Six rules about protein absorption, counted out

Counted by Bodi · Last counted 4 August 2026 · 6 sources

Every rule about protein comes with a study underneath it, a table of numbers inside that, and somewhere past the references a declaration of interests. We read all of it. What was measured is, with some regularity, not quite the rule it now supports. Four of the six declarations carry a company name, printed beside the question it belongs to.

See the price per 100 gram of protein

Can you only absorb 25 gram of protein at a time?

No. Out of 100 gram of protein, 53 gram reached the bloodstream as amino acids within twelve hours.¹ Out of 25 gram, only 16.¹ Amino acids are the parts protein breaks down into during digestion. On the larger serving, absorption was still going when the twelve hours ran out.¹

What this does not say: that 100 gram in one sitting is a sensible idea. The measurement is 36 healthy young men, once, after a session in the gym.¹

Where the 25 gram came from

The ceiling comes out of studies that measured something else. They measured how fast the body builds new muscle protein, over six hours at most, at servings no larger than 45 gram.¹ Above roughly 20 gram, that rate stopped climbing.¹ On that result rests the advice to keep every main meal to 20 or 25 gram.¹

A serving can fail to push that rate any higher and go on delivering amino acids for hours afterwards. Those are two different questions. The second one was measured later.

We could not read the two studies that put the ceiling there ourselves. Both sit behind a paywall and both returned an error at the publisher on 4 August 2026. What stands above is how the authors of source 1 summarise their own field.¹

The study that did follow absorption gave 36 men 0, 25 or 100 gram of milk protein after a training session.¹ The 100 gram was picked as the most the researchers thought a person could get down in one go.¹ Twelve men drank it.¹

The protein was labelled inside the cow itself, with an isotope, which is a version of an atom a laboratory can pick out again.¹ Every amino acid from that protein could therefore be traced.¹

Of the 25 gram, 51, 62 and 66 percent had reached the bloodstream at 4, 8 and 12 hours.¹ That series flattens. Of the 100 gram it was 26, 44 and 53 percent, and that series was still climbing at the end.¹

In muscle tissue, 18 percent of the 25 gram had been built in after twelve hours.¹ For the 100 gram it was 13 gram, or 13 percent, and it went in on a nearly straight line for all twelve hours.¹

One author on that study is employed by FrieslandCampina.¹ The declaration states the company had no role in the funding, the data, the analysis or the writing.¹

A second study put 20 gram of whey next to 40 gram in 30 trained men, each of whom did both servings.² On 40 gram, muscle protein was built about 20 percent faster: 0.059 against 0.049 percent per hour, with p at 0.005.² The lower that p, the smaller the chance that coincidence explains the gap.

That study was paid for by a grant from GlaxoSmithKline Consumer Healthcare, and one of the authors worked there.² The same paragraph states that none of the authors had a conflict of interest.²

The authors point at a difference with the two earlier studies that found nothing between 20 and 40 gram. Those trained the legs only, this one the whole body.² How much muscle you have just worked therefore helps decide what a serving can still be used for.²

Do you just burn off the rest?

Barely. Amino acid burning did go up, but it was negligible next to the rise in protein building across the whole body.¹ The researchers recalculated two earlier studies and arrived at less than 15 percent of the extra intake.¹

What this does not say: that more protein gets you more. This is one serving, followed for twelve hours. Not a day, not a week, and not an outcome anyone can spend.¹

What happened to the 100 gram

The whole idea of a ceiling rests on the surplus being burned off. That is a measurable claim, and it has now been measured.

Burning rose on 25 gram and on 100 gram. Set against the rise in building across the whole body, it vanished.¹

The authors then went back to two earlier studies that had put several servings side by side.¹ They expressed the burn figures from those as a share of the extra intake.¹ It came to less than 15 percent, leaving more than 85 percent going into tissue protein.¹

What this study does not do is issue a recommendation. It measures where the protein ends up, not what anyone has left after six months.¹

The authors call their own figures a lower bound, because several measurements were still running when the twelve hours ended.¹ They add that people who move less, or are in poorer health, may well hit a ceiling.¹

Do you have to eat within half an hour of training?

