Methodology

Every calorie tool on the site walks the first four steps below, in this order. The keto macro tools then add the last three:

1. 1.Normalize your inputs to metric. Pounds become kilograms at 0.453592 per pound; feet and inches become centimeters at 30.48 and 2.54.
2. 2.Estimate basal metabolic rate (BMR), the energy your body uses at complete rest.
3. 3.Multiply BMR by an activity multiplier to get total daily energy expenditure (TDEE).
4. 4.Add the adjustment for your goal, then apply a floor, to get a daily calorie target.
5. 5.Set protein in grams from your lean body mass. It does not depend on the calorie target, so it does not shrink when you cut.
6. 6.Set net carbs in grams from the keto variant you chose.
7. 7.Give every remaining calorie to fat, at 9 calories per gram. Percentages are worked out from the grams afterwards, for display only.

BMR is carried unrounded through step 3 and rounded only at the end. It is a small detail, but rounding early shifts a daily target by tens of calories, and the same inputs should always give the same answer on every tool here.

Mifflin-St Jeor is the default. It needs only weight, height, age, and sex:

It is the default because it was derived on a general adult population and it asks for nothing you need a lab to measure. Mifflin and colleagues published it in 1990 as a replacement for the older Harris-Benedict equations, which tend to overestimate on modern body compositions.

Katch-McArdle takes over the moment you enter a body fat percentage above zero. Instead of guessing at lean mass from height and weight, it works from lean mass directly:

Muscle and fat burn energy at very different rates, so once lean mass is known the estimate improves, and it improves most for lean and muscular people, whom Mifflin-St Jeor tends to underrate. There is a catch worth stating plainly: Katch-McArdle is only better than the default if your body fat number is measured rather than guessed. A DEXA scan or a careful caliper reading is worth entering. An eyeballed figure from a photo chart is not, and leaving the field blank will serve you better.

Body fat does a second job on the macro calculators: the same lean-mass figure sets your protein target (see below). Leave it blank and BMR stays on Mifflin-St Jeor while lean mass for protein is estimated from height and weight instead.

TDEE is BMR multiplied by one of five activity factors. These are the long-standing values used across exercise physiology references, and the site uses them unchanged:

Pick the one that matches a normal week, not your best week. Overstating activity is the single most common reason a calorie target comes out too high, and the gap between 1.375 and 1.725 on a 1,600 kcal BMR is 560 calories a day.

Your goal then shifts that number:

One safeguard sits at the end. The result is floored at 1,500 calories for men and 1,200 calories for women. A 1,000-calorie deficit applied to a small person produces a number no calculator should hand out, so the floor catches it. If your target is sitting on the floor, that is the tool telling you the deficit you asked for is too large for your size, and the keto macro calculators now say so in a note under your result. The floor is the last thing applied to calories, so no later adjustment can take a target below it.

Protein is anchored to lean body mass, not to a percentage of calories. A percentage falls every time calories are cut, which is exactly when holding on to muscle matters most. A lean-mass target stays put whatever goal you pick.

1.76 g per kg of lean mass is about 0.8 g per pound, inside the 0.7 to 1 g per pound of lean mass range the site quotes. Phinney and Volek describe 1.5 to 1.75 g per kg of reference body weight for a well-formulated ketogenic diet, and a 2018 meta-analysis by Morton and colleagues found the benefit of extra protein for muscle tapers off around 1.6 g per kg of body weight. Because lean mass is always lower than body weight, the site’s figure lands at or below those levels for most people rather than above them.

The Boer equations were developed to estimate lean mass in clinical settings. They are an estimate, and like every formula here they will be off for people far from average build. A measured body fat percentage replaces them as soon as you enter one.

Carnivore is the one exception: it uses 2.2 g per kg of lean mass (1 g per pound), the top of the site’s range, because an all-animal diet is protein-heavy in practice.

Net carbs are a fixed number of grams for each variant, not a percentage. Fat then takes every calorie left after protein and carbs. Calories become grams at 9 per gram of fat and 4 per gram of protein and carbohydrate.

