There's something ironic about the Mediterranean diet used in the Lyon diet heart study, the one that dramatically reduced participants' risk of heart attack and all-cause mortality relative to the prudent diet control group: it wasn't actually a Mediterranean diet.
The concept of the Mediterranean diet as protective against heart disease may have originated in Dr. Ancel Keys' Seven Countries study, in which he compared the food habits and cardiovascular mortality statistics both between and within seven European countries. Countries surrounding the Mediterranean, and in particular the Greek island of Crete, had the lowest cardiovascular death rates. The Cretan diet is high in monounsaturated fat, relatively low in saturated fat, low in omega-6, and high in omega-3 fatty acids, including fat from seafood and the plant omega-3 alpha-linolenic acid. It also includes abundant green vegetables. This became the inspiration for the modern American concept of the "Mediterranean diet". The part about low omega-6 tends to be omitted.
Of course, if you look at modern heart attack mortality statistics by country, France is the lowest in Europe. France is a Mediterranean country, yet happens to have a very high intake of saturated fat per capita. So the cardiologist-approved version of the Mediterranean diet isn't exactly accurate.
The Lyon study departs even further from the traditional Mediterranean diet. Neither the Cretan nor the French diet are low in fat, yet participants were encouraged to reduce their fat intake. The Cretan diet includes some animal fat and eggs, while Lyon participants were encouraged to avoid these. And finally, the margarine. You could be guillotined for using margarine instead of butter in France, and I'm sure the Cretans aren't too fond of it either. Yet the margarine used in the Lyon study was rich in omega-3 alpha-linolenic acid, a critical factor.
Previous intervention trials such as MRFIT, the Women's Health Initiative (WHI) dietary modification trial, and others, exhaustively tested the hypothesis that reducing total fat intake reduces cardiovascular mortality. It doesn't. A dozen trials have also tested the idea that reducing saturated fat reduces cardiovascular mortality. It doesn't. Increasing fiber doesn't, according to the DART trial. Increasing fruit and vegetables modestly doesn't, according to WHI.
So what's left that's unique about the Lyon trial? It was the only trial to dramatically reduce omega-6 consumption, to below 4% of calories, while increasing omega-3 consumption from plant and seafood sources. In my opinion, that combination is the only plausible explanation for the large reduction in heart attacks and total mortality. That combination also happens to be a consistent feature of the real Mediterranean diet. In both Crete and France, omega-6 intake is relatively low, and omega-3 intake is relatively high. They also eat more real food than processed food in general, a factor that I don't underestimate.
Where do we go from here? Obviously I'm not going to recommend eating omega-3 enriched margarine. Mediterranean countries don't need industrial goop to avoid a heart attack, and neither do you. Anyone who's been to France knows they don't deprive themselves over there. They eat real food and they enjoy it.
The way to preserve the essential elements of the Mediterranean diet without becoming an ascetic is to eat fats that are low in omega-6, and find a modest source of omega-3. That means eating full-fat dairy if you tolerate it, fatty meat if you enjoy it, organs, seafood, olive oil in moderation, coconut oil, butter, lard, and tallow. Along with a diet that is dominated by real, homemade food rather than processed food. Some people may also wish to supplement with small doses of high-vitamin cod liver oil, fish oil or flax. I consider the latter to be inferior to animal sources of omega-3, but it can be useful for vegetarians.