There is no single best diet for insulin resistance, and the honest answer is more useful than a winner: Mediterranean, DASH, low carbohydrate and the structured lifestyle programmes modelled on the Diabetes Prevention Program all tend to help, and they help for overlapping reasons. Each one lowers the insulin load of an ordinary meal, mostly by cutting refined starch and sugar, adding fibre and protein, and replacing packaged food with food you recognise. Pick the one you can actually run for a year, then make your own version of it score better meal by meal.

That is what the Insulin Fix Score is for. It runs 0 to 100, and a higher score means a lower insulin load. It weighs glycemic load, protein, fibre, processing level, fat quality, liquid calories and the order you eat things in. Every diet below moves those same dials. They just argue about which dial to grab first.

What does the Mediterranean diet actually change?

It changes the fat and it changes the base of the plate. Olive oil instead of seed oil and margarine, fish and legumes instead of processed meat, vegetables and nuts in volume, bread and pasta demoted from the centre of the meal to the side of it. Fat is not restricted, which is why it feels livable to most people.

In a large Mediterranean diet trial, adding extra virgin olive oil was associated with fewer major cardiovascular events than a low-fat control diet (Estruch et al., 2018, New England Journal of Medicine, republished after the 2013 version was retracted). Worth knowing the history: the original publication was withdrawn because randomization was not carried out properly for about a fifth of participants, the authors re-analysed the data without assuming clean randomization, and the effect barely moved. That is reassuring rather than airtight, so the fair word is association.

In scoring terms, this diet wins on fat quality and real food, and it is neutral to weak on glycemic load, because bread, pasta and rice are still on the table. A Mediterranean meal with a fist of pasta and a lot of vegetables scores well. The same meal with a bowl of pasta and a token salad does not.

What does DASH change, and is it useful for insulin resistance?

DASH was built to lower blood pressure, and it shows: vegetables, fruit, beans, whole grains, low-fat dairy, less sodium, less added sugar, less processed meat. For insulin resistance, the sodium part is beside the point and the rest of it is not. Stripping out ultra-processed food and sugary drinks moves the processing and liquid-calorie dials hard, and those are two of the heaviest inputs into an insulin load.

Its weakness is the same as Mediterranean's, with one extra: DASH is generous with fruit juice and low-fat dairy products that often carry added sugar, and it is cautious about fat in a way the current evidence does not demand. A low-fat fruit yoghurt and a glass of orange juice are a DASH-shaped breakfast that scores badly, because liquid sugar arrives fast and leaves you unfed. Eat the orange.

Does low carb beat the others for insulin resistance?

For some people, probably. Low carb attacks glycemic load directly, which is the single largest driver of a post-meal insulin rise, and it tends to raise protein and fat quality at the same time. That is three dials at once.

But the person matters. A year-long Stanford trial comparing low-carbohydrate and low-fat eating found that how much insulin someone secretes may be part of who does better on which approach, with higher insulin secretors responding better to the lower-carb side (Gardner et al. DIETFITS, 2018, JAMA, Stanford). If you have been told you are insulin resistant, you are plausibly in that group. That is a reason to try it, not a guarantee.

The failure mode of low carb is not metabolic, it is social. Diets that ban a food group tend to end at a birthday party. Which brings us to the fourth option.

What about the DPP lifestyle programme, and does the diet choice even matter?

The Diabetes Prevention Program lifestyle arm was never a diet in the branded sense. It was coaching: modest weight loss, regular walking, weekly sessions, someone checking in. No forbidden foods, no named cuisine.

And the pooled evidence suggests that scaffolding is doing a lot of the work. Across randomized trials in people with prediabetes, lifestyle weight loss raised the chance of returning to normal blood sugar and lowered the chance of progressing to type 2 diabetes, and the review found no significant difference between diet, exercise, or both together (Jayedi et al., 2024, American Journal of Clinical Nutrition). Two things in that review are worth carrying into your kitchen. The benefit rose steadily from about 1% to 9% of body weight lost, with no threshold you have to clear before anything starts happening, so a small change is not a wasted one. And since diet and exercise landed in a similar place, the argument about which diet is correct matters less than whether you are still doing it in March.

So the interesting question is what the weekly session was actually supplying, because most people reading this will never be enrolled in a programme. It was feedback. Something outside your own head, at regular intervals, telling you how the last stretch really went rather than how you remember it going. Insulin is invisible, which is exactly why that outside reading matters: without one you are guessing at the thing you are trying to change.

Scoring a meal is the smallest version of that same loop. A number from 0 to 100, where higher means a lower insulin load, attached to the actual plate in front of you, plus one change worth making to it. It is not coaching and it is not a programme, but it does the one job the programme did that you cannot do from memory: it makes the invisible part of a meal visible while you can still do something about it. Run it on three or four meals you eat every week and you will know which of these diets your real cooking already resembles.

What do all four of them share?

Strip the branding and the same four moves are underneath:

There is a fifth move none of these diets names, and it is free. Eating vegetables and protein before the carbohydrate portion of the same meal sharply blunted the glucose and insulin rise, cutting the glucose spike by roughly three quarters and the insulin spike by about half compared with eating the carbs first (Shukla et al., 2015, Diabetes Care, Weill Cornell). Same food, same plate, different order. It works inside any of the four diets above, which is the clearest sign that the thing they share matters more than the thing they argue about.

How do I tell which version of this works for my actual meals?

This is where the comparison stops being useful, because you do not eat a diet, you eat Tuesday dinner. The question that matters is whether your Mediterranean bowl or your low-carb lunch is carrying a hidden insulin load, and the only way to answer it is to look at the specific plate.

Type your meal in a line, or photograph it, into the Insulin Fix Scanner, and you get a score from 0 to 100, where higher means a lower insulin load, plus the single change worth making to that meal. Most people find the change is smaller than they feared: halve the rice, add the eggs, drink the coffee black, eat the salad first. Scoring a meal is free and does not need an account.

What if I am already on medication?

If you are taking metformin, or on a GLP-1 such as semaglutide, or weighing one up, that is a conversation with your doctor and nothing here should delay it. Medication is often the right call. The food work holds either way: the order you eat in, the fibre and protein you put first, the liquid sugar you skip, all behave the same whether or not there is a prescription in the picture, and they are the part you keep when the prescription changes.

Frequently asked questions

If I have to pick one, which should it be? Pick the one that overlaps most with how you already cook, because the pooled evidence does not crown a winner and adherence is the variable you control. If you cook with olive oil and vegetables, go Mediterranean. If bread and pasta are the problem in your week, go lower carb. Then improve the score of individual meals from there.

Do I need a structured programme, or can I do this alone? Plenty of people do it alone. What the programme supplied was regular outside feedback, and you can approximate that by scoring the handful of meals you repeat most weeks rather than relying on how the week felt.

Do I have to lose weight for this to help? Weight loss helps in proportion to how much there is, starting from very small amounts. But the blood sugar number is worth targeting in its own right, and improving the insulin load of your meals is how you move it.

Does food order really matter if I am eating well anyway? It matters most when the meal contains carbohydrate, which almost every real meal does. It costs nothing, needs no new food, and it is the one change you can make at a restaurant table.