Protein intakes of pregnant women and children in India—protein quality implications

Abstract The recent National Family Health Survey (NFHS‐4, 2016) reports a national average of 18% for low birthweight (LBW) and 38% for stunting in children <5 years. Nutrition and environmental influences (chronic enteric pathogenic exposure through poor water, sanitation, and hygiene) are two critical factors that impact the health outcomes of the populxation. This is particularly relevant for vulnerable age groups such as pregnant women and children <5 years, who bear long‐lasting and intergenerational consequences of impoverished nutrition and suboptimal living conditions. The present review provides, for the first time, an analysis of indispensable amino acid (IAA) requirements for pregnant women, separately for the second and third trimesters, using protein accretion data from a recent Indian study. Furthermore, using these estimates for pregnancy, and the current IAA requirements for young children, the quality of protein was assessed in Indian diets consumed by pregnant women and children (1–3 and 4–6 years) from national representative rural National Nutrition Monitoring Bureau survey. The assessment was considered in the context of an adverse environment and in relation to outcomes such as LBW, stunting, and underweight. Finally, an assessment was made of the proportion of the surveyed population at risk of dietary quality protein inadequacy and implications for planning nutrition intervention programmes. Specifically, state‐wise estimates of the risk of quality protein inadequacy are provided, in addition to evaluations of additional dietary supplementation, which could inform the policy of supplementary nutrition programmes to improve health outcomes.


| INTRODUCTION
The prevalence of poor birth outcomes and chronic undernutrition continues to be high in low-and middle-income countries, with India reporting a national average of 18% low birthweight (LBW) and 38% stunting in children <5 years (National Family Health Survey-4, 2016).
Diets in India are predominantly cereal based and lack diversity (Agrawal et al., 2019;Nithya & Bhavani, 2018), which increases the risk of macronutrient and micronutrient inadequacy in its population.
This risk of inadequate dietary intake could specifically impact vulnerable populations such as pregnant women and children, leading to poor outcomes of birth and nutritional status (Ghosh, 2016). Proteinrich foods are also a good source of diverse nutrients, and surveys consistently show that overall nutrient adequacy of a diet improves with inclusion of such foods from various sources (de Gavelle, Huneau, & Mariotti, 2018;Phillips et al., 2015).
Dietary proteins, particularly its constituent indispensable amino acids (IAAs), are required in adequate quantity and proportion to allow for body tissue protein synthesis along with various enzymatic and structural activities (Wu, 2010). Specifically, IAAs are essential for maternal health, in promoting positive birth outcomes (Ghosh, 2016), supporting linear growth in young children (Millward, 2017), and maintaining tissue health, specifically skeletal muscle, in adults and the elderly (Burd,  Although recent national dietary surveys in India show that the daily total protein intake is adequate (National Sample Survey Office, 2014), these estimates are based on the protein density in foods, without considering the digestibility and availability of IAA from these foods. To assess the IAA availability from the diet, a protein quality score needs to be calculated, which depends on the IAA composition of a food in relation to the age-specific daily IAA requirement (the amino acid score), as well as the specific IAA digestibility of the food, as measured in humans (Food and Agricultural Organization (FAO), 2013). The available protein quality scoring methods are the protein digestibilitycorrected amino acid score (PDCAAS), which uses a common oro-faecal balance-based digestibility factor for all IAA, and the digestible indispensable amino acid score (DIAAS), which is a more accurate metric of protein quality, as it uses digestibility factors that are based on digestion and absorption at an ileal level and is specific for each IAA (FAO, 2013). In addition, the effect of the environment on the digestive and absorptive functions of the intestine should be considered (Kurpad, 2006) when defining the IAA requirement, which will in turn impact its scoring pattern. This is of particular relevance to young children (<5 years), wherein, when safe environmental conditions were ensured, high quality protein intake was shown to be more efficacious for linear growth (Millward, 2017). This review will discuss the IAA requirements for pregnant women and children <5 years, protein quality (DIAAS) assessment of their diets in relation to nutritional outcomes from a national representative rural survey (National Nutrition Monitoring Bureau, 2012), environmental influence on the IAA requirements, and the implications of protein quality and environment on planning the provision of protein sources for nutrition intervention programmes.

