This is not medical advice. General information about food tracking, not a treatment plan. If you are under 18, pregnant or breastfeeding, have ever had an eating disorder or a difficult relationship with food, follow a diet that excludes whole food groups, or are managing a medical condition, talk to a doctor or a registered dietitian before you start counting calories or change your intake. The full list is here.

Open any macronutrient tracker and protein is one number. Ninety grams, say, against a target of a hundred and twenty. What the number does not tell you is that those ninety grams came from lentils and bread rather than from chicken and eggs, and that the two are not interchangeable in the way a single figure implies.

How much they differ is a genuinely contested question, and the honest answer is smaller than the internet suggests but not zero. Below is what the scoring systems say, what happened when researchers matched the grams and trained people for twelve weeks, and where the evidence stops.

If you are still deciding how much protein to aim at in the first place, that is a different question and it has its own post: why macro calculator apps disagree about your protein target. This one assumes you have a number and asks what the grams counting toward it are actually made of.

Your tracker counts grams, and grams are not the whole story

Protein is not a nutrient in the way that vitamin C is. It is a delivery vehicle for amino acids, some of which your body cannot make and has to take from food. The FAO calls those the indispensable amino acids, and both of the scoring methods below are built entirely around them: a food is useful in proportion to how well its mixture of them matches what you need.1

A food short in one of the nine limits what the rest can be used for. The shortfall is not made up by eating more of everything else in that food, which is why quality gets scored against the limiting amino acid rather than averaged.

Cereals tend to run short on lysine. Legumes tend to run short on the sulphur amino acids, methionine and cysteine. The two shortfalls are in different places, which is why the FAO's own worked example of a protein quality calculation is not a single food at all but a mixture of wheat, peas and milk powder.1

How protein quality is actually scored

Since 1989 the standard has been PDCAAS, the Protein Digestibility Corrected Amino Acid Score. You take the ratio of the first-limiting amino acid in a gram of the food's protein against a reference protein, then multiply by the protein's digestibility.1 One number, easy to compute, and by the time it came up for review the consultation described it as having been in use for some 20 years.

An FAO expert consultation, held in Auckland over three days in 2011 and reported in 2013, replaced it.1 The report is worth reading for how plainly it states the problems with the method it is retiring, and one of them matters here more than the others.

PDCAAS is truncated at 1.0. Anything scoring above the reference is written down as 1.0. The consultation called truncation "important and controversial" and gave the reason: truncation "removes any nutritional differences between high protein foods such as milk and soya, although actual concentrations of important dietary indispensable amino acids, which may be limiting in some diets, are higher in milk than in soya."1 So the familiar move of pointing at soy and milk landing on the same PDCAAS figure is pointing at a ceiling, not at a finding of equivalence. The consultation's own view is that the two are not equivalent and that the method was hiding it.

The replacement, DIAAS, does three things differently. It scores every indispensable amino acid and takes the lowest. It uses true ileal digestibility, measured at the end of the small intestine, per amino acid rather than for the protein as a whole. And it does not truncate above 100%, except when the subject is a mixed diet or a sole-source food.1

One detail in the recommendation is worth knowing before you treat any published score as a hard measurement. The digestibility figures should come from humans where possible, and where that is not possible, "in growing pigs or in growing rats in that order".1 The FAO's own example calculation for whole milk powder is footnoted as using human values predicted from pig data. There is a pig somewhere behind a good many of these numbers.

There is a second recommendation that will matter later: for foods liable to damage in processing, the report says to measure "reactive" rather than total lysine, because lysine that has gone through the Maillard reaction is still lysine by chemical assay and is no longer available to you.1

The same food scores 69 or 143 depending on who eats it

A league table of protein quality implies a food has one score. The FAO report's first worked example shows it does not.

Whole milk powder, scored with DIAAS against each of the three reference patterns, comes out at 69 for infants, 122 for children aged six months to three years, and 143 for older children, adolescents and adults.1 Same powder, same amino acids, three scores that differ by more than a factor of two.

The limiting amino acid moves too. For infants it is tryptophan; for the other two groups it is the sulphur amino acids.1

Nothing about the food changed. What changed is the requirement pattern it was measured against, and for adults that pattern is a good deal easier to satisfy than the breast-milk pattern used for infants. So a quality score describes a food and the person eating it together, and quoting one without saying which reference pattern produced it means very little.

Plant protein vs animal protein when the trials matched the grams

All of the above is about amino acid arithmetic. The question most people actually want answered is whether it shows up in a body over months, and there is a trial designed to answer exactly that.

