Time to eat our words?

A 'Mediterranean diet' is said to be good for you. But does it deserve its healthy tag? asks Helen Macbeth in the latest in our series on food

The Mediterranean diet is much lauded these days. Everyone knows how it guards against heart disease, strokes, even cancer. Everyone knows, too, what a Mediterranean diet comprises. Pasta, fish, vegetables, olive oil ....

But is it so straightforward? My research suggests a much more complex picture.

European colleagues have found 18th-century references to the phrase "Mediterranean diet", but I prefer a mention in the late 1950s, when the epidemiologist Ancel Keys linked the infrequency of cardiovascular disease in Mediterranean Italy to diet. Following a visit to Italy, Keys wrote about "the Italian diet" and went on to create a recipe book on Mediterranean cooking.

His writing sparked off four decades of medical literature preoccupied with the notion. This included discussion of the benefits of olive oil, of marine oils, of fish compared with meat, of cereals, especially breads, pasta and rice, of fibre, of fresh salads and fruit, of garlic and so on. In 1993 came the welcome suggestion that red wine was beneficial. Now it is the low consumption of full-fat milk, but no doubt next year there will be some other item to discuss. Western science continually updates itself, frequently disproving earlier hypotheses.

The worst element in medical literature which discusses the benefits of a Mediterranean diet is the plethora of academic papers emphasising the advantages - or the harmful effects - of a single substance in regard to such multifactorial conditions as cardiovascular disease and cancer. The aetiology of such diseases depends on several nongenetic factors as well as on genetic susceptibility. Diet, let alone one food item, cannot stand alone as the sole "protection" against - or "cause" of - such illnesses.

Yet the belief systems of western society are greatly influenced by popular concepts of scientific "truth". The beneficial properties of a so-called Mediterranean diet have been incorporated into our folk beliefs about healthy living. Look through your health-centre handouts, recipes in women's magazines or in the magazine produced by your local supermarket, and you will find some version of nutritional literature on healthy eating, most likely including items recognisable as the Mediterranean diet.

So, I undertook a small local study of what people around Oxford thought the Mediterranean diet was. I asked them to list the ten food items that they most associated with the phrase "Mediterranean diet". Almost 90 per cent mentioned fish, seafood or shellfish (sometimes two of these words), while meat was rarely referred to - there was an occasional reference to goat and one mention of lamb. Over 80 per cent mentioned olive oil and an equally large majority mentioned olives, tomatoes and pasta among their ten items. Between a quarter and a half mentioned bread, herbs, cheese, salads or particular vegetables, while about a third mentioned wine, usually red wine. For those familiar with the work of Ancel Keys, it will be obvious how closely this reflects his ideas and the medical literature that followed his publications.

To take the inquiry further, I started to analyse ingredients in the recipe books and magazine articles that claim to portray Mediterranean recipes. Again the same pattern emerged. Using one supermarket (Sainsbury's) cookbook, entitled Mediterranean Cooking, a division between the frequency of meat and fish dishes shows 45 per cent meat and 55 per cent fish. There are recipes based on pasta and others on rice. Red peppers, tomatoes, onions and garlic are included frequently, and the expectations of a "Mediterranean diet" listed by the Oxford respondents are well reflected.

The Mediterranean is, of course, a sea. It has many islands and is surrounded by coastline winding through some 19 different nation-states with many regional and language divisions. To complicate the issue further, a country such as Portugal, with no Mediterranean coastline, is considered by many to be "Mediterranean" in this dietary context. By contrast Egypt is forgotten, despite its Mediterranean coastline. Amid such diversity can the popular idea of a single healthy Mediterranean diet stand close scrutiny?

I researched food intake in coastal Catalonia to examine further Mediterranean food. However one defines Catalonia - a political issue - it includes many kilometres of Mediterranean coastline. In March 1995 I carried out a seven-day food-intake frequency study. Data on every item consumed were collected from over 500 French and Spanish residents of the coastal strip between Port Vendres in France and Llansa in Spain.