| IAA requirements in children and pregnant women
The current daily estimated average requirement (EAR) of IAA for children was derived from a factorial approach, which is a sum of the mean maintenance IAA requirement and IAA requirement for growth.
For the growth component, the protein deposited during growth corrected for efficiency of utilisation is used (World Health Organization, Food and Agricultural Organization & United Nations University, 2007). The scoring pattern of the requirement is then calculated as a ratio of total IAA requirement (mgÁkg −1 Áday −1 ) to the calculated growth plus maintenance protein requirement (gÁkg −1 Áday −1 ; World Health Organization, Food and Agricultural Organization & United Nations University, 2007). The factorial estimates of IAA requirement (mgÁkg −1 Áday −1 ) compared well with empirical estimates of the requirement for some IAA in Indian children, using the stable isotopebased indicator amino acid oxidation method (Kurpad & Thomas, 2011;Pillai & Kurpad, 2012 • The protein quality of dietary intakes in children (1-3 and 4-6 years) from 10 Indian states of the rural National Nutrition Monitoring Bureau survey showed a significant inverse association with the state-wise prevalence of stunting and underweight.
• Supplementary food intake through nutrition programmes should be optimised for its protein quality, by using either animal source foods or a legume/cereal mixture along with animal source foods.
women implies the intake of high-quality protein, which could supply all the required additional IAA for optimal foetal growth. However, the report did not explicitly comment on the total or additional daily IAA requirement of pregnant women. It is therefore difficult to specifically evaluate the diets of pregnant women for their protein quality; this is relevant for operationalising feeding and subsidy programmes.
To compute the additional daily IAA requirement during pregnancy (separately for second and third trimesters), a factorial approach was used, with recent estimations of total body potassium accretion based additional protein requirements in well-nourished, middle socio-economic stratum South Indian pregnant women with a total gestational weight gain (GWG) of 10.7 kg (Kuriyan et al., 2019). In the present analyses, using the factorial approach, the additional IAA requirements were calculated as a sum of IAA requirements for growth (newly deposited tissue) and maintenance of the total GWG (including maintenance of the newly deposited tissue). Therefore, first, the additional protein requirement (gÁkg −1 Áday −1 ) for growing tissue (foetal and maternal tissue deposition) that had been adjusted for an efficiency of utilisation of 42% (World Health Organization, Food and Agricultural Organization & United Nations University, 2007) was multiplied by the IAA composition (mg g −1 protein) of foetal tissue (Widdowson, 1980). Because foetal protein accounts for a major portion of protein accretion (42%; Rasmussen & Yaktine, 2009), and IAA composition of different maternal tissues were not available, it was reasonable to use IAA composition of foetal tissue (Widdowson, 1980) to represent the overall tissue deposition. Moreover, IAA composition of mixed (World Health Organization, Food and Agricultural Organization & United Nations University, 2007) and foetal tissue (Widdowson, 1980) are reasonably similar in humans. Second, the GWG maintenance protein requirement (gÁkg −1 Áday −1 ) was multiplied by the maintenance IAA requirement pattern (mg g −1 protein) for adults (World Health Organization, Food and Agricultural Organization & United Nations University, 2007). In the foetal tissue IAA composition (Widdowson, 1980), the concentration of tryptophan and cysteine was not available; for the sulphur containing amino acids (SAA), a 1:1 ratio of methionine: cysteine was assumed. For the second and third trimesters, the additional IAA requirements for tissue deposition were calculated separately using mean values of foetal tissue IAA composition from 12-24 and 28-40 weeks, respectively (Widdowson, 1980). The scoring pattern was calculated as the ratio of the body weight-corrected IAA requirement (mgÁkg −1 Áday −1 ) and the body weight-corrected total protein requirement (gÁkg −1 Áday −1 ) for the pregnant women studied (Kuriyan et al., 2019

