Quite a bit of nutrition research is simply useless. They ask a bunch of people what they ate and then try to correlate that with health outcomes. You will never learn anything that way, though it might give you areas for further research.

The next step up is a randomized controlled trial, but instead of choosing health outcomes they choose measurements of lab tests as their endpoints. Someone’s cholesterol went up or down by some small number. What did we learn? Nothing.

Unfortunately good research is outrageously expensive, so it doesn’t get done.