Not in these numbers. Across 23 studies with 525 participants, the advantage of protein near the session vanished once total daily intake was counted in.³ That daily total was the strongest predictor in the analysis.³

What this does not say: that no window exists. The authors write that the control groups varied too much to establish one, and they leave four to six hours open.³

What the correction did

Uncorrected, there was a small effect on muscle size: 0.24, interval 0.04 to 0.44, p at 0.02.³ On strength there was nothing.³

The full model corrected for training experience, blinding, sex, age, body mass, duration and total intake.³ What survived was 0.16, interval minus 0.07 to 0.38, p at 0.18. For strength, p came out at 0.49.³

A second model kept only the factors that still counted for something.³ There, total daily intake was the strongest predictor: 0.41, interval 0.14 to 0.69, p at 0.004.³

The same gap shows up in what the participants ate. In the studies without matched daily totals, controls averaged 1.33 gram of protein per kilo per day, the other group 1.66.³

Three studies did match that daily total. Two of them found no benefit from timing.³ Three is too few to say anything firm, and the authors write so themselves.³

The studies were of decent quality, averaging 8.7 on the PEDro scale.³ That scale is a fixed checklist for the design of a trial, and 5 was the cut-off for inclusion here.³

The analysis was supported by a grant from Dymatize Nutrition.³ The authors declare no competing interests.³

Should you spread protein across the day?

The Dutch Health Council sets no standard for spreading protein across the day. Its advice does not address the question, because there is as yet insufficient research on it.⁵

What this does not say: that distribution makes no difference. It says the question has not been researched, which is another thing entirely.⁵

What it actually says

The sentence reads, in translation: “This advice does not address the question of whether the distribution of protein consumption across the day matters in relation to the dietary reference value. There is as yet insufficient research on this.”⁵

The report fences itself off a second time: “The values are aimed at people with average physical activity. The protein requirement of people who exercise extremely may be higher, but this advice does not address that.”⁵

So the advice to take 20 to 25 gram per meal does not come from an authority. It rests on the assumption that a serving has a ceiling.¹ That assumption was measured in the 100 gram study, and it did not hold.¹

What has been written down about daily intake, and by whom, sits in the protein intake calculator. Ten published standards side by side, and they are not answering the same question.

Is fast absorbed better than slow?

Whey does reach the blood faster than casein. That speed produced no demonstrable advantage in this study. Whey gave a high short peak and casein a low long one.⁴ What arrived in the bloodstream from the food differed too little to tell apart from chance: 356 against 300 micromol per kilo.⁴ What was measured is leucine, one of the amino acids.⁴ The leucine eaten minus the share burned came out, across seven hours, at 141 micromol per kilo on casein and 11 on whey.⁴

What this does not say: buy casein. This is 16 people across five protocols, after a single meal with no carbohydrate, fat or training.⁴

The study that invented fast and slow

The whole distinction between fast and slow protein comes out of one study from 1997.⁴ There too the marker sat in the cow, so what had been eaten could be told apart from what the body made itself.⁴

After whey, total leucine appearance in the blood rose from 1.49 to 4.21 micromol per kilo per minute.⁴ That figure adds up the leucine from the food, from the body’s own protein and from the infusion.⁴

Six hours later that inflow was back at its starting value.⁴ Casein stayed low and flat, and was still above its starting value at six hours.⁴

That is the difference in speed. The 356 against 300 above covers only the leucine that came from the food.⁴ That gap was too small to tell apart from chance.⁴

Protein building across the whole body rose 68 percent on whey and 31 percent on casein.⁴ That gap was too small to tell apart from chance as well.⁴ What casein did and whey did not was hold down the breakdown of the body’s own protein.⁴

The leucine balance across seven hours was positive on casein at 141 micromol per kilo, against 11 on whey.⁴ That 11 could not be told apart from zero.⁴

So the slow protein came out ahead here. Fast, in this measurement, is not a synonym for better.⁴

One author on the study is attached to the Nestlé research centre, another to the dairy laboratory of the French institute INRA.⁴ The work was paid for by the French ministry of education and research, the Auvergne region and INRA.⁴

The authors add that the effect may come out smaller in a mixed meal.⁴ Fat and carbohydrate change how fast the stomach empties.⁴

Is the protein figure on the tub correct?

