by ANTONIA TRICHOPOULOU, MD, Professor of Nutrition and Biochemistry, Athens School of Public Health, and Director, World Health Organization Collaborating Center for Nutrition Education in Europe; and DIMITRIOS TRICHOPOULOS, MD, Vincent L. Gregory Professor of Cancer Prevention and Epidemiology, Harvard School of Public Health
How do we know that people in a particular area of the world have a healthier diet than our Western diet? Epidemiology, the study of disease patterns, provides the answer.
The frequency of most diseases varies substantially around the world. For diseases of infectious or occupational origin the importance of the environment, whether physical, chemical, or biological, is clear. But for many other common chronic diseases, including those of the circulatory system and cancer, the role played by the environment has not always been obvious. Studies of migrants have shown that whatever the disease pattern in their country of origin, they tend to acquire, sooner or later, the disease pattern of their host country, even when they remain relatively isolated within ethnic communities. For instance, the frequency of cancer of the large intestine is much lower in Japan than in the United States but rises to American levels within 20 years among Japanese immigrants to this country. We can conclude, then, that outside factors, collectively termed environment, whether imposed by others or created by individuals themselves through personal choices about food, behavior, or lifestyle, are critical determinants of disease. This is not to imply that heredity is not important; it is, but genes are usually evenly distributed among nonisolated population groups and thus influence who within a population will develop disease. And most genes exercise their effects through interactions, whether simple or complex, with the environment.
Among the many environmental factors, qualitative or quantitative deviations from the ideal diet represent, as a group, the most important factor in the genesis of the most common killer diseases, including coronary heart disease and several forms of cancer. This should not be surprising. Only factors that vary widely among countries and population groups can account for the remarkable variability of disease patterns around the world, and few aspects of human conditions or behavior vary as much as diet. Of course, the human diet is unusually complex, involving hundreds of chemical components, and from the evolutionary point of view it is frequently challenging, since humans are the only animals who process and cook their food.
Nobody knows for sure what the ideal diet is, but there are good reasons to believe that the Mediterranean diet may come closer to it than any other realistic diet. What exactly is the Mediterranean diet? What evidence do we have that adopting this diet can benefit our health?
The Mediterranean diet is a loose term, and some doubt remains as to what precise dietary patterns apply. However, there are common elements in the diets of most Mediterranean people, and to the extent that they share lower rates of diet-related diseases, this looseness of definition represents an advantage: several variations on the Mediterranean diet may be equally beneficial. As a rule, the Mediterranean diet is low in saturated fat with added fat mostly in the form of olive oil; high in complex carbohydrates from grains and legumes; and high in fiber, mostly from vegetables and fruits. Total fat may be high—around 40 percent of total energy intake in Greece, as mentioned in the chapter Making the Change—or moderate (around 30 percent of total energy intake, as in Italy). In all instances, however, the ratio of monounsaturated to saturated fats is high (usually two to one or more). This is because olive oil, which contains 60 percent or more of monounsaturated oleic acid, is the principal fat in the Mediterranean region. The large quantities of fresh vegetables and of cereals and the abundance of olive oil guarantee high intakes of beta-carotene, vitamin C, tocopherols (vitamin E), various important minerals, and several possibly beneficial nonnutrient substances such as polyphenols and anthocyanins.
Since Homeric times, the diet of the Mediterranean has been based on wheat, olives and olive oil, legumes, green vegetables, seasonal fruits, and wine. Other components of the diet include onions and garlic, cheese and yogurt (mainly from goat’s milk), and, to a certain extent, fish, fowl, and eggs. Because of limited availability, red meat has been used infrequently and, as a rule, in small quantities. Few adults in the Mediterranean region will find it difficult to recognize this pattern as the essence of their traditional diet. In comparison to the average American, the average Italian consumes 3 times as much pasta, bread, and fresh fruit; almost twice as much tomatoes and fish; 6 times more wine; and 100 times more olive oil—but 30 percent less meat from any source, 20 percent less milk, and 20 percent fewer eggs.
Any overview of the actual foods and dishes consumed in the Mediterranean region must emphasize the large quantities of whole-grain bread that accompany every meal. Pasta (not just in Italy but also in Greece and in North Africa in the form of couscous) is regularly used, frequently in addition to bread, as an added component in several soups, as a supplement to various dishes, or as a separate dish. Abundant complex carbohydrates, as well as plant proteins, are contributed by whole grains and cereals and also by legumes. In Greece bean soup, prepared with large quantities of olive oil, is considered the national food, credited with the survival of Greeks in their poverty-stricken country through the millennia.
