I like Rhonda Patrick, but a lot of her recommendations seem to be speculation on what should work, given what she understands about a subject and relating it to a new study. Not that it's not valuable, but it's the sort of thing Reddit would class as "science" rather than realising it is a hypothesis that needs testing.