Nobody weighed the protein on that tub. Protein contains nitrogen, so a laboratory measures that instead and works backwards.⁶ That figure goes times 6.25, and the answer is printed as protein.⁶ The fixed factor rests on the assumption that protein is about 16 percent nitrogen by weight.⁶ The figure on the label can therefore be 15 to 20 percent away from the real content.⁶

What this does not say: that a manufacturer is doing something wrong. The factor is written into European labelling law. We have measured no tub ourselves.⁶

Why 6.25

The factor rests on two assumptions. First: protein is about 16 percent nitrogen by weight. Second: all the nitrogen in food comes from there.⁶

The first assumption is slightly wrong for every source. The nitrogen content of individual amino acids runs from 7.7 to 32.2 percent.⁶ The correct factor therefore differs per source: 5.70 for soya up to 6.32 for milk.⁶

European labelling law uses 6.25 for all of it anyway.⁶ Milk needs 6.32, so the figure on the label lands below the real content. Soya needs 5.70, so it lands above. Those two sums are ours and are not stated that way in the review.

Which is why this site applies no absorption correction to protein. Long before absorption enters the argument, the figure on the tub already carries a margin of 15 to 20 percent.⁶

That puts the rest of this page in proportion. Anyone weighing 20 against 25 gram per shake is working from a label figure with 15 to 20 percent of slack in it.⁶

What we do with it

We divide the price by the grams of protein on the label. That is the whole operation.

No protein source gets an absorption bonus or an absorption penalty from us. Not whey, not casein, not soya. In the six sources above there is no figure with which such a correction could be built.

For every tub in the table, that label figure is the same conversion from nitrogen. One rule for all of them keeps a comparison checkable. How the division runs from there is on how we count.

What is not here

Nothing about what protein does in a body. That is not what this site is for.
Not that 100 gram in one sitting is sensible. It was measured in 36 young men, once, and it is not advice.
Not that spreading protein across the day makes no difference. There is no standard, which is another statement.
Nothing about a brand or a particular tub. We have measured none of them.
Nothing about how much protein you need per day. That is in the protein intake calculator, with ten published standards next to each other.

What is here is one measurement per question, with the participants, the running time and the declaration at the bottom alongside it.

Sources

Open a source and you see exactly what was used from that study, which sentence on this page it supports, and what it explicitly does not support.

1Trommelen J, van Lieshout GAA, Nyakayiru J, Holwerda AM, Smeets JSJ, Hendriks FK, van Kranenburg JMX, Zorenc AH, Senden JM, Goessens JPB, Gijsen AP, van Loon LJC. The anabolic response to protein ingestion during recovery from exercise has no upper limit in magnitude and duration in vivo in humans. Cell Reports Medicine 2023;4(12):101324. DOI 10.1016/j.xcrm.2023.101324. Open access.randomised controlled trial with isotope tracers, three parallel arms · full text read via the PubMed Central copy (PMC10772463), including results, discussion, the limitations section and the declaration of interests · read 2026-08-04

