How can sleep stages be accurate?

[Rupertbear]

I have been very happy with Fitbit's sleep monitoring. It's shown accurately when I've woken up and when I've been restless, but the new colourful sleep stages? Oh dear!

I slept like a log last night, half an hour longer than my usual seven hours, I didn't even dream it was such a good sleep, yet it tells me I was awake six times (I didn't wake up at all) in REM sleep five times at all hours, and in deep sleep for two very short periods.

This is not helpful to me at all. I don't suppose there is a way I can opt for how it was? (MUCH simpler!)

Is this yet another change for change's sake? I do wish they'd just leave a good thing alone!

[HipsDontLy · Accepted solution]

I'm a little amused because the past couple of days have been reading threads about people begging for the sleep stages to roll out, me being one of them!

Regarding the accuracy of sleep stages, it's not uncommon to be technically awake during the night several times (technically meaning your brain is in stage 1, there is alpha wave activity, etc). Your body will physiologically reflect these changes; you're awake, but there's a high chance you're not going to remember this because you're not awake enough/completely. 

Regarding opting out, I guess keep an eye out for moderators? But assuming the worst case scenario is you can't opt out, my first instinct would be to turn off heart rate monitoring for bed. Since the Charge 2 can't detect brain waves (or CAN IT? Are you not telling us something, Fitbit?), my guess is that the only way for them to detect changes in sleep stages is through heart rate and movement. Disclaimer: If you turn off the heart rate monitor, I don't know how accurate the sleep tracking will be with only movement detection. I'd say try it for a night, and if that night it was your regular bed time and you feel as you normally do waking up, to take that sleep log and compare it to the rest of your older ones. See if they're workable equivalents. 

[Rupertbear]

Thanks that's very helpful, although as I said the sleep stages are a disappointment to me as now I can't tell what time I was ACTUALLY awake.

However I will do as you suggest, once I find out how to turn off the heart rate monitor! 

Thanks

[HipsDontLy]

Keep us posted! You can turn off the heart rate monitor by hitting the button on the upper right from the app screen (the icon that lets you enter the screen with alarms, shuffling around exercises, etc). The heart rate icon has "on," "off," and "auto." That's the one you're looking for!

[Mabelsmum]

I completely agree! The new sleep stages are inaccurate, messy and useless. The old system gave me a much  clearer picture of my night. I looked on the forum expecting to see other complaints but the majority of people seem to want stages. Did turning off the heart monitor work- I'm thinking no?

[lapaloma]

Again this morning the time going to sleep and time awakening are wrong.  Graph that is sometimes there showing sleep duration, periods awake, periods of deep sleep is non functional again!  The change to the new format is NOT getting corrected.  This advertised function of the Fitbit is not only unreliable, but exists sometimes, then not.  I was quite happy with my first Fitbit, even at first with this Charge 2, but now it is time to look at other alternatives.

[Rupertbear]

You are correct, it made no difference at all! I'm glad it's not just me
then.
I'm not sure whether they have made a slight alteration but I had a bad
night last night waking up four times, but all the other momentary 'awake'
red lines that were making it so messy were not joined up, they were all
'floating' at the top, so my sleep did show I woke up four times..perhaps
it is not a total loss then.
Is it the same with yours?

Karen

[Tzeni]

Hello all!  Here's the most recent reliable  article I could find… It's worth the read!  I think it's interesting to check Fitbit ( asleep/awake time), but as far as accuracy of sleep stages, --i.e.; deep versus REM, etc. --this article  sums it up well! https://www.google.com/amp/s/amp.livescience.com/42710-fitness-trackers-sleep-monitoring-accuracy.ht...

[bboughton]

Hi all -- some bad info here. The Fitbit study results were presented this summer and are accurate. Most people are awake for almost 45 mins and don't know it. http://www.sleepreviewmag.com/2017/06/study-shows-fitbit-heart-rate-tracking-devices-accurately-trac...

[Rich_Laue]

@HipsDontLy, since the Fitbit Ulta came out, Fitbit has been recording sleep by the movement of the arm without the heartrate being measured.

Yes without heartrate, manually starting/stopping through the app, or adding a sleep record after the fact will give the old style of sleep record.

@Tzeni, i noticed the Livescience article points to a study that is already 6 years old. A lot of changes in technology have happened since then.

