After a few weeks wearing the Charge 5 as a watch, I realised I really don't like wearing a watch 24/7. Fitbit have done a great job for comfort, and there are plenty of straps available but it always gets grimy (doing dishes, showering), sweaty, I fidget with it and due to it needing consistent firm skin contact, it can be uncomfortable - particularly when in bed.
I realised that I don't really care much for the screen and really am only interested in gathering data 24/7 on my health, in particular I am interested in my sleep quality and heart rate throughout the day.
I learned about the Whoop which covers this segment, but I already have a Charge 5 and in reviews they have similar accuracy in sleep tracking results featuring the same sensors.
I know the Whoop can be worn with a bicep band, there is no such band for the Charge 5 so I decided to improvise.
I got a CGM "active" sticker and taped my Charge 5 to my tricep.
It is crazy comfortable. I literally do not feel it and it's just sitting there, beaming data to my FitBit account.
Obviously, the Charge 5 was never designed to be worn on the bicep or tricep so this is outside its normal operation but I am wondering what impact it will have on the device accuracy?
I have noticed that my sleep score has increased in the fitbit app since I started wearing it like this. I assume that's an indication of poor data collection.
Any ideas?
Would be awesome to have an engineer from the Fitbit team give some insight as I can't get the data which Fitbit uses to determine sleep quality - so I don't know how different they are.
Best AnswerWhat sleep information does my Fitbit device track?
If your device doesn’t track heart rate, you see your sleep pattern in the Fitbit app. Your sleep pattern includes your time spent awake, restless, and asleep. Restless sleep indicates that you moved in your sleep, like tossing and turning. If your device detects excessive movement—enough that restful sleep would not be possible—your device records time spent awake.
How does my Fitbit device automatically detect my sleep stages?
Fitbit estimates your sleep stages using a combination of your movement and heart-rate patterns. When you haven’t moved for about an hour, your tracker or watch assumes that you’re asleep. Additional data—such as the length of time your movements are indicative of sleep behavior (such as rolling over, etc.)—help confirm that you’re asleep. While you’re sleeping, your device tracks the beat-to-beat changes in your heart rate, known as heart rate variability (HRV), which fluctuate as you transition between light sleep, deep sleep, and REM sleep stages. When you sync your device in the morning, we use your movement and heart rate patterns to estimate your sleep cycles from the previous night.
Fitbit Charge 5 contains the following sensors and motors:
- Optical heart-rate tracker
- 3-axis accelerometer, which tracks motion patterns
- Ambient light sensor
- Built-in GPS receiver + GLONASS, which tracks your location during a workout
- Multipurpose electrical sensors compatible with the EDA Scan app
- Vibration motor
This seems to indicate that Fitbit only uses the heart rate and the accelerometer to identify sleep stages.
My heart rate readings when the Charge is worn on my tricep are identical to those of my Wahoo TICKR so that's fine.
During night time movements, I can only assume that most movements would be less intense from the perspective of the accelerometer when worn on the upper arm than that of the wrist.
I am unsure of how Fitbit infers the intensity of movement when gathering sleep data.
If they are using ML and the Charge 5 model is trained against a dataset of people who using a Charge 5, then the intensity of movement (or lack thereof when worn on the upper arm) would likely impact how the model understands sleep in those cases.
It's evident from the app that they take your historical data into account when creating scores, so perhaps clearing my sleep history and starting again might retrain the sensor to be more sensitive to movements - though the app hasn't yet asked me how good I felt my sleep was so it seems to be deciding that on its own.
I don't have access to the sensor data from my sleep, so I cannot inspect the difference in accelerometer input to understand how large the difference is.
I have a total of 16 nights of data. "Today" is the first day I have worn the Charge 5 on my tricep, so one day a conclusion does not make but for duration it seems to be within range.
Sleep data for "Today" (worn on tricep)
Sleep data for "15" (worn on wrist)
Sleep data for "14" (worn on wrist)
When worn on my tricep, it seems to indicate a lower number of times awake, but the numbers are in general still within range.
After another 15 days I will plot the two data sets and compare them side by side to see how far they deviate
Best Answer
