Sleep tracking in fitbit versa and too little REM sleep

[kuk_sleeper]

Hi all,

I got a fitbit versa and I have been checking the sleep data in it. I feel somewhat accurate to me but it is showing the amount of REM sleep I get is below the benchmark/average. Granted I have been sleep a bit less after I got the fitbit (6 - 6.5 hours versus the 7 - 7.5). I feel that the fitbit is somewhat accurate ( I tried to correleate the time I was awake and it matches up. I had one time I woke up from a sleep and fitbit show REM sleep at that time.

My question is 1) Have any else here use Fitbit (versa or other models) ? Do you find the sleep tracking accurate?
2) How much REM sleep does usually everyone get (For who ever has data)

For reference I am 34/Male

[Weagle]

First off, I take FitBit's data with a grain of salt. They are only utilizing pulse and movement to get their data, not nearly what's able to be gathered in a professional environment. Also I'm a 43 year old male use a FitBit Charge 2.

That being said, I average the following according to FitBit:
Awake 14%
REM 14%
Light 55%
Deep 13%


Also these were my numbers before starting CPAP therapy. I haven't worn the watch to bed anytime recently.

[chunkyfrog]

Many personal trackers have been programmed to give skewed results to boost sales.
Insiders have ADMITTED this is true.
They are labeled "For entertainment use only" for a reason.
Enjoy, have fun, but "accuracy" is a flexible concept, especially in their business..

[kuk_sleeper]

I wanted to post a update. I have been comparing my Fitbit sleep data with sleepyhead graphs. I believe fitbit's sleep stage tracking (atleast REM sleep) is more accurate that what I would expect from a fitness tracker. I have been having bouts of waking up from vivid dream and whenever I remember the time fitbit seems to suggest that I was in REM sleep.

[c_doe]

I have a FitbIt Alta HR and I find it to be quite accurate...
My deep sleep and REM sleep is also always below the benchmark, or just slightly touching. This corresponds to some nightly video recording I did.

My girl friend also used it for a couple of nights, and here values are perfect, hitting the benchmarks regularly. There can be many reasons for that and the algorithm at work here are just assumptions in the end. But I guess they are quite accurate if you don't look at it on a per night basis, but at the average after couple of weeks. I also find my sleep pretty fragmented, whereas those of my girl friend is pretty regular.

I also feel better on days where my scores are better than on days where they are low. For example the amount of deep sleep corresponds quite good with how tired I feel. And this is not conformation bias. Some times I purposely only look at the data in the evening...

Here is a comparison:

sleep.jpg (74.23 KiB) Viewed 588 times