The insulin resistance diet I follow is boring on purpose. The same breakfast almost every morning, the same architecture at lunch, the same shake at four in the afternoon, the same strange cup of gelled seeds at six. But boring, for insulin, is the point, and the history of how we think about this hormone explains why.

In January 1922, Leonard Thompson, a fourteen-year-old boy dying of type 1 diabetes in a Toronto ward, received the first injection of insulin ever given to a human being. Within weeks he was gaining weight. Within months the drug was in mass production, and insulin went from miracle to commodity in a single year. It remains one of the most reliable life savers in medicine.

That triumph created a blind spot that has lasted a century. We treat insulin as a drug for when the body has none, and almost never as a hormone to keep quiet when there is chronically too much. For the roughly nine in ten metabolic disease cases that are not type 1, the problem runs in the opposite direction: years of insulin levels held higher than the body was designed to tolerate, while everything that appears on a standard lab report looks fine.

I think about this every morning, standing in the kitchen making the same breakfast I have made for the past year. This article describes the diet I built for one purpose: keeping my fasting insulin low and my insulin sensitivity high, using food I am willing to eat every day for the rest of my life. It is my protocol, sized for my body and my goals. It is not a prescription. The physiology behind it is covered in the metabolic health article and the type 2 diabetes article, and I will link to both rather than repeat them.

What insulin resistance actually is

The short version, for anyone who has not read those articles. Insulin is the hormone that tells muscle, liver, and fat cells to absorb glucose from the blood. Insulin resistance is the state in which those cells respond weakly to that signal. The pancreas compensates by producing more and more of it, and it can sustain that compensation for a decade before fasting glucose begins to move. Elevated insulin is the body’s cost of living on credit. The bill only becomes visible years later, when the beta cells tire and the diagnosis arrives.

Two details of the mechanism matter for the diet. The first is that ectopic fat, lipid stored inside muscle and liver cells, is the best-characterized driver of the stalled insulin signal. Anything that reduces that overflow helps. The second is that skeletal muscle is the body’s main glucose disposal reservoir, which means muscle mass and insulin sensitivity are intertwined, a connection covered in the resistance training article and the metabolic flexibility article.

The measurable targets are simple: fasting insulin, HOMA-IR, HbA1c, and triglycerides. If you only measure glucose, you are reading the last chapter of the book.

The insulin resistance diet I actually eat

Here is the disclaimer, up front. This is my protocol, an N of one. It is the product of reading the physiology, testing it against my own blood work, and keeping what worked. I am not a doctor, your metabolism is not my metabolism, and anyone on glucose-lowering medication should talk to their physician before restructuring their diet, because the diet changes how the medication behaves. Copy the principles, not the plate.

With that said, here is a normal day, end to end.

Morning. Greek yogurt, a handful of mixed nuts, blueberries, and a coffee. Greek yogurt because it is the cheapest high-quality protein I can get at breakfast without cooking. Nuts and blueberries because they slow the meal down chemically, which I will explain below. The coffee is not negotiable, and it happens to be evidence-aligned.

While the breakfast settles I make the day’s almond milk for the afternoon shake. It is a small ritual that takes five minutes and removes one more decision from the afternoon, which is the hour when discipline is weakest.

Lunch. Vegetables first, usually broccoli, cauliflower, or whatever green thing is in the fridge, and then a protein: tofu, seitan, or lentils. The order is deliberate. Vegetables enter the stomach before anything dense, and the protein follows them. I eat the carbohydrate portion of the meal last, which at this point is usually just whatever the lentils contribute.

Afternoon, around four. The protein shake: almond milk, two scoops of whey protein, some cinnamon, one spoon of pure cocoa powder. This is the meal that makes the whole structure work. It lands in the gap between lunch and dinner where a normal person eats biscuits, and it carries roughly half of the day’s protein target.

Evening, six o’clock. A cup of psyllium husk and chia seeds. This sounds like punishment and eats like one for the first two minutes, after which it quietly keeps me full for the entire evening and overnight. It is the mechanism that makes a fourteen-hour fast feel effortless instead of heroic. Fiber, hydrated, at the end of the eating day, is the cheapest appetite control I know.