[AlvaroFitbit]

Hi everybody thank you all for sharing your opinions and links to further help this thread. You can always check Fitbit's official explanation here so you have an idea of how these stages are calculated. As always there is room for improvement so rest assured out team is checking on ways to make this even better.

I'll be around!

[Tzeni]

Yes as mentioned it was the most recent article out there. It doesn't change the facts --fitbit does not measure brainwaves-only an EEG is an accurate measure.  Deep sleep cannot be deciphered.  Maybe someday soon it will include a new technology that will have capabilities to measure brainwaves, until then- I'm  certainly not relying on my fitbit to determine how much deep sleep, etc. I have each night.

[Tzeni]

Yes as mentioned it was the most recent article out there. It doesn't change the facts --fitbit does not measure brainwaves-only an EEG is an accurate measure. Deep sleep cannot be deciphered. Maybe someday soon it will include a new technology that will have capabilities to measure brainwaves, until then- I'm certainly not relying on my fitbit to determine how much deep sleep, etc. I have each night.

[WavyDavey]

I think it's implied with a fitness tracker that it's not doing brain wave scans. It's using the information available, such as patterns of movement, to determine sleep stages. It's not medical in nature, and it's not foolproof. I would not try and diagnose anything with it.

[Rich_Laue]

@Tzeni your report dated 2014 was when fitness trackers first started to use opticle sensors to read heart rate.

Sleep was still basically being monitored by arm movements.  Even then people who wore their tracker durring sleep studies where pleasantly surprised to see how close Fitbit was to the professional study, considering the Fitbit is not a multimillion-dollar sleep lab with s technician on duty they were impressed.

The article @bboughton points to is of an Anstract dated less than 4 months ago, not 3 plus years.

The abstract is not based on theory but on real world test on 60 people, it found that using heart rate and movement the sleep could indeed to tracked with way more precision then expected.

Fitbit has a research team that routinely anylizes patterns in the data to refine their algorithms. Yes we all understand they are not pergect nor should be used to teplace your doctor.

Even so Fitbit data has been used by

Doctors to diagnose alowing them to bypass tests an proced to the proper care needed, this saving time and possibly more damage. This happened in a hospital i visit several times a year as part of my job.

Several have called 911 thus saving their own life.

Ladies have noticed they were pregnant.

One case the prosecuter proved the husbamd killed his wife, ran her over with a car

Several have also been able to prove their spouse was cheating.

All because of a little band being worn on the arm.

[Tzeni]

Once again; regardless of how long ago any article was written, it still doesn't change the fact that today's fitness trackers (2017) rely on movement & heart rate monitors to determine sleep stages.  That just doesn't suffice –information on brain waves, eye movements, & muscle tone is needed to definitively distinguish sleep stages.

By the way, I love my "little band I wear on my arm;" I just won't rely on it at this point to determine how many hours of deep sleep vs R.E.M. sleep I'm getting.  

On another note; your  response is full of grammatical errors, lacks scholarly credentials, & valid citations.

I don't know why anyone would want to trust a doctor that would bypass any valid medical testing & instead use Fitbit data to make any determination!  

As an educator I can no longer engage in this rhetoric!  You just lost me.

[Rich_Laue]

The doctor was confident enough in the heart rate readings, based on his interaction with past patients and what he was looking at, that felt the extra tweenty minute test to verify what the Fitbit was telling him was not needed.

Check out the S+ by ResMed Personal Sleep Solution a company that makes medical grade sleep apnea devices. This unit is rated near the top of OTC trackers and it doesnt even touch the body.

So why not validate your possition EKG, eye movements, etc is a must to give an idea of sleep quality, and no sleep OTC sleep tracker is able to give a general idea of sleep quality?

I find a few that seem to not agree.

http://gadgetsandwearables.com/2017/06/05/fitbit-sleep-data/

Spoiler

Fitbit activity trackers monitor light, deep, and rem sleep to a reasonable degree of accuracy says a new independent study.