The clinical ketogenic literature commonly describes 20 to 50g of net carbs a day as the range that keeps most adults in nutritional ketosis, with 20g as the usual strict starting point. Standard keto here sits at 25g, near the bottom of that range. Lazy keto uses the strict 20g as its single flat cap, because one number is the whole point of the approach. Dirty keto shows the same numbers as standard keto on purpose: the difference is where the food comes from, not the macros.

The fat percentage you see is an output, not an input. For a typical adult at maintenance it lands somewhere around 70 to 80% of calories, which is the usual description of a ketogenic diet. It drops on an aggressive cut, because protein stays the same while calories fall.

One safeguard: fat never goes below 20% of the day’s calories. If a low calorie target cannot fit your protein, your carbs, and that much fat, protein is kept, fat is held at 20%, and the day’s calories rise just enough to fit. The result says so in a note rather than quietly showing a negative or near-zero fat figure.

Cyclical keto alternates keto days with 1 or 2 carb-up days a week, so 5 to 6 days are keto days. The calculator defaults to 2 carb-up days. Carb-up days are a choice of their own, not one per training day.

• ▸Keto days are standard keto: lean-mass protein, 25g net carbs, fat fills the rest.
• ▸Carb-up days keep the same calories and the same protein grams.
• ▸Fat drops to 20% of calories on a carb-up day, and carbohydrate fills everything that is left.
• ▸Weekly totals and averages count both day types.

The 20% fat level on carb-up days is a site convention that keeps the day genuinely low in fat; it is not a published constant. Expect the scale to rise for a day or two after a carb-up as glycogen is stored with water.

Targeted keto keeps the standard keto macros every day, training or not. On training days you add one separate dose of 15 to 30g of fast-digesting carbohydrate about 30 minutes before the session. The dose is on top of the daily 25g, never folded into it, so the everyday numbers are identical to standard keto.

Within that range the calculator suggests a starting point of 0.3 g per kg of body weight, held between 15 and 30g. That scaling is a site convention for picking a starting dose, not a published formula; shorter or easier sessions can use the low end.

The IF calculator uses the same calorie target and the same standard keto macros as everything above, then splits them across your eating window: three meals for 14:10, 16:8 and 18:6, two for 20:4, and one for OMAD. The window only changes when you eat, not how much.

5:2 is different because two days a week are fast days, capped at about 500 calories for women and 600 for men. Those days already remove a large amount of energy, so the goal deficit is applied once, across the week, rather than again on the normal days:

That makes the weekly average equal your goal target. Protein and carbs on normal days are the same grams as any other day; the extra calories go to fat.

These three pages run the same engine as every other macro tool, with sex set to female: the female Mifflin-St Jeor equation (or Katch-McArdle with body fat), protein from lean mass, standard keto carbs, and fat filling the rest. There is no separate female macro split and no menopause calorie cut. Mifflin-St Jeor already lowers BMR by 5 calories a day for each year of age, and Katch-McArdle works from measured lean mass, which is where age-related change shows up.

The PCOS calculator changes one number. If you answer yes to diagnosed insulin resistance, net carbs are set to 20g, the strict starting target and the level used in a small pilot study of a ketogenic diet in women with PCOS. If you answer no or unsure, net carbs stay at the standard 25g.

The 1,200 calorie floor for women is applied last, after every other step, so no setting on these pages can produce a lower target. Cycle phase and menopausal status add notes to the result; they do not change the numbers.

The net-carb math is one line, applied per serving:

Fiber comes out in full. On a US Nutrition Facts label fiber is counted inside total carbohydrate, and it is not absorbed as glucose, so subtracting it is what makes the remaining figure useful.

