The famous ‘Mediterranean diet’ is awesome study seems to have had serious randomization issues affecting >20% enrolled participants
Retracted and republished!
nejm.org/doi/full/10.10…
ncbi.nlm.nih.gov/m/pubmed/28580…
To detect fraud, analyzed 1000s articles to see if distribution baseline characteristics consistent with randomization
11 articles inconsistent with randomization & in @NEJM
Check out editors note 👇
Makes me even more suspect of the re-analyzed effect size.
Mediterranean Diet Study Was Flawed. But Was It Wrong? nyti.ms/2MnNckp