What “protein quality” measures — and what it does not. Quality is two things multiplied: amino-acid composition (how well a food’s indispensable amino acids match human requirement — the limiting amino acid caps the score) times digestibility (how much is actually absorbed). It says nothing about how much protein a food carries, nor whether eating higher-quality protein changes any outcome a person cares about. It is a surrogate, and the section at the bottom holds that line.
DIAAS replaces PDCAAS, and the reason matters for plant proteins
The 2013 FAO Expert Consultation recommends a new metric:
«A new protein quality measure (digestible indispensable amino acid score; DIAAS) is recommended to replace PDCAAS. DIAAS is defined as: DIAAS % = 100 x [(mg of digestible dietary indispensable amino acid in 1 g of the dietary protein) / (mg of the same dietary indispensable amino acid in 1g of the reference protein)].» [EXTRACTED (FAO - Dietary Protein Quality Evaluation 2013) chunk 01]
Two DIAAS choices bear directly on plant-vs-animal appraisal:
- True ileal digestibility of each amino acid, measured at the end of the small intestine («preferably determined in humans, but if this is not possible, in growing pigs or in growing rats»), not a single faecal crude-protein figure. This resolves quality at the amino-acid level — where plant proteins are weakest (their limiting amino acid, not their average, sets the score).
- DIAAS is not truncated. PDCAAS capped any score above 100% back to 100%; DIAAS does not (whole milk powder scores 122, below). The FAO expert panel flagged the cost of truncation directly: «Additional benefit of proteins with higher scores in complementing less nutritious proteins is not captured» [EXTRACTED (FAO - Dietary Protein Quality Evaluation 2013) chunk 02]. An uncapped score is what makes complementation legible — a high-quality protein above 100% can lift a lysine-poor cereal, and truncation hid exactly that.
The magnitude — FAO’s own worked example ranks plant below animal
FAO’s Table 6 (illustrating regulatory protein-content claims) is the decision-relevant number set:
| Food | Protein (g/100 g) | DIAAS | FAO judged quality |
|---|---|---|---|
| Wheat | 11 | 40 | Low |
| Peas | 21 | 64 | Low |
| Whole milk powder | 28 | 122 | High |
[EXTRACTED (FAO - Dietary Protein Quality Evaluation 2013) chunk 02, Table 6]
Read it as the maintainer’s claim made concrete. Both plant foods are graded Low quality even though peas carry a respectable 21 g protein per 100 g: the score is capped by the limiting amino acid (cereals are lysine-limited; legumes are the near-mirror, relatively lysine-rich and sulphur-amino-acid-limited — which is why cereal+legume complementation works) and by lower digestibility. Milk powder clears both. FAO’s claim column makes the consequence explicit: wheat and peas are «Eligible for claim based on quantity» but «No, none» on quantity-and-quality; milk is eligible on both.
Does adequate protein from plants cost more mass and carbohydrate? — partly answered, partly a gap
The maintainer’s claim was “getting to adequate/optimal protein from unprocessed plant foods entails high carb load and high bulk.” Decompose it against what the wiki now holds:
- The quality discount is real and quantified (above): a gram of pea protein delivers fewer digestible indispensable amino acids than a gram of milk protein (DIAAS 64 vs 122), so more protein — hence more food — is needed for the same delivered amino acids. Directionally solid from FAO’s own table.
- “Adequate” and “optimal” are two targets on two curves — fix which one first. A population reference intake (~0.8 g/kg/d, set against deficiency) and a contested performance/anti-sarcopenia target (commonly cited ~1.6 g/kg/d for resistance training, Protein and Resistance Training for Muscle and Strength) are different questions; the bulk/carb concern binds far harder at the higher target. The telos is explicit that a recommended intake is set for deficiency coverage and must never be read as an optimum.