Sugar alcohols are not treated as one substance, because they do not behave like one. The coefficient depends on which one you are eating:

Erythritol is subtracted in full because it is largely absorbed and excreted rather than metabolized for energy. Maltitol sits at the other end and is the one that catches people out, which is why a half deduction, not a full one, is the honest default. When a label does not say which sugar alcohol it used, the calculator assumes the cautious 50%. The result is clamped at zero, so a badly transcribed label cannot produce a negative carb count, and rounded to one decimal place.

The verdict badge on a food uses the per-serving figure: 5g net carbs or under reads as excellent, up to 10g as good, up to 20g as moderate, and anything above that as high. Those cut-offs are the site’s own convention for sorting foods at a glance, not a clinical threshold, and they say nothing about how a food fits the rest of your day.

The Ketosis Estimator first checks whether ketosis is a realistic outcome at all. Above 50g of net carbs a day, the top of the 20-50g range commonly cited for nutritional ketosis, it does not give a timeline. It says ketosis is unlikely at that intake and suggests coming down toward 20g.

At 50g or less it runs a simple glycogen-depletion model:

The 130g figure is roughly the glucose an adult brain and body use in a day, the number the carbohydrate RDA is built on, so every gram you eat below it is treated as glycogen drawn down. The result is shown as a window from 70% to 130% of that central estimate, never shorter than one day. For a 70-100 kg adult the formula gives roughly 350-450g of stored glycogen, close to the 400-500g usually quoted for a well-fed adult. Glycogen is stored with water, which is why the first days of keto move the scale more than fat loss alone would.

This is a teaching model, not physiology to the gram. Protein intake, insulin sensitivity and how hard your liver makes new glucose all move the real answer, and none of them is an input. That is why every result says it is an estimate: the only confirmation is a blood ketone (beta-hydroxybutyrate) reading of 0.5 mmol/L or more, with 0.5-3 mmol/L the usual nutritional-ketosis range.

The Electrolyte Calculator gives total daily intake targets: food, broth, salt and any supplement added together. None of the numbers is a supplement dose. Each mineral starts at a baseline for your keto phase, gains fixed steps for activity, sweat and a few symptoms, and is then held inside a fixed range:

Sodium sits above general-population advice because insulin falls on a ketogenic diet and the kidneys excrete more sodium; the 3,000-5,000 mg range follows Phinney and Volek and Virta Health’s guidance for a well-formulated ketogenic diet. Potassium never goes below the NIH Adequate Intake for your sex (2,600 mg for women, 3,400 mg for men, from the 2019 National Academies review), and the result says to get it mainly from food, since US over-the-counter potassium supplements carry at most 99 mg per dose. Magnesium is a total of 300-500 mg, around the NIH RDA of 310-420 mg; separately, the result tells you to keep supplemental magnesium at or below 350 mg a day, the NIH upper limit for supplements, unless a clinician advises otherwise.

Two things the calculator deliberately does not do. It does not raise potassium for heart palpitations, because palpitations can come from too much potassium as well as too little, or from causes unrelated to diet; it tells you to get them checked instead. And it shows every result with a caution: people with high blood pressure, heart failure or kidney disease, or who take diuretics, ACE inhibitors, ARBs or potassium-sparing drugs, should set their sodium and potassium intake with their clinician rather than use these targets.

The common shortcut divides the weight you want to lose by a fixed weekly loss: a 500 kcal daily deficit, 3,500 kcal per pound, so one pound a week forever. It ignores that a smaller body burns less energy, so it over-promises, and the longer the timeline the bigger the miss (Hall and colleagues, The Lancet, 2011). The Keto Weight Loss Calculator projects week by week instead:

Intake stays fixed at the starting figure, never below the 1,200 kcal (women) or 1,500 kcal (men) floor, while TDEE is recalculated from the projected weight every week with the same BMR formula and activity multiplier described above. As weight falls the deficit shrinks, so the weekly loss shrinks with it and the curve flattens. When you give a body fat percentage, Katch-McArdle runs and lean mass is assumed to make up a quarter of each pound lost. The projection stops at your goal, when the weekly loss drops below 0.05 lb (weight has levelled off at that intake), or at three years, and the result shows what the static rule would have said for comparison.