| Adverse environmental influence on IAA requirements of children and pregnant women
Developing countries share greater than 95% of the global burden of stunting in children <5 years (De Onis, Blössner, & Borghi, 2012). This is attributed to the suboptimal quality protein intake and poor environmental conditions, which includes chronic pathogenic exposure (bacterial and parasitic) compounded by family and societal stresses (Crane, Jones, & Berkley, 2015). Prolonged exposures to an impoverished environment can alter the enzymatic secretions in gut, leading to compromised digestion and absorption of the nutrients (Banwell et al., 1967). If the digestion and absorption of protein is low, it is likely that the protein and IAA requirements (especially the limiting IAA, such as lysine) of individuals from poor socio-economic environments will be higher than the recommendations made for a healthy population from safe or "clean" environments. This has been supported by a study on chronically undernourished men from poor socio-economic conditions, where daily requirement of lysine was found to be higher by~50% compared with their age-matched controls (Kurpad et al., 2003). Such increase in the requirement could partly be attributed to the parasitic infections, as a subsequent deworming trial on school children from compromised environmental conditions showed that their lysine requirement was~20% higher in the presence of parasites (Pillai, Elango, Ball, Kurpad, & Pencharz, 2015). Although it is difficult to provide an accurate estimate of the additional IAA requirement imposed by the environmental stress, a potential increase in the EAR of IAA is a necessary consideration. This is of relevance in the current context as the latest NNMB survey Chronic enteric pathogenic exposure leading to subclinical environmental enteropathy in pregnant women has been observed to affect birth outcomes (Lauer et al., 2018). The researchers reported elevation in anti-flagellin and anti-lipopolysaccharide immunoglobulin concentrations, which reflect possible enteropathy, to be associated with shorter gestation and lower birth lengths. This was the first study to show the probability of an environmental impact on pregnancy that in turn relates to adverse birth outcomes (Lauer et al., 2018). A study on Indian, Caribbean, and U.S. women showed that the lactulose: mannitol ratio tended to be higher in Indian non-pregnant non-lactating women, with a lower net mannitol absorption (Kao et al., 2015), suggesting that they could have some intestinal dysfunction and consequently a lower protein digestion and absorption. However, for pregnancy, an increased IAA requirement was not used to calculate DIAAS, because of the lack of data on the environmental interactions and some evidence on the increased efficiency of lysine (and perhaps other IAA) utilisation (by~10%) in pregnant pigs for the second and third periods of gestation (Navales et al., 2019), although there are no similar studies in humans.

| Protein quality evaluation in relation to nutritional outcomes and the risk of protein inadequacy in children and pregnant women from rural India
The relation between dietary protein quality (DIAAS) and nutritional outcomes was evaluated in a representative rural population of Indian for maintenance and foetal growth (average of the second and third trimesters) to the daily maintenance IAA requirement of the non-pregnant non-lactating woman. Finally, the scoring pattern (mg g −1 protein) was calculated as a ratio of mid-pregnancy IAA and protein requirement.
The DIAAS was calculated for children aged 1-3 and 4-6 years and pregnant women based on the food intake (g day −1 ) records of 10 states, using IAA digestibility values (Table 1) of one representative food from each food group. In the present analysis, four out of seven food groups, namely, cereals and millets, legumes, milk and milk products, and other animal source food (ASF; fish, egg, meat, and poultry), that contribute predominantly to the daily protein intake were selected. The representative foods were rice for cereals, finger millet for millets, whole mung bean for legumes, and hen's egg for ASFs (Kashyap et al., 2018;Kashyap et al., 2019;Shivakumar et al., 2019). to account for environmental stress.

