Cultural Models of Eating

Traditional Mediterranean (Greek & S.Italy) & Asian (Japan) Food Patterns of the 1960's.

The Mediterranean & Asian diet pyramids were developed by Professor Walter Willet and colleagues at Harvard University. They reflect the traditional food patterns in the 1960s in Greece and Japan which have been subsequently linked with longevity and low rates of heart disease and cancer. The evidence behind these pyramids originated from the 7 countries study. Unlike other pyramids/food guides, these pyramids indicate frequency of servings (qualitative) across a day, week and month. Red meat is recommended a few times per month and fish weekly. Oils are no longer at the top of the pyramid because they are not severely restricted. Also, legumes and nuts are separated from vegetables and fruits and consumption is encouraged daily.  This paper will review the evidence regarding these cultural models of eating.

Mediterranean Diet Pyramid  (Willet et al. AJCN 1996)


Traditional Asian (Japanese) Diet Pyramid


The 7 countries Study 1960's.

The Mediterranean diet which is currently promoted as a model for 'good health' comes from the Greek island of Crete and Southern Italy around the 1960's when the prospective longitudinal study (30 year mortality follow-up) began - known as the 7 countries study. Since the 1960's, the foods and dietary patterns of Greece and Southern Italy have changed considerably with the increased use of meat, dairy products, polyunsaturated margarines and less olive oil, legumes and nuts.

The 7 countries study instigated by Professor Ancel Keys documented the diets of 13,000 middle-aged men aged 40-59 living in Southern Italy, Greece (Crete, Corfu), Yugoslavia, Finland, Netherlands, Japan and US (about 70% of the men lived in rural areas).

This study showed clearly that populations who ate the most saturated fat (Finland) had the highest levels of blood cholesterol and the most Coronary Heart Disease (CHD). This study also awakened the interest that the Cretan and Japanese diets may be related to the lower incidence of CHD and colon, prostate and breast cancer in these countries.

The Cretan diet was one of the highest fat diets in the study (40% energy as fat, but low in saturated fat and high in monounsaturated fat), but the Cretan Greeks eating it came out with some of the lowest rates of CHD and all-cause mortality and consequently had the longest life expectancies in the world at that time. They were even living longer than mainland Greeks and the Japanese. 

In the first five years of the study there were only two heart attacks in Crete and one of these was the local town butcher. The cohort from Finland had the highest rates of CHD, being 30 times higher than Crete even though they had a similar intake of fat to the Cretans. However, most of their fat was derived from cow's milk products and was very high in saturated fat.

The Cretan diet was also relatively low in carbohydrates (45% energy) with most of the carbohydrates consumed being of low glycaemic index. Even though the diet was high in fat, it paradoxically had a low energy density because of the large volume of plant foods consumed. 

The Cretan diet in the 1960's consisted of a high intake of vegetables (360g/day i.e. 2-3 cups/day), especially green leafy types, high intakes of legumes (30g/day), nuts (30g/day), fruit (460g/day i.e. 2-3 fruits /day) and wholegrain cereals (450g/day i.e. 8 slices of bread/day).

Vegetables were used as an integral part of meals, not served ďon the side? Breads were made by allowing dough to ferment slowly, resulting in a lower glycaemic index. In contrast, many breads we eat today use the rapid dough fermentation procedure resulting in breads with a higher glycaemic index.

They also had moderate intakes of fish (40g/day), alcohol (20g/day, with meals i.e. 2 standard drinks/day) and low intakes of red/white meat (35g/day) and fermented sheep/goats milk products (cheese/yoghurt consumed weekly). Unfermented dairy products i.e. fresh sheep/goats milk was reserved for children. The average daily intakes of some of these foods converted to the real-life situation of weekly consumption, would be: a 200g serving of fish twice a week, a 200g serving of chicken/pork once a week, a serving of lamb once a week or less, and bean/lentil soups or casseroles about 2-3 times a week.