Two things are left out on purpose. Early water loss, the 3-7 lb many people see in the first week or two as glycogen and its stored water go, is shown as a range beside the curve, not built into it, because it varies widely and returns if carbs do. And there is no predicted plateau week: stalls are common, but when they happen is individual, so the calculator says so instead of naming a week.

A fat fast is a short, very-low-calorie, very-high-fat protocol, and the Fat Fast Calculator keeps it inside fixed bounds: 2 to 4 days, 1,000-1,200 kcal a day, and 80-90% of calories from fat. The daily target is half your estimated maintenance, rounded to the nearest 50 kcal and then clamped to that calorie band, so body size nudges it but can never push it outside the protocol. Calories split 90% fat, 8% protein and 2% carbs, divided evenly over four or five small meals.

There is no clinical trial that defines a “fat fast”; it is a convention from the low-carb community, and the calculator treats it as one. Much of the quick drop is water. The result carries a caution that it is not for anyone taking insulin or other glucose-lowering medication, anyone pregnant or breastfeeding, or anyone with a history of an eating disorder, and that people with medical conditions should talk to a clinician first.

The Metabolic Health Calculator does not estimate anything about your body. It takes numbers from a blood test and applies published formulas to them. Lab values entered in mmol/L are converted to mg/dL first (glucose x 18.0, triglycerides x 88.57, cholesterol x 38.67), and insulin in pmol/L is divided by 6.0, because the cutoffs below were set in those units.

Each result is placed against the cutoffs in the research it comes from. A TG/HDL ratio of 3.0 or more is the threshold McLaughlin and colleagues used to identify insulin-resistant overweight adults. TyG cutoffs in the literature run from roughly 8.3 to 9.0 depending on the population, and HOMA-IR has no universal cutoff (studies commonly use about 2.0 to 2.9), so for those two the calculator says whether a value is below, within, or above that published range rather than calling it normal or abnormal. Fasting glucose and HbA1c use the American Diabetes Association ranges (prediabetes from 100 mg/dL or 5.7%, diabetes from 126 mg/dL or 6.5%), and non-HDL cholesterol is desirable under 130 mg/dL. An LDL of 190 mg/dL or more is flagged because the 2018 ACC/AHA cholesterol guideline treats it as needing clinical evaluation on its own.

The metabolic syndrome check follows the 2009 harmonized definition: any three of a large waist, triglycerides of 150 mg/dL or more, HDL under 40 mg/dL (men) or 50 mg/dL (women), blood pressure of 130/85 or more, and fasting glucose of 100 mg/dL or more, with medication for a criterion counting as meeting it. Waist cutoffs are population-specific, so the calculator offers the AHA/NHLBI US and European values (102 cm or 40 in for men, 88 cm or 35 in for women) or the lower Asian values (90 and 80 cm). None of this is a diagnosis, which the result panel says in as many words.

When LDL is entered, the result also notes whether the reading matches the pattern Norwitz and colleagues (2022) described in some low-carb eaters: LDL of 200 mg/dL or more, HDL of 80 mg/dL or more, and triglycerides of 70 mg/dL or less, the so-called lean mass hyper-responder pattern. That is a description, not a risk category; the long-term significance is unresolved, and an LDL of 190 mg/dL or more is still flagged for a clinician’s review.

TheGKI Calculatoruses the formula from Meidenbauer, Mukherjee and Seyfried (2015): blood glucose divided by blood beta-hydroxybutyrate (BHB), both in mmol/L. Glucose in mg/dL is divided by 18.0 first, and BHB in mg/dL by 10.41 (its molar mass is 104.1 g/mol). The glucose-to-ketone ratio some clinicians use, glucose in mg/dL divided by BHB in mmol/L, is the same number multiplied by 18; the calculator shows it but does not grade it, because the targets quoted for it do not come from peer-reviewed research.

