While looking for material for another thread, I ran across Snoozehunter's Sleep Studies, and saw some great examples of pressure-induced central apneas.  After talking it over with Snoozehunter, who is eager to share any information that might prove helpful to the group, we present them here.

As you may remember, Snoozehunter had a heck of a time, starting out with a barely need-to-treat AHI of 8.7 and oxygen desaturation nadir of 90.6%.  A couple of very aggressive CPAP/BiPAP titrations later, with CPAP pressures of 19 and BiPAP pressures in the neighorhood of 19/15, things were out of hand:

Snoozehunter's Sleep Saga

I mean, if your NREM AHI is 6.3 with nothing, how can it get to 82.4 with 15/11 of BiPAP?

A review of the CPAP titrations showed NO obstructive apneas, they were all central, and the decision was made to stop pressure therapy.

Snoozehunter reports that things are doing well without pressure support:

Update

So we'll see what happens.

Anyway, as it turns out, this is a great example of how to generate central apneas with pressure support.  In all fairness to CPAP, all of the centrals seen here were on BiPAP, low level CPAP never really got a fair shot.

This is a 300-second window, BiPAP in the neighborhood of 15/11.  Heart Rate Variation is noted in the ECG R-R channel:



This next histogram shows the respiratory events, with the oxygen desaturations, in relation to the sleep stages.  This was a split study, with the first part of the night showing some desaturation.  But what was not clear then, and is still not clear now, is the source of these desaturations.  But all these events also appear to be central.

What is important to note is the relation of central apneas to sleep stages.  These types of centrals are generally a NREM phenomena, and a Stage 2 phenomenon at that.  You can see when SWS (Stage 3/4) and REM appears, the centrals disappear.  Whole night summary:



And to show how abrupt this change is, here's a progression from stage 2 to stage 3 sleep.  Watch the CPAP FLOW channel and how the r-r interval and oxygen saturation stabilize.   1000 second window:



It is so tempting to call this CBSD, but in order to have that, you have to have the "C"-- complex disorder, or both central and obstructive components.  And the obstructive part is just not that well-defined.

But if you ever want to try a CS-2...
sleepydave

Quote:

This was a split study, with the first part of the night showing some desaturation. But what was not clear then, and is still not clear now, is the source of these desaturations. But all these events also appear to be central.

OK, this is the beginning of the study.  And there's a couple of things wrong.  The biggest problem is that the Airflow channel is blown.  Or not even hooked up.  So all you get is a bunch of artifact.

There are respiratory events though.  The Abdomen and Chest channels show severe reduction in effort compared to the episodic "normal" breath efforts.  Enough to call ithem apneas.  And even if that weren't the case, this is a 300-second window, and there are quite a few 10-seconds-es with absolutely no movement that you can call CAs.

Even without the benefit of the Airflow channel, you could probably assume them to be central apneas anyway, if there's no effort, you can be assured that there is no flow.  Calling them hypopneas was a "safe" bet.

Interestingly, that these can be substantiated somewhat as respiratory events can be seen in the EKG channel, where there are periods of high amplitude that correspond to the respiratory events, separated by reduced amplitude, or the periods of "normal" breathing.



Let's say they were obstructive apneas or hypopneas, tho.  Maybe the belts didn't pick up the effort signals properly, whatever.  But as the night wore on and centrals starting to come out of the woodwork, you gotta say, "Hey, maybe there isn't any obstruction, or not as much as we thought."  It's a heckuva lot more plausible that this is simply periodic breathing associated with sleep-onset.  The desats are always of concern, but the failure to put any kind of a dent in them suggests they have nothing to do with obstruction.  Further, desaturations in periodic breathing perpetuate the phenomenon, another reason that an aggressive BiPAP titration was doomed from the start.
sleepydave

sleepydave wrote:

This next histogram shows the respiratory events, with the oxygen desaturations, in relation to the sleep stages.  This was a split study, with the first part of the night showing some desaturation.  But what was not clear then, and is still not clear now, is the source of these desaturations.  But all these events also appear to be central.

What is important to note is the relation of central apneas to sleep stages.  These types of centrals are generally a NREM phenomena, and a Stage 2 phenomenon at that.  You can see when SWS (Stage 3/4) and REM appears, the centrals disappear.  

The histogram shows the relation of centrals to sleep stages really well. Unfortunately, on bipap I was spending 70-75% of the time in stage 2 sleep. Looking at this, I can see why I never woke feeling rested. REM was almost non-existent and I got very little stage 3/4 where everything stabilized.

Thanks for taking another look at these, Dave.

