Fitabase is proud to announce the publication of two new research studies that used Fitbit activity tracking devices to measure sleep. Since the launch of the Fitbit Classic in 2009, users have been able to track and better understand their sleep patterns. Recent advances to device firmware have made it even easier through the implementation of automatic sleep detection and tracking. Researchers have long been interested in how well Fitbit devices track sleep. In fact, one of the earliest published studies 1 using Fitbits focused on sleep, not steps!

Before we dive into these studies, let’s stop here for a moment and learn about how sleep is measured in research settings. Typically, sleep researchers rely on a thorough combination of measurements called Polysomnography (PSG) to measure sleep. Often you’ll see PSG referred to as a sleep study. During overnight PSG monitoring, participants are  hooked up to a variety of different measurement devices in order to track heart rate, limb movement, eye movement, brain waves, breathing rate/effort, blood oxygen levels, and even snoring. As you can see in the image below, participating in a sleep study doesn’t look like a whole lot of fun.

Pediatric Polysomnogram. Photo by Robert Lawton (CC BY-SA 2.5)

Pediatric Polysomnogram. Photo by Robert Lawton (CC BY-SA 2.5)

In the two studies published last week, researchers wanted to see if different Fitbit devices were able to detect sleep in adolescents and children as well as the standard PSG recording.

In the first study2, published in the Journal of Otolaryngology Advances, researchers at UCLA Children’s Hospital and the University of Southern California recruited 14 children aged 3 to 11 years who were already scheduled for an overnight PSG (due to a prescribed evaluation for sleep disordered breathing). The children wore a Fitbit Flex on their non-dominant wrist and their parents helped to properly set the Flex for sleep mode. Data from the Fitbit was then compared to PSG readings for total time asleep, waking time, sleep efficiency, and movement.

Minute-level data from the Flex was also used to compare epochs of sleep, awake, and movement to PSG recordings. The researchers found the Flex to be very accurate for measuring total sleep time, only showing a difference of 0.2 minutes on average, but concluded this was due in large part to how the data was cleaned. While the results of the epoch (minute-level) comparison showed that the Flex had a hard time accurately distinguishing a waking period form a sleeping period, the researchers maintained that the Flex was beneficial for tracking movement in children with sleep disordered breathing,

“Our novel findings of a significant correlation between Fitbit and PSG related movements highlight the possibility that Fitbit measurements of movement might be used as a means to evaluate sleep disturbance in children with [sleep disordered breathing].”

In the second study3, accepted for publication in Physiology & Behavior, researchers from numerous institutions conducted research on the accuracy of the Fitbit ChargeHR for measuring sleep in 32 healthy adolescents (17 yrs old on average). After screening for sleep disorders, the adolescents wore a ChargeHR on their non-dominant wrist while also undergoing a standard PSG. Both aggregate and minute-level data was compared to PSG data for each participant.

Results of the comparison between the ChargeHR and the PSG data indicated that the ChargeHR accurately detected total time asleep, time awake, and sleep efficiency, with approximately 90% of the participants meeting the clinically set minimum satisfactory differences for these measures. As was found in the children’s study, the ChargeHR had a low level of accuracy for detecting waking minutes. The researchers also compared the heart rate data from the ChargeHR to the clinically ECG included in the PSG. They found that for all sleeping periods, the ChargeHR was as accurate as the ECG, finding with a less than one beat per minute difference on average.

Sleep is an integral part of our lives, and being able to understand how much sleep we get without having to step into a lab is important. As both studies highlight, there is a great deal of potential in using low-cost personal wearable devices to support large-scale research and evaluation studies.

Fitabase is proud to have supported the data collection efforts of both the above mentioned research teams. Our unique platform allows researchers to easily access, download, and explore sleep, activity, and weight data from all the Fitbit models.


Fitabase makes it very easy for us to get fine-grained data from Fitbits. Researchers like us want ever-better measurement capabilities, because ultimately, that it is what will allow us to more fully understand the factors that influence health. In our validation study, we were able to get minute-level data or better to examine how close the Fitbit Charge HR is to our gold-standard for measuring sleep. Our study would not have been as robust as it is if we did not use Fitabase. -- Job Godino, UCSD Center for Wireless and Population Systems


If you’re using Fitbits in your research, or would like to, we’d love to hear from you. Get in touch!


  1. Montgomery-Downs, H. E., Insana, S. P., & Bond, J. a. (2012). Movement toward a novel activity monitoring device. Sleep and Breathing, 16(3), 913–987. doi:10.1007/s11325–011–0585-y ↩
  2. de Zambotti, M., Baker, F. C., Willoughby, A. R., Godino, J. G., Wing, D., Patrick, K., & Colrain, I. M. (2016). Measures of sleep and cardiac functioning during sleep using a multi-sensory commercially-available wristband in adolescents. Physiology & Behavior. doi:10.1016/j.physbeh.2016.03.006 ↩
  3. Osterbauer, B., Koempel, J. A., Davidson Ward, S. L., Fisher, L. M., & Don, D. M. (2016). A Comparison Study Of The Fitbit Activity Monitor And PSG For Assessing Sleep Patterns And Movement In Children. Journal of Otolaryngology Advances, 1(3), 24–35. doi:10.14302/issn.2379–8572.joa–15–891 ↩