There is no clinical standard for GKI bands. The calculator shows Keto-Mojo’s published chart (9 and above not in ketosis, 6 to 9 low, 3 to 6 moderate, under 3 high) and labels it as a meter maker’s guide. The original paper described values approaching 1.0 as potentially most therapeutic in brain-cancer research, which the result mentions only as research context. The ketone reading itself is placed against the site’s ranges: 0.5 to 3.0 mmol/L nutritional ketosis, above 3.0 not a goal, above 5.0 a reason to seek care. BHB of 3.0 mmol/L or more with glucose of 200 mg/dL or more triggers a warning, because those are the ketone and glucose thresholds in the 2024 consensus definition of diabetic ketoacidosis. Only blood BHB is accepted: urine strips (acetoacetate) and breath meters (acetone) measure different ketones and cannot be converted.

TheCarb Tolerance Plannerdoes no estimating. It lays the step-up method that Virta Health and Keto-Mojo describe onto a calendar: start at the net carbs you are already in ketosis on, add a fixed step (3, 5 or 10 g; 5 g is the default) and hold each level for a set number of days (3 by default), up to a ceiling you choose (50 g by default, the top of the range where most adults stay in ketosis). The last step is clamped to the ceiling rather than overshooting it. The decision rule is the site’s nutritional-ketosis threshold: a morning blood BHB of 0.5 mmol/L or more means move up; below that on two mornings means step back. The plan is printed, not tracked; nothing is stored.

TheHidden Carb Ingredient Checkersplits a pasted ingredient list on commas, semicolons and parentheses (so sub-ingredients count), then matches each item against a curated list of about 50 sugars, starches, sugar alcohols and low-impact sweeteners, longest name first so that “high fructose corn syrup” is one match rather than three. Each match is ranked by how it counts toward net carbs, using the same rules as the net carb calculator, with glycemic notes from Atkinson et al. (2008) and Livesey (2003). Because US ingredient lists run from heaviest to lightest (21 CFR 101.4), a sugar or starch in the first three ingredients is called out, and anything after “contains 2% or less of” is marked minor. It reads what is in a product, never how much; the Nutrition Facts panel is the only source for grams.

TheKeto Baking Converterapplies published ratios; it does not estimate anything about you. Bulk sweeteners convert by volume, granulated for granulated, as cups of sweetener per cup of sugar. Pure stevia and monk fruit extracts convert by weight, because their sweetness multipliers are measured by weight and their powders differ from brand to brand. One sweetener converts to another through its sugar equivalent. Sugar weighs 200 g per cup granulated and 120 g per cup powdered (USDA FoodData Central); grams are offered for a sweetener only where its maker states a weight. Where sources disagree, the converter shows a range rather than one number: erythritol is rated 60 to 80% as sweet as sugar (1¼ to 1⅔ cups per cup of sugar, with 1⅓ as the headline), and allulose 1⅓ cups by the most-cited chart or about 1.43 at 70% sweetness. Branded blends use each maker’s own stated ratio, cited on the result.

Net carbs follow the net carb calculator: erythritol is subtracted in full and other sugar alcohols at 50%, so a cup of xylitol counts about 96 g. Allulose is counted as zero on the basis of FDA’s 2020 labeling guidance, which lets it be excluded from total and added sugars; FDA does not define net carbs. Flour swaps are by volume: 1/4 to 1/3 cup of coconut flour per cup of almond flour, with an extra egg per 1/4 cup; almond flour for wheat 1:1, or a gentler 25% swap; coconut flour for up to 20% of wheat flour. Directions with no sourced ratio, such as coconut flour to wheat, are not offered.

The food checker holds 268 entries. Carbohydrate, fiber, protein, and fat values come from USDA FoodData Central, the US Department of Agriculture’s public nutrient reference. Net carbs for each entry use the same formula above, so a number you see on a food page and a number you compute in the net carb calculator agree.

Whole foods vary. A cup of raspberries is not a fixed object, and USDA values are averages across samples. Brand-name and prepared items vary more, so check the label in front of you when there is one.

