> FFS, The Atlantic should be embarrassed to publish articles with "correlation is causation" fallacies of this magnitude in them. This is standard food woo thinking: "people who heavily scrutinized and evaluated everything they ate because they have a strong adverse reaction to X means that not eating X will make YOU healthy too!"

You basically just described all of epidemiology and why it has been so bad for nutritional research.