“Diets, like clothes, should be tailored to you.” Joan Rivers

According to Maimonides, a Philosopher born in Córdoba in what is now modern day Spain, “No disease that can be prevented by diet should be treated with any other means”. Although he lived almost a thousand years ago, there is more than a grain of truth in this statement. In her excellent TEDX talk, Dr Sarah Hallberg, makes a compelling argument that this is the case with Type 2 Diabetes. As the Medical Director of the Medically Supervised Weight Loss Program at IU Health Arnett, she is in a good position to know. 

Now for the science bit…

Unfortunately, a significant problem with any research into preventing disease is that it is inherently difficult to do. There are a number of reasons for this. Firstly, confounding factors play a big role. For example, a person who does one healthy behaviour – such as doing regular exercise – is more likely to eat a healthier diet. Researchers try to get around this by matching as closely as possible each group of participants in a study. However, it’s difficult to mitigate for everything. Secondly, it is hard to do dietary studies over a long period of time. Some people will drop out; some will get bored of the diet; some will make stuff up in their food diaries to keep the researcher happy. The latter is known as self-reporting bias. Thirdly, there is little money invested in finding out which broad diet might be the healthiest and good studies require lots of resources – and therefore lots of money. Many studies on diets are also prone to recall bias. This occurs with retrospective studies where two groups, one with a disease and one without, are asked to recall whether they ate lots of a certain food, say, cola bottle sweets. The researchers are hoping that those with the disease will say “Yeah, I ate loads of cola bottle sweets.” Then just before they publish their research paper they can hold a press conference and say “Cola bottle sweets cause disease X!”

Recently, an article in the New England Journal of Medicine on the Mediterranean Diet gained a lot of media attention. For a number of reasons, this diet is worth close examination. The purpose of this study was primarily to test how good the Mediterranean Diet was in a head-to-head (also referred to as a randomised control trial) against a low fat diet in preventing heart disease, strokes, and deaths from all cardiovascular diseases (CVD) in otherwise healthy people. This is what is known as a primary prevention study. In contrast, a secondary prevention study would look at people who already had heart disease. The researchers also wanted to test whether two variations of the Mediterranean diet had any impact on health outcomes. One variation placed more importance on olive oil and the other placed more emphasis on nuts. So, these researchers were testing out Maimonides’ hypothesis – that a Mediterranean Diet can prevent heart disease and ultimately prevent the need for treatment by means other than diet.

Who was in this study?

One of the aspects that makes this study so interesting is the group of people the researches included. Although none of the participants had CVD when the study started, all of them had medical conditions or lifestyle behaviours that put them at high risk of CVD. The male participants were aged between 55-80 and the women between 60-80. This is an important point. If you are not over 55 and at high risk of CVD then the results are considerably less applicable to you. The researches likely chose patients who were at high risk of CVD for two reasons. Firstly, it is more likely that there will be a beneficial effect in people at higher risk. If they were doing this study in healthy people, they would have needed to enrol far more people to see any effect. Secondly, the people in this study – and in the population in general – really need an effective dietary intervention to prevent CVD. 

A further important point that makes this study impressive is that as all the participants were unhealthy to begin with. Both the Mediterranean Diet and low fat diet were likely to be an improvement on what they were already eating. You might be thinking why, if the researchers want to get impressive results, did they not have one group of participants on their usual diet? This would have introduced what is known as performance bias. At the beginning and throughout the study, participants received coaching from dieticians. Participants also had assessments to check their adherence to the diet. Keeping people on their standard fare diet would have meant that they’d have received less attention than the other participants. Thus, they would have been at even more of a disadvantage. One source of performance bias that was present in the study was that the control group on the low fat diet received less coaching than the two Mediterranean Diet groups. 

As I mentioned earlier, the goal of this study was to see the effect of a dietary intervention on an amalgamation of the number heart attacks, strokes, and deaths from all cardiovascular diseases. This is what is known as a composite end point. This study was geared from the outset to establish the effect of diet on this composite. 

What was in the diet?

One of the attractive aspects of the Mediterranean diet is – in my opinion – its simplicity. The dietary recommendations for the participants in the study were:

Mediterranean Diet

• Extra-virgin olive oil, greater than 4 tablespoons per day
• Tree nuts and peanuts, greater than 3 servings per week
• Fresh fruits, greater than 3 servings per day
• Vegetables, greater than 2 servings per day
• Fish, especially fatty fish such as tuna, salmon and mackerel, greater than 3 servings per week
• Legumes (bean, peas and lentils), greater than 3 servings per week
• Sofrito (diced carrots, celery, onions and garlic fried gently in extra-virgin olive oil) which forms the basis of many sauces in the Mediterranean, greater than 2 servings per week
• White meat rather than red meat
• Wine – yes, wine – with meals, greater than or equal to 7 glasses per week. This was optional for what the researchers called “habitual drinkers.”

Foods that were discouraged included sweet fizzy drinks, commercial bakery goods, sweets and pastries, fat spreads and red and processed meats. 

The low fat diet group were encouraged to eat:

• Low fat dairy products
• Breads, potatoes, pasta, and rice
• Vegetables
• Fresh fruit
• Lean fish and seafood

So who wins in the battle of Mediterranean vs Low Fat?

For the composite end point I mentioned above, the Mediterranean diet resulted in a relative risk reduction of having either one of heart disease, strokes, and deaths from all cardiovascular diseases of 30%. A more useful way of looking at the data is numbers needed to treat. This tells us how many patients I would need to treat to prevent one person from getting one of the diseases in the composite end point. Crunching the numbers shows that, for the Mediterranean nut group, I would need to treat between 46-53 people to prevent one having heart disease, stroke, or death from any CVD; for the Mediterranean olive oil group, it is 56-232. 

It is worth noting that in the subgroups that made up the composite, the two Mediterranean diets only showed a statistically significant benefit in reducing the number of strokes. However, remember the study was only powered to deliver a statistically significant result for the composite end point. 

As mentioned early, dietary trials are difficult to do well. This trial is no different. The reason it gained press attention is that it was retracted and then republished. There were several reasons for the retraction. These included that some members of the same family were in the study but not randomised and that some people in the low-fat group ate a Mediterranean diet. After correcting for their errors, the researchers ran the analysis again and came to the same conclusion.

Overall, the Mediterranean diet certainly shows some benefit in preventing CVD for people who are at risk. This study is certainly an important one. However, it is one of several that confirm the benefits of a Mediterranean diet. So if this diet appeals to you and is aligned with your dietary beliefs I think it is the optimal diet for health. A diet needs to align with your beliefs for you to stick with it. Hence, diets should be, to an extent, personalised. Diet is, though, only part of the picture. It is likely that the reason for longevity in so many Mediterranean communities is a composite. In addition to diet, daily exercise, fasting, sense of community and socialisation, spirituality, sunshine and clean air all contribute.

Dr J Hugh Coyne