Epix 2 vs Sleep Study: SpO2 off, wake & deep undercount

When you say "medical sleep study" what do you mean? Were you at a sleep clinic monitored by a physician/tech and hooked up to an EEG?  From their web site, it looks like WatchPAT is an at home test, and you just use a wrist heart rate strap and a pulse oximeter on your finger. That's not a legit medical test, and it's just algorithm based just like the Garmin watches. It won't tell you whether Garmin's data is accurate, just how their algorithm compares. The only way you can truly determine what stages you are in is through use of an EEG. 

The study was prescribed by my Primary Care Physician, in a huge US health-care provider. It was a one-night, at-home assessment. I didn't pay attention to the brand of the device. Later, I did some research and found it is the same as https://www.itamar-medical.com/professionals/watchpat-300/, and the report looks exactly the same, so I am pretty sure it was this device.

I didn't see a wrist-based heart rate sensor, and I don't know where you found this information.

Here is a link to the device, which has been clinically validated to correlate at 89% with polysommography analysis.

www.itamar-medical.com/.../WatchPAT300-Brochure-Final.pdf

This is all good for the Itamar device and I am going to trust my doctors on this.

It would be great if Garmin published some validation of their SpO2 and sleep detection technology.

For heart rate and GPS we have now a plethora of information and comparisons out there, but for these health-related metrics, it is very limited.

No watch

marco bologni said:

correctly recognize sleep phases and naps in a precise way,

They all use algorithms to determine sleep phases. What some people perceive to be correct on one watch, others perceive to be incorrect on the same watch. There aren't too many watches that do actually track naps well and most of those that do only work for naps in excess of 30 mins. 

marco bologni said:

Why do the huawei (from 200 €) correctly recognize sleep phases and naps in a precise way, while Garmin (much more expensive) has to be set to an "approximate" time and is not able to recognize naps?

If you're taking a nap, you can turn on Sleep Mode and shut it down when done.

you are all right, I questioned the reliability of Garmin's sleep recognition, but not the quality, which as far as a watch is questionable.
If I set a time eg. 23.00 - 08.00 and I go to bed later, it often marks me that I am in REM phase while I am awake like last night I was having dinner with friends.
Instead with huawei it automatically recognizes when I lie down for half an hour on the sofa for a nap or go to bed at different times, making more waves in the day based on when I rest or go to bed quite precisely without the need to set anything or a DND. I think Garmin quality is a little overrated

@Marco Bologni

on the F6 using older versions of the firstbeat sleep tracking it did hold more to the sleep times, but the current versions on Epix definitely do not. The F6 would pick up sleep while resting on the couch if past start of sleep time, however I have noticed that the Epix definitely doesn't do this, so I am in agreement with @Miaden T - sleep schedule is now more of a profile for DND and Battery saver and to put your watch settings into the sleep configuration. 

Would agree that you should try a restart to see if that resolves the issue.

They REALLY should re-name the "Sleep Schedule" to be "Night Mode Schedule" or something (and give the ability to define it based on offset from sunrise/sunset).

It's nothing much to do with sleeping at all!

I would say that the "sleep schedule" is still helping to keep the sleep detection active after longer awake periods...they should just add some extra explanation on how the sleep detection exactly works.

I've read many times other users saying that the sleep detection could start/end 2-3 hours earlier/later than the sleep schedule or stuff like that.

Today, i discovered also this part on Garmin https://support.garmin.com/en-HK/?faq=aYDeOCoVgd4PW3A4mvejP6 but i don't know if it's something new or old, and present only on some devices.
Opening the part related to "manual activation of sleep mode" we get to the following page
https://support.garmin.com/en-HK/?faq=plCzU0wYvU95ShBoA9afv6

that doesn't include Epix for example....and we all know that we can enable/disable, manually, the "sleep mode" on the epix but i never got any extra time on Connect during the few tests that i did.
Could retry the next time that i have a chance to get a nap during the day.

Etupes25 said:

It would be great if Garmin published some validation of their SpO2 and sleep detection technology.

They did. Better told FirstBeat (owned by Garmin), the provider of the technology, did. There are plenty of documents on their website. One of them (but possibly not the most up-to-date) is for example the paper A Sleep Analysis Method Based on Heart Rate Variability. They also claim high accuracy based on comparison with polysomnography, just as anyone who sells such technology would do, anyway, too. If you want to be sure, definitely do rather a true polysomnography test, and not a gimmicky test by another watch app.

Ah, I see, my 90 minute naps in my lunch breaks are not the common case :-)