Link between OSA and the "beer gut"

Gottcha. Yeah, Pdi can be calculated either way.

split_city wrote:Pdi = Pga - Pes. For simplicity, if Pga remains constant and Pes goes down, overall Pdi will go up. If Pes goes up, Pdi will go down.

split_city wrote:I have simply derived a new channel showing Pdi. The Pes channel has a lot of ECG artefact so it's easy to identify deflections. The Pdi clearly shows the oscillations mirror that of the Pes channel.

How did you calculate Pdi? That waveform looks to be simply the inverse of Pes and does not take into account those huge Pga swings. Can you put up the pressure grid?

split_city wrote:I did send Steier an email. I have spoken to him in person (conference in San Fran a few years back) and also via email. I will be sure to add his thoughts once I get a reply.

I wonder if we can get him to put a pressure grid on his waveforms, cause his Pdi doesn't seem to calculate correctly either.

Muffy

Muffy wrote:

split_city wrote:Pdi = Pga - Pes. For simplicity, if Pga remains constant and Pes goes down, overall Pdi will go up. If Pes goes up, Pdi will go down.

split_city wrote:I have simply derived a new channel showing Pdi. The Pes channel has a lot of ECG artefact so it's easy to identify deflections. The Pdi clearly shows the oscillations mirror that of the Pes channel.

How did you calculate Pdi?

Simply calculated Pga - Poes

Muffy wrote:That waveform looks to be simply the inverse of Pes and does not take into account those huge Pga swings. Can you put up the pressure grid?

Sorry, should have put the grids up earlier. Those Pga swings aren't very large.

Split_City: I have a proposal for you............if you search the threads, you will see a recent discussion questioning the possibility of feline OSA....My Himalyan, Sami, with a structural nasal/oral breed smushed in face, snores-n-sleeps by my side is an subject of question.........I think a study of feline OSA would be much more innovative than beer gutted men...which begs the question of OSA.........so whatcha think? OSA and the Feline: I'll volunteer my subjects as long as they sleep in their own bed!!!

elg5cats

elg5cats wrote:Split_City: I have a proposal for you............if you search the threads, you will see a recent discussion questioning the possibility of feline OSA....My Himalyan, Sami, with a structural nasal/oral breed smushed in face, snores-n-sleeps by my side is an subject of question.........I think a study of feline OSA would be much more innovative than beer gutted men...which begs the question of OSA.........so whatcha think? OSA and the Feline: I'll volunteer my subjects as long as they sleep in their own bed!!!

elg5cats

Interesting...the Bulldog is the only other animal that I know of that has confirmed OSA.

The English bulldog: a natural model of sleep-disordered breathing

J. C. Hendricks, L. R. Kline, R. J. Kovalski, J. A. O'Brien, A. R. Morrison and A. I. Pack
Department of Clinical Studies, School of Veterinary Medicine, University of Pennsylvania, Philadelphia 19104.

To establish a natural model of sleep-disordered breathing, we investigated respiration during wakefulness and sleep in the English bulldog. This breed is characterized by an abnormal upper airway anatomy, with enlargement of the soft palate and narrowing of the oropharynx. During sleep, the animals had disordered respiration and episodes of O2 desaturation. These were worst in rapid-eye-movement (REM) sleep, with most bulldogs having O2 saturations of less than 90% for prolonged durations. In contrast, control dogs never desaturated. In REM sleep, the bulldogs had episodes of both central and obstructive apnea, the latter being associated with paradoxical movements of the rib cage and abdomen. During wakefulness, the bulldogs were hypersomnolent as evidenced by a shortened sleep latency (mean of 12 min compared with greater than 150 min for controls). This animal model should facilitate studies of the natural history of the sleep apnea syndrome and its complications.

split_city wrote:

Muffy wrote:How did you calculate Pdi?

Simply calculated Pga - Poes

Muffy wrote:That waveform looks to be simply the inverse of Pes and does not take into account those huge Pga swings. Can you put up the pressure grid?

Sorry, should have put the grids up earlier. Those Pga swings aren't very large.

OK, cause since the inverse of Pdi results in a virtual mirror-image overlay of Pes:



it suggested that Pdi = 0 - Pes was used in error, especially if you look at the overlay



there's all kinds of Pga hanging out in those last couple of breaths because baseline Pga drop (horizontal line) is not reflected in Pdi and there's no offsetting Pes values to otherwise account for that.

Can you make all the pressure grids the same scale and filter out those COs?

Muffy

The max Pga difference between the horizontal line and the nadir in Pga of the last breath is ~0.5cmH2O.

I haven't yet had to filter Poes channels thus far for any of my data analysis so I am unsure of the procedures. My supervisor is on holidays at the moment but will have limited access to email during this time. I have sent him an email to see if he can help. Otherwise, it will have to wait until after the 18th Jan.

Here's a picture with all pressure channels at the same scale. Don't know whether that helps or not.

split_city wrote:Here's a picture with all pressure channels at the same scale. Don't know whether that helps or not.

Yes, I think it clears everything up quite nicely, thanks very much.

Muffy

While the calculation of transdiaphragmatic pressure (Pdi = Pga - Pes) has the appearance of a "take away" math problem, it makes more sense conceptually to consider it more like a "Sum of Forces". This effect can be clearly seen in the Pankow example, where the Pga swing was about 10 cmH2O. Pga augments the visually inverted Pes waveform:



With the original waveforms showing mongo Pga waves:



their effect would have been easily discernible (underscoring now how important knowing what the y axis values are to be able to actually calculate Pdi).