When it comes to sleep stages tye result was not to different then the results you would get from a slepp laboratory. This was presented June 6 7 at the Sleep conference. 

 https://arstechnica.com/gadgets/2017/05/sleep-it-off-how-and-how-well-wearables-can-track-sleep/

Spoiler

Tracking heart rate at night can lead to more useful sleep data than tracking motion alone. Heart rate fluctuates depending on the sleep stage you're in, so an accurate heart rate monitor can better estimate the amount of time you spend in light, deep, and REM sleep. It may be harder for heart rate monitoring devices to differentiate between light and deep sleep, however, since both stages are characterized by a decrease in heart rate. But a good device can monitor the constant and steady decrease in heart rate over dozens of minutes and into hours. At the same time, a good heart rate monitor will detect heart rate increases when your body enters REM sleep. That doesn't mean every wearable with a heart rate monitor will be a good sleep tracker, especially since not all of them mix heart rate data with sleep tracking motion data. But the ones that do will provide a better picture of your night's sleep since they're not just relying on fickle motion data.

All of this is dependent on the heart rate monitor being accurate and the user wearing it correctly. Like exercise data, sleep data is generally best captured with a chest strap heart rate monitor. These straps don't move much while wrapped around your chest, and they stay right up against your body and close to your heart throughout the entire night. But there aren't many chest straps that monitor sleep because many people find them uncomfortable to wear even during the day while working out. Most sleep trackers with heart rate monitors come in the form of wristbands, and all regular rules of daytime wear apply if you want good heart rate data while you sleep. The device must be secure and fairly tight around your wrist (not so snug as to inhibit circulation, but snug enough not to move up and down the wrist easily) and positioned just above the wrist bone or about two-fingers' width from where your hand meets your arm.

 https://wearablezone.com/amp/news/how-accurate-is-fitbit/

Spoiler

Fitbit received help earlier this year from sleep scientists from Stanford, Johns Hopkins, and the University of Arizona in order to improve the accuracy of sleep tracking.

 http://www.sleepreviewmag.com/2017/06/study-shows-fitbit-heart-rate-tracking-devices-accurately-trac...

Spoiler

Fitbit has validated the ability of wrist-worn trackers that incorporate movement and cardiac sensors, like Fitbit Alta HR, Fitbit Blaze, and Fitbit Charge 2, to accurately determine light, deep, and REM sleep stages. The results of Fitbit’s study, which were scored independently by polysomnography technologists, demonstrate that these devices can be used to track sleep stages with a reasonable degree of accuracy in normal adult sleepers, according to Fitbit. Having the ability to gather reliable sleep stage data on wrist-worn devices can help simplify sleep research and increase public knowledge about sleep.

 https://www.sporttechie.com/fitbit-study-fitness-devices-accurately-track-sleep-stages-unlike-calori...

Spoiler

More info on Fitbits study and possition 

http://www.mensfitness.com/life/gearandtech/fitbit-sleep-stages-review-strong-sleep-tracker-plenty-p...

Spoiler

From a users point of view

.

Its true i was not watching the auto correct very well, and really wasnt trying to validate, a simple search of the web would be all that was needed, and those instances have already been documented on these forums They are not about sleep and happened last year.

I have noticed you havent validated your statement yet. 

[oldpal]

@Rich_Laue 

I’m not particularly concerned with the accuracy of the sleep stages, I just want to see them at all. My wife and I acquired our Charge 2s the first week in February and everything worked as advertised. Around the 27th of March Fitbit announced that the Charge 2 would report the sleep stages. No sleep stages were appearing on my tracker. Moderators on this forum kept telling the users that were not seeing sleep stages to just be patient, that Fitbit was rolling out this feature gradually and everyone would soon be seeing sleep stages. About a month went by with no sheep stages for me. Then on April 28th sleep stages started on my tracker and lasted pretty much continuously until July 30th. That was about the time that I upgraded the firmware to 22.54.6. Since then the Charge 2 has reported sleep stages only four times. Since July 30th I have tried my original tracker plus two replacement units. Each tracker reports BPM results consistent with two other devices. I have followed the instructions for wear and worn the tracker in every imaginable way but still no luck. 

I think that the problem exists with the latest firmware, but don’t know for sure. Does anyone here have any suggestions?

[Rich_Laue]

Is your app up yo date @oldpal, if so the app itself will tell you why your not seeing sleep stages. Just go to the detail page of the sleep in question.

[BobM2]

 I gave up on wearing mine at night. Almost every night, it credits me for a hundred of more steps (sometimes over 200) I didn't take, and assumes I'm awake then. Except some nights it says I never slept at all. Combine the meaningless data with it lighting up multiple times a night even with quick view off, and it being bright enough to wake me up.....

It's just not worth it.