| RESULTS
The additional IAA requirements (mgÁkg −1 Áday −1 ) and scoring patterns (mg g −1 ) are presented in have to be consumed in greater quantities. For example, to meet the second trimester quality-corrected protein requirement (~8 g day −1 ), the pregnant women's diet should include an extra 250 ml of milk (2½ glasses) or 70 g of egg (1½ eggs) or 120 g of legumes. As an additional intake, the quantity of legumes appears to be daunting but can be more reasonable when consumed in combination with ASF, for example, 30 g of legumes along with either 100 ml of milk or 25 g of egg, or a mix of 50-g legume and 50-g rice.
The mean daily estimated average protein requirement (gÁkg −1 Áday −1 ), IAA requirement (mgÁkg −1 Áday −1 ), and scoring pattern (mg g −1 protein) for children 1-3 and 4-6 years and pregnant women are presented in Table 4. The DIAAS, crude, and DIAAS-corrected protein intake data of 10 states are presented for children 1-3 and 4-6 years in Table 5 and for pregnant women in Table 6  Abbreviations: AAA, aromatic amino acids; IAA, indispensable amino acid; Lys, lysine; SAA, sulphur amino acids; Thr, threonine. a Mean growth protein requirement (gÁkg −1 Áday −1 ) was calculated as an average growth requirement of children 1, 1.5, 2, 2.5, and 3 years from WHO (2007); maintenance protein requirement (gÁkg −1 Áday −1 ) was taken from WHO (2007); daily lysine requirement (mgÁkg −1 Áday −1 ) was increased by 20% to account for environmental stress. b Protein requirement (gÁkg −1 Áday −1 ) for growth and maintenance was reproduced from the requirements of 3-10 years (WHO, 2007); daily lysine requirement (mgÁkg −1 Áday −1 ) was increased by 20% to account for environmental stress. c Daily IAA requirements were calculated based on the sum of the (a) product of the mean second and third trimester pregnancy maintenance protein requirement and the maintenance IAA pattern and (b) the product of the mean second and third trimester protein deposition for foetal growth and the IAA composition of the fetus; additional mean second and third trimester protein requirement (mgÁkg −1 Áday −1 ) for foetal growth was calculated using midpregnancy body weight, taken as a sum of body weight at the end of the first trimester and half the gestational weight gain over the second and third trimesters (Kuriyan et al., 2019); maintenance protein requirement of non-pregnant non-lactating women was taken from WHO (2007). d Scoring pattern of IAA requirement (mg g −1 protein) for children 1-3 and 4-6 years and pregnant women was calculated as a ratio of each IAA requirement (mgÁkg −1 Áday −1 ) and the protein requirement (gÁkg −1 Áday −1 ).
T A B L E 5 Dietary crude protein intake, DIAAS of the diets, DIAAS-corrected protein intake, and risk of crude and DIAAS-corrected protein inadequacy in children aged 1-3 and 4-6 years across states from NNMB rural survey data  Risk was calculated with a 20% increase in estimated average requirement of protein to account for the increased demand of protein due to poor environmental conditions. generated true ileal IAA digestibility values for commonly consumed foods in Indian meal matrices (Kashyap et al., 2018;Kashyap et al., 2019;Shivakumar et al., 2019). Furthermore, the issue of chronic environmental pathogenic exposure leading to impaired intestinal function needs consideration. A study on a group of healthy children from West Africa (Dakar and Abidjan province) showed a significantly lower concentration of all pancreatic enzymes, electrolytes, and water compared with their age-and sex-matched controls from France (Marseille province), in duodenal aspirates after pancreatic stimulation by pancreozymin (Sauniere & Sarles, 1988). The study also reported an elevated lactoferrin levels in the children from West Africa, which indicated a subclinical pancreatic damage and demonstrated a silent pancreatic insufficiency in the population (Sauniere & Sarles, 1988). In addition, adverse environmental exposure leads to morphological changes in the intestinal epithelium such as villous atrophy, low villus to crypt ratio, increased epithelial permeability, and inflammation (Owino et al., 2016). This impairs absorption and causes subsequent nutrient loss from the gastrointestinal tract (Keusch et al., 2014). Further, chronic stimulation of immune system by enteric pathogens may lead to nutrient partitioning towards supporting the immune response, leaving less available for growth and maintenance. A murine model showed that a poor diet with low protein and fat content (7% and 5%, respectively), compounded by an environmental pathogenic insult (from a bacterial pathogen cocktail), developed features of enteric dysfunction within 3 weeks with altered small intestinal microbiota, permitting greater colonisation of enteric pathogens and increasing susceptibility to enteric infections (Brown et al., 2015).
The recent plant protein IAA digestibility values from Indian studies (Kashyap et al., 2018;Kashyap et al., 2019;Shivakumar et al., 2019) when utilised for DIAAS assessment suggest that cereal : legume complementation may not be sufficient to optimally improve protein quality in children and pregnant women; in this specific instance, an increase in food protein quantity is required but with the risk of excess energy intake. As the volume of food also increases, this will be difficult to achieve in young children with relatively low gastric  , 2000) in NNMB rural survey data (NNMB, 2012); estimated average requirement of protein (gÁkg −1 Áday −1 ) for pregnant women was taken as 0.87 gÁkg −1 Áday −1 (addition of protein requirement for maintenance and foetal growth) from Table 3, which was multiplied by body weight of pregnant women. For the body weight, half the gestational weight gain (GWG) of 4 kg (below) was added to the non-pregnant nonlactating reference body weight of 55 kg to calculate total protein requirement and risk of protein inadequacy in diets; a GWG of 8 kg was previously reported (National Nutrition Monitoring Bureau, 1984); however, in the absence of data on the trimester of recruitment in the latest NNMB survey (NNMB, 2012), the reported GWG was halved (4 kg).