This revealed not only significant international differences between adjacent Mediterranean populations but also a stark mismatch between the actual food intake in these coastal populations and the British concept of a "Mediterranean diet". If one simply takes the percentage of those who ate some sort of shellfish (France 24 per cent, Spain 36 per cent) and compares it with the proportions who ate beef (France 82 per cent, Spain 57 per cent) or even lamb with its highly saturated fatty acids (France 33 per cent, Spain 42 per cent) the pattern scarcely fulfils British assumptions. Similar demonstrations can be made for other food items; such as the numbers eating pasta at least once in the week (France 78 per cent, Spain 72 per cent) compared with the percentages eating potatoes (France 86 per cent, Spain 69 per cent). There were no vegetarians in either sample. Everyone ate meat at least once in the week, but not all ate fish even once. Perhaps the meat/fish situation can best be summarised by comparing average frequencies for the consumption of both: meat accounted for 77 per cent of the proportion of the total meat and fish consumed in France; 84 per cent in Spain. Yet 90 per cent of Oxford respondents mentioned fish or seafood as an essential component of a Mediterranean diet.

The Italian nutritionist Ferro Luzzi argues that the "Mediterranean diet" debate has not emphasised vegetables and fruits enough, nor has it examined those consumed in southern Europe. Tomatoes, olives, aubergines and salads were listed in my Oxford survey and are often mentioned in the Sainsbury recipe book. They were eaten, but not that frequently, according to my Catalan study. On both sides of the border olive oil was considered expensive and usually replaced by other vegetable oils. There was a greater use of butter on the French side than on the Spanish.

My research revealed an internal coherence in the anglophone concept of the "Mediterranean diet", linking medical literature, popular publications and items listed in the Oxford survey. But my Catalan study shows a very different diet, at least in one small area of the western Mediterranean shoreline.

At a congress in Naples this spring, entitled "The road of food habits in the Mediterranean area", while one or two southern European speakers still referred to "a diet characterised by high fish and low meat consumption", others demonstrated changes in the past two or three decades. No one produced comparative data to contrast diets and cultures from around the Mediterranean or questioned the assumption that there exists a single diet deserving of this title. The anglophone ideology of this "Mediterranean diet" seems to rest on descriptions of food habits, primarily Italian and Greek, observed in the 1950s but not at the time analysed quantitatively in many regions of southern Europe, let alone around the eastern and southern shores of the Mediterranean. Even among those southern European countries where mutually intelligible Romance languages are spoken, there are international and regional variations in food habits.

Nevertheless, the anglophone image of this diet has led to valuable biochemical research relating some foods to lower risks for coronary heart disease and cancer. I do not wish to throw any doubts on this research. The evidence, for example, in favour of health benefits from olive oil, marine foods and fresh fruit and vegetables is compelling. But should the title "Mediterranean diet" be linked to this information?

Other research has been based on geographic differences in epidemiology and causes of mortality; results for some southern European regions over a generation ago showed lower frequencies of cardiovascular disease. But to suggest that such results are relevant to countries all around the Mediterranean is a presumption. In some of these countries causes other than coronaries account for more deaths at younger ages. Meanwhile, many other lifestyle factors or genetic data have not received the quantity of research or finance the Mediterranean diet has received. Smoking tobacco has received attention, but what about other factors in the previous five or six decades? Issues perhaps deserving of more attention in terms of whether or not they protect against illnesses such as cancer and heart disease include: poverty, war, hunger, levels of physical activity in the workplace, levels of pharmaceutical intervention, eating communally, mobility and recognition of socioeconomic status.

To glimpse such angles one has sometimes to step back from specialist research. As academic knowledge becomes ever more detailed, disciplines fragment further to allow sufficient specialism to master the accumulated detail. The drawback is that specialists can become blinkered to wider views. The good news is, however, that realisation of this shortcoming has led to an increase in topic-oriented, cross-disciplinary courses, although finance for multidisciplinary researchers remains rare.

In such a way biochemists and nutritionists have perhaps not questioned, or needed to question, what their "Mediterranean diet" really consists of or whether it is appropriately named. Once a name is applied there are consequences. If fish and olives are considered "Mediterranean", research will continue on marine and olive oils. If a cookbook is published under the title "Mediterranean" all the recipes are popularly considered to be healthy. No one stops to ask what recipes should be included and whether they are indeed all beneficial to everyone under all circumstances. Crema Catalana is delicious, Mediterranean and featured in such recipe books, but I would hesitate to recommend it for reducing the risks of heart attack. Associating the word Mediterranean with foodstuffs has become a marketing ploy in northern Europe, on the basis of its healthy profile, and the sequel is that this anglophone concept has recently been invading southern Europe. In Barcelona, Marseilles and Naples, one can talk to ordinary folk who have heard that their diet is "healthy", but they are vague about details. In this way they too are recreating their "traditions", perhaps to conform to the northern ideology of a "Mediterranean diet".

Helen Macbeth is principal lecturer in anthropology, Oxford Brookes University.