Design: Randomised trial with a placebo arm, run at Maastricht University between June 2019 and March 2020, prospectively registered with the Netherlands Trial Register as NL7700. Milk protein was used in which the label had been built into the cow itself, so the amino acids from that protein can be followed wherever they end up. Four tracers at once, with blood and muscle sampling across a twelve-hour window after ingestion.
Studied in: 36 healthy young men, 12 per arm, after a single bout of whole-body resistance exercise.
Compared with: Three arms side by side: 0 gram protein (placebo), 25 gram protein and 100 gram protein, each as a single serving after training.
Dose and duration: A single dose of 0, 25 or 100 gram milk protein.
What was measured: The pre-registered primary outcome was the rate at which muscle protein was built over the full twelve hours. Alongside it: how much of the ingested protein reached the bloodstream as amino acids, how much was built into muscle tissue, how much was burned, and whole-body balance.
What was found: Of 100 gram of protein, 53 plus or minus 7 gram arrived in the bloodstream as amino acids within twelve hours, against 16 plus or minus 1 gram from 25 gram of protein. As a share of what was eaten: for 25 gram, 51, 62 and 66 percent at 4, 8 and 12 hours, so levelling off; for 100 gram, 26, 44 and 53 percent, so still climbing at twelve hours. Incorporation into muscle tissue was 12, 15 and 18 percent at 4, 8 and 12 hours for 25 gram; for 100 gram it was 13 gram, or 13 percent, and that incorporation ran almost linearly across the whole twelve hours. Muscle protein was built roughly 20 percent faster on 100 gram during the first four hours and roughly 40 percent faster over the eight hours after that, compared with 25 gram. Amino acid burning did rise, but it was negligible next to the rise in whole-body protein building; the authors say so in those terms. The authors also recalculated the burn figures from two earlier dose-response studies and arrived at less than 15 percent of the extra intake, leaving more than 85 percent going into tissue protein.
Supports on this page: That 100 gram of protein in one sitting delivers more amino acids into the bloodstream than 25 gram, and by how much: 53 against 16 gram in twelve hours. That absorption from 100 gram had not finished after twelve hours. That the assumption a per-serving ceiling rests on was measured here and did not hold. And that the fraction burned is small next to the fraction built in.
Explicitly does not support: 36 healthy young men, once, after a resistance training session, and the measurement stopped at twelve hours while several outcomes were still running; the authors call their own figures a lower bound because of it. They write that they do not know whether this extends to other groups, and that people with lower physical activity or poorer health may well run into a ceiling. The study covers a single serving and not a daily total, so it yields no figure for how much anyone needs per day. It says nothing about health, nothing about a supplement and nothing about a brand. The line in the introduction that current advice is to cap each main meal at 20 to 25 gram is these authors describing their own field, not a standard set by an authority; we could not read the two studies they cite for it.
Interests and funding: One author, G.A.A. van Lieshout, is employed by FrieslandCampina. The declaration states the company had no role in funding, data collection, analysis or preparation of the manuscript. For the two lead authors the declaration points to a public overview of career funding on the Maastricht University website; we have not inspected that overview.

https://doi.org/10.1016/j.xcrm.2023.101324

2Macnaughton LS, Wardle SL, Witard OC, McGlory C, Hamilton DL, Jeromson S, Lawrence CE, Wallis GA, Tipton KD. The response of muscle protein synthesis following whole-body resistance exercise is greater following 40 g than 20 g of ingested whey protein. Physiological Reports 2016;4(15):e12893. DOI 10.14814/phy2.12893. Open access.randomised crossover trial with isotope tracers and muscle biopsies · full text read via the PubMed Central copy (PMC4985555), including results, discussion, the funding statement and the conflict of interest statement · read 2026-08-04

