Reading Sleep State from Respiratory Rate

[CapnLoki]

Recently I've been trying to track my sleep state by watching the SleepyHead data, in particular the "Resp. Rate." On "well behaved" nights (roughly 2/3 of the time) a particular pattern emerges. It starts with a brief "ragged" period, with a lower than average resp rate, then a long period, about an hour, of steady breathing at about 17-19 bpm, followed by a 15-20 minute period of more rapid and ragged breathing. This repeats 4-5 times through the night.

I interpret the first ragged period as being awake, the second smoother as steady sleep, and the third (ragged, rapid breathing) as REM sleep. I can pretty much verify the first two assumptions from experience, but without brain wave monitors I can't verify REM. I can say that I think I'm having strong dreams during the presumed REM sleep, and frequently get up for a bathroom break after. I would add these patterns became more apparent as I focused more on quality sleep and tuned the equipment (pressures, mask, humidity, mattress, pillows, sleep position etc) with this in mind.

This is a SleepyHead chart from a recent night. This was average for me, AHI about 1.6 - mostly "H's" scattered randomly, not correlated with sleep state. Any thoughts on my REM state theory?

[Wulfman...]

Recently I've been trying to track my sleep state by watching the SleepyHead data, in particular the "Resp. Rate." On "well behaved" nights (roughly 2/3 of the time) a particular pattern emerges. It starts with a brief "ragged" period, with a lower than average resp rate, then a long period, about an hour, of steady breathing at about 17-19 bpm, followed by a 15-20 minute period of more rapid and ragged breathing. This repeats 4-5 times through the night.

I interpret the first ragged period as being awake, the second smoother as steady sleep, and the third (ragged, rapid breathing) as REM sleep. I can pretty much verify the first two assumptions from experience, but without brain wave monitors I can't verify REM. I can say that I think I'm having strong dreams during the presumed REM sleep, and frequently get up for a bathroom break after. I would add these patterns became more apparent as I focused more on quality sleep and tuned the equipment (pressures, mask, humidity, mattress, pillows, sleep position etc) with this in mind.

This is a SleepyHead chart from a recent night. This was average for me, AHI about 1.6 - mostly "H's" scattered randomly, not correlated with sleep state. Any thoughts on my REM state theory?

Some thoughts.......
I'm not sure what it is you're trying to accomplish.
I think it might help to compare THAT graph to your "pressure" graph for the night to see how those breathing rates compare to the pressure changes.
A person can dream in essentially ANY sleep stage, so if you're trying to identify when you may be dreaming it would be somewhat futile.
If you're frequently awakening and/or getting up for a potty break, that disrupts the "data stream" and really makes things convoluted.

If you were using straight pressure with virtually no potty breaks or awakenings, then you may have some credible data to look at. But, again, you can dream in almost any sleep stage.
Did you ever think that the pressure changes may be contributing to your awakenings?


Den

.

[CapnLoki]

I should clarify two points: first, my bathroom breaks which used to be frequent (often 4 a night) have been greatly reduced to 2 or sometimes 1 a night. On this particular night the first was following the third "sleep cycle" and the second was following the next cycle. This has been a common pattern of late - I seem to get over 4 hours of good sleep before the first break. I should have said the when I do get up for a break, its usually following an apparent REM cycle.

The other point is that on my good nights, the pressure stays almost totally flat. My low pressure is 8.5 the the high for the night was 10. Probably 80% of the time was at 8.5. On some bad nights, the pressure escalates to 11 or 12 and I start having leak issues that can wake me up, but that is not typical, and it didn't happen on this night. On the contrary, this was a rather quiet, uneventful night that I think would be ideal for mapping sleep states.

My goal in this is to understand the sleep cycles on the premise that 3 to 5 uninterrupted sleep cycles should lead to a restful sleep.

[HoseCrusher]

You may want to step back a moment and do some research.

The data from the xPAP machines is based upon air flow. Sleep stages are determined by brain wave activity. You need to establish a correlation between those two before you can have any confidence in your observations.

[mellabella]

You may want to step back a moment and do some research.

The data from the xPAP machines is based upon air flow. Sleep stages are determined by brain wave activity. You need to establish a correlation between those two before you can have any confidence in your observations.

I'm also not convinced it's much more than reading tea leaves to try to figure out sleep stages from Sleepyhead data (without EEG data backup), but there is a robust correlation between REM sleep and respiratory rate (it increases during REM) and tidal volume (it decreases during REM). As well as becoming more shallow and more rapid, respiration also becomes more variable in REM. This is according to every sleep medicine textbook and article I've been able to find online.

Articles I've found about trying to deduce sleep state (NREM vs. REM) based on respiratory data alone rely heavily on trying to come up with mathematical formulas for changes, so I don't know this is something you could necessarily eyeball from a Sleepyhead graph. (I'd love a plug-in that would map out and sync up percentage changes in flow rate and tidal volume and minute ventilation--but it would still be guesswork. And you'd better believe if there was a solidly reliable algorithm, ResMed/Respironics would have already built this reporting into their data reports for professionals.)

