Misaligned incentives are certainly part of the problem, but this article points to it being more than just that - even if you want to run a good study, it is incredibly difficult to do so.
I have spent a bunch of time on that site and read Dr. Greger's book "How Not to Die". He certainly works hard to understand how studies could be flawed, but he is still limited by the fact that really good underlying studies are hard to come by.
A few things stood out to me reading that book:
1) So, so many studies are hopelessly flawed.
2) Even ones that seem relatively good often have other studies showing the opposite result.
3) The evidence for a plant-based diet seemed fairly overwhelming, but some of the specific recommendations felt more questionable.
4) If you combine all of the claimed increases/decreases in disease risk for a given set of foods, you can fairly quickly get to a ridiculous result where you have something like a 1 in a billion chance of getting a relatively common ailment or are almost guaranteed to get 3 different cancers tomorrow.
I have spent a bunch of time on that site and read Dr. Greger's book "How Not to Die". He certainly works hard to understand how studies could be flawed, but he is still limited by the fact that really good underlying studies are hard to come by.
A few things stood out to me reading that book:
1) So, so many studies are hopelessly flawed.
2) Even ones that seem relatively good often have other studies showing the opposite result.
3) The evidence for a plant-based diet seemed fairly overwhelming, but some of the specific recommendations felt more questionable.
4) If you combine all of the claimed increases/decreases in disease risk for a given set of foods, you can fairly quickly get to a ridiculous result where you have something like a 1 in a billion chance of getting a relatively common ailment or are almost guaranteed to get 3 different cancers tomorrow.