With the waveforms now on a standardized scale, although the Pga effect in this Pdi calculation is far more subtle and somewhat obscured by the cardiac oscillations, they can nonetheless be seen, in some breaths more clear than others:



Muffy

split_city wrote:Here's a picture with all pressure channels at the same scale. Don't know whether that helps or not.

Although it would have been a little more helpful to post that corrected grid a little sooner. It's a little tough trying to perform algorithmic calculations using "Paint":



Muffy

Still trying to work on removing CO. Macros aren't behaving at this stage but hopefully the supervisor fixes the bug, despite being on holidays . I think removing the CO from the below Pes channel maybe a little difficult given that the oscillations don't always appear to line up with the CO. For example, compare Pes data...



with another example...



Basically we cut out the artefact in the Pes channel and join the two points with a straight'ish line.

OK great, but even with the COs, I don't think it will make a difference in your overall calculations because your waveforms are entirely over the 0 baseline. However...

Muffy wrote:

split_city wrote:I did send Steier an email. I have spoken to him in person (conference in San Fran a few years back) and also via email. I will be sure to add his thoughts once I get a reply.

I wonder if we can get him to put a pressure grid on his waveforms, cause his Pdi doesn't seem to calculate correctly either.

Just to complete that thought, since the Steier raw value Pdi < Pes (and assuming that he has used standardized pressure grid), it can be graphically calculated that about 19.3% of the Pes waveform is > 0 cmH2O (which is easily plausible, cause this is another fat guy), represented thusly:



However, what now makes this pertinent is that as you can see, all of the positive Pes is composed of cardio-oscillatory artifact, which must be filtered out before Pdi can be accurately calculated. This will result in a dramatic change in Pdi value.

Muffy

I guess at this point, a discussion about the interpretation of Pdi when the Pes waveform hangs over both sides of the 0 cmH2O threshold might be a little "over the top" for a discussion forum that includes such references as "Let's tape the exhalation ports closed and see what happens!"

Muffy

split_city wrote:So iPEEP is the pressure difference between end-expiratory Pes and the onset of flow. But I guess there isn't any flow to begin with because he is initially obstructed. So should iPEEP represent the pressure difference between end-expiratory Pes and when (expiratory) flow reaches zero? Please correct me if I'm wrong!

The definition of iPEEP is the pressure in the airway at the point of inspiration. To make this differentiation clear, let's call it iPEEP-AW.

What y'all are measuring is actually pressure in the abdomen (iPEEP-ABD) and saying "Well, this must be happening in the airway too."

BTW, how can one call it iPEEP if the whole Pes is subatmospheric such that the point of iPEEP is < 0 cmH2O?

So anyway, in answer to your question, when iPEEP is measured in the airway, the critical characteristic will be seen, namely, there is is still expiratory flow when the ensuing inspiration begins.

split_city wrote:But it would certainly be interesting to know what's obstructing his airway during expiration. Time to stick a camera down his airway I reckon.

Didn't Mary Morrell already do that?

Muffy

Muffy wrote:

split_city wrote:So iPEEP is the pressure difference between end-expiratory Pes and the onset of flow. But I guess there isn't any flow to begin with because he is initially obstructed. So should iPEEP represent the pressure difference between end-expiratory Pes and when (expiratory) flow reaches zero? Please correct me if I'm wrong!

The definition of iPEEP is the pressure in the airway at the point of inspiration. To make this differentiation clear, let's call it iPEEP-AW.

So why do Pankow et al, Steier et al, Lotti et al and O'Donoghue et al all define (dynamic) iPEEP as the pressure difference between end-expiratory Pes/start inspiratory effort and the onset of flow? Is that different to static iPEEP? I'm reading up on the two...

"PEEPi is usually measured under static conditions by
occluding the airway opening at end-exhalation (PEEPi-
stat). It is considered to represent mean PEEPi
in the lungs with inhomogeneous regional PEEPi.
Dynamic PEEPi (PEEPi-dyn) is the airway pressure required
to overcome expiratory flow. It is considered to
represent the lowest regional value of PEEPi and reveal
a lower value than static PEEPi in the presence of
significant time constant inequalities."

Taken from Fujino et al 1999

Muffy wrote:What y'all are measuring is actually pressure in the abdomen (iPEEP-ABD) and saying "Well, this must be happening in the airway too."

Well, if there is an airway obstruction, then increased abdominal muscle contractions appears to influence Pes.

Muffy wrote:BTW, how can one call it iPEEP if the whole Pes is subatmospheric such that the point of iPEEP is < 0 cmH2O?

Good point. But isn't that true for "normal" people? Doesn't Pes go above atmospheric pressure in say COPD patients?

Muffy wrote:

split_city wrote:But it would certainly be interesting to know what's obstructing his airway during expiration. Time to stick a camera down his airway I reckon.

Didn't Mary Morrell already do that?

Muffy

Yeah she has, but we haven't in our lab. Hey, we compress the abdomen, stick wires in the genioglossus, insert large nasal catheters in to measure gastric and esophageal pressure, plus diaphragm EMG, why not get a camera down as well?

Oh and HAPPY NEW YEAR! You're logged in pretty early though......or haven't you gone to bed yet?