long slow decay in inspiratory flow rate

[cathyf]

One of the striking things that I see in my data is that I have these multi-hour periods where the top of the flow rate curve decays slowly and steadily, punctuated all along by short spikes that my fitbit sees as arousals. But the bottom of the curve is jaggy up and down, but not rising or falling in the big-picture view.

Here's a particularly striking example from a couple of nights ago. First the whole-night view:
[attachment=33156]

here's what the Fitbit made of the night:
[attachment=33157]

I'm focused on what happened during the period between falling asleep and just after 2am. I start with the inspiratory flow rate trundling along at ~15/16. By three hours later I've decayed down to ~5/6/7 or so. Here's a closeup of the end:
[attachment=33155]

Now when this happens suddenly it's an event -- to quote the AASM hypopnea rules

AASM Hypopnea rule IA: A 50 % or greater fall (but less than 90%) in the nasal pressure signal excursion for at least 90% of the event duration from pre-event baseline, the duration of which lasts at least 10 seconds, and is associated with a 3% or greater desaturation or EEG arousal and continued or increased inspiratory effort throughout the entire period.

AASM Hypopnea rule 1B: A 30 % or greater fall (but less than 90%) in the nasal pressure signal excursion for at least 90% of the event duration from pre-event baseline, the duration of which lasts at least 10 seconds, and is associated with a 4% or greater desaturation with continued respiratory effort.

Clearly here "baseline" means the time immediately preceding the event. But what does it mean (if anything) when it's a long slow steady decay that leads me to a flow rate that's 1/3 of what it was when I first was asleep, and it's followed by a sharp jump up as opposed to a sudden collapse?

Those first three hours are pretty darn good sleep for me. If I believe the Fitbit, good solid deep sleep at the beginning, and a nice REM period later -- and not very many arousals. But the flow rate decay is peculiar!

[Ratchick]

Not an expert on this but I'm gonna guess that as your flow limitation increases, you roused enough to take a deep breath and start over again before it caused a hypop or RERA... and towards the end of the night, you simply didn't rouse and that's when the events popped up. That'd be my guess anyway.

[cathyf]

Yeah things more towards the morning look like REM sleep, and I was on a bathroom break from 5:46-5:56 where the desaturation is, so there's too much going on there to make any sense out of it. That's why I'm focusing on just the 3 hours at the beginning of the night that I find so curious -- while in some ways it's the picture of excellent sleep, the long slow decay in inspiratory flow is peculiar...

Here's a big-picture view of the three hours, with the fitbit sleep stages glued on the bottom:
[attachment=33174]

It's really pretty great sleep -- but the tops of the inspiratory flow rate just steadily marches downward. My Fitbit decided sleep onset was 22:22 and deep sleep started a minute and a half later:
[attachment=33175]


Then at exactly 2am the fitbit thought that I went from REM to light, and then logged an arousal from 2:03:30 to 2:04:
[attachment=33177]

(The Fitbit resolution is 30 seconds, and I keep my CPAP clock pretty tight to the atomic clock, so I'd say to the resolution that the fitbit has it's pretty darn good for a device that's only connection to me is my wrist!)

At 22:22, the top of the flow rate curve was 16.56. By 2am until that arousal, it's right around 5. And if you look at the end of the first picture, between the 2:04  and 2:16 arousals I'm back up around 16.

If that all had happened over minutes rather than hours, it would have been scored as hypopneas. But because it was slow motion it only got scored as massive flow limits. The point is that scoring the hypopnea relies on 30% or 50% from "baseline" -- and that must be local baseline because the only way that moving from 16-1/2 to 4-1/2 is not a 70-75% drop is if the baseline is moving. Also I don't think that the arousals would fit the definition of RERA because they don't "work" in that after most of the arousals I'm as flow-limited as before.

And of course "hypopnea" is about oxygen starvation, and I'm getting plenty of oxygen all the way through.

I know that the definition of UARS is kind of vague and a little sloppy -- but is this showing the subtlety of UARS, or is it just nothing in particular interesting? (As in me seeing patterns that aren't really patterns, kind of how we look at clouds and see faces and objects.)

Ultimately I'm trying to answer the question why I'm fighting against falling asleep all day. I can create truly amazing positional apnea pretty much on command and so I can get a diagnosis of sleep apnea, but I can also totally eliminate all that via the cervical collar. So apnea is not the reason (or the only reason) that I'm sleepy! For example, a couple of days ago it was literally when I was standing up, in my friend's kitchen, as we brought her adorable puppy home for the first time. Being sleepy in that context is not normal -- why is it happening? Are my ugly flow limits a cause? An effect?

[Ratchick]

I mean, it really does seem to line up beautifully with the Fitbit data. LOL

Honestly, the one thing that I've found using OSCAR and my SpO2 data is that the AHI is only representative of half of my problem. I have multiple SpO2 drops from large and small Flow Restrictions (as per OSCAR'S basic definition of 50$ and 20% over 8 seconds but the number of "short" apneas that don't get flagged because they're 9 seconds but still cause an SpO2 drop just... it's crazy. It would at least double my AHI if I included all of the other "mini events" that trigger a drop for me.