Lipid Luminations: PREDIMED Commentary

The recently published PREDIMED trial is a landmark study assessing the efficacy of a dietary intervention in the primary prevention of cardiovascular disease. The authors used a parallel group design and randomized subjects from multiple centers in Spain. Subjects were men ages 55 to 80 and women ages 60 to 80, with either diabetes or three or more risk factors for coronary artery disease, without diagnosed coronary heart disease at baseline.

The subjects were randomized to one of three groups: a Mediterranean diet supplemented with nuts, a Mediterranean diet supplemented with extra virgin olive oil, or a control diet. The primary outcome was a composite of myocardial infarction, stroke or death from cardiovascular causes. There were 288 primary outcome events in 7,447 subjects, driven primarily by the rate of stroke. The trial was terminated early (median follow-up of 4.8 years) when an interim analysis showed a statistically significant difference between groups with a hazard ratio of 0.7 for both Mediterranean diets compared to the control diet.^1,2,3

The authors of this trial are to be commended. To date, randomized, controlled outcomes data on dietary interventions are sparse and limited to secondary-prevention populations. The study was well-powered to show a difference in outcomes among the groups and sought to differentiate benefit obtained from individual components of the Mediterranean diet, including nuts and olive oil. The hazard ratio of 0.7 from a lifestyle intervention is extremely impressive, although it is difficult to know where that benefit was derived.

A careful analysis of the food-frequency questionnaires completed by patients listing the composition of the three dietary groups shows a great deal of overlap, making it unclear whether there was a true control group. The intervention groups consumed only slightly more legumes, fish and sofrito (minced tomato, garlic and onion simmered with olive oil), and less butter, cream or margarine than the control group. In addition, although the control group was planned to be a low-fat diet, it included 37% fat, with no difference in saturated fat, making it very similar to the overall fat content in the intervention groups⁴. Given the diet structure actually followed, it seems as though it is only the presence versus absence of sofrito, olive oil and/or nuts that makes up the dietary difference across the three groups. It also is difficult to know whether the results of this trial, which were noted in the setting of nuts and oil being given to patients as gifts, would have been replicated in a population that was expected to purchase and consume these foods in large quantities.

Aside from the different dietary patterns between groups, there was far less counseling provided to the control group. Patients in the Mediterranean diet groups received two individual motivational interviews and one educational group session every three months throughout the trial. Patients in the control group only received these sessions after the third year, when it was recognized as a limitation. The less frequent contact with the dietitian during the first three years of the trial may have been the reason for the higher dropout rates and lower compliance with dietary recommendations noted in the control group. In addition, the sessions may have had other psychological benefits that could have independently contributed to the change in outcomes noted.

The PREDIMED trial is the first randomized controlled trial to assess the use of a Mediterranean diet in primary prevention of cardiovascular disease. Although there are limitations to the trial, the potential benefits of a diet based on hard clinical outcomes evaluated in a prospective randomized study design is tremendous.⁵ Understanding the individual components of diet that improve health is a challenging task, but there appears to be ample evidence to recommend a balanced Mediterranean diet to our patients at risk for cardiovascular disease.

Disclosure statement: Dr. Gianos has no disclosures to report. Dr. Carral has no disclosures to report.

Editor’s Note:
The authors were cautious that this is in a Mediterranean country for persons at high CVD risk and limitation of generalization needs to be considered.

The editor of this article, Dr. Wild, thanks the Lipid Luminations authors for turning the article around on a short deadline.