I find it interesting that in livestock antibiotic use research, studies that look for an effect on human outcomes tend to find a relationship, but studies that look for a mechanism can’t find anything. I’m assuming that it is because of several weak effects, but there also might be a white whale