Design: Crossover in random order: every participant did both servings. The rate at which muscle protein was built was measured with a labelled phenylalanine infusion and muscle biopsies, across the five hours after training.
Studied in: 30 resistance-trained young men, split into a group with less lean body mass (65 kilo or below, 15 men) and one with more (70 kilo or above, 15 men).
Compared with: 20 gram against 40 gram of whey protein isolate, within the same person, after a bout of whole-body resistance exercise.
Dose and duration: A single dose of 20 or 40 gram whey protein isolate.
What was measured: The rate at which muscle protein was built, in percent per hour, over the first five hours after training.
What was found: Across both groups combined: 0.059 plus or minus 0.020 percent per hour on 40 gram, against 0.049 plus or minus 0.020 percent per hour on 20 gram, at p = 0.005. That is roughly 20 percent. Lean body mass made no difference at these two doses: the men carrying more of it did not respond differently from the men carrying less. In the discussion the authors point out that the two earlier studies which found no difference between 20 and 40 gram trained the legs only, whereas this one trained the whole body, and that the amount of muscle worked is in their view the most likely explanation for the difference in outcome.
Supports on this page: That 40 gram of whey built muscle protein roughly 20 percent faster than 20 gram in this trial, and that the answer to how much protein a serving can use depends in part on how much muscle the training session worked.
Explicitly does not support: 30 trained young men, no women, and a five-hour measurement window. The study measures the rate at which muscle protein is built, not absorption in the gut, and it does not measure breakdown, so it says nothing about net balance. Two earlier studies in comparable men found no difference between the same two doses; we could not read those two ourselves. It offers no per-meal recommendation, no daily total, and it says nothing about health, about a supplement or about a brand. That 40 beats 20 also says nothing about where the curve does flatten: that was not measured here.
Interests and funding: The study was funded by a grant from GlaxoSmithKline Consumer Healthcare to two of the authors. One author, C.E. Lawrence, was an employee of GlaxoSmithKline Consumer Healthcare during the study. The same paragraph states that none of the authors had a conflict of interest.

https://doi.org/10.14814/phy2.12893

3Schoenfeld BJ, Aragon AA, Krieger JW. The effect of protein timing on muscle strength and hypertrophy: a meta-analysis. Journal of the International Society of Sports Nutrition 2013;10:53. DOI 10.1186/1550-2783-10-53. Open access.systematic review with meta-analysis and multi-level meta-regression · full text read via the PubMed Central copy (PMC3879660), including the table of included studies, the full results section, the limitations, the competing interests statement and the acknowledgement · read 2026-08-04

Design: Systematic search of PubMed and Google Scholar up to March 2013, with pre-set inclusion criteria and a quality assessment on the PEDro scale; only studies scoring 5 or higher were admitted, and the average score was 8.7. Then a meta-regression correcting per study for training status, blinding, sex, age, body mass, duration and total protein intake.
Studied in: 478 participants across 20 studies for strength, and 525 participants across 23 studies for size. Mostly people without training experience; only four studies used trained participants.
Compared with: Protein within an hour before or after training, against a control group given no protein within two hours either side. At least 6 gram of essential amino acids, and at least six weeks of resistance training.
Dose and duration: Varies by study. In the studies that did not match daily totals, controls averaged 1.33 gram protein per kilo per day and treatment groups 1.66. In the three studies that did match, it was 1.81 against 1.91.
What was measured: Effect size on maximal strength (1RM) and on muscle size, the latter as cross-sectional area or as fat-free mass.
What was found: Without correcting for confounders, timing produced a small effect on size, 0.24 plus or minus 0.10 with an interval of 0.04 to 0.44 at p = 0.02, and no effect on strength. In the full corrected model that effect vanished: 0.16 plus or minus 0.11 with an interval of minus 0.07 to 0.38 at p = 0.18 for size, and p = 0.49 for strength. In a second, reduced model only total protein intake, study duration and blinding survived as significant covariates; there, total daily protein intake was the strongest predictor of the effect on size: 0.41 plus or minus 0.14, interval 0.14 to 0.69, p = 0.004. That 0.41 therefore comes from the reduced model and not from the full model. Of the three studies that matched daily totals, two found no benefit from timing.
Supports on this page: That the measured advantage of protein taken close to training disappeared once total daily protein intake was accounted for, and that the daily total was the strongest predictor in this analysis.
Explicitly does not support: This does not disprove that a window exists. The authors write plainly that control groups differed a great deal between studies in when they got their protein, and that this makes it impossible to establish whether a window exists; they leave open that it may be four to six hours wide. Only three of the included studies matched daily totals, which is too few for a firm conclusion. Most participants had no training experience, and trained people may respond differently; four studies were not enough to say. It says nothing about health, nothing about a supplement and nothing about a brand.
Interests and funding: The authors declare no competing interests. The acknowledgement states the study was supported by a grant from Dymatize Nutrition of Dallas, a manufacturer of protein supplements.