- The carbohydrate/bulk arithmetic itself is a
Ggap. Computing grams-of-food and the protein:carbohydrate ratio of a realistic plant food-set needs a food-composition database the wiki does not hold — FAO gives DIAAS and protein-per-100 g for three illustrative foods, not the composition tables. Do not fabricate the arithmetic.-> ingest-queue.md(a food-composition / FAO INFOODS-class source).
”Unprocessed” may be the binding constraint — not “plant”
FAO names the digestibility drag on plant proteins explicitly:
«Many foods, including novel protein sources, may contain high levels of known antinutritional factors, which may be naturally occurring (e.g. tannins, phytates, trypsin inhibitors, glucosinolates, isothiocyanates), formed during processing (e.g. D-amino acids, lysinoalanine)» [EXTRACTED (FAO - Dietary Protein Quality Evaluation 2013) chunk 02]
— and draws the line that scopes this page’s synthesis: «Many of these affect digestion and will be taken into account in the determination of true ileal amino acid digestibility but others, such as glucosinolates, isothiocyanates, etc., have more general metabolic effects» [EXTRACTED (FAO - Dietary Protein Quality Evaluation 2013) chunk 02]. The digestion-acting ones show up as a lower DIAAS (so the remedy is «proper processing to minimize their levels» [EXTRACTED (FAO - Dietary Protein Quality Evaluation 2013) chunk 02]); the goitrogenic ones act elsewhere and are not part of the protein story. Traditional processing — soaking, fermentation, sprouting, and the transformations that make tofu and tempeh — lowers antinutrients and raises digestibility, i.e. raises DIAAS. So the constraint the maintainer felt as “plant” may be substantially “unprocessed”: a hypothesis this page holds open, pointing at a different intervention (prepare differently) than the claim implied (eat animal protein) -> Antinutrients in Plant Foods.
Processing cuts both ways, though. FAO recommends that «for foods susceptible to damage from processing, ‘reactive’ rather than ‘total’ lysine» be used [EXTRACTED (FAO - Dietary Protein Quality Evaluation 2013) chunk 01] — heat/Maillard damage destroys available lysine. Processing that removes antinutrients can also degrade the limiting amino acid; the net effect on DIAAS is the object, not either half.
Synthesis — the antinutrient concern and the digestibility discount are one variable
A move present in neither source alone. FAO reaches plant-protein digestibility from the protein-quality side: phytate, tannins and trypsin inhibitors depress true ileal amino-acid digestibility, so they are the DIAAS discount. Petroski reaches the same compounds from the toxicity side, and shows they are largely defused by the same step:
«Phytates should not significantly impair mineral status when included as part of a diverse and balanced diet, especially if using traditional processing methods such as soaking, germinating, fermenting, and cooking.» [EXTRACTED (Petroski - Antinutrients Narrative Review 2020) section 6.6]
So the maintainer’s two separate questions — “is soaking worth it?” and “is plant protein
bioavailable?” — are one variable (antinutrient load) with one lever (traditional processing) seen
from two sides. The lever raises protein quality and lowers the toxicity concern at the same time.
[INFERRED (FAO - Dietary Protein Quality Evaluation 2013; Petroski - Antinutrients Narrative Review 2020) — each source states one side; the identity is the wiki's] -> Antinutrients in Plant Foods.
The caveat that keeps this from becoming a slogan: the identity holds for the digestibility-acting subset (phytate, tannins, trypsin inhibitors). Antinutrients acting elsewhere — oxalate on mineral absorption, goitrogens on the thyroid — are not part of the protein story and must not be swept in.
Decision — protein powders: whey/casein vs plant (a small lever, correctly ranked)
Applying the score to a common decision (Challenge #13): whey and casein out-rank most single plant-protein powders on exactly the determinants above — amino-acid composition (dairy is not limited in lysine; leucine-rich, and leucine triggers the saturable muscle-protein-synthesis response held on Protein and Resistance Training for Muscle and Strength) and digestibility. Dairy powders sit near milk’s DIAAS 122; most single plant proteins sit in the pea/wheat 40-64 range. The challenge’s exception is right and is the same mechanism: a carefully-blended plant isolate (legume-lysine + cereal-methionine, e.g. pea+rice; or soy isolate) uses complementation to close much of the gap — which is what the untruncated DIAAS was built to reward.
