The Mediterranean diet has better epidemiological support than any other named pattern. It also bears little resemblance to what most people picture when they hear the phrase, which is why the pattern keeps outperforming attempts to reproduce it.
What was observed
The populations studied in the mid-twentieth century ate a great deal of olive oil, vegetables, legumes and fish; modest amounts of dairy, mostly fermented; small amounts of meat; and wine with food. What they did not eat is at least as important: almost no refined seed oil, almost no sugar in liquid form, and almost nothing that came out of a factory.
They were also, by any modern standard, physically active, and they ate in a social rather than a grazing pattern. Both of those travel poorly into a nutritional recommendation, and both were probably doing real work.
The version that gets sold
Pasta with a jar of sauce, a supermarket bread basket, a drizzle of oil that may or may not be olive, and a glass of wine. That is not the studied pattern. It is a high-refined-carbohydrate diet wearing the name of one.
Nearly every large dietary trial that reported benefit from a Mediterranean pattern provided olive oil or nuts directly. The interventions were closer to supplement trials than to teaching people to eat pasta.
Which components appear to carry the effect
- Olive oil in quantity — the trials that supplied it saw the clearest results, and extra virgin specifically, for the polyphenols
- Oily fish, for the omega-3 content and the displacement of other proteins
- A very low intake of industrial seed oil, which is the mirror image of the modern diet
- Vegetables and legumes in volume, eaten as food rather than as a side
- Fermented dairy rather than fresh, in modest amounts
How it fits with everything else here
This site argues that most chronic disease runs through insulin resistance, and a diet with substantial carbohydrate looks at first like an odd fit. It resolves once you look at what kind. Legumes and vegetables with fat and protein produce a very different insulin profile from bread and sugar, and the studied pattern contained almost none of the latter.
For someone who is already insulin resistant, a lower-carbohydrate approach will move the markers faster. For someone metabolically healthy who wants a pattern with long observational support and no restriction they will resent, this is the best-evidenced option available — provided they eat the version that was studied.
No comments yet