https://doi.org/10.1186/1550-2783-10-53

4Boirie Y, Dangin M, Gachon P, Vasson MP, Maubois JL, Beaufrere B. Slow and fast dietary proteins differently modulate postprandial protein accretion. Proceedings of the National Academy of Sciences 1997;94(26):14930-14935. DOI 10.1073/pnas.94.26.14930. Open access.controlled feeding trial with isotope tracers, five sub-protocols · full text read via the PubMed Central copy (PMC25140), including the methods section, the tables, the full results section, the discussion and the acknowledgement with the funding · read 2026-08-04

Design: Milk protein with the label built into the cow, combined with an infusion of a second labelled amino acid, so the eaten protein can be followed separately from the body's own. Blood and breath sampled every 20 minutes for seven hours after a single meal, following a ten-hour fast. Five separate protocols with three to six participants each.
Studied in: 16 healthy young adults, average age 24.
Compared with: Whey against casein, as the standard example of a fast and a slow protein. The two meals were matched on leucine rather than on nitrogen; an extra protocol with three participants did it the other way round.
Dose and duration: A single 30 gram serving of whey against 43 gram of casein, both supplying 380 micromol of leucine per kilo of body weight. No carbohydrate, no fat, no exercise.
What was measured: The rate at which amino acids from the eaten protein appear in the blood, whole-body protein building and breakdown, leucine burning, and the balance across seven hours. The methods section defines that balance as the leucine eaten minus the integrated area under the curve of total leucine oxidation over 420 minutes, assuming all administered leucine was absorbed.
What was found: Whey produced a high, short peak: TOTAL leucine appearance in the blood (Total Leu Ra) rose from 1.49 to 4.21 micromol per kilo per minute and was back to baseline within six hours (360 min). NOTE, the first draft of the article got this wrong: 1.49 and 4.21 are NOT dietary leucine alone. The methods section defines Total Leu Ra in so many words as the sum of dietary leucine (Exo Leu Ra), leucine from the breakdown of the body's own protein (Endo Leu Ra) and the intravenously infused labelled leucine. Casein rose less (1.55 to 2.05 at 80 min) and was still above baseline at six hours (1.74), by which time whey had fallen back. Dietary leucine alone (Exo Leu Ra, area under the curve) did not differ significantly: 356 plus or minus 35 micromol per kilo for whey against 300 plus or minus 35 for casein. Whole-body protein building rose 68 percent on whey and 31 percent on casein (average across 40 to 140 min), a difference that was not significant. Casein suppressed the breakdown of the body's own protein and whey did not. The LEUCINE balance across seven hours (420 min) was positive on casein at 141 plus or minus 96 micromol per kilo, and indistinguishable from zero on whey at 11 plus or minus 36, at p below 0.05 between the two.
Supports on this page: That whey reaches the blood faster than casein, and that this is a difference in speed and not in quantity: dietary leucine delivery did not differ significantly. And that in this trial the seven-hour leucine balance came out better on the slow protein.
Explicitly does not support: 16 participants spread across five protocols, so three to six per measurement, which is small. A single meal with no carbohydrate, no fat and no training, in people who had fasted for ten hours; the authors write that the effect may be smaller in a mixed meal, because fat and carbohydrate change gastric emptying and the insulin response. It measures whole-body balance and not the muscle, and it runs seven hours rather than a day. It says nothing about how much protein anyone needs per day, nothing about training, nothing about health and nothing about a brand. No recommendation to pick casein over whey follows from this study.
Interests and funding: One author, M. Dangin, is attached alongside the university to Nestec Ltd, the Nestle research centre, and another to the dairy technology laboratory of the French agricultural research institute INRA. The study was funded by the French ministry of education and research, the Auvergne region and INRA. There is no separate conflict of interest statement; in 1997 this journal did not yet require one.