The Cretan diet was also low in animal fats (butter/margarine rarely eaten) and saturated fat  (8% energy), but high in total fat (40% energy) and monounsaturated fat (25% energy) because the major source of fat was olive oil (about 4 tablespoons/day), which was nearly always consumed as 'extra virgin' or the first pressing of the olive and nearly always consumed with plant food. 

It was also high in omega-3 linolenic acid (probably about 1% energy intake) due to the high intake of nuts (especially walnuts), wild greens (especially purslane) and legumes.

Furthermore, animals (sheep, goats, pigs, chickens) were not always fed on grains (which is relatively rich in n-6 fats), but were also allowed to graze on wild plants and herbs, resulting in fat deposits which were higher in n-3 fats. So the little meat they ate tended to be leaner and the fat of these animals was higher in n-3 fats and probably phytochemicals.

The diet also had a favourable n-6: n-3 ratio (<5:1) due to the absence of polyunsaturated oils in the diet, the high intake of olive oil and foods high in n-3 fatty acids (including fish). In the 7 countries study, Cretan men had the lowest incidences of CHD and cancer, followed by Japan. Even though the fat content of the two diets was very different (40% vs. 10% total energy respectively) both were rich in omega-3 fatty acids (Simpoulos Prostaglandins, Leukotrienes & Ess Fatty Acids 1999; 60: 421-9). Cretan men were found to have 3-fold higher intakes of n-3 linolenic compared with the cohort from the Netherlands and 21% lower intakes of n-6 linoleic acid (Sandker et al. Eur J Clin Nutr 1993; 47: 201-8).

The Lyon Diet Heart Study 1980's & 90's.

More evidence in support of the health benefits of the Mediterranean diet has emerged recently from Lyon in France, rural Greece and Melbourne (see next section).

In the Lyon Diet Heart Study 600 middle aged Frenchmen who had suffered and survived a heart attack were placed in two different groups: 
Group 1.
Prudent diet recommended by the American Heart Association  
Group 2.
Mediterranean diet. 
There was no difference in the medications or referral for invasive cardiac testing between these groups.

Both groups had a similar content of fat (30% energy) but group1 compared to group 2 had a higher daily intakes of: saturated fat (12% vs. 8% energy); almost double the intake of n-6 linoleic acid (5.3% vs. 3.6%); one third of the intake of n-3 linolenic acid (0.3% vs. 0.84%); a lower intake of monounsaturated fats (10.8% vs. 12.9%); lower intake of fruit (203g vs. 251 g), bread (145g vs. 167g), legumes (10g vs. 20g), canola oil margarine (5g vs. 20g); higher intakes of meat (60g vs. 40g) and butter/cream (16g vs. 3g); and similar intakes of fish (40g) and oil (16g).

To mimic the omega-3 content of the 1960's Cretan diet, a canola oil margarine was developed for the study, but subjects were also encouraged to use olive oil. The study was meant to last for 5 years, but after looking at the rates of death, heart attacks and angina it was stopped after 2 years because of a marked difference between the two groups. Within months of starting the diets, plasma levels of vitamin C, E and omega-3 linolenic acid (mainly from canola oil) were higher and those of omega-6 lower in the Mediterranean diet group. This group had a 70% reduction in all cause mortality, including heart attack and angina, compared with people on the prudent diet. These differences in mortality occurred despite similar levels of blood lipids, blood pressure, body mass index and smoking (~17% smokers) in the two groups (De Lorgeril et al. Lancet 1994; 343: 1454-9; Circulation 1999; 99: 779-85).

