The PREDIMED trial, Mediterranean diet and health outcomes: How strong is the evidence?

Auteur(s) :
Estruch R., Corella D., Klepp KI., Martinez-Gonzalez MA., Guasch-Ferré M., Salas-Salvadó J., Rosa E., Fíto M.
Date :
Juil, 2017
Source(s) :
Nutrition, metabolism, and cardiovascular diseases : NMCD. #27:7 p624-632
Adresse :
Human Nutrition Unit, University Hospital of Sant Joan de Reus, Department of Biochemistry and Biotechnology, Faculty of Medicine and Health Sciences, IISPV, Rovira i Virgili University, Reus, Spain; Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, MA, USA; CIBER Fisiopatología de la Obesidad y la Nutrición (CIBEROBN), Instituto de Salud Carlos III, Madrid, Spain.

Sommaire de l'article

To address potential controversies on the health benefits of the Mediterranean diet (MedDiet) after PREDIMED, a randomized trial of MedDiet for primary cardiovascular prevention. We have focused on: a) the PREDIMED study design, b) analysis of PREDIMED data and c) interpretation of its results.

Regarding the design of the trial, its early termination and between-group differences in the intensity of the intervention are potential causes of concern. The planned duration was 6 years but the trial was prematurely stopped when an interim analysis at 4.8-year provided sufficient evidence of benefit for the two MedDiets. In the MedDiet groups supplemented with extra-virgin olive oil or mixed-nuts, the primary composite endpoint (myocardial infarction, stroke, or cardiovascular death) was reduced by 30% and 28% respectively, as compared with the control group. Final results did not change after taking into account the different intensity of educational efforts during the trial. Other potential doubts related to data analysis (e.g., intention to treat versus a per-protocol approach, and consequences of dropouts) should not be causes of concern. Finally, we addressed alternative interpretations of the effect on all-cause mortality. The protocol-defined primary endpoint was a composite cardiovascular endpoint, not all-cause mortality. To analyze total mortality, we would have needed a much larger sample size and longer follow-up. Therefore, the PREDIMED results cannot be used to draw firm conclusions on MedDiets and all-cause mortality.

The PREDIMED study was designed to overcome three major problems of previous nutritional research: a) residual confounding, addressed by using a randomized design; b) single-nutrient approaches, by randomizing an overall dietary pattern; and c) the limitations of assessing only intermediate risk markers, by using hard clinical end-points.

Source : Pubmed