Estimated oxygen variation graph: when to see a doctor

[GoodGuy5000HD]

I can understand why the FDA was hesitant to approve the Estimated oxygen variation (EOV) graph: with little guidance about what constitutes a "problem," it could/would create anxiety in a lot of people, and needlessly fill-up doctors' schedules as well as sleep labs (good for the sleep labs). If I bring it up with my doctor, I'm certain he'll take a "why not?" approach to a sleep study costing me thousands (even on forced socialized medicine).

Since my EOV graph has been activated, most days show a squiggly line in the low range. However, over the past 3 weeks, I seem to get a single spike past the "high" threshold once every 2-3 nights. Is this enough of a "problem" to schedule a sleep study?

Generally, what does the EOV look like for people who are eventually prescribed a CPAP machine (before and after the CPAP)? Or, have users seen some spikes, scheduled a one-night sleep study and were told that their sleep was normal?

Perhaps if Fitbit won't (or cannot because of the FDA) give guidance about what a problematic EOV graph looks like, we can crowd-source where that "see a doctor" threshold really is.

[Dgreener]

This is the next night. I might have central sleep apnea which is not obstructive or causes Snoring. Your brain just doesn't tell you to breathe. 

Any more pics out there?

 

[mkohn]

I continue to read viewers' comments about EOV and the ridiculous Fitbit graph which provides absolutely no pragmatic information with incredulity.  Why not simply show actual O2 levels with the standard medical preface as to accuracy and that it is not intended to be any type of predictor, if that's within the scope of the device.  Or, at least, overlay the graph with the sleep chart.  Or, just remove the graph altogether so we don't have to continue this thread.

[datasimple]

So after doing some thinking and testing, here is my take on this graph:

The Fitbit's estimated oxygen variation (EOV) graph displays a 5-minute snapshot of your blood oxygen saturation at a specific moment in time. The actual graph and color are related, but not correlated. The Fitbit takes samples every 2 seconds in the background and averages them between the 5-minute samples. If the 2-second variations haven't changed much, the graph is green. If it has gone up and down a lot (e.g., 92% to 99%), the graph is yellow.

Holding your breath can imitate an apnea event, causing your oxygen level to drop to 92%. Taking deep breaths afterward can quickly raise your oxygen level to around 98%, which happens over a short period (around 3-4 seconds). If you continue to do this several times between two 5-minute samples, the graph will be yellow. Normal variations in your O2 level will change gradually, indicating to the Fitbit that it is normal, such as changes in sleep phases. However, rapid rises and falls in O2 levels usually occur when you stop breathing and start gasping to raise your O2 levels.

[GreenThumb79]

Super helpful. I have yellow as often as I have green every single night. I have a bad feeling I have sleep apnea, and I sooooo don’t want to face that reality. Thanks for the info. 

[mkohn]

Have passed the dotted line before but have never been this close to the top line.  Has anyone ever gone to or through the top line?

[Technobrat]

Doesn’t look good, eh? 

[MikeR-PDX]

It would be nice if you defined what both "high variation" and "frequently" are?  Does "high variation" mean O2 levels are going below 95%, or 90% or some other number?  Does "frequently" mean more than once a night, or more than 2, 3, 4 times a night?  It is impossible to understand what constitutes a level of concern that warrants talking to a doctor.

[Stuckonone]

I am totally confused.

1. I get a couple yellow spikes per night, every night. Is this frequent?

2. The spikes often seem to roughly coincide with "awake" time, often when I get up and hit the bathroom. (I can't tell exactly, because there is no time on the first graph, and the scale looks different. ) Is this significant?

How much or frequent yellow is normal for people whondon't have sleep apnea or similar problem?

[Maysy1]

Hi read your post. Can uou tell me I have spikes where you have one can you tell me did you get tested did you have sleep apnea or not.

[Maysy1]

Hi did you ever find out why your graph was like that was it apnea, did you get tested. Would be grateful to hear from you. I was diagnosed years ago no cpap said mild. I now have pacemaker through apnea caused atrial fibliration. Still having apnea tests saying no apnea but having oxygen drops whilst awake. Did you get checked out. Please reply.

[MikeR-PDX]

Yes, I was tested and have sleep apnea BUT, my CPAP machine has solved my
sleep apnea problem, AND I still have occasional nightly "high variation"
spikes according to Fitbit even when I use my CPAP.

[JMark235]

I think that the estimated oxygen variation information, at least by itself, is fairly worthless even assuming that the SP02 sensor is accurate. FitBit measures your O2 once every minute and then essentially averages the data. Consider the 2 following scenarios;

1) Person 1 : Their average oxygen saturation as reported on the health metrics screen is 93% and  because their oxygen level fluctuates between 92% and 94% (+/- 1%) they show very little 'Estimated Oxygen Variation' as shown on the sleep screen for today.

2) Person 2: Their average oxygen saturation as reported on the health metrics screen is 98% and  because their oxygen level fluctuates between 96% and 100% (+/- 2%) they show more or higher 'Estimated Oxygen Variation' as shown on the sleep screen for today.

Would you rather be person 1 with a significantly lower O2 level but very little fluctuation in O2 level or person 2 with a much higher average O2 level that fluctuates more between high O2 and very high O2.

I think that the 02 level analysis should incorporate more heart rate data. If you are entering or leaving a period of REM sleep (Rapid Eye Movement) and your dream results in a lot of agitation, your heart rate and presumably your O2 level will increase significantly as you enter REM and then decrease after you leave REM which will probably result in a significant O2 variation but might just indicate you have a lot of bad dreams rather than a health issue.

You might want to read the interview with Dr. Conor Heneghan, Fitbit’s Lead Sleep Research Scientist to explain everything you need to know about FitBit's SP02 data.

https://www.wareable.com/fitbit/fitbit-estimated-oxygen-variation-explained-7878 

I do think that significant changes in the average O2 level over a longer period of time might indicate a change in health that should be investigated but that the nightly variations have too little data to be worthy of concern. I would prefer to see a graph of the nightly O2 sensor readings for each minute rather than just an average for the entire night and a summary of the nightly range. I have never used FitBit premium and based on my research FitBit premium does not provide this level of granularity. If I am incorrect on this point then please do correct me.

[Staceyg456]

I would like to know what is considered "frequent ". That is not even a little bit clear. I often get one or two spikes a night. Is that something that I should be concerned about?

[ChaoticGootral]

This was so insanely helpful. Thank you for posting.

The last time I charged and put on my Fitbit it felt looser and I wondered if that was the reason for recent spikes.

[MrCut]

If you read the disclaimer right under this graph it says many fluctuations are normal. Its when there are many fluctuations in the high area is there concern. 

I would read your graph as being completely normal and not having sleep apnea. But im not a doctor.

[SunriseRun56431]

Wow that is a lot of variations. Mine goes flat to a little variations to one yellow line. You have great awake time. Mines never less than 12 or 13% of the night

 

[SunriseRun56431]

 

[sbf54]

This was mine last night. I'm wondering the same thing.

 

[Jaimew4]

I envy the other graphs people post. Mine would scare them to smitherins. Yours is similar to mine. I’m wondering if it varies across devices. I have. Fitbit charge 5 that I’ve come to loathe. I didn’t have these issues as much with versa 

[MaryCharMar]

Hello, FitBit Moderator... Could you please clearly define the Y-axis of the oxygen variation graph? Without that information, the graph is nearly useless.  A time indicator on the X-axis would also be helpful, in case the company is considering modifications that would allow users to correctly interpret this information.