Hevia-Larrain and colleagues took 19 habitual vegans and 19 omnivores, all young men, put them through the same supervised twice-weekly resistance training programme for 12 weeks, and adjusted both groups to 1.6 grams of protein per kilogram per day using supplements: soy protein isolate for the vegans, whey for the omnivores.2

Both groups gained. Leg lean mass rose 1.2 kg in the vegans and 1.2 kg in the omnivores. Rectus femoris cross-sectional area rose 1.0 and 0.9 square centimetres. Leg-press one-rep max rose 97 kg and 117 kg. There were no statistically significant differences between the groups on any measure.2

Read their conclusion with its conditions attached, because they wrote it carefully: protein source "does not affect resistance training-induced adaptations in untrained young men consuming adequate amounts of protein".2 Untrained, young, male, and eating enough. That is 38 people. The 1.6 g/kg they were held at is roughly where a meta-analysis of 49 resistance training trials put the point that extra protein stopped adding lean mass, though that breakpoint has a wide interval and the sibling post works through how wide.8 Either way, the trial tested whether source matters when quantity is already generous, and answered no.

A mechanism study points the same way from underneath. The usual explanation for plant protein's supposed disadvantage is leucine, which most plant proteins carry less of. Corn protein is the exception the authors went looking for: its leucine content is the highest among most plant proteins and exceeds the level in animal proteins such as whey. Pinckaers and colleagues gave 36 young men 30 g of corn protein, 30 g of milk protein, or a 30 g blend of 15 g corn plus 15 g milk, and measured myofibrillar protein synthesis over five hours. Corn and milk came out at 0.053 and 0.053 percent per hour, with no difference between them or between the blend and milk.4

If leucine is the mechanism, a plant protein carrying enough of it should behave like an animal protein, and in that study it did.

Where the meta-analysis disagrees, and by how much

I would be doing the thing this site exists to avoid if I stopped at the two studies that agree with each other.

Lim and colleagues screened 3,081 articles, selected 18 and meta-analysed 16, comparing animal against plant protein for lean mass and muscle strength. Their headline is that protein source did not affect changes in absolute lean mass or muscle strength. But they also found a favouring effect of animal protein on percent lean mass, and in adults under 50 an advantage on both measures: a weighted mean difference of 0.41 kg of absolute lean mass (95% CI 0.08 to 0.74) and 0.50 percentage points (95% CI 0.00 to 1.01).3

A few things about those numbers. The percent figure's confidence interval has a lower bound of exactly 0.00, which puts it on the boundary rather than clearly past it. The absolute figure, 0.41 kg over the length of a training study, is real but small next to the 1.2 kg both arms of the trial above put on in twelve weeks. And the authors note that total protein intakes in the pooled studies were generally above the recommended dietary allowance at baseline and at the end.3

Their own summary is that animal protein "tends to be more beneficial for lean mass than plant protein, especially in younger adults".3 I would quote it as written rather than round it off in either direction. There is a small edge in the pooled data and it is smaller than most of the arguments conducted about it.

What a macronutrient tracker can see, and what it cannot

None of this reaches the app on your phone, mine included.

No mainstream macronutrient tracker scores protein quality. There is no DIAAS field, no amino acid breakdown, and no distinction between a gram of protein from whey and a gram from wheat. Your tracker shows you grams because grams are what a food composition database holds, and for most foods a quality figure either does not exist or traces back, like the FAO's own milk powder example, to digestibility coefficients predicted from pig data.

That is a real limitation rather than a scandal, and it is worth being precise about its size. On the evidence above, the correction a quality score would apply to your number is small if you are eating enough protein in total and drawing it from more than one plant family. It is larger if you are near the bottom of your protein range and getting most of it from one source.

There is a second blind spot and it is bigger than the first. A review of vegan athletes concluded that current evidence does not support a vegan diet enhancing performance, adaptation or recovery, but equally suggests an athlete can follow one without detriment. The caveat they attach is that vegan diets eaten spontaneously "may induce suboptimal intakes of key nutrients, most notably quantity and/or quality of dietary protein and specific micronutrients", naming iron, calcium, vitamin B12 and vitamin D.6

MyPlate, the app I build, tracks calories, protein, carbohydrate and fat. It tracks no micronutrients at all, so it can tell you nothing about B12 or iron, and it has no food database and no barcode scanner either. If you have moved to a mostly plant-based diet, the things most likely to go wrong are the ones a macro tracker is structurally unable to show you.

What the evidence does not settle

I went looking for answers to the following and did not find them.