The Mediterranean diet is frequently, and rightly, associated in the minds of many Americans with that lively salad made from fresh vegetables, olive oil, and feta cheese, known in this country as Greek salad. Usually prepared with tomatoes, cucumbers, onions, and olives, the original Greek salad is different from its American variant; in the Greek version, vegetables are fresher and tastier, olive oil is much more abundant, and the serving is much larger. A Greek fresh vegetable salad can also be made from just greens, herbs, olive oil, and lemon juice; again, feta cheese, made from goat’s or sheep’s milk, is frequently added. It is worth noting that cheese is consumed in large quantities along the northern Mediterranean shore—in fact, Greeks and French lead the world in per capita cheese consumption.
In Greece, vegetables are usually cooked, especially in olive oil, and vegetable combinations include eggplant, zucchini, beans or okra, and almost always tomatoes, onions, and herbs. Vegetables also dominate many meat dishes, and, in the absence of meat, feta cheese is regularly added to most vegetable stews.
As might be expected, fish has been an important food for Mediterranean people since antiquity. Meat, however, has not been consumed in large quantities in the region until recently, mainly because it’s been expensive and in short supply. The changing pattern has been implicated in the increasing incidence of diverticular disease and perhaps cancer of the large bowel.
Wine consumption is high in the European Mediterranean countries, but widespread abuse of hard liquor is not. For millennia wine has been consumed in moderation, almost always during meals and as a rule in the company of friends—the ancient Greek word symposium means “drinking in company,” but with a connotation of intellectual interchange. The philosophy that shaped the Mediterranean attitude toward wine consumption is best expressed in a passage from Plato’s Symposium that many modern epidemiologists would approve: “I prepare but three kraters for prudent men; the one is for health, the one they drink first; the second is for love and pleasure; the third for sleep. The fourth is not ours, but belongs to licentiousness.”
By the late 1950s the hypothesis linking diet in general to coronary heart disease was gaining widespread support, although reliable data linking dietary habits of individuals to their risk of developing coronary heart disease were lacking. The existing evidence was indirect: diet, in particular the composition of dietary fat, had been shown to be a critical determinant of serum cholesterol, with saturated fat increasing and polyunsaturated fat reducing serum levels; serum total cholesterol had been established as a major risk factor for coronary heart disease. At that time no distinction was made between low- and high-density lipoproteins (LDL and HDL or “bad” and “good” cholesterol).
In an effort to evaluate the hypothesis of a relationship between diet and coronary heart disease and to explore several other aspects of the etiology of this disease, Ancel Keys, MD, and his colleagues set out to undertake what turned out to be one of the most celebrated studies in modern epidemiology, the Seven Countries Study. The study involved 12,763 men aged 40 to 59 years. The men were enrolled between 1958 and 1964 in 16 study groups or cohorts: 2 in Greece, 3 in Italy, 2 in Croatia and 3 in Serbia (both then part of Yugoslavia), 2 in Japan, 2 in Finland, 1 in the Netherlands, and 1 in the United States. The men were interviewed and examined so that information was available for virtually all characteristics that later were identified as major risk factors for coronary heart disease, including blood pressure, serum cholesterol, tobacco use, physical activity, dietary habits, and several others. After a follow-up period that lasted for at least 10 years, the study generated several important results; findings continue to be reported by investigators active in some of the original study centers. Among the most important findings were that the Mediterranean groups had lower mortality rates from all causes together than the northern European and American groups; that the difference in mortality and incidence was particularly striking with respect to coronary heart disease; that the mean percentages of calories from saturated and polyunsaturated fats in the diets of the groups could account to a considerable extent for the differences in mean levels of serum cholesterol among them; and that the mean percentage of calories from saturated fats could account to a considerable extent for the differences in the incidence of coronary heart disease among the groups (an effect that could have been mediated through the detrimental effect of saturated fats on serum cholesterol).
Results subsequently reported from other investigations provided direct support for many of the findings of the Seven Countries Study and indirect support for others. Data assembled by the World Health Organization have confirmed the low rates of coronary heart disease in Mediterranean countries not only in the late 1950s but even now, when the traditional diet is not so closely followed. In 1990 the annual mortality from this disease per 100,000 persons of a “standard” age was 243 among men and 132 among women in the United States, whereas it was, respectively, 139 and 64 in Italy, 106 and 47 in Spain, 137 and 59 in Greece, and 91 and 40 in France. Mediterranean countries are also characterized by lower rates, as compared to those in the United States and northern Europe, of several nonsmoking-related cancers, including those of the large bowel, breast, prostate, and ovary. The consequences of these disease patterns create surprising contrasts. According to the most recent (1992) World Health Organization data, men at the peak of their lives (45 years) have longer life expectancy in Greece than in any other European or North American country, even though Greek men are notorious for their high tobacco consumption, they rarely exercise in a systematic way, and they are served by a rather modest health-care system.