Food data is baked into the site when it is built and republished. There is no background job quietly rewriting numbers between deploys, which means what you read is what was reviewed.

Each of the 43 restaurant guides takes its numbers from the chain’s own published nutrition information: its website, nutrition calculator, or nutrition PDF. Every guide lists the pages it used and the month we checked them. We show total carbohydrate and fiber for every item, and net carbs are carbohydrate minus fiber, rounded to the nearest gram.

Many keto orders are modified orders, such as a burger without the bun or a bowl without rice. For those we start from the chain’s published values and subtract the published values of the part you leave off, or add up the published values of each component you choose. The note on each item says which. If a chain does not publish a component, we do not estimate it; we pick a different item.

Chains reformulate, portions differ between franchises, and a sauce swapped at one location changes the count. Use the guides to decide what to order before you walk in, then confirm against the restaurant’s current nutrition information if a few grams either way matters to you.

The keto drinks database lists 252 drinks across beer, hard seltzer, wine, spirits, cocktails, soda and mixers, energy and sports drinks, milk and milk alternatives, and coffee and tea. Every row links to its source. Brand drinks use the maker’s published nutrition facts. Generic drinks, such as a glass of dry red wine, a shot of vodka, a cup of whole milk, or a classic margarita, use USDA FoodData Central (SR Legacy and FNDDS), scaled from USDA’s per-100-gram values to a stated serving using USDA’s own portion weights.

A few low-carb builds, such as vodka with soda and lime, are the sum of their USDA components, and the row says so. Net carbs are total carbohydrate minus fiber, to one decimal. Sugar alcohols and allulose are noted where a label lists them, not subtracted. The verdict on each row is a share of a strict 20 g day: low carb is 3 g or less per serving, plan for it is up to 7 g, and anything higher is high carb.

Every formula on this site is a population average wearing a personal answer’s clothing. Mifflin-St Jeor was fitted to a group of people and then handed to you. Measured resting metabolic rate in real individuals can sit meaningfully above or below what any predictive equation returns, and no equation asks about the things that move it.

Nothing here can see:

• ▸Thyroid function, medications, or any condition that changes how you use energy
• ▸A long history of dieting, which can leave expenditure lower than a formula expects
• ▸Sleep, stress, and how much you move outside of deliberate exercise
• ▸Menstrual cycle phase, perimenopause, and other hormonal shifts
• ▸How much of your body weight is muscle, unless you supply a measured body fat percentage

Treat any output here as a starting point, not a verdict. Run it for two to three weeks, weigh the trend rather than a single morning, and adjust from what actually happened. Real feedback beats a better equation every time. If your results and the calculator disagree, the calculator is the one that is wrong.

None of this is medical advice, and no clinician reviews it. Our content review process sets out exactly what we check, what we do not, and when to bring a dietitian or physician into the conversation.

