Measuring protein quality from plant-based foods

Concerns over climate change highlight current food systems and agriculture as major contributors to greenhouse gas emissions⁽²⁷⁾, with animal production generating significantly more emissions than plant production. This has led to recommendations for reducing the proportion of dietary protein derived from ASF, in particular. While there are various alternatives to animal protein, plant protein production already plays a significant role in agriculture. Compared to animal production, crop production generates far fewer greenhouse gases, uses less land, requires fewer agrochemical inputs and consumes less water^{(Reference Bryant28)}. Additionally, nitrogen-fixing legume crops, which are excellent sources of plant protein, have the unique ability to reduce the need for nitrogen-based fertilisers, a significant driver of emissions^{(Reference Guo, Yang and Yu29)}. Yet, plant proteins are often viewed less favourably than animal proteins as they typically lack sufficient quantities of one or more of the nine indispensable amino acids (IAA) required for bodily function and growth^{(Reference Hertzler, Lieblein-Boff and Weiler30)}. However, most plant proteins are rarely consumed in isolation and are eaten in diverse mixed meals, where the amino acid deficiencies of one plant protein can be complemented by another. This raises protein quality by improving the amino acid score of the meal.

The importance of plant-based and novel foods (e.g. insects, algae, mycoproteins, etc.) lies in their potential to reduce environmental impact and improve human health. As scrutiny grows over the environmental costs of producing animal protein, there is an increased awareness of the lower environmental impacts of plant proteins. This has led to recommendations that plant proteins should make up a larger portion of total protein intake. With the shift towards reducing animal protein consumption – with such proteins described as being highly digestible and bioavailable – there is a need to understand the digestion of protein and the availability of amino acids from plant-based and novel foods to support their adoption as part of a sustainable diet.

Protein digestion is a complex process involving a cascade of enzymatic reactions and motility control in the gastrointestinal tract. Given this complexity, the FAO has recommended that protein digestibility should be assessed in vivo ^{(Reference Gilani, Tomé and Moughan31)}. With time, extensive studies have been conducted in adults using intrinsic labelling of foods with ¹⁵N and collection of ileal samples, where the disappearance of amino acids at the terminal ileum serves as a measure of protein digestibility^{(Reference Mariotti, Mahe and Benamouzig32–Reference Oberli, Marsset-Baglieri and Airinei34)}. Stable isotope tracers are ideal for studying the rate and extent of absorption and metabolism in vivo, as there is no harm in their use for human studies. Some plant-based foods (e.g. legumes, cereals) have been grown in the presence of modest quantities of ¹⁵N, which labels the plant and its proteins uniformly. The disappearance of the labelled IAA from the intestinal lumen provides an accurate measure of digestion and absorption, unaffected by endogenous protein digestion or unlabelled proteins in a meal. However, the use of this method requires an invasive intubation procedure to collect ileal samples and is, therefore, unsuitable for children, older adults and those who do not wish to volunteer for such a procedure.

A dual tracer stable isotope technique was developed to eliminate the need for intubation by measuring protein digestibility through the appearance of IAA from a test protein in plasma, compared to the IAA appearance from a standard protein with a known digestibility^{(Reference Devi, Varkey and Sheshshayee35)}. The standard protein is also intrinsically labelled but with a different isotope. Spirulina whole cells, a highly digestible, commercially available ¹³C-labelled source of single-cell protein, are commonly used as the standard protein, but other standard proteins are also possible. Test proteins can be of plant, microbial or animal origin. Plant-derived test proteins are produced during plant growth, often under controlled environments, with the use of modest amounts of ²H₂O or ¹⁵N-labelled fertilisers added to the soil to label IAA during seed development^{(Reference Kashyap, Bellam and Preston36)}. To avoid interference from transamination affecting the appearance of ¹⁵N-IAA in plasma, the method uses ¹³C for the standard protein and ²H for the test protein. As with ¹⁵N, such isotopes are also stable and are safe to use in human subjects. Both the test and standard proteins are incorporated into a standardised meal, with postprandial blood samples collected at multiple time points to quantify the isotopic ratios for each IAA. When the meal is taken in small, frequent boluses, the steady-state ratio of IAA enrichment of the test protein in plasma is compared to that of the standard protein using mass spectrometry^{(Reference Devi, Varkey and Sheshshayee35,Reference Kashyap, Bellam and Preston36)} . A plateau feeding protocol was developed to minimise the number of blood samples required^{(Reference Devi, Varkey and Sheshshayee35)}.

