There is a precise moment when you realize you have wandered into a religious war. Someone at the table orders pasta and somebody else looks at them as though they have made an irreversible mistake. Someone brings carbohydrates into the conversation. Someone explains, with the certainty of a person who has read three articles and watched two podcasts, that they no longer eat grains. Or sugar. Or food cooked above 42 degrees Celsius.

Diets have become identities. And as with identities, defending them has become more important than understanding them.

Let us try the opposite.

Seven dietary models have dominated public debate in recent years. They are not all the same. They do not all have the same scientific evidence behind them. They do not all produce the same effects over time. It is worth looking at them one by one, with the data, without the enthusiasm of someone who has just lost three kilos.

Looking at the evidence, one thing stands out: the Mediterranean diet is not the strongest option because somebody decided it should be. It has the most robust body of scientific evidence — a large controlled trial, dozens of meta-analyses and observational cohorts spanning decades. The American Heart Association, WHO, EFSA and Italy's CREA guidelines converge, each in their own language, on a very similar pattern. It is not often that institutions this different point in the same direction.

The only model that comes close in terms of evidence quality is a fully plant-based diet — but only when it is built around whole foods, not ultra-processed soy burgers and bags of vegan-certified crisps. In that form, it has good evidence for body weight and cardiovascular health, and it can outperform the Mediterranean diet for weight loss in the first six months. Its most concrete limitation is vitamin B12: supplementation is mandatory because reliable dietary sources are overwhelmingly animal-based and the body cannot do without it. A Mediterranean diet does not normally require routine supplementation.

The other five — ketogenic, paleo, low-carb, intermittent fasting and high-protein diets — can work in the short term for weight and some blood markers. Some work very well over the first six to twelve months. Then the evidence becomes thinner, dropout rates rise, and studies on long-term mortality are absent or inconsistent. That does not mean these diets are inherently dangerous. It means we know less than the confidence of the public debate often suggests, and much of what we do know concerns relatively short time horizons.

There is one thing science says clearly about restrictive diets in general, however, and it is not mentioned often enough in podcasts: a large proportion of people regain lost weight within two to five years after stopping the diet. Not because they are weak. Because restrictive diets change the rules temporarily without necessarily changing the relationship with food. And when the rules end — because sooner or later they do — people often return to where they started, sometimes with a little more weight and a great deal more frustration.

That is where the most important difference lies, and it is one that no comparison of macronutrients can fully capture.

“Going on a diet” is a temporary intervention with a beginning and an end. It works while you follow it. Stop following it and its effects often fade. “Adopting a dietary pattern” is different: it is a stable way of eating integrated into real life, habits and the cultural context you live in. The Mediterranean diet did not begin as a therapeutic regime invented by a specialist. It emerged from the way people living around the Mediterranean ate — people who, not coincidentally, already showed lower rates of cardiovascular disease than many other populations when Ancel Keys began studying them in the 1950s.

It is not a diet. It is an architecture for eating.

Eating well does not mean eating little and depriving yourself every time you want something. It means building a balance that can survive over time — at a family lunch, at dinner out, while traveling, during difficult weeks. A balance that does not require constant calculation, weighing or guilt. One in which Sunday dessert is not a violation of the rules but a normal part of the picture.

The eating patterns that work in the long term are not necessarily the most restrictive. They are the ones you can live with without feeling constantly burdened by them. And the pattern with the deepest long-term evidence base is probably one you already know, even if you have never thought of it as a “diet.”