If I need more. On heavy training days, or whenever the day’s protein lands short, I add one more meal, same rules: vegetables, then protein. No rules are broken by adding a meal. The structure bends, it does not break.

The one number that governs the whole day is protein. The target is 1.6 to 2 grams per kilogram of desired body weight, not current body weight. That range sits exactly where the best meta-analysis of protein supplementation and resistance training found benefits plateauing, around 1.6 g/kg per day.[^[1]](#ref-1) It is the range that protects muscle while losing fat, and muscle is the glucose sink that keeps insulin sensitivity high.

Here is roughly where the protein comes from on a standard day:

Meal	Protein
Greek yogurt, 200 g	~20 g
Mixed nuts, 30 g	~6 g
Lunch: tofu, seitan, or lentils	25 to 35 g
Whey shake, 2 scoops	~45 g
Psyllium and chia	~5 g
Standard day total	~100 g

Against a target of 1.6 g/kg for a desired body weight of 70 kilograms, that is 112 grams, so the standard day sits slightly short and the optional extra meal is not a fallback, it is part of the design. On rest days the standard structure is close enough. The point of running the number is that the diet is not a vibe. It has a target, and the day is built backward from it.

Why each food earns its place

Nothing on the list is decorative. Each item is there because of what it does to insulin, glucose, or the willingness to keep eating this way for decades.

Greek yogurt. The largest dose-response meta-analysis of dairy intake and diabetes incidence found that dairy is associated with a lower risk of developing type 2 diabetes, with fermented dairy showing the most consistent signal.[^[2]](#ref-2) Yogurt also concentrates protein in a form that requires no cooking and no willpower. Fermentation eats most of the lactose, which keeps the carbohydrate load of breakfast low without feeling like sacrifice.

Nuts. Fat, fiber, and protein in one package, with almost no digestible carbohydrate. They slow gastric emptying and blunt the glycemic impact of everything eaten with them. In the trial literature, nut-rich Mediterranean patterns produced some of the strongest dietary effects on metabolic outcomes we have.[^[3]](#ref-3)

Blueberries. A double-blind, placebo-controlled trial in obese, insulin-resistant adults found that a daily blueberry smoothie improved insulin sensitivity on the gold-standard measure, the hyperinsulinemic-euglycemic clamp, in six weeks.[^[4]](#ref-4) One food group where dessert and medicine overlap. They also make breakfast worth eating, which matters more to adherence than any micronutrient.

Coffee. The dose-response meta-analyses of habitual coffee intake keep finding the same thing: each daily cup is associated with a lower risk of type 2 diabetes, and the association holds for decaffeinated coffee too, which points at the polyphenols rather than the caffeine.[^[5]](#ref-5) I would drink it regardless. The evidence is a bonus.

Cruciferous vegetables at lunch. Broccoli and cauliflower are mostly water, fiber, and volume, which means they occupy stomach space before the dense part of the meal arrives. They carry almost no glucose of their own. They are the physical implementation of a low glycemic environment, the concept David Jenkins turned from observation into instrument with the glycemic index.[^[6]](#ref-6)

Tofu, seitan, lentils. The three protein anchors of a vegetarian diet that is also low carb, a Venn overlap smaller than people assume. Seitan is nearly pure wheat gluten, absurdly protein dense, and almost carbohydrate free. Tofu is complete, cheap, and versatile. Lentils carry carbohydrate, roughly twenty grams per cooked cup, but they earn it: they have one of the lowest glycemic indexes of any staple, and a randomized trial in type 2 diabetes found that a legume-focused low glycemic diet improved HbA1c more than a whole-wheat high fiber diet over three months.[^[7]](#ref-7) Rotation matters here. A diet you can vary within rules is a diet you keep.

Whey. The afternoon shake is whey protein in almond milk. Whey is the most studied protein for glucose control, and the timing is not incidental: a randomized trial in type 2 diabetes found that consuming whey as a pre-load before a meal increased insulin secretion and substantially lowered the post-meal glucose excursion compared with the same meal without it.[^[8]](#ref-8) A shake four hours after lunch and two hours before the last fiber meal is the structural midpoint of the day, and it doubles as insurance for the protein target.

