I don't think that's true at all. Capnography, the measure of carbon dioxide partial pressure is wholly separate from pulseox:
> Pulse oximeters have some limitations. They can only employ light at two wavelengths. Thus the devices can only distinguish between hemoglobin and oxygenated hemoglobin. When carboxyhemoglobin and methemoglobin are also present, there are two additional wavelengths required for differentiation. In the presence of elevated carboxyhemoglobin levels, pulse oximetry overestimates the true saturation of oxygen as carboxyhemoglobin binds with a higher affinity than oxygen. In the case of carbon monoxide poisoning, the absorbance spectrum of carbon monoxide is very similar to hemoglobin, which results in a falsely high level of oxygen (overestimation of oxygen saturation) ...
I don't think that's safe to assume at all, for two reasons:
1. CO2 has effects on the human body of its own that aren't simply a lack of oxygen, and vice-versa. [0]
2. The baseline proportions involved aren't close, so even doubling CO2 isn't going to show up easily as a large swing in in oxygen%.
For example, the article references a study where the CO2 proportion going from 0.04% -to 0.25% correlates to mental problems.
Even if the watch could sample atmosphere directly, is it sensitive enough to detect a shift from 21.00% -> 20.79% oxygen?
As it's estimating oxygen in the owner's blood, it might not detect anything different at all... not if the owner's body has already compensated by breathing harder or by "underclocking" their brain to make dumber decisions.
> The article references a study where the CO2 proportion going from 0.04% -to 0.25% causes mental problems. In other words, a difference in 0.21% of the air.
I'm finding that pretty difficult to believe, to be quite honest with you.
And before you say "aha, carbon dioxide brain fog!" consider that I'm about a mile from the sea with a 40mph onshore breeze. This air is about as oxygenated as it gets.
It makes a lot of sense actually. You get severe symptoms when CO2 makes up only a couple % of the air. And can become fatal at like 5%. There’s not like a hard line where you suddenly die, it’s a gradual thing. It very much makes sense that we’d notice minor symptoms at a few thousand PPM when it only takes like ten thousand to feel it severely.
I got a monitor as we had an old apartment with bad ventilation. When I started feeling it I would check and it was always around 1200ppm and would open a window for a bit. Outside air is around 420ppm, but that's not the problem, enclosed and badly ventilated rooms are if you spend a few hours in there.
The air in your lungs sits around 40,000ppm or 4% carbon dioxide.
In every breath you remove about 25% of the oxygen from the air in your lungs, which is why mouth-to-mouth resuscitation works, at all. Most of the oxygen is still in there.
To be clear, that 25% represents a change in oxygen level from around 21% to around 16%, so the few tenths of a percent change in carbon dioxide just isn't a huge amount.
I really must learn to quote the full context of a reply. OP said:
> just due to lower oxygen levels in your blood
So my comment is in response to that.
PS. The downvote button isn't a disagree button. (Although how you can disagree with the fact that Apple watches have a blood oxygen sensor I don't know).
> Pulse oximeters have some limitations. They can only employ light at two wavelengths. Thus the devices can only distinguish between hemoglobin and oxygenated hemoglobin. When carboxyhemoglobin and methemoglobin are also present, there are two additional wavelengths required for differentiation. In the presence of elevated carboxyhemoglobin levels, pulse oximetry overestimates the true saturation of oxygen as carboxyhemoglobin binds with a higher affinity than oxygen. In the case of carbon monoxide poisoning, the absorbance spectrum of carbon monoxide is very similar to hemoglobin, which results in a falsely high level of oxygen (overestimation of oxygen saturation) ...
https://www.ncbi.nlm.nih.gov/books/NBK539754/