I like this: from JAMA; even though findings are not statistically significant, reporting and interpretation of those same findings in research papers can make it appear so.

This confirms what I keep finding. When we blame journalists for bad reporting of health stories, we may just be shooting the messenger.