Protein calculator
How the calculator works
Every formula, range and cap behind the calculator, with its source. Last updated .
Anyone can read the exact maths our calculators run, check it against the sources, and tell us where we are wrong. The code is in src/lib/protein.ts and it runs entirely in your browser.
The protein calculator, in short
- Work out a reference weight from your weight, height and sex.
- Pick a g per kg range from your situation: losing weight or not, training or not, age.
- Multiply, then apply floors and caps.
- Divide by your number of meals to get a per-meal range.
1. Reference weight
| Your BMI | Weight used |
|---|---|
| Under 30 | Your actual body weight |
| 30 or above | An adjusted body weight |
BMI is your weight in kg divided by your height in metres squared.
Adjusted body weight is ideal body weight + 0.4 × (actual weight − ideal body weight), the adjustment used in dietetic practice for people with obesity. Ideal body weight uses the Devine formula: 50 kg for men, 45.5 kg for women, plus 2.3 kg for every inch over 5 feet [4]. "Prefer not to say" uses the lower of the two, which gives the more conservative target.
Why adjust at all? Body fat needs very little protein. Multiplying a high body weight by 1.6 g/kg produces a number most people cannot eat, and that no guideline intends. The 2025 joint advisory makes the same point when it notes that a target based on fat-free mass is more accurate but needs a body scan most people do not have [1].
2. The ranges
| Situation | g per kg of reference weight a day |
|---|---|
| Not losing weight, not training | 0.8–1.2 |
| Not losing weight, training | 1.2–1.6 |
| Losing weight (including on a GLP-1) | 1.2–1.6 |
| Losing weight and training | 1.4–1.8 |
Where these come from:
- 0.8 g/kg is the recommended dietary allowance for adults, quoted in the 2025 joint advisory [1].
- 1.2–1.6 g/kg during active weight loss is the range that advisory says has been proposed [1]. A 2025 expert consensus written specifically for GLP-1 care lands in the same place at 1.2–1.5 g/kg of actual body weight [2].
- Higher with training follows the International Society of Sports Nutrition position stand, which puts 1.4–2.0 g/kg as the range for people doing resistance exercise [3].
- Age 65 and over: the floor is raised to 1.0 g/kg, or 1.2 if you also strength train. A 2025 practical review of weight loss and musculoskeletal health puts older adults at 1.2–1.6 g/kg, and at 1.6 or more when diet and exercise are combined [5].
If you take a GLP-1 medication, the calculator treats you as losing weight, because that is what these medicines are prescribed to do.
3. Floors and caps
| Rule | Value | Why |
|---|---|---|
| Never below | 0.8 g/kg, and never under 50 g a day | The RDA floor [1] |
| Never above | 2.0 g/kg, and never over 200 g a day | The advisory says prolonged intake at or above 2 g/kg a day is best avoided [1] |
| Per meal | Shown as your daily range divided by your meals | 20–40 g per eating occasion is the usual effective amount [3] |
If a per-meal figure lands under 20 g or over 40 g, the result says so and suggests changing the number of meals.
4. Rounding
Daily totals are rounded to the nearest 5 g, and per-meal figures to the nearest gram. Protein in real food varies by brand and cut anyway, so a range rounded to 5 g is honest about the precision available.
What this calculator does not do
- It does not use body composition. A target based on fat-free mass is more accurate [1], but it needs a DXA or a reliable body scan. We use an adjusted weight instead and say so.
- It does not count what you eat. The optional "what do you eat now" box compares one number you type against your range. It is not a food log.
- It does not personalise for disease. Kidney disease, liver disease, pregnancy, breastfeeding, eating disorders and anyone under 18 are out of scope. The tool says this on the page and gives no number for under-18s.
- It is not medical advice, and it never comments on medication or dosing.
- It does not store or send anything. Every field stays in your browser. Nothing you type is sent to our servers or to analytics.
Known limits and open questions
- The evidence for per-meal distribution comes mostly from short studies of muscle protein synthesis, not long-term body composition trials. The 2025 review we follow is explicit that total daily protein is the main driver, and per-meal splitting is a practical way to reach it [5].
- The 1.2–1.6 g/kg range is not settled. A 2022 meta-analysis found the clearest gains in lean mass at 1.6 g/kg or above in adults under 65, and 1.2–1.59 in older adults [6].