https://doi.org/10.1073/pnas.94.26.14930

5Health Council of the Netherlands (Gezondheidsraad). Voedingsnormen voor eiwitten: referentiewaarden voor de inname van eiwitten. The Hague: Gezondheidsraad, 2 March 2021, publication no. 2021/10. Open access, Dutch language.advisory report from a national scientific council · full text read, 26 pages, PDF fetched from gezondheidsraad.nl and extracted with pdftotext; both quoted passages were located verbatim in that text · read 2026-08-04

Design: Advisory report by a standing committee, written after reviewing the existing dietary reference values from EFSA, the WHO, the Nordic countries and the DACH countries, with two background documents including a systematic review.
Studied in: Healthy people in the Netherlands at a healthy weight with average physical activity.
Compared with: The EFSA values from 2012, the Nordic recommendation from 2012 and the DACH reference values from 2017.
Dose and duration: Not applicable. This report sets reference values for intake per day, not per meal.
What was measured: Reference values for protein intake, expressed in gram per kilo of body weight per day.
What was found: On distribution across the day the report states, in translation: this advice does not address the question of whether the distribution of protein consumption across the day matters in relation to the dietary reference value. There is as yet insufficient research on this. On exercise it states: the values are aimed at people with average physical activity. The protein requirement of people who exercise extremely may be higher, but this advice does not address that.
Supports on this page: That the Dutch advisory council sets no standard for spreading protein across the day, and why not: there is too little research. And that the values it does set apply to people with average physical activity.
Explicitly does not support: Having no standard is not the same as establishing that distribution does not matter: the report says the question has not been researched, not that it has been answered. The report does not cover athletes, people with obesity or frail older adults; it says so itself. It says nothing about a supplement, nothing about a brand and nothing about absorption in the gut.
Interests and funding: Independent scientific advisory body to the Dutch government. No commercial funding.

https://www.gezondheidsraad.nl/documenten/2021/03/02/voedingsnormen-voor-eiwitten

6Geirsdottir OG, Pajari AM. Protein: a scoping review for Nordic Nutrition Recommendations 2023. Food & Nutrition Research 2023;67:10261. DOI 10.29219/fnr.v67.10261. Open access.scoping review underpinning a regional dietary reference value · full text read via the PubMed Central copy (PMC10770649); the quoted passage on the conversion factor was located verbatim in that text · read 2026-08-04

Design: Scoping review summarising the scientific basis under the 2023 Nordic nutrition recommendations.
Studied in: The population of the Nordic and Baltic countries. Not applicable to the passage used here, which concerns the measurement method behind a nutrition table.
Compared with: Not applicable for the passage used.
Dose and duration: Not applicable for the passage used.
What was measured: For the passage used: the nitrogen-to-protein conversion factor, and the error that comes with it.
What was found: The protein content of food is not measured as protein but calculated from nitrogen content, by default with a factor of 6.25. That factor rests on the assumption that protein is about 16 percent nitrogen by weight and that all nitrogen in food comes from protein. The nitrogen content of individual amino acids ranges from 7.7 to 32.2 percent, so the correct factor differs per source: 5.70 for soya up to 6.32 for milk. The review states that multiplying by the default of 6.25 is imperfect and can lead to a 15 to 20 percent error in the actual protein content, and that European Union labelling legislation continues to use that default.
Supports on this page: That the protein figure on a label is a conversion from nitrogen and not a measurement of protein, that a laboratory measures the nitrogen because protein contains nitrogen, that the fixed factor 6.25 rests on the assumption that protein is about 16 percent nitrogen by weight, that European labelling legislation prescribes that factor, that the correct factor ranges from 5.70 to 6.32 depending on the source, and that the resulting error in protein content can be 15 to 20 percent.
Explicitly does not support: This is a statement about the measurement method behind nutrition figures in general, not about a specific product, tub or brand. It is not a claim that any manufacturer is doing something wrong: the factor 6.25 is prescribed. We have measured no product ourselves, and neither did this review. The review says nothing about absorption in the gut, nothing about a supplement and nothing about how much anyone needs.
Interests and funding: Scientific underpinning commissioned by the Nordic Council of Ministers. No commercial funding reported.

https://doi.org/10.29219/fnr.v67.10261