For good or ill, we repeatedly and frequently test top athletes extensively for taking things that ought to improve their performance less, and give them multi-year bans if they show any hints of having taken banned substances.

People don't wait until all the records have been broken by those who use performance-enhancing substances to ban those substances.

That's the context, and in that context the burden of proof would normally be on those wanting an exception: showing convincingly that an intervention that from first principles seems like it ought to provide an advantage is, in fact, not an advantage.