I'm thinking that any research into a controversial subject will be used by one party or the other to promote their point of view. Words like "link" as in "researchers find a link between carrots and multiple sclerosis" usually just imply that they found a correlation which was sufficient to test directly, in order to see if there is a causal link or another factor at play. But the layperson reads "link" and assumes, by default, that it is causal. Special interest groups love the word "link." They can use it to promote their cause, or disparage research that might indicate otherwise (at which point they respond to the word "link" with "correlation does not equal causation!").
Take note, in the OP's link, of how many times words like "associated", "correlated", "linked" and similar words are used in the article headings or abstracts. Then see if you can find how many of the claims are strengthened or corroborated by additional testing and research. My guess is that a pattern will emerge. But that's just correlation.