SnoozeHunter

sleepydave wrote:

Let's say they were obstructive apneas or hypopneas, tho.  Maybe the belts didn't pick up the effort signals properly, whatever.  But as the night wore on and centrals starting to come out of the woodwork, you gotta say, "Hey, maybe there isn't any obstruction, or not as much as we thought."  It's a heckuva lot more plausible that this is simply periodic breathing associated with sleep-onset.  The desats are always of concern, but the failure to put any kind of a dent in them suggests they have nothing to do with obstruction.  Further, desaturations in periodic breathing perpetuate the phenomenon, another reason that an aggressive BiPAP titration was doomed from the start.
sleepydave

I've seen references to periodic breathing associated with sleep-onset before and it seems to be considered either normal or at least not unusual. But, for someone like me, who spends a lot of time in the sleep onset phase, isn't it as problematic as regular sleep apnea? Why does it occur?

How do you differentiate between pressure induced centrals and true central sleep apnea?

And how come my r-r response looks different up there?

Sorry. I'm full of...questions.  

SnoozeHunter

SnoozeHunter wrote:

I've seen references to periodic breathing associated with sleep-onset before and it seems to be considered either normal or at least not unusual. But, for someone like me, who spends a lot of time in the sleep onset phase, isn't it as problematic as regular sleep apnea? Why does it occur?

How do you differentiate between pressure induced centrals and true central sleep apnea?

And how come my r-r response looks different up there?

Hi snoozehunter:
The problem, of course, is that we weren't able to get a hold of the June 2003 sleep study, the one on ambient pressure that showed only an AHI of 8.7, and this last split study has a bunch of problems associated with it.  Since sleep-onset central apnea is a phenomenon of periodicity (as is pretty much all central sleep apnea, as opposed to "central apneas", which may occur post-arousal or as a normal consequence of REM), if you had them only at the beginning of the study at an index of 60, well an hour of that and there's the AHI for the night.  A key would have been to see if you were (are) able to ever achieve stable non-SWS NREM (stage 2) sleep.

If your diagnostic study has no (or few) centrals, and the titration has a million centrals, then voila!  Pressure-induced centrals.

Jean Krieger in her chapter on Breathing During Sleep in Normal Subjects in Principles and Practice of Sleep Medicine by Meir H. Kryger summarizes the incidence of periodic breathing at sleep onset from 40-80% of normal subjects, and lasting from 10-60 minutes.  The trigger is hypercapnia, as opposed to the hypocapnia that creates all the havoc in CSBD and CSR.
Oh yeah, r-r intervals...
sleepydave



Last edited by sleepydave on Thu Jul 06, 2006 12:50 pm; edited 2 times in total

Hi again!

Well, you're right!  There is something a little odd with the r-r intervals!  In the example from the other thread on on Heart Rate Variability, the obstructive apneas show a progressive increase in the r-r interval, or a slowing of the heart rate:



Yet in yours, the central apneas show a progressive decrease in the r-r interval, or an increase in the heart rate:



I dunno there, snooze, I'm beginning to think you're just a very contrary person.

Well, seeing as how we can juggle these waveforms around, let's shift 'em a little bit and see what we can see.  Let's move up the O2 saturation waveform:



Wha...?
sleepydave

sleepydave wrote:

Yet in yours, the central apneas show a progressive decrease in the r-r interval, or an increase in the heart rate:

Is that normal with centrals?

Quote:

I dunno there, snooze, I'm beginning to think you're just a very contrary person.

LOL! I'm beginning to think so, too.

Quote:

Well, seeing as how we can juggle these waveforms around, let's shift 'em a little bit and see what we can see.  Let's move up the O2 saturation waveform:

Hey...that looks like a perfect match. Is it supposed to do that? If it's cause and effect, which is the cause and which is the effect?

SnoozeHunter

There are a number of factors at work here, but let's cover 3 anyway.

First of all, these waveforms are not necessarily all real-time relative to each other.  For instance, changes in ventilation take a little time to show up in the pulse oximetry, which is typically measured at the finger, like maybe 20 seconds later.  And pO2, pCO2 and pH values need to travel to chemoreceptor sites to actually generate the response in HR.  

But that's one heckuva coincidental tracing.

Of course, there is research on this, and in a study of Cheyne-Stokes central apnea involving fixing the O2 saturation with supplemental oxygen, the HRV oscillations were not affected:

Influence of Cheyne-Stokes Respiration on Cardiovascular Oscillations in Heart Failure

But, if you read the article, what was interesting was...

...that correcting the pCO2 does.

C'mon, you all knew we were headed back there anyway, didn't you?