Cinnamon. The honest version: cinnamon lowers fasting glucose modestly in pooled trials but shows no significant effect on HbA1c, and the trial literature is inconsistent about dose and species.[^[9]](#ref-9) I include it because it is harmless at culinary doses, useful for making plain things taste like dessert, and its evidence is exactly as strong as a garnish should be. It is also why cassia cinnamon should stay at spice doses, since its coumarin content is not meant for spoonfuls.

Cocoa. Pure cocoa powder, unsweetened, carries flavanols, and meta-analyses of randomized trials find small improvements in insulin sensitivity and blood pressure at flavanol doses comparable to a spoonful or two of good powder.[^[10]](#ref-10) Mostly, though, cocoa is there to make the shake taste like a decision rather than a duty.

Psyllium and chia. The evening cup is the most important meal of the day nobody talks about. Psyllium is a gel-forming soluble fiber, and the meta-analytic evidence shows it lowers fasting glucose in type 2 diabetes in proportion to how impaired glycemic control was to begin with, with the effect strongest when it is taken before or with meals.[^[11]](#ref-11) Chia adds more fiber and plant omega-3s. Together, hydrated in a cup at six, they produce satiety that lasts until morning. An overnight fast you do not have to fight is an overnight fast you keep.

Time-restricted eating

The day above runs from roughly eight in the morning to six in the evening. A ten-hour eating window, a fourteen-hour fast, every day, without exception. The diet does not state this as a rule because it does not need to: the structure produces it.

The evidence for time-restricted eating is genuinely interesting and honestly mixed, and both halves matter. The interesting half: in a controlled feeding trial in men with prediabetes, confining meals to a six-hour window before mid-afternoon improved insulin sensitivity, lowered blood pressure, and reduced oxidative stress even though participants ate enough calories to maintain their weight. The timing itself, aligned with circadian rhythm and early in the day, appeared to do work that weight loss was not doing.[^[12]](#ref-12) The mixed half: the best-known randomized trial of time-restricted eating found no significant advantage over simple calorie restriction for weight or metabolic outcomes, suggesting much of what the clock does is help people eat less.[^[13]](#ref-13)

My reading of the two together: the window is not magic, but a consistent window with an early start and a long overnight fast is the easiest scaffolding I have found for keeping total intake right without counting anything. The overnight component also lines up with the fasting physiology covered in the autophagy article and the mTOR and AMPK article: the low-insulin hours are when cellular cleanup and fat oxidation actually run. Insulin suppresses both. A day with fewer and lower insulin excursions is a day with more hours in the other state.

Food ordering

The most underrated free intervention in the entire field is the sequence in which you eat a mixed meal.

In a randomized crossover trial in type 2 diabetes, eating vegetables and protein before the carbohydrate portion of an identical meal lowered the post-meal glucose area under the curve by roughly half and lowered the post-meal insulin area by about a quarter, compared with eating the same carbohydrate first.[^[14]](#ref-14) Same food, same calories, same people. Only the order changed, and the glucose response halved.

The mechanism is intuitive once stated: fiber and protein slow gastric emptying and stimulate the gut hormones that blunt the arrival of glucose, so the carbohydrate lands on a buffered system instead of an empty one. Whey works as a pre-load through the same physiology, which is why the shake sits where it sits in my day.

On a diet where carbohydrate is already low, ordering matters less at most meals, but it matters enormously at the meals where carbohydrate sneaks back in: restaurant dinners, holidays, the bread basket before lunch arrives. Vegetables and protein first is the rule that makes imperfect meals survivable without compensation arithmetic. It is the cheapest piece of the protocol and the one I would keep if I could only keep one.

Supplements

Supplements are the third lever, after what you eat and when you eat it, and the honest ranking is that they move the needle least. A few have real evidence.

