Protein: who genuinely needs more of it, and what the extra does not do

Protein is the nutrient with the widest gap between how much is sold and how much is needed. Most adults in wealthy countries already eat above the requirement. The two groups with a real case for more are older adults losing muscle and people training hard, and the case for everyone else is thinner than the packaging suggests.

How much you need

• The Daily Value for protein on a food label is 50 g a day. — adults and children 4 years and older Evidence E source
• In resistance-trained women measured directly after a workout, protein needs levelled off at about 1.5 g per kg a day, nearly double the general adult recommendation. — resistance-trained women aged about 23, studied in the luteal phase after a bout of whole-body resistance exercise, over roughly eight hours of recovery Evidence B source
• In the same meta-regression the benefit stops at a total intake of about 1.6 g of protein per kilogram a day, and it shrinks with every year of age. — healthy adults in the same 49 randomised trials; subgroups by age and by training experience Evidence A source
• To get 20 g of protein from faba beans a trial had to serve 234 g of cooked bean mash, which is a full bowl rather than a spoonful. — test meal served to healthy adults in a metabolic study Evidence B source
• Europe's food safety authority puts the adult protein intake at 0.83 g per kilogram of body weight a day, which is about 58 g for someone of 70 kg, and sets no upper limit at all. — adults in the European Union, including older adults Evidence E source
• Grain is slowly getting less protein: across thirty years of open-air experiments, raising carbon dioxide over a field lowered the nitrogen and protein in the seeds of every crop except legumes. — 18 C3 crops and maize and sorghum grown in open fields, not people Evidence D source

How well your body absorbs it

• Cooked faba beans give up 84 per cent of their nitrogen by the end of the small intestine, and the amino acids themselves come through better than that, at about 90 per cent. — 8 to 9 healthy adults aged 33 plus or minus 10 years, BMI 18 to 30, no food allergy, fitted with a naso-ileal tube after an overnight fast Evidence B source
• The quality score of one and the same bean moves from 0.67 to 0.77 depending only on the number used to turn nitrogen into protein, so two honest labs can publish two different answers about the same food. — cooked dehulled faba bean mash fed to healthy adults Evidence B source
• Measured the same way and in the same country, cooked meat gives up more of its amino acids than whole boiled egg, and whole boiled egg more than egg white alone. — healthy non-smoking Indian adults aged 18 to 45, BMI 18.5 to 25, both sexes equally represented, no antibiotics in the previous four weeks Evidence B source
• Red kidney beans in an ordinary cooked meal hand over about 79 per cent of their essential amino acids, and threonine does worst of all at 69 per cent. — 10 healthy adults, 5 male and 5 female, eating a typical Jamaican meal prepared with the labelled beans Evidence B source
• The official reason plant protein counts for less is not the amount of protein but its amino acid pattern together with how much of it gets absorbed. — general, as a definition used to frame protein requirement research Evidence E source
• Running chickpeas and yellow peas through an extruder lifted amino acid absorption to about 88 per cent, roughly a fifth better than the same beans pressure-cooked. — 28 moderately stunted South Indian schoolchildren aged 6 to 11, height-for-age z score between minus 3 and minus 2, low to middle income households Evidence B source
• Ultra-high-temperature milk left 8 per cent less of its nitrogen in the body after a meal than microfiltered or pasteurised milk did. — healthy adults aged 20 to 68 across four included studies, of which the nitrogen retention result comes from one study of 25 participants using 15N-labelled milk Evidence B source