Almost everything above was measured in young men. The trial was 38 untrained young men, the corn protein study 36 young males, and the only age subgroup the meta-analysis reports an effect for is adults under 50.243 It does not report a corresponding figure for older adults, so I cannot tell you from it whether the edge grows, shrinks or reverses with age. That is the group where protein quality is most often argued to matter most, and it is the group this evidence says least about.

Nobody has tested the version of this question most readers actually have. The trial supplemented its vegans with soy protein isolate, which is a purified, high-quality, deliberately chosen product. That is not the same as a diet whose protein arrives as bread, rice, beans and peanut butter, in the ratios a person happens to eat them. The FAO's whole framework says the mixture is what determines the score, and no training trial I found has randomised people to a realistic mixed plant diet against a realistic omnivorous one without supplements on either side.

And the processing question is genuinely open in a direction that cuts against the usual story. Bandyopadhyay and colleagues measured the metabolic availability of lysine in seven healthy young Indian men and got 91.9% for spray-dried cow milk powder, 86.6% for a habitual cereal and legume vegetarian meal, and 69.9% for the same milk powder after heat treatment.5 Heat treatment cost the milk 22% of its available lysine and took it well below the vegetarian meal.

Seven men, one meal, one processing condition. It does not overturn anything. But it is a measured result in which how a food was handled mattered more than whether it came from an animal, and it is the reason the FAO asks for reactive lysine rather than total lysine in the first place.1

What to do if most of your protein comes from plants

The practical advice that follows from all of this is duller than the debate.

Hit your total first. The trial that found no difference by source held both groups at 1.6 g/kg,2 which is where a meta-analysis of 49 resistance training trials put the point that extra protein stopped adding lean mass,8 and the pooled studies in the animal-versus-plant meta-analysis were generally above the recommended dietary allowance throughout.3 Every null result here was obtained in people already eating plenty. Source is a refinement on top of an adequate total, not a substitute for one. If your tracker says you are at 0.9 g/kg, the amino acid profile of what you are eating is not your problem.

Spread it across families rather than chasing single foods. Cereals are short on lysine and legumes on the sulphur amino acids, so a day containing both covers both gaps, and this is what the FAO's mixed-diet worked example is demonstrating.1 The old advice about combining proteins within a single meal is stricter than anything in the report, which scores mixtures over a diet.

Consider whether the upper end of your range is the safer place to sit. A plant-heavy diet is the case where the pooled small edge and the incomplete profiles point the same way, and the cost of aiming higher within the 10 to 35% of energy that the Institute of Medicine's acceptable range allows is low.7

Then look at the things your tracker cannot show you. The vegan athlete review names vitamin B12, iron, calcium and vitamin D as the intakes a spontaneously adopted vegan diet is liable to leave short.6 None of those will ever appear in a macro tracker's protein number, and all of them are a better use of attention than the difference between soy and whey.

Common questions

Is plant protein as good as animal protein?

For building muscle, when the total is adequate and the sources are varied, the best trial says yes. Nineteen habitual vegans and nineteen omnivores trained twice a week for 12 weeks with both groups held at 1.6 grams of protein per kilogram per day, and gains in leg lean mass (1.2 kg in each group), muscle cross-sectional area and leg-press strength showed no significant difference by source.2 The authors limited that conclusion to untrained young men eating adequate protein, which is who they studied. A meta-analysis of 16 trials is slightly less generous: no difference in absolute lean mass or strength overall, but a small advantage to animal protein on percent lean mass, and in adults under 50 an advantage of 0.41 kg of lean mass with a confidence interval of 0.08 to 0.74.3

What is DIAAS, and why did it replace PDCAAS?

DIAAS is the Digestible Indispensable Amino Acid Score, recommended by an FAO expert consultation in 2013 to replace the PDCAAS method that had been standard since 1989.1 It scores every indispensable amino acid and takes the lowest, uses true ileal digestibility measured at the end of the small intestine for each amino acid separately, and does not truncate scores above 100% except for mixed diets and sole-source foods. Truncation is the main reason for the change: capping PDCAAS at 1.0 "removes any nutritional differences between high protein foods such as milk and soya", in the consultation's words, even though milk carries more of some indispensable amino acids.1 So the familiar claim that soy and milk both score a perfect 1.0 is a ceiling effect rather than a finding of equivalence.

Do I need to combine proteins at every meal?