The use of stable isotopes reduces the burden on study participants compared with invasive intubation methods but introduces greater complexity and costs to the laboratory phase. Measuring the enrichment of two isotopes currently requires separate mass spectrometers and trained personnel for sample collection, preparation and analysis. In addition, producing isotopically labelled proteins is an expensive undertaking, which may limit the accessibility of this method. Yet, the method itself has proven to be manageable in different countries^{(Reference Kashyap, Shivakumar and Varkey37–Reference Thomas, Riley and Devi39)} and has even been used in studies involving young children^{(Reference Devi, Varkey and Dharmar40,Reference Shivakumar, Kashyap and Kishore41)} . Several crop varieties, including chickpea, mung bean, faba bean, yellow pea, kidney bean, pinto bean and rice, have been studied, consistently showing that plant-based foods may have lower protein digestibility than ASF such as egg, milk and meat. This lower digestibility is often attributed to factors such as reduced bioaccessibility, the presence of anti-nutritional compounds or amino acid modification during food processing. However, processing methods – such as de-hulling – can improve the digestibility of some plant-based foods^{(Reference Kashyap, Varkey and Shivakumar42)}. Additionally, food preparation techniques – including extrusion – have been shown to enhance protein digestibility in plant-based foods comparable to that of ASF^{(Reference Devi, Varkey and Dharmar40)}.

There are numerous opportunities for the use of this stable isotope method. One key gap is the lack of information on protein digestibility from a wider range of economically significant crops, with a broader geographic representation. For instance, data on the protein digestibility of common leguminous crops from Africa are significantly underrepresented in existing studies. Expanding the diversity of crops studied could provide a more comprehensive understanding of not only protein digestibility but also of crops in different geographic areas, which could have significant environmental and economic potential as part of local sustainable food systems. As protein digestibility from plant-based foods is generally lower than ASF, it may be useful to explore the utility of additional processing and preparation methods to enhance the potential of plant-based alternatives^{(Reference Day, Cakebread and Loveday43)}. Another area worth exploring is whether the bioavailability of other nutrients could be assessed within the same experiment. When crops have been intrinsically labelled with ²H₂O (or ¹³CO₂) as they grow, all organic molecules become labelled. As such, organic nutrients other than IAA, such as essential fats and vitamins, also become labelled. It may be beneficial to measure the availability of these other nutrients in conjunction with protein digestibility. Given the considerable effort and expense involved in conducting intrinsic labelling studies, measuring the bioavailability of multiple nutrients in a single experiment could offer value in understanding both the nutrient composition and overall bioavailability of plant-based foods and how essential nutrient supply may meet demand.

Taken together, the dual tracer stable isotope technique for measuring protein digestibility is a safe, minimally invasive method that is able to quantify the digestibility of IAA from plant-based and novel food sources in humans. This approach is suitable for use in vulnerable populations, including children, older adults and individuals with chronic diseases, to assess protein digestibility throughout the life cycle and under various health conditions. Additionally, this method has the potential to fill knowledge gaps with respect to understanding the protein digestibility from a diverse variety of crops grown globally as well as novel sources (including meat alternatives) and how various processing and preparation techniques can impact the protein digestibility of plants-based foods, all of which could inform how the diet may supply daily protein recommendations.

