Since OSA as a condition is said to have a statistical association with heart disease and diabetes and, in turn, heart disease and diabetes presently seem to have a statistical association with serious complications from COVID-19, then scientifically speaking, there IS a statistical relationship between OSA and serious COVID-19 problems, whether we like it worded that way or not in harmony with our desire to prevent people's assumptions about causation.
I therefore choose to give the poster and the doc that was quoted a pass.
All summary statements are a form of shorthand that mean different things from different perspectives. Just because a statement is loosely worded and easily misunderstood does not make the statement completely false or purposely misleading. Expected degrees of exactness on a public board should have some wiggle room, I think.
Especially my statements.

Remember: Statistically speaking, taking a a painkiller makes it more likely you will be in pain, if you go only be the association when you report it.
Hey, just me.