What foods improve your insulin sensitivity?

Whole grains, fibre-rich carbohydrates that raise blood sugar gradually, plenty of non-starchy vegetables and fresh fruit, and few simple sugars and sugary drinks: this dietary pattern improves insulin sensitivity. The Mediterranean diet and the DASH diet (rich in vegetables, fruit and whole grains, low in salt) fit this picture well and have been assessed favourably in multiple studies. You do not need to cut out dairy to achieve this.

Calorie restriction is the best-supported approach for reducing insulin resistance. The mechanisms are well described and the effect on blood sugar and cardiometabolic risk has been robustly demonstrated. One caveat: how safe prolonged, strict calorie restriction is over the long term has not yet been sufficiently studied.

When you eat matters at least as much as what you eat. An early eating window, in which you take your last meal before 15:00 and then fast for around 18 hours, improved insulin sensitivity in men with prediabetes in a randomised trial, even without weight loss or a reduction in food intake. Concentrating calories and carbohydrates in the morning and early afternoon, and avoiding a late evening meal, has a similarly beneficial effect. This aligns with the idea that your body processes carbohydrates more efficiently earlier in the day.

The order of foods on your plate also makes a difference. Eating vegetables or soup first, then protein, and only then starchy foods lowers the blood sugar peak and the insulin response after a meal compared with other sequences. This is an easy adjustment you can apply straight away.

For the ketogenic diet, the effect on insulin sensitivity stems almost entirely from the weight loss the diet brings about, not from the low carbohydrate intake itself. In a controlled study in which body weight was kept stable, insulin sensitivity did not change at all in people with overweight and type 2 diabetes. Foods such as vinegar, yoghurt, whey protein, peanuts and nuts are mentioned in the literature, but the evidence for them is limited and predominantly observational: treat these as a supplement to your approach, not as its foundation.