Understanding iron absorption and loss from whole diets and interventions

Micronutrient deficiencies are often the result of inadequate intakes or excessive losses of vitamins or minerals^{(Reference Gregory, Wahbi and Adu-Gyamfi44)}. They are also caused by malabsorption, which is commonly related to the presence of infection, inflammation or chronic disease^{(Reference Bhaskaram45–Reference Saboor, Zehra and Qamar47)}. Termed ‘hidden hunger,’ the occurrence of micronutrient deficiencies is typically less apparent than energy or protein malnutrition^{(Reference Gregory, Wahbi and Adu-Gyamfi44)}. Yet, it is projected that over 2 billion people experience some form of micronutrient deficiency, far outweighing those affected by macronutrient malnutrition^{(Reference Lowe48)}. Micronutrient deficiencies can result in both minor and severe conditions, including the aggravation of diseases and infections, blindness, cognitive impairment, diminished growth and a general loss in human potential^{(Reference Ritchie and Roser49,Reference Stevens, Beal and Mbuya50)} .

Fe deficiency is reported to be the most common micronutrient deficiency^{(Reference Han, Ding and Lu51)} and is closely related to the development of anaemia, where an estimated 50 % of all anaemia cases are caused by Fe deficiency, with certain variations between regions⁽⁵²⁾. Fe deficiency, even without anaemia, manifests in a variety of symptoms including fatigue, lethargy, reduced concentration, dizziness, tinnitus, pallor and headache^{(Reference Pasricha, Tye-Din and Muckenthaler53)}. Even though it is acknowledged that a large proportion of anaemia is amenable to Fe supplementation, and thus due to Fe deficiency, population studies measuring Fe status biomarkers beyond haemoglobin are scarce. To a large extent, this is due to the complexity in interpreting those biomarkers, especially in conditions of clinical or sub-clinical inflammation^{(Reference Pasricha, Tye-Din and Muckenthaler53)}.

Conventional Fe status biomarkers – such as serum ferritin and soluble transferrin receptor – reflect Fe metabolism rather than Fe balance and thereby only indirectly inform Fe status. Furthermore, such biomarkers can be influenced by physiological factors, specifically infection and inflammation^{(Reference Lynch, Pfeiffer and Georgieff54)}. The Fe isotope dilution technique was therefore developed to directly determine Fe turnover, Fe balance, as well as the response to Fe interventions^{(Reference Speich, Mitchikpè and Cercamondi55–Reference Cai, Ren and Zhang58)}. This methodology was based on earlier radioisotope studies, which applied a similar concept^{(Reference Green, Charlton and Seftel59,Reference Hunt, Zito and Johnson60)} .

The Fe isotope dilution technique relies on the fact that after equilibration, the concentration of the isotopic tracer in all body compartments is equal^{(Reference Charlton, Fatti and Lynch61)}. After tracer administration, a first enrichment is seen in red blood cells, which is then followed by a re-distribution of the tracer in the body after red blood cell senescence. It has been shown that equilibration is reached after approximately 12 months^{(Reference Green, Charlton and Seftel59)}, which can make this approach time-intensive, as it requires prolonged follow-up to ensure full equilibration as well as retention of the labelled Fe cohort. At this point, any Fe lost has the same isotopic composition as the total body Fe and will therefore not lead to a change in isotopic ratios. Thus, any change in isotopic tracer ratios or concentration can occur only by the addition of Fe with a natural isotopic composition in the form of Fe absorbed from the diet or supplements. In other words, a decrease in the concentration of the Fe isotope tracer in circulation is proportional to the amount of absorbed Fe. At the same time, a decrease in the absolute quantity of Fe isotope tracer can only occur if Fe, and thereby the tracer, is lost from the body. This concept can then be used to determine Fe balance, Fe absorption over time from whole diets, Fe losses, as well as the response to Fe interventions. Furthermore, the data can be used to accurately determine Fe requirements in different life stages.

