I've put people in the position of reading an instruction, and following it. Everybody, 100%, wqants to "pass the test." Everybody, 100%, will act as though they "get it," even if they don't. It's human nature, plus we are conditioned to want to "pass the test."
If you really wonder if your proposed language is comprehended, put it to the test. Figure out a way to put it in front of people, and have them express the meaning of each phrase in their own words. Don't lead them to the answer YOU want.
I've done that with "documents" of under 25 words, and have been humiliated to find that words perfectly clear to me are perfectly confounding to the general public.
Just a thought or two or three to expand. This task is hard. You can't be honest with the test subject. They'll keep asking "Did I get that right?" "How am I doing?". The challenge is to draw them out to find the impression the words really make in their mind. Any encouraging and coaxing has to be content neutral. You definitely can't have more than one test subject at the same time, they will interact and tend to develop a "team" answer. If there is a larger group, the dynamic of being a member of the "winning" (bigger) team kicks in.
Legislators and standards committes that compose statutory language for contracts and/or instruciotns and/or warnings don't do this ;-) The result is predictable. The text is something "the insiders" understand perfectly. They ought to, they've been immersed in it for some time. But "outsiders" seeing the writing for the first time are confused, but they (generally) won't admit it.
Point taken. "Hydration" could be replaced with fluids, and the results could be reworded slightly: "...result in death from starvation", "...result in death from dehydration (thirst)", "...result in death from asphyxiation (lack of oxygen)".
Even if people don't get "asphyxiation", or even "oxygen", I think they'd figure out "death".