Insulin resistance happens when the body doesn’t respond well to insulin, a hormone that helps regulate blood sugar. Because of that, the pancreas has to make more insulin to keep blood sugar in a normal range.
No dietary pattern has been established as the single best diet for insulin resistance. Several patterns are supported by major guidelines, and they overlap far more than their labels suggest.
One reason it’s hard to name a best diet is that most diet trials measure body weight, fasting glucose, A1C, blood lipids, or an estimate calculated from fasting blood tests, rather than insulin sensitivity itself. Those are not the same, and a diet can look better on one and no different on another.
The 2026 American Diabetes Association guidance on preventing or delaying diabetes states that no ideal percentage of calories from carbohydrate, protein, and fat applies to everyone, and it recognizes several appropriate eating patterns rather than a single one (ADA 2026).
This guide explains what the major comparison trials actually measured, in whom, and for how long.
Is there a best diet for insulin resistance?
No single diet has been proven to be best for improving insulin resistance. The 2026 ADA Standards of Care does not recommend one ideal ratio of carbohydrates, protein, and fat for preventing diabetes. Instead, it recommends individualizing the eating pattern based on a person’s preferences, health needs, and ability to maintain it over time (ADA 2026).
Several eating patterns can fit within that approach, including (ADA 2026):
- Mediterranean-style
- Plant-based
- DASH (Dietary Approaches to Stop Hypertension)
- Low-carbohydrate
In its formal recommendation for people with prediabetes, the ADA advises clinicians to prescribe an evidence-based eating pattern and names Mediterranean and low-carbohydrate as examples (ADA 2026). That recommendation is about preventing type 2 diabetes, not evidence that these two patterns improve insulin sensitivity more than the others.
What the evidence-supported patterns have in common
Several eating patterns are recommended options for diabetes prevention, including Mediterranean-style, plant-based, DASH, and low-carbohydrate approaches. Although the names are different, they often include many of the same foods, such as vegetables, fruits, whole grains, beans, and nuts, while limiting highly refined and processed foods (ADA 2026, Shai et al. 2008).
That overlap is important. In practice, the quality of the foods you eat may matter as much as the name of the diet you follow.
Even lower-carbohydrate eating patterns can emphasize high-fiber, minimally processed carbohydrate sources. This is why diabetes prevention is not simply a question of whether carbohydrates are “good” or “bad.”
Blood sugar response is also influenced by the rest of the meal. For example, research shows that adding protein to a carbohydrate-containing meal can change both the rise in blood sugar and the amount of insulin released, although the effects vary among people and by protein source (Wolever et al. 2024).
This is one reason a single measure, such as the glycemic index of an individual food, cannot tell you everything about how a complete meal will affect blood sugar.
The main takeaway is that a healthy eating pattern is defined less by its label and more by the quality and balance of the foods it contains.
The main dietary patterns, compared
What this looks like in practice
Most supported patterns point in the same direction (ADA 2026, Shai et al. 2008):
- Meals built mostly around vegetables, beans, whole grains, fruit, and nuts
- Fiber-rich, minimally processed carbohydrates in place of refined ones
- A pattern that fits your preferences, health needs, and routine, so you can keep it up over time
Head-to-head trials exist for some of these patterns, and even those measured different outcomes in different populations over different durations. The table summarizes what each pattern is and, where a comparison was made, what it actually examined.
| Pattern | Short description | Comparison covered here | What was measured | Population | Duration |
|---|---|---|---|---|---|
| Low-fat | Reduced total fat as the main restriction | DIRECT trial (Shai 2008, Schwarzfuchs 2012) | Body weight (full cohort); HOMA-IR, fasting glucose, fasting insulin (T2D subgroup) | 322 moderately obese adults; 36-person T2D subgroup | 2 years (+ 4-year follow-up) |
| Mediterranean | Vegetables, legumes, fish, unsaturated fats | DIRECT; Keto-Med crossover (Gardner 2022) | Body weight; HOMA-IR, FPG, insulin (subgroup); A1C and lipids in Keto-Med | As above; 40 adults with prediabetes or T2D in Keto-Med | 2 years; 12 weeks per phase |
| Low-carbohydrate | Reduced carbohydrate as the main restriction | DIRECT trial | Body weight; HOMA-IR, FPG, insulin (subgroup) | As above | 2 years |
| Ketogenic | Substantially greater carbohydrate restriction | Keto-Med crossover | A1C (primary), LDL, triglycerides | 40 adults with prediabetes or T2D; 33 completed | 12 weeks per phase |
| DASH | Pattern originally developed for blood pressure | Recognized in 2026 ADA prevention guidance | No head-to-head comparison covered here | N/A | N/A |
| Plant-based | Predominantly plant foods, with variable animal foods | Recognized in 2026 ADA prevention guidance | No head-to-head comparison covered here | N/A | N/A |
The DIRECT trial, one of the largest direct comparisons, was a two-year randomized study in Israel that assigned 322 adults to low-fat, Mediterranean, or low-carbohydrate diets (Shai et al. 2008). Average weight loss after two years was 2.9 kg with the low-fat diet, 4.4 kg with the Mediterranean diet, and 4.7 kg with the low-carbohydrate diet.
