What is a Mediterranean diet? I had always believed that it involves eating shawarma three times a day while sitting on a beach in Tel Aviv, just because that’s my diet when I visit the Mediterranean. I was astounded to learn that this is not the case. A Mediterranean diet includes a lot of fruits, nuts, vegetables, olive oil, and cereals. It includes moderate intake of fish and poultry, and very little dairy, red meat, and sweets. Wine is included in moderation and consumed with meals.

Like the low-carbohydrate (Atkins) diet and the low-fat diet, the Mediterranean diet has passionate adherents and advocates. A debate between proponents of different diets quickly resembles one between zealots of different religions – there is much heat but little light. That’s because virtually no high-quality studies have directly compared one diet to another. So in the face of weak data, each camp highlights the data that confirms their bias and disregards the rest.

A study published in this week’s issue of The New England Journal of Medicine (NEJM) tried to compare a Mediterranean diet to a low-fat diet. The study took place in Spain and randomized over 7,000 people who did not have cardiovascular disease at the start of the study but had risk factors for cardiovascular disease, like diabetes or high cholesterol. The people were randomized into three groups. The first group was instructed to follow a Mediterranean diet and told to add about four tablespoons of extra-virgin olive oil (which was provided by the study) to their diet daily. The second group was also instructed to follow a Mediterranean diet. They were instructed to add about a quarter cup of mixed nuts (also provided) to their diet daily. The third group was instructed to follow a low-fat diet.

The three groups were followed for an average of 5 years. Strokes, heart attacks, and other episodes of cardiovascular badness were tallied.

The three groups had similar numbers of heart attacks, but the two groups following the Mediterranean diet had significantly fewer strokes than the group instructed to follow the low-fat diet. The statistics suggest that for every 60 people in one of the Mediterranean diet groups instead of the low-fat diet group one stroke is prevented every five years. That’s pretty impressive, and is better than some medications used for stroke prevention.

So does this mean that a Mediterranean diet is better at stroke prevention than a low-fat diet? Not at all. As an accompanying editorial makes clear, what is important is what the three groups actually ate, not what they were supposed to eat. The first two groups were pretty good at keeping a Mediterranean diet, but the third group which was supposed to eat a low-fat diet, didn’t  Most of the people in the third group, despite being instructed to eat a low-fat diet, ate pretty close to a Mediterranean diet, which is what they were eating before the study. That makes sense. Spain is Mediterranean, and it’s very hard to change people’s eating habits.

So this study doesn’t teach us anything about the benefits of a Mediterranean diet, but you wouldn’t know that from the headlines in the popular press coverage (links below). This study taught us much more about how ineffective it is to instruct people to change what they eat, and much less about whether one kind of diet is healthier than another.

This study does suggest that in people eating a Mediterranean diet, adding olive oil or mixed nuts decreases stroke risk, which in itself is very interesting. Does that mean that olive oil and mixed nuts might prevent strokes in the rest of us? Maybe. I certainly wouldn’t object to my patients adding nuts and olive oil to their diet, and I’m busily trying to figure out how to order that with my next shawarma.

Learn more:

Olive Oil Diet Curbs Strokes (Wall Street Journal)
Mediterranean Diet Shown to Ward Off Heart Attack and Stroke (New York Times)
Mediterranean diet over low fat? Well, at least it’s more fun (Los Angeles Times opinion)
Primary Prevention of Cardiovascular Disease with a Mediterranean Diet (NEJM article)
Did the PREDIMED Trial Test a Mediterranean Diet? (NEJM editorial)