Let's throw in another hold your horses.  At this point, the situation is iatrogenic.  This is a problem that was generated by aggressive BiPAP.  This problem could very well be solved by removing the first "solution", the overly aggressive BiPAP titration.  This mechanism can be likened to the increase in ventilation by PAV in the first look at loop gain in TMOAT.  I have little doubt that the BiPAP has made the problem exponentially worse.

But let's review that thought carefully-  "Made the problem..."  That these horrendous numbers of central apneas could be generated gives great support that there is an underlying issue, at least that of a very low central apnea  threshold.  All you need is the trigger to set it in motion, and then it tends to perpetuate itself.  And if there is any obstructive component in there, well, the cascade is set in motion.

And finally, the heart rate variation (HRV).  Given different circumstances, heart rate can increase and decrease, and for a heap of different reasons.  However, the increases/decreases can take on a repetitive pattern, and this is what HRV is all about.  Examples of HRVs are in the other thread

HRV

But now let's give them some names:

There are 4 HRV bands, based on frequency:

High Frequency (HF) (0.15 to 0.4 Hz)  This wave would have a length of 2.5 - 6.7 seconds.  In sleep, this activity will often translate into the phenomenon of Respiratory Sinus Arrhythmia (RSA), where the HRV would correlate directly with each respiration.

Low Frequency (LF) (0.04-0.15 Hz)  This wave would have a length of 6.7 - 25 seconds.  In sleep, this activity will correlate with the respiratory events.

Very Low Frequency (VLF) - not my department

Ultra Low Frequency (ULF) - really not my department

Now, if we get a copy of the pulse oximetry on room air and ambient pressure, perhaps we can see if the the desaturation pattern persists past sleep onset.  Unfortunately, without sleep staging, then the assumptions take over, we haven't the slightest inkling of sleep stage, or even if you're asleep or not.  And tracking HRV is difficult at best, because you really need to look at r-r intervals, and the downloads from pulse oximeters usually signal average over a window, say 4 seconds, so the signal will be dampened a ton, even if we're only looking for LFs.
sleepydave

So eventually, Snoozehunter was able to get on low level CPAP (8 cmH2O) and have nocturnal oximetry performed.  It looked like this:



And before you jump up and say, "Aha, desats and LF HRVs!" note that by log periods of wake were reported at 0100 (briefly) and from 0300 forward.  The areas of apparent sleep are quite stable.  So what we have here is at least the potential of some pretty effective therapy.  Without PSG, we don't know for sure, but compare this with the BiPAP oximetry and we're talking night and day here.

So if less is more (more or less), then a trial without CPAP was attempted:



This night was a little tough, the areas of sleep are underscored in orange.

Hmmm.  Y'know, overall there is some good stuff.  The oximetry looks pretty good (maybe some stuff at approximately 0045 and 0200) but again, when sleep occurs, it seems fairly continuous.  On the other hand, this night is such a wreck, the best thing to do would be to repeat it until a full night of sleep (or "normal" anyway) is obtained.  If you're looking for a clean bill of goods, this won't be the one to do it.

Regardless, getting off those horrendous BiPAP pressures was critical, and whether the Snoozehunter ends up on low-level CPAP or CPAP-free, we're certainly headed in a much better direction than before.

But we'll still keep the dead space handy if we need it.
sleepydave

sleepydave wrote:

Of course, there is research on this, and in a study of Cheyne-Stokes central apnea involving fixing the O2 saturation with supplemental oxygen, the HRV oscillations were not affected:

Influence of Cheyne-Stokes Respiration on Cardiovascular Oscillations in Heart Failure

But, if you read the article, what was interesting was...

...that correcting the pCO2 does.

C'mon, you all knew we were headed back there anyway, didn't you?

N-N-Nooo! Not THE MOAT!! I'll breathe! I swear I will!

Quote:

On the other hand, this night is such a wreck, the best thing to do would be to repeat it until a full night of sleep (or "normal" anyway) is obtained. If you're looking for a clean bill of goods, this won't be the one to do it.

Umm...that IS a normal night for me. That was me on Lunesta. I don't sleep very well.

I was supposed to follow-up with the pulmonologist in 6 weeks, but it turns out he doesn't have any openings until the 10th week out. In the meantime, I'm going to call the sleep lab and tell them I've changed my mind about the psg. I declined the doc's invitation to do a diagnostic, but now I'm wondering if any apneas would show up in the absence of pressure support. Looking at the graphs, the night on low level cpap looks a little better to me. SpO2 is good during sleep periods both nights, but maybe cpap helps me sleep better.

SnoozeHunter

Quote:

Not THE MOAT!!

Yes, the MOAT!!

