Proteins with different weak spots balance each other out. In a pig study, a mix of pea protein and whey protein reached a higher DIAAS than either ingredient alone. Grains and legumes go well together, but not every combination helps, and the ratio counts. Whether it has to be the same meal is an open question. A new study recommends it.
The principle: filling each other's gaps
Every protein is limited by its scarcest essential amino acid. In plants, according to Herreman et al. 2020, the pattern is fairly consistent: grains have little lysine but plenty of methionine and cysteine. Legumes have plenty of lysine but little methionine and cysteine. Eat both together and each source fills the other's gap.
The DIAAS values are not averaged. You add up how much of each digestible amino acid the mix contains, divide by the total amount of protein and check which amino acid is now the scarcest. That's how the FAO 2013 describes it in its report, and that's how the proteinscore calculator works too.
Measured: the mix beats both ingredients
Stroebinger et al. 2024 tested in six pigs whether this calculation holds up in practice. They fed pea protein isolate, whey protein concentrate and two mixes of the two, and measured the DIAAS.
Line = 100, fully usable from this point. Measured DIAAS, young child pattern (0.5–3 years). For adults: pea 71.3, mix 75:25 99.0, whey 119.8, mix 25:75 130.4. Data: Stroebinger et al. 2024, Table 7.
Two results stand out:
- The mix with 25 % pea was better than either component. Pea is limited by methionine and cysteine, whey by histidine. Together, neither gap carries much weight anymore.
- The prediction from the individual values was right. The calculated values were 82.9 and 104.2; the measured ones were 84.4 and 104.3. For the individual amino acids, the deviation averaged 1 to 2 %. The authors write that the prediction is therefore reliable.
According to the authors' calculation, any mix between 57 : 43 and 11 : 89 (pea : whey) reaches a DIAAS of at least 100 (young child pattern). The authors name one limitation themselves: they tested protein powders, not a real meal.
Which plant combinations work?
For purely plant-based mixes, there are mainly calculations. Herreman et al. 2020 used data on 12 plant proteins to simulate the best mixing ratio for each pair. The simulation was capped at a maximum of 100.
Line = 100, fully usable from this point. Simulated DIAAS, young child pattern (0.5–3 years), ratio based on protein, simulation maximum 100. Data: Herreman et al. 2020, Table 3 and Fig. 3.
The authors stress that not every legume and grain combination brings much. If lysine is only the second or third scarcest amino acid in a legume, the score barely rises. Their example is lupin. Faba bean with corn also only reaches 64 in their table.
None of the studies reviewed measured classics like rice and beans directly. But they can be calculated with the study values in our database (adult reference pattern):
| Combination | separately | together | |
|---|---|---|---|
| 200 g rice + 150 g kidney beans | 71 and 69 | 85 | open |
| 200 g rice + 150 g lentils | 71 and 88 | 98 | open |
| 250 g pasta + 120 g lentils | 54 and 88 | 91 | open |
| 150 g tofu + 200 g rice | 89 and 71 | 96 | open |
Does it have to be the same meal?
This is where opinions differ.
Over the day is enough. Hertzler et al. 2020 refer to an estimate by Young and Pellett from the 1990s. According to it, you don't need to eat complementary proteins in the same meal as long as the meals are about 3 hours apart.
Better in every meal. Soh et al. 2026 analyzed the food diaries of 193 vegans from New Zealand, meal by meal. Of just over 3,700 meals, only 822 reached a DIAAS of 75 or more. Around 78 % were below that. Lysine was the most common gap (in 57 % of meals), followed by methionine and cysteine (21 %).
What set the better meals apart:
- They contained legumes more often, in 70 % of cases compared with 40 % for the weaker meals.
- The ratio was about 2 parts legumes to 1 part grains.
- The weakest group consisted mainly of grains with few legumes and nuts (average DIAAS 48).
The authors write that a sufficient daily total can mask gaps in individual meals. How long the body can "save" amino acids from one meal for the next is not known, they say. They therefore recommend deliberately combining in every meal. One study they cite found no significant difference in muscle protein synthesis between complete and incomplete meals with the same amount of protein. After the incomplete meal, however, blood sugar and insulin rose more.
If you want to play it safe, combine in the same meal. It costs nothing, and it's common anyway with dishes like dal with rice, chili with beans or hummus with bread.
Limits of the studies
- The measurements come from pig studies with protein powders. For real meals made of grains and legumes, there are mainly calculations.
- Moughan & Lim 2024 consider interactions between foods in normal meals to be fairly small. Where strong effects are suspected, the meal should be measured as a whole.
- The analysis by Soh et al. is an observational study. It shows associations, not causes, and digestibility was only applied per food group.
- The powder study was funded through a consortium of food companies coordinated by the Global Dairy Platform.
Frequently asked questions
Do I have to combine proteins in the same meal?
The evidence is not clear-cut. An older estimate assumes that a gap of about 3 hours between meals is not a problem. A new analysis of vegan food diaries, on the other hand, recommends combining in every meal where possible, because the time window is unknown.
Which proteins complement each other well?
Grains and legumes complement each other, because grains have little lysine but plenty of methionine, and legumes the other way around. Calculated values are, for example, 85 for pea with wheat and 90 for soy with wheat. Not every mix helps: faba bean with corn only reaches 64.
Can you simply average DIAAS values?
No. You add up the digestible amino acids of all ingredients and calculate a new value from them. A pig study showed that this prediction is accurate to within about 1 to 2 %.
How does your meal score?
Open the calculatorSources
- Stroebinger, N., Hodgkinson, S. M. & Moughan, P. J. (2024): Pig true ileal digestible amino acid contents of individual ingredients are additive in mixed meals and allow accurate predictions of digestible indispensable amino acid score values. Journal of Nutrition 154, 3323–3330. doi:10.1016/j.tjnut.2024.09.020
- Herreman, L., Nommensen, P., Pennings, B. & Laus, M. C. (2020): Comprehensive overview of the quality of plant- and animal-sourced proteins based on the digestible indispensable amino acid score. Food Science & Nutrition 8, 5379–5391. doi:10.1002/fsn3.1809
- Soh, B. X. P., Vignes, M., Smith, N. W., von Hurst, P. R. & McNabb, W. C. (2026): Beyond daily totals: meal-level digestible indispensable amino acid score reveals how food groups shape protein quality in vegan diets. Frontiers in Nutrition 13, 1752697. doi:10.3389/fnut.2026.1752697
- Hertzler, S. R., Lieblein-Boff, J. C., Weiler, M. & Allgeier, C. (2020): Plant proteins: assessing their nutritional quality and effects on health and physical function. Nutrients 12, 3704. doi:10.3390/nu12123704
- Moughan, P. J. & Lim, W. X. J. (2024): Digestible indispensable amino acid score (DIAAS): 10 years on. Frontiers in Nutrition 11, 1389719. doi:10.3389/fnut.2024.1389719
- FAO (2013): Dietary protein quality evaluation in human nutrition. Report of an FAO Expert Consultation. FAO Food and Nutrition Paper 92, Rome. PDF
You can find values for individual foods and where they come from under Data & sources. How proteinscore calculates is explained under Method.