The findings related to insulin resistance came from a much smaller subgroup of just 36 participants who had type 2 diabetes. In that group, the Mediterranean diet improved HOMA-IR more than the low-fat diet, while the low-carbohydrate diet did not differ significantly from low-fat (Shai et al. 2008).
HOMA-IR is an estimate based on fasting glucose and insulin rather than a direct measure of insulin sensitivity. That matters because the different outcomes in the study did not all point in the same direction. The diet associated with the greatest average weight loss was not the same one with the strongest HOMA-IR result.
After the 2-year intervention ended, DIRECT trial participants were followed for 4 more years (6 years total) (Schwarzfuchs et al. 2012). By then, adherence was no longer controlled, and only about two-thirds remained on their assigned diet. Even so, net weight loss from baseline remained significant for the Mediterranean (3.1 kg) and low-carbohydrate (1.7 kg) groups and not the low-fat group (0.6 kg), and several lipid improvements persisted. Notably, this follow-up measured only weight and blood lipids; it did not reassess insulin resistance, so it tells us nothing about whether the earlier HOMA-IR improvement lasted.
Another trial, Keto-Med, compared a ketogenic diet with a Mediterranean-style diet in adults with prediabetes or type 2 diabetes (Gardner et al. 2022). A1C did not differ significantly between the two diets. LDL cholesterol was higher during the ketogenic phase, while triglycerides decreased more. Importantly, Keto-Med measured blood sugar control and cholesterol, not insulin sensitivity itself.
Taken together, these trials show that different eating patterns may perform differently depending on which outcome is measured. That is not the same as showing that one diet is best overall.
Meal composition and order
Blood sugar after a meal depends on more than just how much carbohydrate you eat. Protein, fat, fiber, and even the order in which foods are eaten can affect the rise in blood sugar and insulin after a meal.
For example, a systematic review found that adding protein to a carbohydrate-containing meal could reduce the rise in blood sugar, although it often increased the amount of insulin released (Wolever et al. 2024). That is an important distinction: a smaller blood sugar spike after one meal does not necessarily mean the body has become more sensitive to insulin.
The order in which foods are eaten may also affect blood sugar. A 2026 systematic review of six studies involving 144 people with diabetes found that eating carbohydrate-containing foods later in the meal generally led to smaller blood sugar rises (McKenzie et al. 2026). However, the studies were small and mostly measured post-meal glucose responses. They do not show that changing food order improves insulin resistance over the long term.
Macronutrient order is also different from meal timing. Eating carbohydrates later within a meal is not the same as changing when you eat during the day or following intermittent fasting or time-restricted eating.
The main takeaway is that a smaller blood sugar spike after one meal is a short-term effect. It should not automatically be interpreted as an improvement in insulin resistance.
What has held up, what hasn’t, and what remains disputed
What has held up. Diet can affect several health measures related to insulin resistance, including body weight, blood sugar, A1C, and cholesterol (Shai et al. 2008, Gardner et al. 2022). The 2026 ADA recognizes several eating patterns for diabetes prevention, including Mediterranean-style, plant-based, DASH, and low-carbohydrate approaches (ADA 2026). These patterns differ, but they also share many features, especially an emphasis on overall food quality rather than one specific ratio of carbohydrates, protein, and fat.
What has not been established. No single eating pattern has been proven to be best for insulin resistance. Many diet studies measure outcomes such as weight, fasting glucose, or A1C rather than insulin sensitivity itself. Even when insulin resistance is estimated using HOMA-IR, that is still an indirect measure. In one major trial, the HOMA-IR finding came from a subgroup of only 36 participants with type 2 diabetes (Shai et al. 2008).
It is also not clear how well findings from people with prediabetes or type 2 diabetes apply to people who have insulin resistance but have not developed diabetes.
What is still being studied. People can have very different blood sugar responses to the same foods. One large study using continuous glucose monitoring showed substantial variation in how individuals responded to identical meals (Zeevi et al. 2015).
