Heh, without reading the article, I thought "correlation isn't causation" and clicked into the comments to see if anyone else who read the article thought so, too. And I see you at the top of the comments. It's nice to know I'm not alone.
I hoped that the article itself would make at least some attempt to deal with the issue, but it does not. It’s a meta study (meta meat?) that includes a set of studies. Those studies do not randomly assign people to eat/abstain from meat. It’s as bad as it sounds. Further, they use their poor study to criticise other, poorer, studies.