking, running, and cycling. In terms of margin of error, the devices were most accurate for cycling and least accurate for walking. The devices were more accurate for heart rate than energy expenditure: best performers while walking were the Apple Watch (2.5% median error), PulseOn (4.9% median error), and Microsoft Band (5.6% median error). Error for energy expenditure was much higher, and not within acceptable range as defined by the investigators. Median error rates ranged from 27.4% for Fitbit Surge to 92.6% for PulseOn. Overall, the study concluded Apple Watch had the most favorable error profile while the Samsung Gear S2 and PulseOn had the least favorable error profile.
According to a post by MobiHealthNews, some of these activity tracking devices have gone to the “graveyard.” The article notes that the rate-limiting step is the difficulty in changing people’s behavior to form new habits. As written in the post, Apple is trying to innovate new methods to motivate users with supportive rather than competition-focused community sharing, activity rings, personalized “nudges” and monthly challenges.
data-animation-override>
“I don’t personally own a fitness tracker. My number one reason is cost.”
When it comes to the impact of fitness trackers on health outcomes, research in this area is still emerging. The main disease states for which activity trackers may be helpful are those where exercise and weight loss have shown benefits in improving outcomes, such as heart disease and diabetes. However, a patient page in JAMA Cardiology stated that no long-term studies have yet been completed to test fitness trackers and their relationship with the prevention or treatment of heart disease.
The devices’ cost and notorious reputation for accuracy (especially during more intensive exercise) are additional limiting factors for a physician recommending their use to patients.
I don’t personally own a fitness tracker. My number one reason is cost. The Apple Watch, which from my research above is the most accurate, ranges from $250 to $600. And even though the accuracy is one of the best, it is still so variable and not validated very well. The step count can be variable and the estimation of calories is probably even less accurate, since most of this depends on genetics and the user’s own personal effort, and is not just predetermined for a certain activity someone does.
Fitness trackers may help in encouraging regular exercise and weight loss, but long-term health benefits remain unknown and require habitual use of the device. Fitness and activity trackers, to me as a resident, are an interactive way to motivate people to exercise and provide reminders to get up and walk around. If they work for someone in that way, then that is great. But if not, what will these devices change?
A survey by Endeavor Partners, quoted by MobiHealthNews, showed that a third of users with fitness wearables of any kind stopped using their device after six months. Some digital health companies and apps, such as Omada and the Corrie app, are starting to not only focusing on the user themselves, but instead beginning to channel their physician or family member by sharing the user’s data. Some are also using incentives such as monetary payments. This gives the user an external motivation, rather than just focusing on their own internal self-motivation. However, until these devices are validated, more accurate, and real results can be seen in hard numbers, then clinical application is going to be limited.
An activity tracker may have a role in helping patients get moving, but ultimately, users will need to draw upon internal self-motivation rather than external reward to achieve self-improvement.
Written by: Neil Shaw
Interested in adding the iHub as your next resident elective?
Sources:
- http://www.mobihealthnews.com/content/jay-blahnik-what-separates-apple-watch-other-fitness-trackers
- http://www.mobihealthnews.com/35697/wearable-device-adoption-is-growing-but-engagement-still-drops-after-six-months
- https://www.theatlantic.com/health/archive/2017/08/how-well-do-health-trackers-really-work-for-chronic-conditions/537816/
- https://www.cnet.com/news/smartwatch-step-counter-and-distance-tracker-accuracy/
- Kaiser DW, Harrington RA, Turakhia MP. Wearable Fitness Trackers and Heart Disease. JAMA Cardiol.2016;1(2):239. doi:10.1001/jamacardio.2016.0354
- Shcherbina, A., Mattsson, C M., Waggott, D., Salisbury, H., Christle, J W. et al. Accuracy in Wrist-Worn, Sensor-Based Measurements of Heart Rate and Energy Expenditure in a Diverse Cohort. Journal of Personalized Medicine. 2017;7(2): E3. https://doi.org/10.3390/jpm7020003