This method has already been successfully applied in several studies to address Fe requirements across different population groups. In an early study by Fomon et al. (2003), Fe requirements were determined to be 1·46 mg/d for males and 1·15 mg/d for females, based on two measurements taken at around 13 and 17 years of age. Isotopes were administered at around 10 years of age as part of an Fe absorption study^{(Reference Fomon, Drulis and Nelson57)}. In a more recent study, Cai et al. (2018) determined the physiological Fe requirements of Chinese adults^{(Reference Cai, Ren and Zhang58)}. In this study, participants were given an intravenous dose of ⁵⁸Fe and were followed for 1155 days. Participants were found to be fully equilibrated after 425 days, and the four blood samples collected thereafter, days 515, 605, 767, and 1155, were used to determine requirements. The estimated daily Fe requirements in this population were found to be 0·96 mg/d for males and 1·10 mg/d for females^{(Reference Cai, Ren and Zhang58)}. More recently, the method was used to determine Fe requirements in African children^{(Reference Speich, Brittenham and Cercamondi56)}. Children, recruited at an average age of 6·3 years after having participated in Fe absorption studies 5 years earlier, were followed for 2 years with blood samples taken every 3 months. The study confirmed that the WHO/FAO recommended Fe requirement of 32 µg/kg of body weight/day was adequate for Fe-sufficient children, but Fe-deficient children would need higher amounts to correct their deficit^{(Reference Speich, Brittenham and Cercamondi56)}.

The isotope dilution method has also been applied in determining Fe balance and the amount of Fe absorbed from whole diets or supplements. In a study by Speich et al. (2021), Fe balance was assessed in women from Switzerland and Benin before, during and after an Fe intervention, with each period lasting 3 to 4 months. The study revealed differences between the two population groups in terms of both Fe balance and the bioavailability of dietary Fe^{(Reference Speich, Mitchikpè and Cercamondi55)}.

The long equilibration period limits the use of this method for infants under 1 year of age unless they are already born in equilibrium. This was demonstrated in a recent study in which mothers were labelled with stable Fe isotopes during pregnancy, thereby resulting in their infants being born equilibrated^{(Reference Stoffel, Cepeda-López and Zeder62)}. By analysing blood samples collected at 2 days, 3 months and 6 months after birth, it was possible to calculate the amount of Fe absorbed and gained during this period based on feeding practices (breastfed, formula fed or mixed feeding). The data also allowed the determination of Fe bioavailability from the three different diets^{(Reference Stoffel, Cepeda-López and Zeder62)}.

The direct approach of measuring Fe absorbed, Fe lost, and thereby Fe gained, rather than using surrogate measures of Fe metabolism, opens doors for further applications in the future. One area of growing interest is the calculation of Fe absorption from whole diets. Traditional Fe absorption studies using stable isotopes typically focus on individual meals or specific Fe compounds, which do not reflect long-term Fe absorption from a certain diet. Other methods to estimate bioavailability from whole diets, such as algorithms based on intake data, have also faced challenges in practical application^{(Reference Lynch, Pfeiffer and Georgieff54,Reference Collings, Harvey and Hooper63)} . By directly determining Fe absorption and losses over extended periods of time, the isotope dilution technique can overcome these issues and allow for a more accurate estimate of Fe absorption from a given diet.

Table 1. Overview of the stable isotope techniques used to support sustainable food systems

²H₂O, deuterated water or deuterium oxide; IAA, indispensable amino acid; ICP-MS, inductively coupled plasma mass spectrometry; TIMS, thermal ionisation mass spectrometry; FTIR, Fourier-transform infrared spectroscopy.

With the aim to better understand the nutritional quality of crops and the nutrients they provide for human health, another approach could involve measuring total Fe bioavailability (e.g. Fe absorbed, lost and gained) after consuming crops designed to be nutritionally adequate, through biofortification or other methods. This approach would offer valuable insights as to how much Fe is absorbed, lost and retained by the body from crops cultivated either for higher yields or for enhanced nutritional quality. It could also increase our understanding of how efforts to improve food systems, such as through biofortification to increase the nutritional adequacy of crops, can impact Fe status and overall human health.