I get the urge to try to make sense of it visually, though--data is fun, and we're all trying to figure out how to improve sleep and correlate sleep quality markers beyond just AHI. Again, it's all tea leaves (at this point in technology, at least--and there are a couple potential Zeo replacements on the horizon for crude, at-home EEG and it will be interesting to add that to the guessing game).

[Sludge]

[...there is a robust correlation between REM sleep and respiratory rate (it increases during REM) and tidal volume (it decreases during REM). As well as becoming more shallow and more rapid, respiration also becomes more variable in REM. This is according to every sleep medicine textbook and article I've been able to find online.

Maybe in phasic REM (to the tune of ~75% of the time), but respiration in tonic REM tends to be pretty regular.

[mellabella]

[...there is a robust correlation between REM sleep and respiratory rate (it increases during REM) and tidal volume (it decreases during REM). As well as becoming more shallow and more rapid, respiration also becomes more variable in REM. This is according to every sleep medicine textbook and article I've been able to find online.

Maybe in phasic REM (to the tune of ~75% of the time), but respiration in tonic REM tends to be pretty regular.

And God help us if we're trying to figure out the difference between tonic REM and phasic REM on Sleepyhead! I greatly simplified (and debated whether to include variability at all, but you can detect periodic variability on my graphs)--but the two most "useful" (as far as that goes, and probably not very far) metrics on sleepyhead are respiratory rate (speed, not variability) and tidal volume. Still holding out for my space-age, affordable home EEG!

[HerbM]

There is an open source hardware (and software) project named "OpenEEG" with schematics and full build plans for a 2 (;up to 6) channel EEG.
http://openeeg.sourceforge.net/doc/hw/

I don't want to go to the trouble to assemble such a device (and have to radically improve my soldering skills since it apparently is not simply 'stick and heat' with some of the delicate components.

However there is also a company that is selling pre-assembled (still build at you own risk) boards and even a packaged device.
https://www.olimex.com/Products/EEG/Ope ... e-hardware

Looks like the full thing for a 2-channel device is about US $200 (it's in euros and you have to buy a few things) and you probably would need to sew it into a head band etc..

Of course you would need an attached PC which means a laptop would probably be most convenient for moving to the bedroom (although in principle any unit with USB would work.)

Serious, consideration but first I need to get my recording Pulse Ox to work (driver issues.)

[mellabella]

There are at least 3 consumer devices in development that I know of that are aiming to do the same kind of simple EEG (at least one with eye movement sensors as well) that the Zeo was advertised as providing, before that company went out of business. They look much more consumer-friendly and I'll probably hold out for that, but that's an interesting option you found!

[DoriC]

[...there is a robust correlation between REM sleep and respiratory rate (it increases during REM) and tidal volume (it decreases during REM). As well as becoming more shallow and more rapid, respiration also becomes more variable in REM. This is according to every sleep medicine textbook and article I've been able to find online.

Maybe in phasic REM (to the tune of ~75% of the time), but respiration in tonic REM tends to be pretty regular.

What's phasic and tonic?

[Pugsy]

What's phasic and tonic?

Normal sleep is divided into non–rapid eye movement (NREM) and rapid eye movement (REM) sleep. NREM sleep is further divided into progressively deeper stages of sleep: stage N1, stage N2, and stage N3 (deep or delta-wave sleep). As NREM stages progress, stronger stimuli are required to result in an awakening. Stage R sleep (REM sleep) has tonic and phasic components. The phasic component is a sympathetically driven state characterized by rapid eye movements, muscle twitches, and respiratory variability. Tonic REM is a parasympathetically driven state with no eye movements. The REM period length and density of eye movements increases throughout the sleep cycle.[1]

http://emedicine.medscape.com/article/1188226-overview

[Islandwoman]

Wow, lots of input I have been caring for my grand-dog(recovering from surgery) all afternoon. I still have her for a bit but I scanned quickly and saw some interesting things. Compared Capn's graph with mine and I have been interpreting pretty much the same. It is so enjoyable to know more I can't seem to stop. My AHI last night Was .46 with 4 Hypopneas and 30 RERAs total. For any Newbies it took me 10+ months of cpap to get below a 1.0. I am happy with the dream info thanks all and will now concentrate on RERAs 90% seem to be awake or probably dreaming. I now have long strips of flow graphs in unsullied black. When I started I could hardly view the flow itself, it looked like a long very hairy caterpillar! Thanks again to everyone on cpaptalk You are all the reason I am doing as well as I am.

[Islandwoman]

Ouch, Pugsy I read what was in your link and unfortunately am probably among the "elderly" at least I was. I believe am doing way better on cpap so will move myself down a notch to adult. Perhaps cpap is the air fountain of youth!

