AHI

[Webbie73]

When checking my AHI in the morning I am finding ( I think )central apneas.
Example
AHI 2.5
AI 2.1
central AI .1

Does this mean I am having central apneas? If so, should I be concerned?

[retired_guy]

Nope. Not according to those numbers. They look just fine. Now the thing about centrals: they kind of get a bad rap around here. They really aren't any more weird than obstructives. We want to control them, keep them to a minimum or completely gone away even, but we're always going to see some. And, not all centrals are equal. The machines can and often report a central event when it is not actually a central apnea event.

So, relax and sleep well. You're doing great.

[trish6hundred]

Hi Webbie73,
What R_G said.
Sounds like you are off to a good start, just keep on goin'.

[NickDanger]

Agree with the above: 0.1 = no problem. 1.0 = no problem. 5.0 = maybe a problem (but may not actually be central apneas). CPAPs do a decent job of distinguishing central apneas from obstructive apneas, but they aren't perfect. They frequently get confused by irregular breathing while awake or while on the border between sleeping and waking. So, if you had 50 to 100 times as many CAs showing up, then you'd want to look closely at your detailed data (using SleepyHead or ResScan software) to examine the patterns of your events. At 0.1 CAs and 2.5 AHI, it really isn't worth spending time on the analysis if you are feeling better. (My numbers are actually a little bit better than yours, but I still like looking at the pretty graphs - but then, I'm a math/science geek).

Great start!!!

[herbm]

Everyone else is correct. 0.1 is practically nothing. 2.5 is excellent (even were it all CAs, but it isn't).

You are doing great, if those numbers hold just keep on doing what you are doing.

[Ghost1958]

ive noticed both sleepy head and encore with my machine will show occasional CAs right over a perfect waveform where nothing was going on at all except steady breathing. One software is a bit worse for doing that than the other one.