Exploring the first food system: human milk intake

The first food system is broadly defined as the system that provides foods for children aged 0–36 months^{(Reference Baker, Santos and Neves15)}. As the ‘first food’ for children, breast milk is a key component within this system. Breast milk provides optimal levels of nutrients and energy personalised to the needs of the infant by the mother–infant dyad^{(Reference Victora, Bahl and Barros64)}. For the first 6 months of life, breast milk alone suffices to meet the nutritional needs for growth and development, and it is a recommended staple food until 2 years of age^{(Reference Victora, Bahl and Barros64,65)} . The health benefits of breast milk and breastfeeding are clear for infants as well as mothers^{(Reference Rollins, Bhandari and Hajeebhoy17,Reference Victora, Bahl and Barros64,Reference Horta, Rollins and Dias66)} .

However, the reality is that it is a global challenge to optimise breastfeeding rates^{(Reference Pérez-Escamilla, Tomori and Hernández-Cordero19,Reference Victora, Bahl and Barros64,Reference Neves, Barros and Baker67,Reference Sarki, Parlesak and Robertson68)} . Infant and young child feeding behaviours are shaped by numerous interlinked individual, political, social, cultural and economic determinants^{(Reference Pérez-Escamilla, Tomori and Hernández-Cordero19)}. Although the influence or importance of these factors may vary among populations, breastfeeding is at risk in many countries due to the aggressive and inappropriate marketing of CMF^{(Reference Baker, Santos and Neves15)}. The CMF industry undermines breastfeeding through inappropriate marketing that influences food environments and consumer behaviours^{(Reference Rollins, Piwoz and Baker69)}.

The threat CMF pose to breastfeeding is related to wealth^{(Reference Baker, Santos and Neves15,Reference Victora, Bahl and Barros64,Reference Neves, Gatica-Domínguez and Rollins70)} . Country-level data show that CMF sales follow a gradient across income levels, with sales being highest in high-income countries and lowest in low-income countries^{(Reference Baker, Santos and Neves15)}. At the individual level, wealth (i.e. socio-economic status) is also related to breastfeeding, although the relationship varies depending on the context^{(Reference Victora, Bahl and Barros64,Reference Neves, Barros and Baker67,Reference Sarki, Parlesak and Robertson68,Reference Neves, Gatica-Domínguez and Rollins70)} . In high-income countries, breastfeeding is more common among those of higher socio-economic status, whereas the opposite is reported for low-income countries. The positive association of breastfeeding and socio-economic status in high-income countries may be indicative of a ‘new’ phase of returning to breastfeeding. At the same time, the rapidly growing CMF sales in middle-income countries suggest that these regions are in the middle of a dietary transition towards CMF, and it can be expected that low-income countries will follow soon if no action is taken.

It is therefore important to accurately monitor breast milk intake across countries and population groups to support breast milk intake as part of a sustainable food system. The deuterium oxide dose-to-mother technique – developed by W. A. Coward and colleagues in the UK in the early 1980s – is a precise, objective and safe tool to measure the average daily volume of breast milk consumed by infants⁽⁷¹⁾. It is an alternative to test weighing, where the infant’s weight difference before and after every breastfeed amounts to the volume consumed, a method that is imprecise, time-consuming and disturbs the normal feeding behaviours of mothers and infants^{(71,Reference Savenije and Brand72)} . The dose-to-mother technique requires the mother to consume a single oral dose of deuterated water, after which it mixes with her body water and a portion is transferred to the infant via breast milk. Deuterium oxide appears in the saliva or urine of mothers and infants and is repeatedly sampled over a period of 2 weeks, a sampling scheme and duration that could impact participant retention. Deuterium enrichment is then analysed using FTIR or mass spectrometry, enabling the calculation of water intake as human milk as well as water from sources other than milk. Besides the knowledge on the volumes consumed, information on both sources of intake additionally indicates whether breastfeeding was exclusive⁽⁷¹⁾. Such information cannot be captured using test weighing and is commonly measured through maternal recall, which has been shown to overestimate exclusive breastfeeding rates^{(Reference Winichagoon, Pongcharoen and Fadjarwati73)}. The dose-to-mother technique is therefore also a useful tool to evaluate policies and interventions aiming to protect, promote and support exclusive breastfeeding.