But while there's a lot of stuff floating in there, we have to be careful what we pull out.

While almost everybody was concerned with "adding CO2", there were a bunch of other important points, including the role of an enhanced chemoresponse system, the location of the central apnea threshold and the importance of having a lung-to-controller lag that made ventilatory response 180 degrees off, so you ended up having a response piled on top of another response.  And while that happened in the BiPAP titrations, it may be too soon to call this CSBD, because there weren't any obstuctive events seen.  Course, they didn't get a very good chance to appear, and we may eventually find that the obstructive component initially found in June 2003 is valid and still there.

But let me re-emphasize:

Quote:

Jean Krieger in her chapter on Breathing During Sleep in Normal Subjects in Principles and Practice of Sleep Medicine by Meir H. Kryger summarizes the incidence of periodic breathing at sleep onset from 40-80% of normal subjects, and lasting from 10-60 minutes.  The trigger is hypercapnia, as opposed to the hypocapnia that creates all the havoc in CSBD and CSR.

That's a whole different central apnea mechanism.  So in the PSG on ambient pressure, we're going to be looking at the type and location of events.  (BTW, it would be nice to track ETCO2 during this time, but that's probably asking for a lot).

As you fall asleep, pCO2 rises as ventilation decreases.  Sleep onset is associated with a 2 - 7 mmHg increase in pCO2.  If you have a whole bunch of central events in the Wake-Stage1-Stage 2 transitions, then hypercapnia may be the culprit, triggering the cascade.  If they're purely centrals, you have to leave them alone.  When stable NREM (stable Stage 2 and SWS) and REM occur, they should disappear.

If, once you fall asleep, and you have obstructive events (and we would be looking closely at REM) then you could treat that with low-level CPAP.

If, once obstructive apneas start, AND you're a little hypocapneic, AND central and/or mixed events start (with or without pressure support), then it may be CSBD and dead space may be attempted.

If, once you fall asleep, central apneas start and persist in NREM (they probably won't appear in REM), then that could be purely central apnea (idiopathic, probably, if you're OK cardiac-wise) and again, some dead space may be attempted.  But that would be unlikely, since they weren't there in June 2003, and nothing has really changed since then.

So until one of those guys answers the question

Quote:

if your NREM AHI is 6.3 with nothing, how can it get to 82.4 with 15/11 of BiPAP?

then I think you're right, get a good baseline PSG (no more splits!) and really try to figure out what's what.

Make sure they use a pressure transducer to look for the subtle changes associated with UARS.

Just picking a card out of the deck, tho, yeah, that 8 cmH2O nocturnal oximetry looked pretty good, low-level CPAP might end up the way to go.  
sleepydave

sleepydave wrote:

C'mon, you all knew we were headed back there anyway, didn't you?

SnoozeHunter wrote:

N-N-Nooo! Not THE MOAT!! I'll breathe! I swear I will!

sleepydave wrote:

Yes, the MOAT!!
----
But while there's a lot of stuff floating in there, we have to be careful what we pull out.
sleepydave

ROTFL!  You two!!     

In case readers are scratching their heads, going "Wha....wha...???  What's "THE MOAT?" ...  sleepydave christened this extra, extra, EXTRA long thread on another board, "The Mother Of All Threads."  Lotsa' flotsam and jetsam there, along with tasty morsels:

Nov 06 2005   subject: Sleep Disordered Breathing - a New Category. Excellent artic

Another thread, on yet another message board, has some interesting speculations about CSDB (complex sleep disordered breathing) by a poster nicknamed -SWS:

Jun 28, 2006   subject: APNEA -v- HYPOPNEA

Actually, that MOAT may have something interesting in it.  CSDB - I had read before that Complex just meant both Central and Obstructive, not that it was a unique condition to be treated differently.

As a person with lots of hypopneas (and a surprising, though not serious, number of Centrals) I hope that's not what I have.  It sounds much harder to treat  

Now where'd that snoozehunter go?
sleepydave

sleepydave wrote:

Now where'd that snoozehunter go?
sleepydave

Here I am, over here caught in the loop!

Actually, I just got back from a sleep study. The doc's office called me at work yesterday and said they had a cancellation so I got in unexpectedly last night. I don't have a feel for how it went and the tech that turned me loose this morning was not the one there during the night. When I asked her how it went she said she'd just gotten there and didn't know. Then the standard bit about it being scored later and the doctor will go over it with me next week.

That tracing looks the way I feel when I sleep. Doze off...stop breathing...surface...doze off...stop breathing...surface...

How do I get out of the loop?

Gotta go scrub paste out of my hair.

SnoozeHunter