Magnesium is the one I consider closest to foundational. Prospective cohort meta-analyses consistently find that higher magnesium intake is associated with lower risk of developing type 2 diabetes,[^[15]](#ref-15) magnesium is the cofactor for the enzymes that make insulin signaling run, and most modern diets underdeliver it. It is in the nuts and seeds above, and I supplement on top of food.

Vitamin D, if you are deficient, mostly corrects a deficit rather than adding a benefit. Enough people running low insulin protocols train indoors at northern latitudes that it is worth checking rather than assuming.

Omega-3 from fish or algae is for cardiovascular risk more than insulin sensitivity, and I take it for the former without pretending it does much for the latter.

Berberine is the interesting frontier: small trials in type 2 diabetes found glucose-lowering effects comparable in magnitude to metformin,[^[16]](#ref-16) but the trials are small, quality varies, and long-term safety data are thin. It is on the watch list, not in the daily stack.

The pattern across all of them: food first, then timing, then molecules. Anyone selling a stack that replaces the first two is selling the third lever as a forklift.

The honest opposing case

The fair objections to this protocol are real, so here they are.

Vegetarian and low carb is a narrow Venn overlap, and it requires the industrial food system to work: whey protein, seitan, packaged almond milk. A diet that needs processed inputs to hit its protein target is not what our ancestors ate, although it is also not trying to be.

The monotony is real. A protocol this repetitive trades variety for reliability, and not everyone’s psychology can pay that price. The literature on adherence is blunt about this: the best diet for insulin resistance is the one the person can hold for years, and some people will hold a varied Mediterranean pattern longer than a repetitive low-carb one. The Da Qing study, cited in the type 2 diabetes article, succeeded with more vegetables and more activity, not with macronutrient extremism.

Time-restricted eating partly works by making people eat less, and its trials show no advantage over plain calorie restriction where intake is actually controlled. If a rigid window makes you miserable, the window is the wrong tool for you.

And protein, the backbone of my target, is calibrated against the resistance training literature. If you do not lift, your needs sit lower and 1.6 g/kg of desired body weight is generous.

What would make me change this protocol: an HbA1c that creeps despite the numbers, a fasting insulin that stops responding, or evidence that moves. The protocol serves the biomarkers, not the other way around.

If you build your own version

Do not copy my plate. Copy the algorithm, which fits in five lines.

Pick a protein target at 1.6 to 2 g/kg of the body weight you want, and check the number on the scale, not the assumption. Anchor every meal with protein and vegetables before anything else. Set an eating window you can hold daily, early rather than late. Put a fiber obstacle between your afternoon self and your dinner, and let the overnight fast run itself. Then measure: fasting insulin and HOMA-IR every few months, the full case for which is in the blood tests article, and a continuous glucose monitor or a wearable for a month if you want feedback loops tighter than quarterly labs, the tradeoffs of which are in the wearables article. Adjust what the data tells you to adjust, and keep the rest boring.

The whole scheme compresses into a sentence: make insulin’s job easy for most hours of most days, feed the muscle you keep, and let the boring consistency compound.