No. The FAO framework scores a mixture, and its own worked examples cover mixed dishes and whole diets rather than requiring the pairing to happen within one sitting.1 What matters is that the gaps get covered, and they are predictable ones: cereals run short on lysine, legumes on the sulphur amino acids methionine and cysteine. A day that includes both covers both. The stricter version of this advice, that you must pair them in the same meal, asks for more than the scoring method it is usually justified by.

Does any macro tracking app show protein quality?

Not that I am aware of, including the one I build. Mainstream trackers show grams of protein because grams are what food composition databases hold; there is no DIAAS field, no amino acid breakdown, and no way to tell a gram from whey apart from a gram from wheat. If you want a macro tracking app to be useful on a plant-heavy diet, judge it on whether it makes hitting a total easy rather than on any quality feature, because none of them have one. It is worth knowing what else is missing: MyPlate tracks calories, protein, carbohydrate and fat and no micronutrients at all, so it can say nothing about the B12, iron, calcium and vitamin D that a review of vegan athletes flags as the likelier shortfalls.6

Should I eat more total protein if I am vegan or vegetarian?

Aiming at the upper part of your range is a reasonable hedge, and it is cheap. The pooled evidence gives animal protein a small edge on lean mass in younger adults, and plant sources are individually incomplete in ways that a narrow diet may not cover, so both point the same direction.3 The Institute of Medicine's acceptable range runs from 10 to 35% of energy, which leaves plenty of room to move up inside it.7 What the evidence does not support is a specific multiplier: I could not find a trial that established how much extra, if any, a plant-based eater needs, and the one that matched groups at 1.6 g/kg found they did not need any.2

Hit the total first, then worry about the source

The gap between plant and animal protein is real in the amino acid arithmetic and mostly disappears in the outcome trials, provided the total is adequate and the sources are not all from one family. That is a less satisfying answer than either side of the argument usually offers, and it is the one the studies support.

Which means the number your tracker already shows you is doing most of the work. Getting from 0.9 to 1.6 grams per kilogram is worth more than any rearrangement of where those grams came from, on the two figures this post has put beside each other: the breakpoint where extra protein stopped adding lean mass,8 against 0.41 kg for source in the pooled trials.3 Hitting a total is the thing a macronutrient tracker is actually built to help you do.

Then spend the attention you have left on the nutrients no macro tracker reports. That is where a plant-based diet is most likely to be genuinely short, and it is the part none of this arithmetic touches.6

If reading this has you wanting to audit every meal you ate last week for its amino acid profile, that is the point to stop rather than go further. Tracking that has become something to get right instead of something that helps is worth raising with a doctor or a registered dietitian.

Sources

Every figure above traces to one of these. If you find a number that doesn't match the source it claims, tell me and I'll correct it.

  1. Food and Agriculture Organization of the United Nations (2013). Dietary protein quality evaluation in human nutrition: report of an FAO Expert Consultation. FAO Food and Nutrition Paper, 92. Read the FAO report →
  2. Hevia-Larrain V, Gualano B, Longobardi I, et al. (2021). High-protein plant-based diet versus a protein-matched omnivorous diet to support resistance training adaptations: a comparison between habitual vegans and omnivores. Sports Medicine, 51(6):1317-1330. Read on PubMed →
  3. Lim MT, Pan BJ, Toh DWK, Sutanto CN, Kim JE (2021). Animal protein versus plant protein in supporting lean mass and muscle strength: a systematic review and meta-analysis of randomized controlled trials. Nutrients, 13(2):661. Read on PubMed →
  4. Pinckaers PJM, Weijzen MEG, Houben LHP, et al. (2024). The muscle protein synthetic response following corn protein ingestion does not differ from milk protein in healthy, young adults. Amino Acids, 56(1):8. Read on PubMed →
  5. Bandyopadhyay S, Kuriyan R, Shivakumar N, et al. (2020). Metabolic availability of lysine in milk and a vegetarian cereal-legume meal determined by the indicator amino acid oxidation method in Indian men. Journal of Nutrition, 150(10):2748-2754. Read on PubMed →
  6. West S, Monteyne AJ, van der Heijden I, Stephens FB, Wall BT (2023). Nutritional considerations for the vegan athlete. Advances in Nutrition, 14(4):774-795. Read on PubMed →
  7. Institute of Medicine (2005). Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. The National Academies Press, Washington DC. Read the report →
  8. Morton RW, Murphy KT, McKellar SR, et al. (2018). A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength in healthy adults. British Journal of Sports Medicine, 52(6):376-384. Read on PubMed →

The formulas and limits behind MyPlate's own numbers, with their citations, are on the health sources page.