- The Devine formula is old (1974) and was written for drug dosing, not nutrition. It is still the basis of the adjusted body weight used in dietetic practice, which is why it is here.
- Sex is used only for ideal body weight. We do not vary the g/kg ranges by sex, because the guidance we follow does not.
The muscle loss calculator
Separate tool, same rule: it reports a range, never one number, and never a range outside what the trials found.
| Step | What it does |
|---|---|
| Input | Weight lost, strength training, protein, and whether you are 65 or over |
| Overall range | 24% to 39% of the weight lost, the published 95% confidence interval |
| Where you sit in it | One of three bands: lower 24–29%, middle 27–35%, upper 33–39% |
| Output | Lean mass lost as a range, the fat side as the remainder, and what pushed the estimate each way |
Where the range comes from. A 2026 meta-analysis of 36 randomised trials, 5,316 participants, found lean mass made up 31% of total weight lost, 95% confidence interval 24% to 39% [7]. Individual trials land across that span: about 25% in the SURMOUNT-1 body-scan sub-study of tirzepatide, where the placebo group lost a similar proportion, and about 39% in the STEP 1 sub-study of semaglutide (5.3 kg of 13.6 kg).
How your answers move it, and how much to trust that. Resistance training scores −2 when done three or more times a week and −1 for once or twice; hitting a protein target scores −1 and being below it +1; no training scores +1, and being 65 or over scores +1. A total of −2 or lower puts you in the lower band, +1 or higher in the upper band, anything between in the middle.
That weighting is a judgement about order, not magnitude. What the evidence supports is the direction: trials that added resistance training and adequate protein report less lean mass loss than diet alone [5], and older adults lose more for the same weight loss [5]. What no trial gives us is a formula that converts your habits into a precise percentage. So the tool sorts you into a third of the published interval rather than pretending to a decimal place, and the page says so where the reader can see it.
What it deliberately does not do.
- It does not convert lean mass into muscle. Lean soft tissue on a DXA scan includes water and organ tissue. The share that is muscle varies widely between people, and any fixed multiplier claims a precision nobody has.
- It never reports outside 24% to 39%. No combination of answers can produce a number the trials did not find. There is a unit test for all eighteen combinations.
- It does not measure anything. It applies trial averages to what you type.
- It does not use how fast you lost the weight, or your starting weight, because the published trials do not support adjusting for either with any confidence.
The water calculator
Lives inside the guide on how much water to drink on a GLP-1 rather than on its own page, because the useful part is situational and a bare number would mislead.
| Step | What it does |
|---|---|
| Total for the day | 35 ml per kg of body weight, with a floor of 2 litres |
| Water from food | Subtracts 20% of the total, the share that normally arrives inside food |
| Eating less | Reduces that food share by 25% if you are eating less, 50% if much less, and adds the difference to what you must drink |
| Today | Adds 500 ml for a training day, 750 ml for a sick day |
| Output | What to drink, with the total, the food share and every adjustment shown separately |
Where the numbers come from. The 2025 expert consensus gives "more than 2 L of fluid intake per day" and cites 2 to 4 L, or about 35 ml per kg [2]. The 20% food share is the US National Academies' figure, whose 3.7 L and 2.7 L targets are explicitly total water from all beverages and foods [8]. The 2025 joint advisory gives no number, only "ensure adequate fluids" [1].
Why the food subtraction exists at all. Every consumer hydration calculator we checked quotes a total-water figure as though it were water to drink, which overstates the drinking target by roughly a fifth for an ordinary eater. On a GLP-1 the error runs the other way and gets worse: eating much less removes part of that food water, so the amount you have to drink goes up.
What it deliberately does not do.
- It does not give a medical fluid restriction or target. Anyone with heart or kidney disease is told on the page to use the figure their clinician gave them instead.
- It does not model electrolytes. No GLP-1 guidance we found recommends routine supplements, and we are not going to invent a formula for one.
- It does not claim to prevent anything. The labels' kidney-injury warnings are about volume depletion; whether hitting a fluid target prevents those events has not been tested.
