The seed oil shills citing "muh mEtAanalYsiS" are lying to you on purpose. They're not as stupid as they pretend to be.
Another meta-analysis was put together by the guy who made the Tufts "Food Compass" algorithm, which ranks Cheerios and Count Chocula as more healthy than milk or ground beef.
This one looked at just 7 clinical trials, and did suspicious things:
- Lumping together trials that measured Ω-6 Ω-3 fats together as "Polyunsaturated fat" generically.
- Every single trial was confounded by changes in trans-fat intake, which goes unstated.
- Oddly weighted trials, 5 of 8 of which showed no significant relationship at all, with 2/3 of the remaining ones just barely significant.
4/