[wilsonintexas]

There are some ZEO units available for about $400 new on amazon. It will give pretty accurate sleet stage data for you. There is also a viewer available that will let you download the data and see graphs in 30 second intervals. (recent post in this forum, if you can not find it pm me and I will track it down. ) It is called ZEOVIEWER. It can export a CSV file, but it is in 30 second intervals and will not currently load into sleepyhead.

I find zeo very helpful in assessing my sleep stages directly, without having to infer.

From my original research before I bought it it was between 80-90% accurate when compared to a full sleep study.

You will need replacement "headbands" these are really sensor pads and there are several posts out there on how you can adapt normal eeg pads or make your own pretty cheaply.

It is more money but makes it easy to see what is happening in your stages of sleep.

here is a sample of the display:

http://i.imgur.com/rw7gdzw.png

[Pugsy]

Ouch, Pugsy I read what was in your link and unfortunately am probably among the "elderly" at least I was. I believe am doing way better on cpap so will move myself down a notch to adult. Perhaps cpap is the air fountain of youth!

Hey, I never said I understood it...I just put it out there for reading.
Got question "what is it" so I found the definition and put it out there.. I didn't understand very much of it myself.

I don't really worry so much about when REM shows up or if I can spot it or really spend much time thinking about it beyond maybe seeing a little cluster and it sort of looks like probably REM time frame...shrug my shoulders and move on.
I can't do a thing about it one way or the other. I either get REM or I don't and I can't change anything about it so for me to spend an inordinate amount of time sorting through miles of respiratory rates and flow graphs isn't something I am interested in doing. It changes nothing except maybe satisfy a curiosity itch and I don't have that itch. No if you and the Capn have that itch...hey, go for it. Have fun.

To be honest I have enough on my plate as it is without trying to figure it out and making out like I know what I am doing and then someone would want me to "teach how to read it" and I have no clue...it's guessing what I do. Educated guess for sure but without EEG brain wave data to support and give us some sort of baseline to go by it's kinda a theory with nothing to compare to. Now if I had my hands on a night with EEG data...and then I had my hands on the Sleepyhead graphs so I could put both under the microscope and get an idea exactly what confirmed REM and other sleep stages might look like on the same SH graphs...yeah..maybe I could come up with more than a guess but I don't have that EEG data and it's doubtful I ever real and I don't really need it anyway.

I am more of a problem solver.. Give me a problem and I try to fix/solve it. That's where my strengths lie. Keep trying ideas until something works or I run out of ideas.
I have zero control over REM...can't make it happen when it doesn't...can't make it not happen when it does. There's nothing about it that is "fixable" so to me if I can't fix something not much sense in me worrying about it or trying to pick out which 20 minutes was REM and which wasn't.
I am good with "close enough for government work" when it comes to stuff that can't be fixed. Just being curious about when REM occurs isn't a problem to me but it might be a really big itchy problem for others and if that's the case then by all means put all those graphs under the microscope but without a baseline EEG and same night SH reports...it can't be proven if what we think about those respirations are REM related or not and I need proof or some ability to compare known proven baseline data before I want to go down that road.

Sludge would be the one to help in that area. He's seen enough EEGs and respirations side by side that I have no doubt he can probably do more than guess. I have never seen any and again...if there is a problem that can't be fixable... that's a lot of work to learn how to interpret something that I can't do anything about anyway.

But you know what...everyone is different in all sorts of ways and what's important to me is not necessarily important to someone else and vice versa. It never ever hurts anyone to take the time and effort to expand their knowledge in areas that interest them. Right now my interests lie in ASV therapy...that's a big elephant for me to eat but I am slowly eating it up small bite by small bite. Maybe once that elephant is consumed I might find that I want to tackle trying to figure out sleep stages by respirations and flow rates or maybe you can save me some work and get it done before I am ready to start on that elephant and you can shorten my work load. Nothing would make me happier.
I would need a known baseline to start with though...that would actually make the analyzing easier along with being more accurate. It's just me...I need some sort of known baseline to start with and that's why I use the normal sleep stage cycles...it's proven. Yeah, lots of room for variables but that's actually quite common in the medical community anyway...nothing is ever truly cut and dry or black and white.

If you catch Sludge on a good day he might point you to some materials to help you sort things out...but beware...when he gives homework...it's never easy homework. Been there and done that and he doesn't give good hints unless he is in a really good mood.

Learning is hardly ever easy...most of the time it takes a lot of hard work...and that's why it is so rewarding when we learn something and understand it with confidence.
Look how far you have come since you've started and look at what all you are understanding and now want to tackle?
Would you have had even the remotest idea a year ago that you would be here trying to sort out sleep stages using these graphs? Feels good doesn't it?

I don't know your age but I am 62 now..learning new stuff takes a LOT more effort now than it did 35 to 40 years ago. Those old brain cells don't like being woke up. That's sort of why I limit the elephants I want to eat...I just can't handle too many of them at one time anymore.

Should you happen to stumble across some good baseline EEG and sleep respiration and flow rate side by side comparisons...send me the link if you think about it. I will save it for the next elephant buffet reading material.