That finding has helped drive interest in personalized nutrition. However, it has not yet been shown that choosing foods with a prediction algorithm improves insulin resistance or long-term health more than following broadly supported healthy eating principles.
Conclusion
There is no single best diet for insulin resistance. Several eating patterns can be appropriate, and they tend to share important features, including an emphasis on overall food quality, fiber-rich foods, and fewer highly refined foods.
Studies comparing different diets have also measured different things, such as weight, blood sugar, A1C, and estimates of insulin resistance. A diet that performs better on one of those measures will not necessarily perform better on all of them.
People also vary in how their blood sugar responds to the same foods. However, that has not yet led to a proven method for choosing one ideal diet for each person.
The best-supported approach is to focus less on finding the “perfect” diet and more on following a balanced eating pattern that is nutritious, practical, and sustainable.
Frequently asked questions
Do you have to cut carbohydrates to improve insulin resistance?
No. Low-carbohydrate eating is one supported option, but not the only one. Across the supported patterns, the common thread is carbohydrate quality: more fiber-rich, minimally processed sources and fewer refined ones.
Is a ketogenic diet different from a low-carbohydrate diet for insulin resistance?
Yes. A ketogenic diet usually restricts carbohydrates much more than a general low-carbohydrate diet.
In the Keto-Med trial, adults with prediabetes or type 2 diabetes followed both a ketogenic diet and a Mediterranean-style diet. A1C did not differ significantly between the two approaches. LDL cholesterol was higher during the ketogenic phase, while triglycerides decreased more (Gardner et al. 2022).
The study did not directly measure whether one diet improved insulin sensitivity more than the other.
Can diet alone improve insulin resistance, or is exercise needed too?
Diet can improve blood sugar, weight, and other measures related to insulin resistance. Exercise can help as well.
However, many lifestyle studies combine diet, physical activity, and weight loss, which makes it difficult to separate the effect of each one. The 2026 ADA guidance recommends a broader lifestyle approach rather than focusing on diet alone (ADA 2026).
Is an insulin resistance diet different from a diabetes diet?
There is no formally defined “insulin resistance diet.” Most of the research used to guide dietary recommendations comes from studies in people with prediabetes or type 2 diabetes (ADA 2026, Gardner et al. 2022).
In practice, the same broadly supported eating patterns are often used to improve blood sugar control, support a healthy weight, and reduce diabetes risk.
Do you need to count carbohydrates or track what you eat?
It’s not necessary for everyone. Tracking food intake can help many people understand what they are eating and stick with a plan, but it’s not necessary for everyone. If you are able to follow a healthy diet without tracking what you eat, then you don’t need to. One exception is that those living with type 1 diabetes often need to count carbohydrates in order to know how much insulin they should use for that meal.
References
- American Diabetes Association Professional Practice Committee. 3. Prevention or Delay of Diabetes and Associated Comorbidities: Standards of Care in Diabetes—2026. Diabetes Care. 2026;49(Supplement 1). doi:10.2337/dc26-S003
- Shai I, Schwarzfuchs D, Henkin Y, et al. Weight Loss with a Low-Carbohydrate, Mediterranean, or Low-Fat Diet. N Engl J Med. 2008;359(3):229-241. doi:10.1056/NEJMoa0708681
- Wolever TMS, Zurbau A, Koecher K, Au-Yeung F. The Effect of Adding Protein to a Carbohydrate Meal on Postprandial Glucose and Insulin Responses. J Nutr. 2024;154(9):2640-2654. doi:10.1016/j.tjnut.2024.07.011
- Schwarzfuchs D, Golan R, Shai I. Four-Year Follow-up after Two-Year Dietary Interventions. N Engl J Med. 2012;367(14):1373-1374. doi:10.1056/NEJMc1204792
- Gardner CD, Landry MJ, Perelman D, et al. Effect of a ketogenic diet versus Mediterranean diet on glycated hemoglobin in individuals with prediabetes and type 2 diabetes mellitus: Keto-Med. Am J Clin Nutr. 2022;116(3):640-652. doi:10.1093/ajcn/nqac154
- McKenzie R, Sterling C, O’Hara H, Woodside J. The Impact of Macronutrient Ordering on Postprandial Glycaemic Control in Diabetes: A Systematic Review. Endocrinol Diabetes Metab. 2026;9(3). doi:10.1002/edm2.70228
- Zeevi D, Korem T, Zmora N, et al. Personalized Nutrition by Prediction of Glycemic Responses. Cell. 2015;163(5):1079-1094. doi:10.1016/j.cell.2015.11.001