2017). Several observational and intervention studies have shown
positive associations between ASF intake and linear growth even after adjusting for confounders such as socio-economic status, morbidity profile, parental education, and nutritional status (Allen & Dror, 2011;Iannotti et al., 2017;Millward, 2017;Neumann, Harris, & Rogers, 2002). In further support, PDCAAS-corrected quality protein intake showed positive impact on the rate of weight gain in severely acute malnourished Peruvian children of 6-32 months (Arsenault & Brown, 2017). A recent observational study of South Indian pregnant women showed a significant positive association (P < .001) between protein intake from milk products and birthweight (Mukhopadhyay et al., 2018). The state-level analyses of rural NNMB dietary intake data show that, unlike plant source foods, the consumption of ASF in pregnant women has a significant inverse correlation with LBW prevalence (rs = −.72, P < .05). A similar inverse but non-significant relation was observed between ASF intake and prevalence of stunting in children (rs = −.48, P = .13 for 1-3 years and rs = −.47, P = .14 for 4-6 years). It is worth noting that such inverse association can be attributed to different macronutrient and micronutrient contents in their diets and corroborates with the observed significant positive association between ASF intake and height for age in preschool children from developing countries of Central and South America in Demographic and Health Surveys (Ruel, 2012).
It is possible to improve the quality of protein intake with modest additions of ASFs to reduce the risk of quality protein (DIAAS corrected) inadequacy, but a DIAAS of 100% cannot be achieved unless safe environmental conditions are ensured for children. If not, an additional allowance of protein and IAA is required, which in turn raises the amino acid score of the requirement. Adding a combination of 100-ml milk (1 glass) and 50-g egg (1 egg) increased the DIAAS of the diet protein to 92% in 1-3 years and reduced the DIAAS-corrected risk of inadequacy to 9%. Similarly, for 4-6 years, additional 200-ml milk (2 glasses) and 50-g egg increased the DIAAS of the diet to 92% and reduced the quality-corrected risk of protein inadequacy to 5%. However, risk reduction is more difficult in pregnant women, where the addition of 350-ml milk (3½ glasses) and 75-g eggs (1½ pieces) increased the DIAAS of diet to 100% but reduced the risk of DIAAScorrected protein inadequacy only to 20%, because of a habitual low protein intake in this population. The addition of legumes was not as effective due to their poor digestibility; as with addition of 50 and 60 g of legume day −1 , the risk of DIAAS-corrected protein inadequacy reduced only to 20%, 19%, and 58% in children (1-3 and 4-6 years) and pregnant women, respectively. Plant source foods have other beneficial components like fibre and so should not be ignored as important foods, but from a quality protein-provision perspective, they are inadequate. The addition of ASFs in above-mentioned amounts increased the DIAAS-corrected protein energy ratio by~5% in children 1-3 (from 7.1% to 12.2%) and 4-6 years (from 6.6% to 11.5%) and (from 7.4% to 12.4%) pregnant women, which was within (<15%) the current recom- data were based on a single day "total household" recall and not an age-specific individual recall; to get to an individual consumption, ageand sex-specific consumption units were applied (NNMB, 2012), which may lack representation of a regular, weekly, or monthly consumption of food groups. Third, the NNMB dietary intake data were surveyed for the 4-to 6-year age group but correlated with the prevalence of stunting and underweight in children 3-5 years. Fourth, digestibility values for cereals and millets were taken from measurements made in children; these data were used for calculating the DIAAS of pregnancy protein intakes; similarly, egg digestibility values were assumed to be the same as that for milk and flesh foods.
The review presents for the first time to our knowledge the IAA requirements and scoring pattern for pregnant women, based on Indian data. It underscores the need for protein quality evaluation of commonly consumed foods and diets in population, specifically for vulnerable age groups and pregnancy, in a poor environmental setting.
The state-wise estimates of DIAAS-corrected risk of protein inadequacy can be used for planning intervention trials, such that dietary interventions aiming to promote linear growth in young children and positive birth outcomes are optimised for the quality of their protein as well as safe environmental conditions.

ACKNOWLEDGMENTS
We acknowledge Ms. Srishti Sinha and Ms. Saba Naqvi for their input and critiques. A. V. K. was supported by the Margadarshi fellowship of the Wellcome Trust/DBT India Alliance. He is also a senior advisor to the Tata Trusts, Mumbai, India.

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest CONTRIBUTIONS SB and NS collated and analysed the data; NS and AVK conceived the research; and all authors contributed to the writing of the manuscript.
NS had primary responsibility for the final content.