A more practical view is that food is both needed by the body, and at the same time does 'damage' to it. Causing or promoting cancer may be among that damage.

A more practical view is that food is both needed by the body, and at the same time does 'damage' to it. Causing or promoting cancer may be among that damage. Thus a varied diet is good because: a) It limits the damage any 1 foodstuff can do, and b) It 'trains' the body to counteract that damage using the widest variety of inputs. You'd have less damage by eating less, but of course there's a point where the negatives (feeling weak / thirsty, vitamin shortages etc) outweigh the benefits. Optimum depends on a wiiiiide variety of interacting factors like genetics, environment, activity level, source & processing of foods, cooking methods, gut microbiome, just to name a few. TLDR; we know 'nothing' about what exactly food does in the body. Food science is not an exact science. Don't believe me? Go read some food-related scientific studies. The more you dive into the details, the more things become fuzzy. Not to mention in-vitro vs. in-vivo, the difficulty of establishing control groups, and more. In essence it's mostly statistics + theories + lab tests + educated guesswork. And meta-studies. :-) Nothing wrong with that, btw - as long as those elements are well done, you can still come up with useful findings.