What the evidence does and does not show

• In older adults diagnosed with muscle loss, whey protein raised muscle mass and walking speed across ten randomised trials. — older adults with sarcopenia by EWGSOP or AWGS criteria, in community and hospital, 10 trials Evidence A source
• An umbrella review of 33 systematic reviews found that when calories are held equal, eating more protein does not change body weight, fat mass or waist size. — adults, 33 systematic reviews mostly of randomised trials Evidence A source
• Adding a protein supplement to at least six weeks of resistance training gains about 0.3 kg more lean mass and about 2.5 kg more on a one-rep max than training alone. — healthy adults in 49 randomised trials, trained and untrained, young and old Evidence A source
• Multi-ingredient supplements that put creatine or vitamin D alongside protein beat no supplement by about 0.8 kg of lean mass, but they did not beat plain protein. — healthy adults in 35 trials; the commonest combinations were protein with creatine (17 trials) and protein with vitamin D (10 trials) Evidence A source
• High-protein eating showed no link either way with cardiovascular death or stroke in people who started without heart disease. — adults without established cardiovascular disease, across 14 cohorts Evidence C source
• Pooling 30 randomised trials, animal protein built slightly more muscle than plant protein, and the gap disappeared in people over 60 and between soy and milk. — adults in 43 eligible trials, 30 pooled for muscle mass Evidence A source
• Soy protein matched milk protein for muscle, while rice, chia, oat and potato protein came out behind. — adults, 17 trials for the soy comparison and 5 for non-soy plant proteins Evidence A source
• Spreading protein evenly across meals instead of loading it at dinner made no difference to body composition during weight loss. — healthy women aged 20 to 44 with a BMI of 28 to 45, eight weeks of controlled feeding then eight of self-choice Evidence B source
• Whether heat treatment of milk changes what your body does with the protein is still an open question, because only four human studies have ever asked it and no two used the same method. — healthy adult men and women Evidence B source
• The advice to front-load protein at breakfast rests on thinner ground than it sounds: of eleven studies that looked at breakfast protein and muscle mass, fewer than six in ten found anything, and for muscle strength it was half of six. — adults, with the muscle mass signal concentrated in five studies of older people plus two of middle-aged women and young men Evidence C source
• Nobody can pool the studies behind the protein requirement, because every research group measures it a different way, and the agency review says so in as many words. — all population groups across the lifespan, with infants, children, adolescents, pregnant and lactating people named as the least studied Evidence E source

What too much does

• In the OmniHeart feeding trial, a diet with 25 per cent of energy from protein raised kidney filtration rate in healthy adults, and what that means long term is unknown. — healthy adults with prehypertension or stage 1 hypertension, weight held constant Evidence B source
• An earlier meta-analysis put numbers on the same shift: filtration up about 7 ml/min, blood urea up, and 25 mg more calcium a day leaving in the urine. — adults without chronic kidney disease Evidence A source
• There is no agreed safe upper limit for protein or for individual amino acids, and the agency review that says so puts the reason plainly: almost nobody has studied it. — general population across the lifespan Evidence E source

What to do with this

If you eat meat, fish, eggs, dairy, beans or tofu on most days, you are almost certainly above the 50 gram label figure already. The question worth asking is not how to get more but whether you are in one of the groups where more genuinely helps.

Older adults are that group. Muscle loss with age is real, and in trials in people diagnosed with it, protein raised muscle mass and walking speed, with grip strength improving when the protein came alongside resistance training. The training matters as much as the protein.

If you are trying to lose weight, be careful what you expect. When calories are held equal, more protein does not change body weight or fat mass. Whether it helps on a reduced-calorie diet is a question the evidence does not answer cleanly.

If you lift and you buy powder, two numbers are worth having. Adding protein to at least six weeks of resistance training gains about 0.3 kg more lean mass and about 2.5 kg on a one-rep max than the training alone, which is small next to what the training itself does. And the gain stops at a total intake of around 1.6 grams per kilogram a day, shrinking with every year of age.

Multi-ingredient powders are the other thing worth knowing about. Products that put creatine or vitamin D alongside the protein beat taking nothing by about 0.8 kg of lean mass, and they did not beat plain protein.

On the kidney worry: in healthy adults, a diet with a quarter of energy from protein raised filtration rate, and nobody knows whether that matters over decades. If you already have kidney disease, protein intake is a clinical decision rather than a fitness one.

Two numbers from those same studies are worth having in the kitchen. To get 20 g of protein from faba beans, a trial had to serve 234 g of cooked bean mash, which is a full bowl rather than a spoonful. Processing changes the rest. Extruded chickpea and yellow pea reached 89 and 88 per cent amino acid digestibility in moderately stunted schoolchildren, about a fifth better than the same pulses pressure-cooked and measured by the same method in adults.

Milk is the one food where heat treatment has been measured in people. After a meal, UHT milk left 8 per cent less of its nitrogen in the body than microfiltered or pasteurised milk did, with 5 per cent more of it sitting in the urea pool at eight hours. Read the next line before you change what you buy.

Unsettled. Whether heat treatment of milk changes what your body does with the protein is an open question. Four human studies have ever asked it, no two used the same method, and the review that gathered them reached no conclusion.

Unsettled. The habit of front-loading protein at breakfast rests on thinner ground than it sounds. Of eleven studies of breakfast protein and muscle mass, about 58.8 per cent of findings reported an increase, and on muscle strength three studies out of six found anything. Most of the muscle-mass signal comes from five studies in older people.

One last thing, and it is about the food supply rather than about you. Across thirty years of open-air experiments, raising carbon dioxide over a field lowered the nitrogen and protein in the seeds of every crop except legumes, alongside less zinc and iron in most of them. That is plants rather than people, and it is the reason a protein table is a snapshot of one particular decade.