Honest about the weakest part: the 25% and 50% appetite steps are our own, chosen to be round and conservative. Nobody has measured how much food water a person on a GLP-1 actually loses. We show the adjustment on screen rather than folding it invisibly into one number, so you can disagree with it.
What the evidence does not establish
A page listing only what is known is not worth trusting. These are open questions, and we would rather say so than round them off.
- Whether strength and function hold up in the people most at risk. Older adults, or people who already have little muscle. The reassurance so far is about mass, mostly in younger and heavier groups. A 35-trial review found no included study that measured physical function.
- Whether tirzepatide costs more lean mass than semaglutide. One preprint of routine-care data suggests one to two points more, alongside more weight loss. It is not peer reviewed and its authors call it descriptive.
- Whether the medicines add lean loss beyond what the same weight loss would cost anyway. No pair-fed human trial exists. The closest evidence is the SURMOUNT-1 placebo group, which lost a similar proportion.
- Whether splitting protein across meals changes body composition, or only the daily total matters. The per-meal evidence comes mostly from short muscle-protein-synthesis studies, and the 2025 practical review is explicit that the daily total is the main driver [5].
- How much of any of this applies to people over 65, who are under-represented in the trials and have the most to lose.
Corrections
We change numbers when the source says we should. Each change is logged here.
| Date | What changed | Why |
|---|---|---|
| 17 Sep 2026 | The share of weight lost as lean mass moved from "26% to 40%" to 24% to 39%, middle 31% | The old figure leaned on a three-patient case series quoting trial sub-groups. The new one is a meta-analysis of 36 trials and 5,316 people |
| 17 Sep 2026 | Tofu, salmon and cod protein figures in the meal plan corrected | Checked against USDA FoodData Central rather than estimated |
| 19 Sep 2026 | The smart scale guide no longer implies hydration is a major cause of jumpy body-fat readings | A randomised crossover trial found that drinking about 832 ml moved the estimate by less than one percentage point, which its authors doubted was clinically relevant [9]. The scale's own equation matters more |
Corrections and feedback
If you are a dietitian, clinician or researcher and something here is wrong or out of date, email hello@dyadfit.com. Tell us which rule and which source, and we will fix it and note the change on this page.
Sources
- Mozaffarian D et al. Nutritional priorities to support GLP-1 therapy for obesity: a joint Advisory (ACLM, ASN, OMA, TOS). Obesity 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12304835/
- Sievenpiper JL et al. Nutritional and lifestyle supportive care recommendations for management of obesity with GLP-1-based therapies. Obesity Pillars 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12768930/
- Jäger R et al. International Society of Sports Nutrition Position Stand: protein and exercise. J Int Soc Sports Nutr 2017;14:20. https://pmc.ncbi.nlm.nih.gov/articles/PMC5477153/
- Devine BJ. Gentamicin therapy. Drug Intell Clin Pharm 1974;8:650–655, as used for ideal body weight in clinical practice. https://journals.sagepub.com/doi/10.1177/106002807400801104
- Mesinovic J et al. Exercise and dietary recommendations to preserve musculoskeletal health during weight loss in adults with obesity: a practical guide. Rev Endocr Metab Disord 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12534310/
- Nunes EA et al. Systematic review and meta-analysis of protein intake to support muscle mass and function in healthy adults. J Cachexia Sarcopenia Muscle 2022;13:795–810. https://pmc.ncbi.nlm.nih.gov/articles/PMC8978023/
- Meta-analysis of 36 randomised trials (5,316 participants): lean mass loss was 31% of total weight lost, 95% CI 24–39%. Diabetes Metab Res Rev, June 2026. https://pmc.ncbi.nlm.nih.gov/articles/PMC13282017/
- US National Academies of Sciences, Engineering, and Medicine. Dietary Reference Intakes for water: 3.7 L a day for men and 2.7 L for women as total water from all beverages and foods, about 20% of it from food. https://www.nationalacademies.org/news/report-sets-dietary-intake-levels-for-water-salt-and-potassium-to-maintain-health-and-reduce-chronic-disease-risk
- Randomised controlled crossover trial of fluid intake and bioelectrical impedance, 18 participants: about 832 ml, a 2% rise in total body water, changed body fat estimates by less than 1 percentage point. https://pmc.ncbi.nlm.nih.gov/articles/PMC10143694/