---

References

1. Morton RW, Murphy KT, McKellar SR, et al. (2018). A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength in healthy adults. British Journal of Sports Medicine, 52(6), 376-384. https://doi.org/10.1136/bjsports-2017-097608 2. Gijsbers L, Ding EL, Malik VS, de Goede J, Geleijnse JM, Soedamah-Muthu SS. (2016). Consumption of dairy foods and diabetes incidence: a dose-response meta-analysis of observational studies. American Journal of Clinical Nutrition, 104(5), 1233-1241. https://ajcn.nutrition.org/article/S0002-9165(23)11936-8/fulltext 3. Estruch R, Ros E, Salas-Salvadó J, et al. (2018). Primary prevention of cardiovascular disease with a Mediterranean diet supplemented with extra-virgin olive oil or nuts. New England Journal of Medicine, 378(25), e34. https://pubmed.ncbi.nlm.nih.gov/29897866/ 4. Stull AJ, Cash KC, Johnson WD, Champagne CM, Cefalu WT. (2010). Bioactives in blueberries improve insulin sensitivity in obese, insulin-resistant men and women. Journal of Nutrition, 140(10), 1764-1768. https://pubmed.ncbi.nlm.nih.gov/20724487/ 5. Ding M, Bhupathiraju SN, Chen M, van Dam RM, Hu FB. (2014). Caffeinated and decaffeinated coffee consumption and risk of type 2 diabetes: a systematic review and a dose-response meta-analysis. Diabetes Care, 37(2), 569-586. https://pubmed.ncbi.nlm.nih.gov/24201328/ 6. Jenkins DJA, Wolever TM, Taylor RH, et al. (1981). Glycemic index of foods: a physiological basis for carbohydrate exchange. American Journal of Clinical Nutrition, 34(3), 362-366. https://pubmed.ncbi.nlm.nih.gov/6269919/ 7. Jenkins DJA, Kendall CWC, Augustin LSA, et al. (2012). Effect of legumes as part of a low glycemic index diet on glycemic control and cardiovascular risk factors in type 2 diabetes mellitus: a randomized controlled trial. Archives of Internal Medicine, 172(21), 1653-1660. https://pubmed.ncbi.nlm.nih.gov/23089999/ 8. Jakubowicz D, Froy O, Wainstein J, Boaz M. (2014). Incretin, insulinotropic and glucose-lowering effects of whey protein pre-load in type 2 diabetes: a randomised clinical trial. Diabetologia, 57(7), 1314-1319. https://doi.org/10.1007/s00125-014-3305-x 9. Allen RW, Schwartzman E, Baker WL, Coleman CI, Phung OJ. (2013). Cinnamon use in type 2 diabetes: an updated systematic review and meta-analysis. Annals of Family Medicine, 11(5), 452-459. https://www.annfammed.org/content/11/5/452 10. Hooper L, Kay C, Abdelhamid A, et al. (2012). Effects of chocolate, cocoa, and flavan-3-ols on cardiovascular health: a systematic review and meta-analysis of randomized trials. American Journal of Clinical Nutrition, 95(3), 740-751. https://pubmed.ncbi.nlm.nih.gov/22751495/ 11. Gibb RD, McRorie JW, Russell DA, Hasselblad V, Buckley DI. (2015). Psyllium fiber improves glycemic control proportional to loss of glycemic control: a meta-analysis of data in euglycemic subjects, patients at risk of type 2 diabetes mellitus, and patients being treated for type 2 diabetes mellitus. American Journal of Clinical Nutrition, 102(4), 744-748. https://ajcn.nutrition.org/article/S0002-9165(23)27235-4/fulltext 12. Sutton EF, Beyl R, Early KS, Cefalu WT, Ravussin E, Peterson CM. (2018). Early time-restricted feeding improves insulin sensitivity, blood pressure, and oxidative stress even without weight loss in men with prediabetes. Cell Metabolism, 27(6), 1212-1221. https://www.cell.com/cell-metabolism/fulltext/S1550-4131(18)30253-5 13. Lowe DA, Wu N, Rohdin-Bibby L, et al. (2020). Effects of time-restricted eating on weight loss and other metabolic parameters in women and men with overweight and obesity: the TREAT randomized clinical trial. JAMA Internal Medicine, 180(9), 1231-1239. https://pubmed.ncbi.nlm.nih.gov/32539664/ 14. Shukla AP, Andono J, Touhamy SH, et al. (2017). Carbohydrate-last meal pattern lowers postprandial glucose and insulin excursions in type 2 diabetes. BMJ Open Diabetes Research and Care, 5(1), e000440. https://pmc.ncbi.nlm.nih.gov/articles/PMC5604719/ 15. Dong JY, Xun P, He K, Qin L. (2011). Magnesium intake and risk of type 2 diabetes: meta-analysis of prospective cohort studies. Diabetes Care, 34(9), 2116-2122. https://pubmed.ncbi.nlm.nih.gov/21775495/ 16. Yin J, Xing H, Ye J. (2008). Efficacy of berberine in patients with type 2 diabetes mellitus. Metabolism, 57(5), 712-717. https://pubmed.ncbi.nlm.nih.gov/18443365/