For logistical reasons the dietary data in the Seven Countries Study were analyzed only as group averages and not for each individual subject in the study. Although this is not an optimal analysis, there can be no doubt that many of the differences in disease occurrence between Mediterranean groups on the one hand and northern European and North American groups on the other were due to differences in dietary patterns. For age-adjusted data, only major differences in other important disease-causing factors or conditions could explain the observed large differences in disease incidence. Such factors are genes, widespread epidemics, tobacco smoking, and low socioeconomic class. But none of these factors is likely to have played a major role: migrant studies have eliminated a genes-based explanation; no major infectious epidemic was selectively affecting northern Europe and North America; and poverty and tobacco smoking have been, if anything, more common in Mediterranean countries.
The Seven Countries Study confirmed that a diet low in saturated fat, like the Mediterranean diet in virtually all its variations, can reduce total serum cholesterol and risk of coronary heart disease. In the process, however, the Mediterranean diet came to be perceived as just another low-saturated-fat diet. Scientists from the Mediterranean countries have tried to argue that there is much more to the Mediterranean diet than its low intake of saturated fats, but the majority of the educated public was too preoccupied with the polyunsaturated-to-saturated-fatty-acid ratio and the “reduce total fat” commandment to pay much attention. It was only in 1991 that a major editorial in the influential New England Journal of Medicine, by the distinguished scientists Frank Sacks and Walter Willett, revived interest in the Mediterranean diet. These authors pointed out that high levels of serum HDL cholesterol are probably as important for the prevention of coronary heart disease as low levels of serum LDL (and total) cholesterol. They argued further that a diet low in saturated fat but high in monounsaturated fat, which would reduce LDL cholesterol and increase HDL cholesterol, appears to be at least as good as a diet low in saturated fats and high in carbohydrates. Although both monounsaturated fat and carbohydrates are effective in reducing LDL, carbohydrates do not increase HDL cholesterol as much as monounsaturated fats do. Furthermore there is experimental evidence and limited human data justifying some concern about the long-term safety of a diet high in polyunsaturated fats, the diet that has been promoted by the American Heart Association, whereas the safety of a high-monounsaturated-fat diet has been demonstrated through the centuries in the people of the Mediterranean region. Willett and his colleagues and other authors also have provided evidence that partially hydrogenated vegetable fats, the basis of margarine and vegetable shortening, far from being safer than saturated animal fats, may actually increase the risk of coronary heart disease. So, it appears that among the realistic alternatives open to people accustomed to the Western diet, the Mediterranean diet stands out as apparently healthier.
Although most scientists believe that diet can in fact play a role in coronary heart disease by contributing to high cholesterol levels and subsequent atherosclerosis, an increasingly influential minority theorizes that diet may also interfere with the blood-clotting events that trigger acute coronary episodes (mainly myocardial infarction and sudden cardiac death). According to this hypothesis, long-chain polyunsaturated fatty acids, of what is chemically described as omega-3 type, have beneficial effects, including a reduction in blood-clotting tendency and blood viscosity and an increase in the breaking up of blood clots. These fatty acids are provided mainly by fish, but also by several plants, and there is indeed evidence that eating fish may reduce the risk of coronary heart disease. Although fish is not a defining characteristic of the Mediterranean diet, it is obviously an important part of it—certainly more important than red meat in the time-honored traditional diet of the Mediterranean.
There has been a tendency to associate the health effects of the Mediterranean diet with its fat composition, but perhaps equally important is the fact that almost all variants of this diet are high in complex carbohydrates (bread, pasta, legumes) and rich in fresh vegetables and fruits. Complex carbohydrates and fiber, derived from vegetables, fruits, cereals, or legumes, seem to play a major beneficial role in protecting against constipation, diverticular disease, and perhaps colorectal cancer and coronary heart disease. But the protective role of fresh vegetables and fruits is likely to extend to a much wider range of diseases. They have been consistently reported to be protective against, or at least inversely associated with, several common types of cancer, including those of the esophagus, stomach, large bowel, liver, pancreas, lung, bladder, cervix, and even ovarian and breast cancers. Whether the effects of vegetables and fruits are due to their high content of antioxidant vitamins like beta-carotene, vitamin C, and vitamin E, other vitamins like folate, certain minerals and trace elements, or other compounds is not clearly established. Antioxidants are strong candidates as protagonists in the disease-preventing capacity of vegetables and fruits. They are believed to prevent or neutralize the effects of oxidative processes that may be involved in the development of cancer, atherosclerosis, other chronic diseases, or aging itself. In most instances the protective effect seems to be more strongly associated with vegetables and fruits themselves than with any particular constituent nutrient or nonnutrient. Nevertheless, certain vitamins appear to protect against certain diseases in high doses that can be taken only in the form of supplements. This may be particularly true with respect to vitamin E and coronary heart disease. This field of research is rapidly developing, and it is not conclusively established that very high doses of vitamins are required. In any case it would be prudent to assume that whether vitamin supplements are taken or not, fresh vegetables and fruits should be plentiful in the diet, and the Mediterranean diet is as plentiful as any.
