Don’t know about the exact inverse. But disease X could be improved while disease Y risk rises. It’s curious that both studies are about cardiac health. Maybe there’s different types of FAO, or FAO is desirable in some cell types but not in others.
If you have fifty different groups studying roughly the same thing and publishing any interesting-looking results with p<.05, you're probably going to have some spurious results on the topic. If you have 500 groups, you're going to have a lot of spurious and probably some outright contradictory results published.
In many cases, with large enough numbers, the spurious results will outnumber the real ones, especially so since true-null-hypothesis results are often unpublished.
