As the creator of the Zone Diet, I am a strong believer in the Mediterranean diet as a lifetime dietary program for good health. In fact, the Zone Diet can be considered to be the evolution of the Mediterranean diet as it provides even greater anti-inflammatory benefits. That being said, this week's New England Journal of Medicine contained an article on using the Mediterranean diet with high-risk cardiovascular patients that got great press based on some really poor science1. Let's get to the bad science first.
The researchers compared two "Mediterranean" diets (one with extra nuts and the other with extra olive oil) to a low-fat diet. Unfortunately, they were unable to get the subjects to follow a low-fat diet. If you are a follower of Dean Ornish, then a low-fat diet means less than 10% of your calories coming from fat. Using that definition of a low-fat diet, you have to throw out one-third of the subjects because they couldn't reduce their fat intake below 37% of total calories. In fact, at the end of this five-year study, the percentages of protein, carbohydrate, and fat in the diets of all three groups were approximately the same. As a result, you are left with a study with two groups of subjects being compared to another group of subjects who really didn't change their diet that much.
Even Dean Ornish pointed this out in his rebuttal blog in the Huffington Post to this study2. He wrote that if people had followed his low-fat diet, then the results would have been much different. Well, actually when high-risk cardiovascular patients did follow his diet in a study done 15 years ago, those on his low-fat diet had twice the deaths compared to those in the control group3. So maybe it's a good thing that the low-fat group couldn't follow the prescribed low-fat diet.
The reason for adverse effects of a low-fat, low-protein, very high-carbohydrate diet for cardiovascular patients is quite clear. Those subjects following his high-carbohydrate, low-fat, and low-protein diet developed insulin resistance as evidenced by a significant increase in their triglyceride-to-HDL ratio3. If you already have had a heart attack, then an increase in insulin resistance and the accompanying increase in inflammation are almost certain to push you over the edge.
If you really dig deeper into the supplemental material (Table S7 to be exact) of this article (as most journalists neglected to do), you are remarkably unimpressed by the changes in the diet over a five-year period except that the people who got free olive oil and free nuts were consuming more free olive oil and free nuts than those who were not getting free food.
Now, back to the clinical results -- a strange brew of stroke, heart attack, and death. Usually when you include a lot of different clinical end points as your primary goal, it means you are not very confident about seeing any real striking clinical benefit. Stroke is primarily associated with high blood pressure, whereas heart attack is associated with the rupture of small vulnerable plaques leading to blockage of the coronary arteries. I personally like death as a clinical end point since you can't cheat on its definition, thus making it harder to manipulate your statistics to prove your point.
So let's look at the individual clinical endpoints. There was a reduction in strokes that was statistically significant. Unfortunately, there was no statistically significant reduction in either heart attacks or death. For such a large study, these clinical results are not too impressive. Maybe if the researchers had actually gotten the low-fat group to reduce their fat intake to less than 10% of calories (instead of going from 39% to 37% of calories), there might have been more deaths in that group, which would have made the other two Mediterranean diet groups look better.
Virtually every cardiovascular researcher knows that fatty acid composition of the plasma is an important factor in the prediction of future cardiovascular events. Unfortunately, the authors of the New England Journal of Medicine article apparently didn't think so. Obviously, they measured one fatty acid (alpha linolenic acid) in Figure 5S (again buried deep in the supplemental material), but somehow forgot to report the other 34 fatty acids also found in the plasma. Two of the most important of these unreported fatty acids would have included arachidonic acid (AA) and eicosapentaenoic (EPA). The AA/EPA ratio in the blood is the best marker of cellular inflammation that drives heart disease4. You would think inclusion of information on this ratio (or at least providing the fatty acid levels) would be important since a far larger JELIS study demonstrated that the lowering of the AA/EPA ratio resulted in a significant reduction of cardiovascular events5.
In contrast to this poorly executed study, there exists a far more powerful study conducted nearly 20 years ago on the benefits of a stricter Mediterranean diet. This is was the Lyon Diet Heart Study6. The primary clinical difference between this new study and older Lyon Diet Heart Study is that the Lyon Diet Heart Study generated a 65% reduction in overall cardiovascular mortality, a complete reduction in cardiac sudden death, and 44% reduction in all-cause mortality6,7. Those are clinical end points to get excited about. On the other hand, this New England Journal of Medicine article showed no impact on mortality. The only striking difference between the two groups in the Lyon Diet Heart Study was the restriction of omega-6 fatty acids in the experimental group. You find omega-6 fatty acids in vegetable oils like corn, safflower, and sunflower oils. They accomplished this dietary change by giving the subjects in the experimental groups margarines rich in omega-3 fats and trans fats. Although there was a dramatic decrease in death between the two groups in the Lyon Diet Heart Study, there were no differences in weight, BMI, blood pressure, cholesterol (good and bad), and blood lipids between the two groups. In other words, all the usual suspects in heart disease were eliminated. The only differences between the two groups were in the fatty acids, both linoleic acid and the AA/EPA ratio. If you again go back to bowels of the recent New England Journal article (in supplemental Table S7), you find out that the levels of linoleic acid (an omega-6 fatty acid) as analyzed from dietary records of the subjects was between 5 and 6% in both of the Mediterranean diets. In the Lyon Diet Heart Study, the investigators were able to reduce to the linoleic levels to 3.6%, which is similar to levels found in the Japanese (actually Okinawans), who have the lowest cardiovascular mortality in the developed world (8). The subjects in the control group of the Lyon Diet Heart Study had a nearly 50% higher level of linoleic acid in their blood compared to the experimental group8. However, those subjects following the "Mediterranean" diets in the new study had even higher levels of linoleic acid than those in the control group of the Lyon Diet Heart Study. That is the most likely reason there wasn't any change in cardiovascular mortality or overall mortality in the New England Journal of Medicine study. Unlike this more "modern" study, the Lyon researchers further demonstrated that the AA/EPA ratio was reduced by some 30% (from 9 to 6.2) in the active group compared to the control group, and this resulted in a 65% reduction of cardiovascular death.
Bottom line
Unless you dramatically reduce omega-6 intake by reducing your consumption of vegetable oils (such as corn, soy and safflower oils), you will not get clear-cut clinical results (like reduction in death) no matter how much hype the media give to the research.
As I said earlier, the Zone Diet can be considered to be the evolution of the Mediterranean diet because it represents a superior dietary program to control inflammation, the true underlying cause of heart disease. This is because the Zone Diet dramatically reduces white carbohydrates (pasta, bread, rice, and potatoes) and replaces them with increased amounts of colorful carbohydrates (vegetables and fruits). Unlike the New England Journal of Medicine article where the subjects were consuming about 5 servings a day of vegetables and fruits, the Zone Diet recommends 10 servings per day. Rather than keeping the linoleic acid content at 6% of the calories (the American Heart Association recommends 10-15%) or even at the 3.6% level as in the Lyon Diet Heart Study, the Zone Diet recommends fewer than 2% of total calories should consist of linoleic acid. Like the JELIS study, the Zone Diet recommends extra supplemental of omega-3 fatty acids to reduce the AA/EPA ratio to 1.5 or less.
Although the jury may still be out on the Mediterranean diet (especially after this poorly executed study) for the primary prevention of heart disease, the data from secondary prevention studies (5-7) strongly suggest that the Zone Diet may be the dietary approach you want to follow if reducing mortality is your personal clinical end point.
References: