Revised Estimates: Primary Prevention of Cardiovascular Disease with the Mediterranean Diet

Following a partial retraction in 2013 for protocol deviations, authors for the PREDIMED study of the Mediterranean diet have published revised estimates first made in their original study in a recent issue of the New England Journal of Medicine.  

The researchers included 7,447 participants at high CV risk but with no CVD at enrollment to either a Mediterranean diet supplemented with extra-virgin olive oil, one supplemented with mixed nuts, and a control (low-fat) diet. The primary study endpoint was a major CV event (myocardial infarction, stroke, or death from CV causes). Median follow-up was 4.8 years.

https://twitter.com/ACCCardioEd/status/1007617490162192384

According to the updated manuscript, the primary study endpoint occurred in 288 participants (96 in the extra-virgin olive oil cohort (HR=0.69; 95% CI, 0.53-0.91), 83 in the mixed nut cohort (HR=0.72; 95% CI, 0.54-0.95), and 109 in the control group. The results, according to the authors, were similar to those reported in a retracted analysis that had to exclude 1,588 participants for protocol deviations. 

I think Andrew Gelman's last statement is probably what worries most statisticians: what other problems were there with the Mediterranean diet study that have not yet been documented? https://t.co/VDLk52BZsX

— Frank Harrell (@f2harrell) June 15, 2018

About that quasi-retracted study on the Mediterranean diet . . . https://t.co/VuNAKjb1FT

— Andrew Gelman et al. (@StatModeling) June 15, 2018

Source: NEJM