Fitbit Charge 2: Tracking Your Sleep, Including Rem

Fitbit devices with heart-rate tracking can be used to track sleep. The Fitbit Charge 2 uses an accelerometer to measure bodily accelerations of the wrist, coupled with a technique called optical photoplethysmography (PPG) to measure blood flow based on how green light from an LED on the back of the watch is reflected by the body. The Fitbit Charge 2 can track light sleep, deep sleep, and REM sleep. However, it is important to note that there may be some limitations to the accuracy of the device's sleep staging algorithm and its ability to capture sudden heart rate changes.

Fitbit Charge 2's accuracy in tracking REM sleep

The Fitbit Charge 2™ is a multisensory wristband that can be used to track sleep patterns. It uses an accelerometer to measure bodily accelerations of the wrist and optical photoplethysmography (PPG) to measure blood flow based on how green light from an LED on the back of the watch is reflected by the body. Fitbit Charge 2™ can detect sleep/wake states and sleep stage composition by looking at how much the user moves and how often their heart beats.

The Fitbit Charge 2™ has been compared to polysomnography (PSG) in measuring sleep/wake state and sleep stage composition in healthy adults. The results showed that the Fitbit Charge 2™ had an accuracy of 0.74 in detecting rapid-eye-movement (REM) sleep. It also correctly identified 82% of PSG-defined non-REM–REM sleep cycles across the night.

However, the Fitbit Charge 2™ algorithm probably treats REM and deep sleep states differently if they are less than 4.5 minutes long, which may impact the accuracy of the tracker. Additionally, the Fitbit Charge 2™ overestimated REM sleep latency by 29.4 minutes and wakefulness after sleep onset (WASO) by 37.1 minutes. It is important to note that the Fitbit Charge 2™ is not intended to be a medical device, and there may be limitations to its accuracy in tracking REM sleep.

The MCC value, which can be interpreted as a usual correlation coefficient, ranged from 0.11 in light sleep to 0.34 in REM sleep, indicating low to medium strength of correlation and pointing toward room for improvement in the estimation of sleep stages by Fitbit. However, according to a paper published by Fitbit scientists in 2017, the watch’s sensors can classify sleep stages with 69% accuracy in any given 30-second time window.

Fitbit's sleep staging algorithm

Fitbit's sleep-staging algorithm was introduced in 2017. It is incorporated into the Fitbit Charge 2, Charge 3, Alta HR, Versa, Versa 2, Blaze, and Ionic models. This feature uses a combined body movement and heart rate variability (HRV) algorithm to identify and estimate the time spent in individual sleep stages. The algorithm was derived using machine learning methods (linear discriminant classifier) and applied to three types of parameters: motion, HRV, and respiratory rate. The last two parameters are calculated from heartbeat data sensed by photoplethysmography.

Fitbit estimates your sleep stages in a more convenient and comfortable way than traditional methods. It uses your movement and heart rate patterns to estimate your sleep cycles. When you haven't moved for about an hour, your Fitbit tracker assumes you're asleep. It also tracks the beat-to-beat changes in your heart rate, which fluctuate as you transition between light sleep, deep sleep, and REM sleep.

Fitbit's sleep researchers and the National Sleep Foundation describe the following sleep stages: Light sleep, which serves as your entry point into sleep each night, and deep sleep, which typically occurs in the first few hours of sleep. During REM sleep, the final stage, your brain becomes more active, and dreams occur. Your heart rate increases, and your breathing becomes more irregular.

Limitations of the Fitbit Charge 2

The Fitbit Charge 2 can track sleep and heart rate estimates. However, it has several limitations:

Inaccurate Sleep Tracking

The Fitbit Charge 2's sleep-tracking algorithm may not always provide accurate readings. It has been found to overestimate total sleep time by 9 minutes and underestimate sleep onset latency by 4 minutes. It also overestimates light sleep and underestimates deep sleep when compared to PSG-derived sleep stages. These discrepancies could be due to the device's algorithm or the different populations sampled and recording conditions in studies.

Inability to Display Long-Term Averages

While the Fitbit Charge 2 can provide insights into sleep patterns, it does not offer a 30-day average or typical ranges for total time slept. This limitation makes it challenging for users to understand their sleep health over an extended period.

Limited Comparison with PSG Measurements

Fitbit's sleep staging algorithm treats REM and deep sleep states differently, introducing a non-biological discontinuity. Additionally, the “restless” stage on Fitbit does not have a clear PSG equivalent, making meaningful comparisons challenging.

User Experience Factors

The Fitbit Charge 2's sleep tracking accuracy can be influenced by user experience factors. For example, wearing the device too loosely or in an incorrect position can impact heart rate readings. Additionally, the "Begin Sleep Now" option in the Fitbit app might not provide the same accuracy as automatic sleep tracking.

Underestimation of Sleep Dynamics

The Fitbit Charge 2 has been found to underestimate sleep stage transition dynamics, which could impact the understanding of the user's sleep quality and patterns.

Fitbit's sleep sensitivity settings

The Fitbit Charge 2 estimates sleep variables by using a proprietary algorithm. This device can estimate sleep onset latency, total sleep time, light sleep, deep sleep, and REM sleep. However, the Fitbit Charge 2 may not accurately differentiate between REM and deep sleep states lasting less than 4.5 minutes.

Fitbit devices offer two sleep sensitivity settings: "normal" and "sensitive". The "normal" setting provides a basic idea of your sleep patterns, while the "sensitive" setting offers extremely detailed sleep reporting. The "sensitive" setting records nearly all movements as time spent restless or awake, which may be helpful if you frequently wake up feeling tired despite your sleep history indicating sufficient rest.

It is important to note that the sleep sensitivity settings are a legacy feature from before Fitbit devices could detect heart rate. Now that most Fitbit devices can detect heart rate, these settings have become largely redundant, as most users track their sleep using this feature.

To adjust your sleep sensitivity settings, log into your PC Dashboard and change the sleep sensitivity. Then, sync your Fitbit Charge 2 to save these changes.

Fitbit's heart-rate tracking

Fitbit devices with heart-rate tracking capabilities can be used to track sleep. The Fitbit Charge 2, for instance, can be used to track sleep and heart rate. It can also estimate sleep variables such as sleep onset (Son) latency, total sleep time (TST), light sleep, and deep sleep.

The Fitbit Charge 2 uses an algorithm to estimate sleep variables, and while it has been found to be reasonably accurate, there are some limitations to its staging algorithm. For example, it may treat REM and deep sleep states differently if they are shorter than 4.5 minutes, and it may not accurately capture sudden changes in heart rate.

To get the most accurate heart-rate readings while sleeping, it is recommended to wear the device higher on the wrist, securely but not too tightly. Additionally, the device's battery should be sufficiently charged, and the automatic sleep tracking feature should be enabled.

The Fitbit Versa, another device with heart-rate tracking, uses an accelerometer and optical photoplethysmography (PPG) to measure blood flow and bodily accelerations of the wrist. This data, along with heart rate information, is used to infer whether the wearer is awake, in a deep sleep, a light slumber, or dreaming.

While Fitbit's sleep staging has been reported to have an accuracy of 69% in any given 30-second time window, it is important to note that this accuracy is dependent on context. Using the device over multiple nights can provide a better understanding of one's sleep patterns, as it can estimate the duration of light sleep, deep sleep, and sleeplessness without systematic bias.

Frequently asked questions