Diets high in CMF are not only harmful to infant and maternal health but also planetary health. CMF directly contributes to climate change through the resources needed for production, packaging, distribution and preparation, and the greenhouse gases emitted through these processes^{(Reference Karlsson, Garnett and Rollins74,Reference Smith75)} . This is in stark contrast with the direct mother-to-child food chain, which only requires a marginal increase in energy intake for mothers (500 kcal/day when breastfeeding from birth until 6 months)⁽⁷⁶⁾. Sustainability is therefore yet another argument in favour of breastfeeding and monitoring breast milk intake that deserves more recognition.

As of now, the first-food system is not included in the Conceptual Framework of Food Systems for Diets and Nutrition used in the United Nation’s Food Systems Summit^{(77,Reference Oot, Mason and Lapping78)} . It is important to acknowledge the significance of the first-food system by including it within the conceptual framework. A better understanding of breast milk intake, through the use of the dose-to-mother stable isotope technique, will be key to successfully attaining the Summit’s objectives in alignment with the sustainable development goals.

Practical considerations and applications

A key consideration when using stable isotope techniques to inform sustainable food systems is the practical challenges associated with their implementation. Notably, all three of these techniques are more expensive than conventional methods and require instrumentation that is not widely available across laboratories globally and demand skilled personnel to perform the measurements. However, they enable the generation of accurate data that would be difficult, if not highly unlikely, to obtain in humans using other methods. Owino et al. ^{(Reference Owino, Slater and Loechl79)} emphasised the need to make stable isotope techniques more affordable, an aspect that remains crucial for their broader implementation. In addition, scaling up these methods in resource-limited settings is indeed difficult, but using the techniques to validate more field-friendly methods before their implementation is one way to scale up the assessment.

Despite these challenges, the evidence generated by these techniques has played a significant role in shaping practical applications which support sustainable food systems. For instance, the deuterium oxide dose-to-mother technique has provided valuable data verifying that exclusive breastfeeding for the first 6 months of life ensures infants receive sufficient energy for normal growth, thus supporting the WHO/UNICEF recommendation^{(Reference Agne-Djigo, Kwadjode and Idohou-Dossou80,Reference Nielsen, Reilly and Fewtrell81)} . Moreover, a study using the dose-to-mother technique indicated that mothers living in food-insecure households often have limited breast milk output, which in turn leads to reduced caloric and micronutrient intake for their infants^{(Reference Miller, Young and Boateng82)}. Such findings highlight the need for improved screening measures to identify at-risk mothers and provide targeted support to ensure both optimal and sustainable infant feeding practices.

In response to the growing concern about the environmental impact of heavily relying on ASF, the FAO and IAEA have initiated the development of a database on the protein quality of foods^{(Reference Tome, Xipsiti and Shertukde83)}. This resource, informed in part by stable isotope data, can help assess the long-term health effects of transitioning to more plant-based dietary proteins and aid in understanding the balance between meeting nutritional needs and maintaining planetary health. Data on protein digestibility and quality derived from stable isotope methods are already informing dietary recommendations for complementary feeding^{(Reference Shivakumar, Kashyap and Kishore41)}. And as more data becomes available, the Fe isotope dilution technique also holds great promise to refine our understanding of Fe requirements in different contexts and inform the development of food programs or initiatives that meet these needs.