No, trusting scientific sources. If you see someone (including an author) correctly interpreting scientific studies (and not cherry-picking them) then you can think about trusting their arguments.
Still, you have to trust the book's author did not cherry-pick studies in order to make his case for selling the book. This may go against his economic interests.
You don't need to put trust into any one book. Of course it would be foolish to look at just one book or just books from one point of view on a subject. If you look at the evidence from different sources site it will become apparent who is cherry-picking and mis-interpreting. Maybe everybody is, and you make up your own mind.