Where to get it from food

The bottom line

Fifty grams a day is the label figure, and an ordinary diet with meat, fish, eggs, dairy, beans or tofu passes it without planning. Older adults losing muscle are the group with the clearest evidence: in ten randomised trials, whey protein raised muscle mass and walking speed, and raised grip strength when it came with resistance training. If you train, the gain from extra protein stops at a total intake of around 1.6 grams per kilogram a day, and powders that add creatine or vitamin D beat taking nothing without beating plain protein. For everyone else, when calories are held equal, more protein does not change body weight, fat mass or waist size. On kidneys, the pooled trials show filtration rate rising while creatinine stays put, along with about 25 mg more calcium a day leaving in the urine, and whether any of that matters over decades is still unknown. Three things from the digestibility literature belong under all that. The first is that the gap between animal and plant protein is a measured quantity rather than a slogan, and it is narrower than the language around it. Cooked faba beans gave up 84.1 per cent of their nitrogen by the end of the small intestine and 89.8 per cent of their amino acids. Red kidney beans in an ordinary cooked meal gave up 79.4 per cent, with threonine worst at 69.0. Measured the same way and in the same country, egg white came out at 86.3 per cent, whole boiled egg at 89.4 and cooked meat at 92.0. The official reason plant protein counts for less is that pattern of amino acids together with how much of it gets absorbed. The second thing is that the score itself depends on an arithmetic choice. One and the same bean scores 0.67 or 0.77 depending only on the factor used to turn nitrogen into protein, so two honest laboratories can publish two different answers about the same food. The third is that no upper limit for protein exists anywhere. Europe's food safety authority puts the adult intake at 0.83 g per kilogram of body weight a day, about 58 g for someone of 70 kg, and sets no tolerable upper level at all. The agency review that went looking for one on individual amino acids found the evidence sparse and said its own scope did not cover the question. The same review could not pool the requirement studies either, because every research group measures it a different way. An absent ceiling here is a measure of what nobody has studied.

