I'm not sure if these issues are with the methodology in the studies, special interests wanting certain results, or just poor execution. But it seems out of hand.
Small off topic rant. If we want the masses to believe the results from sciences, then we need to be consistent. I'm not advocating lying in order to do this. The results from the studies need to be reproducible and accurate. How can we convince people that something like climate change is happening and real, when the earth is a massively complex system, when we can't even go 2 months without flipping dietary results from sugar is ok, to sugar is bad, back to sugar is ok. Or the other example would be the no amount of alcohol is good for you, when for years the common result from studies was 1 or 2 drinks was a benefit.