Source: Estruch R, Ros E, Salas-Salvadó J, et al; PREDIMED Study Investigators. Primary prevention of cardiovascular disease with a Mediterranean diet. N Engl J Med. 2013 Apr 4;368(14):1279-90. doi: 10.1056/NEJMoa1200303. Epub 2013 Feb 25. PubMed PMID: 23432189.

Efficacy Endpoints: Stroke, heart attack, death

Harm Endpoints: Adverse events potentially caused by the diet

Narrative: Large randomized trials have demonstrated significant benefits with a Mediterranean Diet among patients with known heart disease.1 It was unclear whether this also meant that those at risk, but without a history of heart disease, could also benefit.

The trial summarized here undertook the education of three groups: two were educated on the use of a Mediterranean Diet (one that concentrated on olive oil and the other on nuts and legumes) while a third group was counseled on a nutritional, low fat diet. The research included 7447 subjects followed for an average of roughly five years, and demonstrated a clear reduction in their combined endpoint of strokes, heart attacks, and death (RR=0.7; 0.55-0.89).

Caveats: While this study was extremely rigorous and impressively conducted the most important caveat is that the trial was underpowered (ie too small) to reliably demonstrate statistical significance in each of the endpoints of stroke, heart attack, and death. The result is that the Mediterranean Diet groups were statistically superior to the control group only for the individual endpoint of stroke. In many cases we at TheNNT will not accept composite endpoints for reporting. In this case we did so both because of supporting literature for the Diet and also because the study was clearly powered only for the composite, making separation of the endpoints fraught. To leave this data out of our database based on this strikes us as misleading.

An additional criticism is that the trial was adjusted midway through in order to strengthen the control group education on low fat diet. We felt that this did not represent a major problem in translation to most cohorts, largely because most people considering a Mediterranean Diet are converting from a less healthy lifestyle and diet, rather than from a low fat ‘heart-healthy’ diet. In addition, the control group in this study was culturally Mediterranean and therefore already was engaging in a Mediterranean Diet to some degree. Our sense is that the Mediterranean Diet is therefore likely to be much more powerful in a real world setting than these data would suggest.

Overall, despite the limited data (one large trial only), we have chosen to rate this intervention ‘Green’ because there are no known harms or down sides to the Diet, and because in the absence of significant potential for future profits based on the intervention it seems unlikely that better or more extensive data will trump this study’s findings. Moreover, for the purposes of primary prevention it appears, despite nonsignificance in some of the endpoints, that based on a 30% reduction in events the Mediterranean Diet represents the most beneficial and least harmful intervention we have seen.

Author: David Newman, MD

Published/Updated: November 3, 2013