After 4 years, overall death rates and cancer deaths were respectively 56% (p=0.03) and 61% (p=0.05) lower among patients eating the Mediterranean diet high in omega-3 linolenic acid (De Lorgeril et al. Arch Inter Med 1998; 158: 1181-7) 

In other words, the protective effect of the Mediterranean diet has been shown after 2 years and now 4 years, suggesting it is not only effective but robust. The chances of dying from a heart attack (and cancer) were more than halved in those who had received only a one-hourís instruction on the benefits of eating a Mediterranean diet, especially the increased intake of plant food, omega-3 linolenic acid (canola oil) and monounsaturated fats (olive oil) and decreased intake of saturated fats, linoleic acid, butter/cream and meat.

What is particularly interesting about this study is that it is possible to achieve positive health benefits with relatively little dietary education. Just learning about the diet seemed to be enough to lead to major lifestyle changes, whereas previously it has been thought that dietary change is a complicated process requiring major health educational input. However, the magnitude of this effect, and its robustness over time is enough to suggest that this type of eating can be promoted more widely.

Mediterranean diet and Survival in the Elderly in the 1990's.

Further evidence in support of the Mediterranean food pattern of the 1960s has come from three prospective cohort studies which described the food habits of people aged 70 and over in Greece (Trichopoulou, Kouris-Blazos et al., BMJ 1995; 311 (7018): 1457-1460), Australia (Kouris-Blazos et al. Br J Nutr, 1999; 82: 57-61) and Denmark (Osler  & Schroll, Int J Epidem 1997; 26 (1): 155-9). These cohorts were followed-up 5-6 years later to ascertain survival status. The Australian study included Greek-born and Anglo-Celtic born elderly living in Melbourne. 

Subjects with food patterns consistent with the food patterns found in Greece in the 1960s had a reduced risk of death by about 50%, even as late as 70 years and onwards. Smoking and male gender were not significant predictors of mortality. The food pattern was defined as follows: 
1) high consumption of vegetables; 
high consumption of legumes;
high consumption of fruits; 
high consumption of cereals; 
low consumption of
dairy products; 
low consumption of meat and meat products; 
moderate ethanol
high monounsaturated: saturated fat ratio. 
Subjects achieved greater
mortality advantage if  they followed the entire food pattern, suggesting synergy between food groups. To read the full text BMJ paper on the study in Greece go to:

The traditional Greek diet has also been implicated in protecting middle-aged and elderly Greek migrants in Australia from fatal cardiovascular disease (CVD), even though they appear to have all the standard risk factors for CVD; their blood pressure was found to be similar to people born in Australia, their cholesterol is similar, they are more overweight, and yet they appear to be protected from heart disease and diabetes complications. This phenomenon has been described as a 'morbidity mortality paradox' in Greek born-Australian elderly (Kouris-Blazos et al. Aust J Nutr Diet 1999; 56 (2): 97-107) and as a 'Greek-migrant paradox' in the middle-aged (Itsiopoulos & O'Dea unpublished data).

This phenomenon may be partly explained by the olive oil in the diet, but it is probably also related to their high intake of plant foods and circulating blood levels of a lot of protective antioxidant phytochemicals that are absorbed from fruit and vegetables with the help of olive oil.  This is supported by recent evidence emerging from a Monash University dietary intervention study on Anglo-Australian middle-aged diabetics (put on a traditional Cretan diet for 3 months) and the epidemiological evidence on 10,000 Melbourne Greek migrants "Health Radio 2000," that suggests that "its not what you eat, its how you eat it" (O'Dea 1999, ABC interview).

It is now being hypothesised that the Mediterranean cuisine is a very important component of maximising the protection one derives from eating a lot of vegetables combined with olive oil. For example, the antioxidant lycopene in tomatoes is better absorbed in the intestine if the tomatoes are cooked; absorption is even better if cooked with oil. This makes sense, since these pigments are fat soluble. The glycaemic index concept might help further explain why the high fat low carbohydrate Mediterranean diet, with all the legumes and wholegrains, is healthy.

  To listen to a "Health Report" interview (conducted by Norman Swan) about the Mediterranean diet, click here.

Mediterranean VS  Japanese diet.