There are several other dimensions to the Mediterranean diet, and very few of them are linked to adverse health effects. Total fat intake, including intake of monounsaturated fat, is considered a risk factor for cancers of the large bowel and perhaps those of the prostate and the pancreas. However, the incidence of these cancers is generally lower in the Mediterranean region than in northern Europe or North America. It appears that red meat (and possibly animal protein and fat) is a more important determinant of cancer of the large bowel than monounsaturated fat, although other explanations may also apply. Consumption of red meat has been low in the traditional Mediterranean diet, although recent data suggest rapidly increasing trends. Salt intake is probably unnecessarily high in several Mediterranean countries, and this may contribute to the modestly high incidence of stroke and stomach cancer in some of these countries. Total energy intake is rather high in the European Mediterranean countries, but in the absence of excess prevalence of obesity in the region, this is actually an advantage—it implies higher energy expenditure and therefore higher levels of physical activity. Physical exercise is an important protective factor for coronary heart disease and possibly cancers of the large bowel and the prostate. Finally, the relatively low intake of milk is partly compensated by the high cheese consumption that covers, in most instances, the requirements for osteoporosis-preventing calcium.
The deleterious effects of excessive alcohol intake, both long-term and short-term, are well documented. By contrast, light to moderate drinking has been shown consistently to be associated with reduced risk of coronary heart disease by about 25 percent. Since coronary heart disease is the principal cause of death among both men and women in the developed countries, it is not surprising that light to moderate drinking has been associated with longevity and has been singled out as a likely explanation for the very low coronary mortality of the French, the so-called French paradox. Light to moderate drinking should be interpreted as two to three glasses per day for men or one to two glasses per day for women; a lower level is indicated for women because of their presumed higher sensitivity, as well as concerns that alcohol may slightly increase the risk for breast cancer.
It has been shown in several studies that alcohol increases the levels of “good” HDL cholesterol and, at least in theory, all alcoholic drinks should impart similar degrees of protection against coronary heart disease. This would be true even in lay quantitative terms, since the servings of most alcoholic beverages, including spirits, wine, and beer, contain approximately equal amounts of alcohol. Most epidemiologic studies have supported this hypothesis, but some authors have argued that wine, and in particular red wine, may be more beneficial, possibly because the latter contains compounds with antioxidant properties. Probably more important than the chemical differences between wine and other alcoholic beverages is the way wine is drunk, particularly in the Mediterranean countries—almost always during meals and in the company of family or friends, under conditions that favor moderation and discourage acute intoxication. Mature people and societies can find the balance that maximizes the beneficial health effects of wine and minimizes its adverse effects, but the balance can be a precarious one.
It is not clear whether the identified health-promoting aspects of the Mediterranean diet can fully explain the otherwise unexplained good health of Mediterranean people. Some scientists have proposed that the freshness of plant foods, various interactions among components of diet, or the pattern of eating and drinking may have elusive and difficult-to-identify synergistic effects. Others have argued that the relaxing psychosocial environment in most Mediterranean countries, the preservation of the extended family structure, the stress-releasing afternoon siesta habit, and even the mild climatic conditions may complement the beneficial effects of diet. Nevertheless few, if any, deny the central role of diet in the constellation of favorable conditions surrounding the Mediterranean people.
The Mediterranean diet and lifestyle are not the product of unusual insight or wisdom. They were shaped by climatic conditions and environmental constraints, and in many instances they represent adaptive responses to poverty and hardship. Furthermore, many of their important health-promoting components are not unique to the Mediterranean region but can be found in other areas of the world and other population groups. Still, several variations of the Mediterranean diet can be considered realistic models of a prudent diet that fits our current understanding of healthy nutrition. Indeed, it is the convergence of recent results from methodologically superior nutritional investigations with the dramatic ecological evidence represented by the Mediterranean natural experiment that has created the present momentum toward the Mediterranean diet. It would appear logical for non-Mediterraneans to adopt critical elements of this diet in their food habits and for Mediterraneans to reverse the trends that tend to draw them away from their health-promoting nutritional traditions.
Both authors of this report were born and raised in Greece, but both have traveled and lived in several parts of the world. In the process, we have come to realize that individuals as well as ethnic groups can learn a lot from the experience and the cultural tradition of others. It seems to us that a healthy diet may represent one of the most important contributions of the Mediterranean people to others around the world. It is tasteful, lively, and highly variable—and it can be an integral and important component of a health-promoting lifestyle.