How Accurate is Your Activity Tracker?

By | June 8, 2017

The functionality and popularity of consumer-grade activity trackers (such as Fitbit) appear to be ever-increasing.  If you don’t personally own one, you probably know at least one or two people who do.  In an online survey of 1,000 respondents [PDF], conducted in 2016 by PwC, 45% owned a fitness band, with “Health” being the primary motivator for purchase. Further, adoption of wearable activity trackers appears to be going strong, with 57% of respondents saying they would likely purchase a fitness band in the next 12 months.

With activity trackers becoming so common, have you ever wondered about their accuracy?  Just how correct are those devices when they tell you how many steps you’ve taken or how much sleep you’ve gotten?

This recently published article focusing on a small sample has gotten some press lately, but let’s dig a little deeper. A 2015 article in the International Journal of Behavioral Nutrition and Physical Activity looked the validity and reliability of activity trackers (Fitbit and Jawbone) by systematically reviewing the literature on 5 measures: steps, distance, physical activity, energy expenditure, and sleep. Of the 22 included studies, 21 reviewed the validity of at least one activity tracker, and 7 discussed the interdevice reliability of Fitbit; however, none reported the interdevice reliability of Jawbone. The authors identified the following key points:

  • Step Count: Validity for step count in both Fitbit and Jawbone were high. Interdevice reliability was high for Fitbit; however, interdevice reliability for Jawbone wasn’t evaluated. Hip-worn trackers were typically more accurate at counting steps when compared to devices worn in other locations.
  • Distance: One study discussed the validity and reliability of distance measurement. Distance was generally over-estimated at slow speeds and under-estimated at fast speeds.
  • Physical Activity: One study found that moderate-to-vigorous physical activity was overestimated, while another found it aligned with an accelerometer.
  • Energy Expenditure: Interdevice reliability was high, but validity was low and energy expenditures were often underestimated.
  • Sleep: Similar to energy expenditures, the reliability of sleep was also high, but total sleep time and sleep efficiency were generally overestimated, while wake after sleep onset was underestimated.

To muddy the waters further regarding activity trackers, a class-action lawsuit has been filed against Fitbit bringing into question the precision of Fitbit’s heart rate monitoring technology. Specifically, the claim is that Fitbit is misrepresenting the accuracy of user’s heart rates during physical activity.  While Fitbit denies this, they aren’t the only consumer wearable firm that’s faced accusations of inaccuracy.  Back in 2014, CNET measured five activity trackers (the Garmin VivoFit, the Basis Carbon Steele, the Withings Pulse O2, the Samsung Gear Fit, and the Samsung Galaxy S5) and found inaccuracies across all five devices when measuring heart rate.

Knowing your Fitbit or Jawbone might not be 100% accurate isn’t great news.  Luckily, the review by Evenson and colleagues also discusses what activity tracker owners can do to help make these devices a little more accurate:

  • Wear your activity tracker in the same spot from day to day. If you wear it on your wrist, indicate whether it’s on the dominant or non-dominant side.
  • Make sure to enter your details (such as height and weight) during the initial set-up, and then sync this information with the tracking device.
  • Replace the default stride length by calibrating the device to your stride when possible.
  • Use the journal or diary function to provide additional information to your activity tracker so it can learn what specific activities look like.
  • Obtain updates and use add-on features so your device can have the most accurate algorithm to calculate sleep or physical activity. For instance, Fitbit just released new features associated with tracking user’s sleep.
  • If your device has sleep mode settings, interact with them to help the device learn if you’re sleeping, napping, or awake.

If you’re interested in more information about activity trackers and wearable technology, check out an interview with Robert Furberg (who co-authored the activity tracker systematic review) on the podcast The Measure of Everyday Life.  Also, check out my prior post discussing why people stop wearing their Fitbits.

Alexa Ortiz

Alexa Ortiz

Health IT Scientist at RTI International
Alexa Ortiz graduated from the University of North Carolina at Charlotte in 2009 with a Bachelor of Science in Nursing. Before receiving her graduate degree she was a practicing nurse for five years and has clinical experience in the field of both Cardiology and Neurology. In 2014 she received a Master of Science in Nursing specializing in nursing informatics from Duke University. Presently, she works as a Health IT Scientist at RTI International in the Center for Digital Health and Clinical Informatics. Despite no longer working in a clinical setting, she continues to maintain an active nurse license in the state of North Carolina. Her primary areas of research at RTI International focus on the clinical implementation of health information technology and the evaluation of consumer wearable devices.
Alexa Ortiz

Latest posts by Alexa Ortiz (see all)

Category: General Health behaviors Tags: , , , , ,

About Alexa Ortiz

Alexa Ortiz graduated from the University of North Carolina at Charlotte in 2009 with a Bachelor of Science in Nursing. Before receiving her graduate degree she was a practicing nurse for five years and has clinical experience in the field of both Cardiology and Neurology. In 2014 she received a Master of Science in Nursing specializing in nursing informatics from Duke University. Presently, she works as a Health IT Scientist at RTI International in the Center for Digital Health and Clinical Informatics. Despite no longer working in a clinical setting, she continues to maintain an active nurse license in the state of North Carolina. Her primary areas of research at RTI International focus on the clinical implementation of health information technology and the evaluation of consumer wearable devices.