Sources

1. prot-01 · Position of an expert body · regulatory reference value
   50 g (value from the source's table, not a sentence)
   FDA, Daily Value on the Nutrition and Supplement Facts Labels (updated 2024-03-05) · checked 2026-09-15 · independently re-checked
2. prot-r2-5 · Randomised controlled trial(s) · controlled metabolic crossover study
   A mixed model biphase linear regression revealed a break point (i.e., estimated average requirement) of 1.49 ± 0.44 g·kg·d (mean ± 95% confidence interval) in Ox (r = 0.64) and 1.53 ± 0.32 g·kg·d in NB (r = 0.65), indicating a saturation in whole-body anabolism.
   Malowany et al., Medicine & Science in Sports & Exercise, 2019 · checked 2026-09-17 · independently re-checked
3. prot-t8 · Meta-analysis or systematic review · meta-regression within a systematic review of randomised controlled trials
   The impact of protein supplementation on gains in FFM was reduced with increasing age (-0.01 kg (-0.02,-0.00), p=0.002) and was more effective in resistance-trained individuals (0.75 kg (0.09, 1.40), p=0.03). Protein supplementation beyond total protein intakes of 1.62 g/kg/day resulted in no further RET-induced gains in FFM.
   Morton et al., British Journal of Sports Medicine, 2018 · checked 2026-09-17 · independently re-checked
4. prot-e2-09 · Randomised controlled trial(s) · controlled human feeding study, meal preparation
   The calculated amount of bean mash providing 20 g protein (3.7 g nitrogen, N × 5.4) was 234.1 g.
   Guillin et al., The Journal of Nutrition, 2024 · checked 2026-09-18 · independently re-checked
5. prot-e2-11 · Position of an expert body · expert body statement
   Adults (including older adults) – 0.83 g per kg of body weight per day.
   EFSA sets population reference intakes for protein, European Food Safety Authority, 2012 · checked 2026-09-18 · independently re-checked
6. prot-e2-13 · Mechanistic or laboratory data · meta-analysis of field experiments
   Subsequent study showed reduced levels of nutrients, notably Zn and Fe in most crops, and lower nitrogen and protein in the seeds of non-leguminous crops.
   Ainsworth & Long, Global Change Biology, 2021 · checked 2026-09-18 · independently re-checked
7. prot-e2-01 · Randomised controlled trial(s) · controlled human ileal balance study
   Thus, the true ileal digestibility of protein was 84.1% ± 7.7%.
   Guillin et al., The Journal of Nutrition, 2024 · checked 2026-09-18 · independently re-checked
8. prot-e2-02 · Randomised controlled trial(s) · protein quality calculation inside a human digestibility trial
   Assuming that tryptophan digestibility was similar to that of other AAs, histidine and tryptophan were deficient at the same level, resulting in a score ranging between 0.66 and 0.67 for a N-to-protein conversion factor of 6.25 and 0.76–0.77 for a factor of 5.4.
   Guillin et al., The Journal of Nutrition, 2024 · checked 2026-09-18 · independently re-checked
9. prot-e2-03 · Randomised controlled trial(s) · controlled human dual isotope tracer study
   The mean ± SD true ileal IAA digestibility was 86.3% ± 4.6% for egg white protein, 89.4% ± 4.5% for whole boiled egg, and 92.0% ± 2.8% for cooked meat.
   Devi et al., American Journal of Clinical Nutrition, 2018 · checked 2026-09-18 · independently re-checked
10. prot-e2-04 · Randomised controlled trial(s) · controlled human dual isotope tracer study
    Mean RKB-IAA digestibility (%) was 79.4 ± 0.5, ranging from 69.0 ± 1.2 (threonine) to 85.7 ± 1.7 (lysine).
    Bandsma et al., The Journal of Nutrition, 2023 · checked 2026-09-18 · independently re-checked
11. prot-e2-05 · Position of an expert body · agency systematic review, background section
    Protein from animal sources (i.e., eggs, milk, meat) is considered higher quality while protein from plant sources (i.e., grains, legumes, nuts) is considered lower quality due to amino acid content and digestibility.
    Evaluation of Dietary Protein and Amino Acid Requirements: A Systematic Review, Agency for Healthcare Research and Quality, 2024 · checked 2026-09-18 · independently re-checked
12. prot-e2-06 · Randomised controlled trial(s) · randomised controlled trial
    Extruded chickpea and yellow pea protein had good IAA digestibility in moderately stunted children, which was 20% higher than an earlier report of their digestibility when pressure-cooked, measured by the same method in adults.
    Kashyap et al., The Journal of Nutrition, 2020 · checked 2026-09-18 · independently re-checked
13. prot-e2-07 · Randomised controlled trial(s) · rapid systematic review of human intervention studies
    A postprandial reduction in dietary N retention of 8% was also observed after ingestion of the UHT compared to the two other types of milk
    Frontiers in Nutrition, 2021 · checked 2026-09-18 · independently re-checked
14. prot-t1 · Meta-analysis or systematic review · systematic review and meta-analysis
    In WP group versus (vs.) Isocaloric placebo (PLA)/Routine consultation (RC) group, WP significantly increased the appendicular skeletal muscle mass index (SMD: 0.47, 95%CI: 0.23, 0.71), appendicular skeletal muscle mass (SMD: 0.28, 95%CI: 0.11, 0.45) and gait speed (SMD: 1.13, 95%CI: 0.82, 1.44) in older patients with sarcopenia.
    Li et al., Journal of Nutrition, Health and Aging, 2024 · checked 2026-09-16 · independently re-checked
15. prot-t3 · Meta-analysis or systematic review · umbrella review
    "Possible" evidence exists that the amount of protein does not affect BW, FM and WC in adults under isoenergetic conditions.