Two bounds keep this a small lever, not a headline:
- It is a surrogate advantage (next section). Whey’s edge is largest acutely and per gram; when total daily protein is matched, long-term muscle/strength gains equalize substantially — RT is the driver, and Morton’s supplemental effect is modest (+0.3 kg lean mass).
- It only bites below adequacy. For someone already hitting their protein target from a mixed diet (any dairy/egg/meat, or deliberate complementation), the limiting-amino-acid problem is solved at the diet level and the powder’s source barely matters -> Is the Food Category Doing Any Work.
The same lens on the whole-food animal->plant shift. Reducing red/processed meat (the contested decision on Should Adults Reduce Red and Processed Meat) removes a high-quality protein source, so the substitution’s protein quality is the layer-3 cost to weigh — trivially covered if dairy/eggs/fish remain or complementation is deliberate, more real for a strict plant shift. The three axes point different ways (quality favours animal/dairy; the cancer axis mildly favours cutting red/processed meat; the antinutrient worry is defused by preparation, Antinutrients in Plant Foods) — so the decision is multi-axis, not a single verdict, and the quality axis is the one this page supplies.
The surrogate line — do not let a high score become a recommendation
DIAAS is a surrogate: it predicts how well a protein could meet amino-acid demand. The wiki holds no evidence that choosing higher-DIAAS protein improves a patient-important outcome — muscle, function, or mortality — for a person already meeting their protein requirement from a mixed diet. A score is a target only if its transmission to a real outcome is itself evidenced, which is not shown here -> Surrogate Outcomes. For someone eating a varied diet with any animal protein or deliberate complementation, the limiting-amino-acid problem is largely solved at the diet level regardless of any single food’s score — which is Is the Food Category Doing Any Work applied to protein.
Provenance and limits
- FAO 2013 is guideline-grade (an FAO Expert Consultation), but it is a methods report — it standardises how quality is scored, not what eating high- vs low-DIAAS protein does to people.
- Single hard-outcome evidence: none here. Everything on this page is composition/digestibility
(surrogate).
confidence: low. - Open gaps routed: the food-composition arithmetic (carb/bulk)
-> ingest-queue.md; the requirement level itself (adequate-vs-optimal, WHO/FAO/UNU 2007) is staged, not held.
Self-critique [run 2026-07-29, before commit]
- Over-claim check: the synthesis is tagged
[INFERRED]and scoped to the phytate/tannin/trypsin subset, with oxalate/goitrogen explicitly excluded — guarding against “antinutrients = protein discount” as a blanket identity. Held. - Fabrication check: the carb/bulk arithmetic was NOT computed — flagged as a
Ggap needing a food-composition dataset. The only magnitudes asserted (DIAAS 40/64/122, protein 11/21/28 g/100 g) are FAO’s Table 6 verbatim. - Surrogate discipline: the page states twice that DIAAS moves no patient-important outcome on the evidence held. This is the load-bearing honesty and it runs against the intuitive “eat higher-quality protein” recommendation.
- Laundered independence: none claimed. FAO and Petroski approach from genuinely different sides (quality scoring vs toxicity review); the synthesis is marked as the wiki’s induction, not either source’s.
- Whey-vs-plant powder section (added Challenge #13, 2026-07-29): guarded inline against overstatement — the whey advantage is scoped to the surrogate (DIAAS/acute MPS), bounded by total-protein-matched outcome equalization and by the above-adequacy small-lever point, and the blend exception is attributed to the held complementation mechanism, not asserted anew. No patient-important outcome claimed for whey over plant.