• ▸Mifflin MD, St Jeor ST, Hill LA, et al. A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr. 1990;51(2):241-247.
• ▸Masood W, Annamaraju P, Khan Suheb MZ, Uppaluri KR. Ketogenic Diet. StatPearls. NIH National Library of Medicine, 2023.
• ▸U.S. Department of Agriculture, Agricultural Research Service. FoodData Central.
• ▸Phinney SD, Volek JS. The Art and Science of Low Carbohydrate Living. Beyond Obesity LLC, 2011.
• ▸McLaughlin T, Abbasi F, Cheal K, et al. Use of metabolic markers to identify overweight individuals who are insulin resistant. Ann Intern Med. 2003;139(10):802-809.
• ▸Simental-Mendía LE, Rodríguez-Morán M, Guerrero-Romero F. The product of fasting glucose and triglycerides as surrogate for identifying insulin resistance in apparently healthy subjects. Metab Syndr Relat Disord. 2008;6(4):299-304.
• ▸Matthews DR, Hosker JP, Rudenski AS, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412-419.
• ▸Alberti KGMM, Eckel RH, Grundy SM, et al. Harmonizing the metabolic syndrome: a joint interim statement. Circulation. 2009;120(16):1640-1645.
• ▸McArdle WD, Katch FI, Katch VL. Exercise Physiology: Nutrition, Energy, and Human Performance. Lippincott Williams & Wilkins. Source of the Katch-McArdle lean-mass BMR equation (370 + 21.6 x lean mass in kg).
• ▸Boer P. Estimated lean body mass as an index for normalization of body fluid volumes in humans. Am J Physiol. 1984;247(4 Pt 2):F632-F636.
• ▸Morton RW, Murphy KT, McKellar SR, et al. 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. Br J Sports Med. 2018;52(6):376-384.
• ▸Mavropoulos JC, Yancy WS, Hepburn J, Westman EC. The effects of a low-carbohydrate, ketogenic diet on the polycystic ovary syndrome: a pilot study. Nutr Metab. 2005;2:35.
• ▸Kreitzman SN, Coxon AY, Szaz KF. Glycogen storage: illusions of easy weight loss, excessive weight regain, and distortions in estimates of body composition. Am J Clin Nutr. 1992;56(1 Suppl):292S-293S.
• ▸Hall KD, Sacks G, Chandramohan D, et al. Quantification of the effect of energy imbalance on bodyweight. Lancet. 2011;378(9793):826-837.
• ▸Virta Health. Managing Potassium and Sodium as Part of a Well-Formulated Ketogenic Diet.
• ▸National Institutes of Health, Office of Dietary Supplements. Potassium Fact Sheet for Health Professionals.
• ▸National Institutes of Health, Office of Dietary Supplements. Magnesium Fact Sheet for Health Professionals.
• ▸National Academies of Sciences, Engineering, and Medicine. Dietary Reference Intakes for Sodium and Potassium. Washington, DC: The National Academies Press, 2019.
• ▸Norwitz NG, Feldman D, Soto-Mota A, Kalayjian T, Ludwig DS. Elevated LDL Cholesterol with a Carbohydrate-Restricted Diet: Evidence for a "Lean Mass Hyper-Responder" Phenotype. Curr Dev Nutr. 2022;6(1):nzab144.
• ▸Meidenbauer JJ, Mukherjee P, Seyfried TN. The glucose ketone index calculator: a simple tool to monitor therapeutic efficacy for metabolic management of brain cancer. Nutr Metab (Lond). 2015;12:12.
• ▸Keto-Mojo. Glucose Ketone Index (GKI): what it is and how to calculate it, including the published GKI levels chart. Manufacturer educational material, not a clinical guideline.
• ▸Umpierrez GE, Davis GM, ElSayed NA, et al. Hyperglycemic crises in adults with diabetes: a consensus report. Diabetes Care. 2024;47(8):1257-1275.
• ▸Virta Health. Guidance on adding carbohydrates back while staying in nutritional ketosis (blog).
• ▸Keto-Mojo. Article on finding your carb limit on a ketogenic diet.
• ▸U.S. Food and Drug Administration. 21 CFR 101.4, Food; designation of ingredients (ingredients listed in descending order of predominance by weight).
• ▸Atkinson FS, Foster-Powell K, Brand-Miller JC. International tables of glycemic index and glycemic load values: 2008. Diabetes Care. 2008;31(12):2281-2283.
• ▸Livesey G. Health potential of polyols as sugar replacers, with emphasis on low glycaemic properties. Nutr Res Rev. 2003;16(2):163-191.
• ▸U.S. Food and Drug Administration. The Declaration of Allulose and Calories from Allulose on Nutrition and Supplement Facts Labels: Guidance for Industry. 2020.

Tell us. A formula, a food entry, or a restaurant figure that does not match its source is a defect, and we would rather hear it from you than leave it up. Write to hello@meketo.com and point at the page. Our editorial policy describes how corrections are handled once they arrive.