NEJM Retracts Mediterranean Diet Study


In June of 2018, the New England Medical Journal retracted a study from 2013 that claimed to show dramatically reduced risk of cardiovascular disease following a Mediterranean Diet [1]. This news has caused a stir in the health-conscious world- how can we trust the evidence of scientific research if it can be retracted, a few years later? To settle this concern, we must look at why the study was downgraded from showing a causal relationship between heart health and dietary patterns, to merely a correlation.

The medical students who become our doctors are taught that studies must be randomized in order to be credible. This helps to keep the experiment neutral, and to remove the doubt that the results may be purely coincidental. It can also result in a bias against the benefits of nutrition before pills and procedures [2]. Although this study had originally been passed off as randomized, further investigation has revealed that it was not. This means that people were not randomly assigned to their dietary groups in the study. Husbands and wives were given the same diet. A lead researcher assigned the same diet to an entire village without telling anyone else on the research team.

However, we can still get the most important evidence to suggest causal relationships between nutrition and health regardless of whether a study is randomized. Examples of this include Dr. T. Colin Campbell's epidemiological research in the Cornell-Oxford-China Study that showed how even the inclusion of small amounts of animal-based foods in an otherwise a whole food plant based (WFPB) diet of rice and vegetables raised the risk of such a variety of chronic diseases that any well-qualified statistician would agree that this cannot be mere coincidence [3]. Another example is Dr. Caldwell Esselstyn's clinical studies, which do not show a slightly lower rate of cardiac events, as this Mediterranean Diet showed, but the reversal of heart disease altogether, with before and after angiograms to prove it: 18 patients with a combined history of 49 coronary events prior to the whole food plant based intervention, and 0 events over the course of 25 years for those who continued to eat this way [4].

Such dramatic results are impossible to ignore- no matter how small the study size, and regardless of whether it was randomized. The effects simply could not have happened by chance. This certainty is the whole point of having criteria for research methodology. Does the Mediterranean study in question still have any important evidence to offer, that would negate randomization as a factor? It would seem that it does- Dr. Steven Nissen, the Chairman of the Department of Cardiovascular Disease at the Cleveland Clinic, was among the "experts" who established its reputation as a great success back in 2013. The study became the basis of the mentality that lowering fat consumption is not beneficial to heart health, and that instead we must increase consumption of "good fats." Yet when we look at the results, there were still 288 cardiac events among the 7,447 subjects over the course of the 4.8 years following the dietary changes: the Mediterranean Diet with 4 tablespoons of olive oil each day accounted for 96 events, the Mediterranean Diet group who ate one ounce of walnuts daily experienced 83 events, and the "low fat" group experienced 109 events.

Incidentally, the "low fat" group was not actually low fat. When we want to know whether a food or an overall diet is high in fat, carbohydrates, or protein- the three macronutrients that calories come from- we look at the percentage of total calories that each macronutrient provides. When experts such as Dr. Nissen refer to a low fat diet, and dismiss it as ineffective, we have to ask for specifics. How much of the caloric content of the diet was fat, exactly? This particular low fat diet was 37% fat, which was technically considered a low fat diet because the subjects of the study had been eating a diet of 39% fat prior to the intervention. If we look at healthy people around the world, they are living on sweet potatoes and rice, with some steamed vegetables on top. This adds up to about 5% calories from fat. Indeed, the human requirement for fat is even lower than that, at 1-2% of calories [5]. When we consider a diet of over 30% fat in the context of the nutritional requirement for humans, we see that the low fat diet in the Mediterranean study is actually still a very high fat diet.

So why is a diet of 30% fat considered the target for those who aspire to eat low fat? Dr. Campbell was on the expert panel that came up with the goal of a low fat diet, specifically 30% of calories from fat [6]. Dr. Campbell reveals that although the evidence they gathered suggested a WFPB diet with under 10% of calories from fat, this was met with concern from the members of this expert panel from the USDA. The USDA members felt threatened by the recommendation. To advise no more than 10% fat would imply that people would be healthier without consuming the animal products that comprised their industry. As a result, Dr. Campbell and his panel were obligated to set the fat target for 30%, the lowest they could go without calling for people to reduce animal product intake. This is the only reason why a low fat diet apparently does not work- it does not go low enough to make a difference.

On the other hand, a varied whole food plant based diet, with no oil and small amounts of nuts, seeds, and avocados (if any), easily remains under 10% of fat. The evidence suggests that eating whole plant foods with no added fat is the only way to cut our fat content down to an appropriate level for the human body. After the onslaught of excess concentrations of fat has been ceased, the body is shown to heal itself. Results include reversing heart disease, type 2 diabetes, and shedding excess body fat. On the other hand, replacing fat with animal protein, or refined carbohydrates such as white flour products, has consistently been shown to fail to benefit health. One example of how fat content is evaluated outside the context of whether it is compared to whole plant foods or processed foods or animal protein is the Nurses' Health Study [7-9].

The interpretation of details out of context is a clever way to make low fat diets appear ineffective or harmful, and to encourage the consumption of rich, fatty foods. So why did this Mediterranean Diet study present a 37% fat diet as the low fat group? Why not compare this Mediterranean Diet to eating the 10% fat WFPB? To sell more olive oil and nuts. The study was paid for by olive oil and nut producers in Spain, and the California Walnut Commission. As for the advice to consume wine and fish on a regular basis as part of the Mediterranean Diet, several researchers in the study had ties to food and wine industries.

Interestingly, when the only dietary change was increased nuts, oil, fish, and wine, no one benefitted. The only people whose health improved on the Mediterranean Diet were those who had also increased their consumption of vegetables, fruit, and whole grains. If this study tells us anything, it reinforces what we have already seen. The more people eat whole plant foods and the less they eat other foods, the better off they will be. Why bother to compare various sources of excess fat, animal protein, and refined carbohydrates to each other? Why not replace all of these foods with the best foods, the whole plants? This is what seems to bring people the results they deserve.


1. Estruch, R, Ros, E, Salas-Salvadó, J, et al. Primary presentation of cardiovascular disease with a Mediterranean diet supplemented with extra-virgin olive oil or nuts. N Engl J Med 2018;378:e34. DOI: 10.1056/NEJMoa1800389

2. L E Thomas. How evidence-based medicine biases physicians against nutrition. Med Hypotheses 2013 81(6):1116 – 1119.

3. Chen J, Campbell TC, Li J, et al. Diet, life-style and mortality in China. A study of the characteristics of 65 Chinese counties. Oxford, UK; Ithaca, NY; Beijing, PRC: Oxford University Press;Cornell University Press; People's Medical Publishing House, 1990.

4. Esselstyn, CB Jr. Resolving the Coronary Artery Disease Epidemic Through Plant-Based Nutrition. Prev Cardiol. 2001 Autumn;4(4):171-177.

 5. Sanders, TA. Essential fatty acid requirements of vegetarians in pregnancy, lactation, and infancy. Am J Clin Nutr. 1999 Sep;70(3 Suppl):555S-559S.

6. Committee on Diet, Nutrition and Cancer, Diet, Nutrition, and Cancer (Washington, DC: National Academies Press, 1982).

 7. Willett W. "Diet and cancer: one view at the start of the millenium." Cancer Epi. Biom. Prev.10 (2001): 3–8.

8. Belanger CF, Hennekens CH, Rosner B, et al. "The Nurses' Health Study." Am. J. Nursing(1978): 1039–1040.

9. Marchione M. "Taking the long view; for 25 years, Harvard's Nurses' Health Study has sought answers to women's health questions." Milwaukee Journal-Sentinel July 16, 2001: 01G.

Article written by: Ryan Larmer