    Hoffmann et al., European Journal of Nutrition, 2024 · checked 2026-09-16 · independently re-checked
16. prot-t7 · Meta-analysis or systematic review · systematic review and meta-analysis of randomised controlled trials
    Data from 49 studies with 1863 participants showed that dietary protein supplementation significantly (all p<0.05) increased changes (means (95% CI)) in: strength-one-repetition-maximum (2.49 kg (0.64, 4.33)), FFM (0.30 kg (0.09, 0.52)) and muscle size-muscle fibre cross-sectional area (CSA; 310 µm2 (51, 570)) and mid-femur CSA (7.2 mm2 (0.20, 14.30)) during periods of prolonged RET.
    Morton et al., British Journal of Sports Medicine, 2018 · checked 2026-09-17 · independently re-checked
17. prot-t9 · Meta-analysis or systematic review · systematic review and meta-analysis of randomised controlled trials
    Data from 35 trials with 1387 participants showed significant (p<0.05) increases in FFM (0.80 kg (95% CI 0.44 to 1.15)), 1RM lower body (4.22 kg (95% CI 0.79 to 7.64)) and 1RM upper body (2.56 kg (95% CI 0.79 to 4.33)) where a supplement was compared with all non-MIP supplemented conditions (means (95% CI)).
    O'Bryan et al., British Journal of Sports Medicine, 2020 · checked 2026-09-17 · independently re-checked
18. prot-r6-04 · Observational data · meta-analysis of prospective cohort studies
    A total of 6 studies, including 221,583 participants, reported data about cardiovascular death, without showing a statistically significant difference in the random effect model (odds ratio: 0.94; confidence interval: 0.60-1.46; I2 = 98%; p = 0.77). Analysis of three studies, which included 90,231 participants showed that a high protein diet was not associated with a lower risk of stroke (odds ratio: 1.02; confidence interval: 0.94-1.10; I2 = 0%; p = 0.66).
    Papadopoulou et al., Nutrients, 2023 · checked 2026-09-17 · independently re-checked
19. prot-t4 · Meta-analysis or systematic review · systematic review and meta-analysis
    Compared with animal protein, plant protein resulted in lower muscle mass following the intervention (SMD = -0.20; 95% CI: -0.37, -0.03; P = .02), with stronger effects in younger (<60 years; SMD = -0.20; 95% CI: -0.37, -0.03; P = .02) than in older (≥60 years; SMD = -0.05; 95% CI: -0.32, 0.23; P = .74) adults.
    Robinson-Mitchell et al., Nutrition Reviews, 2025 · checked 2026-09-16 · independently re-checked
20. prot-t5 · Meta-analysis or systematic review · systematic review and meta-analysis
    There was no pooled effect difference between soy and milk protein for muscle mass (SMD = -0.02; 95% CI: -0.20, 0.16; P = .80) (n = 17 RCTs), yet animal protein improved muscle mass compared with non-soy plant proteins (rice, chia, oat, and potato; SMD = -0.58; 95% CI: -1.06, -0.09; P = .02) (n = 5 RCTs)
    Robinson-Mitchell et al., Nutrition Reviews, 2025 · checked 2026-09-16 · independently re-checked
21. prot-t6 · Randomised controlled trial(s) · randomised controlled trial
    Daily dietary protein distribution at a fixed protein level does not appear to influence changes in body composition during weight loss in women of reproductive age.
    Hudson et al., Journal of Nutrition, 2024 · checked 2026-09-16 · independently re-checked
22. prot-e2-08 · Randomised controlled trial(s) · rapid systematic review of human intervention studies
    No conclusive results could be obtained about how heat treatment affects postprandial protein metabolism.
    Frontiers in Nutrition, 2021 · checked 2026-09-18 · independently re-checked
23. prot-e2-12 · Observational data · scoping review
    Approximately 58.8% of findings from 11 studies examining the association between high protein intake during breakfast and muscle mass indicated an increase in muscle mass among participants.
    Nutrition Reviews, 2025 · checked 2026-09-18 · independently re-checked
24. prot-e2-14 · Position of an expert body · agency systematic review, research gaps section
    However, the heterogeneity across these methods makes conducting a meta-analysis impossible.
    Evaluation of Dietary Protein and Amino Acid Requirements: A Systematic Review, Agency for Healthcare Research and Quality, 2024 · checked 2026-09-18 · independently re-checked
25. prot-t2 · Randomised controlled trial(s) · randomised controlled trial
    A healthy diet rich in protein increased eGFR. Whether long-term consumption of a high-protein diet leads to kidney disease is uncertain.
    Juraschek et al., American Journal of Kidney Diseases, 2013 · checked 2026-09-16 · independently re-checked
26. prot-r6-02 · Meta-analysis or systematic review · systematic review and meta-analysis of controlled trials
    30 studies including 2160 subjects met the objectives and were included in the meta-analyses. HP regimens resulted in a significantly more pronounced increase in glomerular filtration rate [MD: 7.18 ml/min/1.73 m2, 95% CI 4.45 to 9.91, p<0.001], serum urea [MD: 1.75 mmol/l, 95% CI 1.13 to 237, p<0.001], and urinary calcium excretion [MD: 25.43 mg/24h, 95% CI 13.62 to 37.24, p<0.001] when compared to the respective LP/NP protocol.
    Devries et al., PLoS One, 2014 · checked 2026-09-17 · independently re-checked
27. prot-e2-10 · Position of an expert body · agency systematic review, research gaps section
    Briefly, evidence on the UL for amino acids appears sparse but has been slowly growing over the past ~2 decades.
    Evaluation of Dietary Protein and Amino Acid Requirements: A Systematic Review, Agency for Healthcare Research and Quality, 2024 · checked 2026-09-18 · independently re-checked