Landmark Study of Mediterranean Diet Was Flawed

2013 study retracted, though updated version reaches similar conclusion
By Arden Dier,  Newser Staff
Posted Jun 14, 2018 1:39 PM CDT
Shrink
This 2011 photo shows the entrance to the editorial offices of the New England Journal of Medicine in Boston.   (AP Photo/Michael Dwyer)
camera-icon View 2 more images

(Newser) – The repute of the Mediterranean diet, considered one of the world's healthiest, took a hit this week: Citing methodology errors, authors retracted a landmark 2013 study published in the New England Journal of Medicine that found people who consumed the diet of fruits, veggies, nuts, fish, and olive oil had a 30% reduced risk of heart attack and stroke. But believers still have something to cheer: The authors replaced it with a "re-analysis" that reached essentially the same conclusion, though with what NPR calls "softer language." The key difference: Those on the diet did indeed appear to have a 30% reduced risk of serious health risks, but the authors no longer assert that the diet is the reason—only that they found an association between the two.

"After all this long work, I am more convinced [of the findings] than ever," study author Miguel Martínez-González of Spain's University of Navarra tells the New York Times. The big problem in the original study is that it was supposed to be randomized—that is, the nearly 7,500 participants were supposed to be assigned one of three diets in the study at random—but that didn't happen in some cases. One example: A researcher would randomly assign one member of a household to adhere to one of the three diets, but then ask everyone else in the household to follow it as well. That means the other members of the household weren't randomly assigned. "The evidence is still strong, but not as strong as a randomized study in which the randomization was executed flawlessly," says an NEJM spokesperson. (Plenty of other studies tout the diet's benefits.)

My Take on This Story
Show results without voting  |  
17%
40%
2%
6%
3%
31%