It's self-reinforcing. Because there is a subset of the population that reads health news and acts on it, they both: A: exercise, sleep 8 hours, avoid ultra-pr

It's self-reinforcing. Because there is a subset of the population that reads health news and acts on it, they both: A: exercise, sleep 8 hours, avoid ultra-processed foods B: take whatever the latest anti-cancer fad supplement is Behavior A actually helps, while B is probably neutral. But studies will show that all the B behaviors will be associated with lower cancer risk, because they're also associated with A behaviors. So this entire class of studies (looking for associations between foods and health) is complete crap.