But surely the Mediterranean diet does not have a monopoly on eating well.  In the seven country study, the other diet that came out with low heart disease rates (but not strokes or stomach cancer) was the Japanese diet high in carbohydrates and low in fat - the exact opposite to the Mediterranean diet.  This suggests that disparate food patterns can result in comparable health i.e. there is more than one ideal diet that can lead to good health. 

Unfortunately, China or other Asian countries (India, Indonesia, Thailand) were not included in the 7 countries-study so we do not have comparable information from the 1960's on these cuisines. 

The Japanese population of Kohama island have the lowest incidence of CVD in Japan and probably in the world. Their high intake of omega-3 and omega-9 fats, low intakes of omega-6 fats have been implicated in providing cardioprotection. Their high intake of n-3 fats is due to their use of canola and soybean oils and high intake of fish (Kagawa et al. J Nutr Sci Vitaminol 1982; 28: 441-53).

There has been some debate about whether the Mediterranean diet is preferable to the Japanese diet. It does seem as though both of these diets can be healthy. However, Professor Willet has one word of caution about the Asian diet

" it tends to be very high in white rice, and that may be something that is tolerated by a population that is extremely lean and active as traditional societies in Asia have been. But when a society starts to work in offices and drive cars instead of working in the fields for many hours a day, they will tend to develop more insulin resistance, and in that case its pretty clear that we can't tolerate high carbohydrate intake nearly as well as a peasant farmer can ... the total percentage of calories from fat in the diet probably does not make too much difference, if itís the right fat. But we need to pay attention to the type of carbohydrate in the diet, and try to minimise the highly refined carbohydrates like white bread and potatoes and instead use whole-grain, high-fibre types of carbohydrates whenever we can. We really have to recognise it's not that carbohydrates are good and fats are all bad, there's good and bad in both of them".

According to Prof Frank Sacks of Harvard University, both cultural models result in low blood cholesterol (LDL cholesterol) levels. Both a low fat and a high fat diet, low in saturates, can reduce LDL cholesterol by 17%. However, low fat high carbohydrate diets reduce the 'good' blood cholesterol (or HDL cholesterol) by 18%, relative to the Mediterranean model.

To relate this HDL effect to coronary incidence, the low fat diet would predict a 16% increase in coronary incidence in men and even more in women. On the Mediterranean model there would be less change. In the short-term at least, low fat diets also increase serum triglycerides of the order of 25%. This would predict an increase in coronary incidence of 16% in women and 6% in men. Replacing one type of fat with another does not affect triglyceride levels. On the low fat diet the benefits of the decrease in LDL are nearly completely offset by the negative impact of reduced HDL and increased triglycerides, particularly in women, and this actually predicts an increase in coronary incidence. These effects on blood lipids may explain why the Greeks in the 7 countries study had much lower death rates from CHD than the Japanese.

One shortfall of the traditional Cretan and Japanese diets of the 1960's was the high content of salt due to high intakes of olives, bread, salted fish, pickles etc. Studies have found a relationship between high salt intakes in these countries and the high rates of stroke and stomach cancer.

Overall, the low intake of meat and saturated fats, moderate intake of alcohol (with food) and high intake of unrefined cereals, legumes, vegetables, monounsaturated fat, fish (especially Japan) and fruit (especially Greece) is believed to have contributed to their longevity in the 1960's. However, the much higher rates of stroke and stomach cancer in Japan are thought to be linked with their much lower intake of fruit, possibly fat and higher intake of salty foods.

  We need to remember that traditional cuisines are not necessarily perfect and can be improved. For example, they evolved when there was limited food availability and no refrigeration; as a result certain foods/dishes developed which may not be so good for ones health e.g. salty pickled vegies/meat/fish.

  Professor Walter Willet concludes in an article in Science magazine (1994) that the evidence is in favour of the traditional Greek diet with respect to long-term safety.

Last Updated: March 28, 2001.