Hostname: page-component-78c5997874-lj6df Total loading time: 0 Render date: 2024-11-09T22:11:29.113Z Has data issue: false hasContentIssue false

Assessment of protein adequacy in developing countries: quality matters

Published online by Cambridge University Press:  01 August 2012

Shibani Ghosh
Affiliation:
Friedman School of Nutrition Science and Policy, Tufts University, 150 Harrison Avenue, Boston, MA02111, USA Nevin Scrimshaw International Nutrition Foundation, 150 Harrison Avenue, Room 232, Boston, MA02111, USA
Devika Suri
Affiliation:
Nevin Scrimshaw International Nutrition Foundation, 150 Harrison Avenue, Room 232, Boston, MA02111, USA
Ricardo Uauy*
Affiliation:
Nevin Scrimshaw International Nutrition Foundation, 150 Harrison Avenue, Room 232, Boston, MA02111, USA Institute of Nutrition and Food Technology (INTA), University of Chile, Santiago, Chile
*
*Corresponding author: Dr Ricardo Uauy, fax +1 617 636 3771, email [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Dietary protein and amino acid requirement recommendations for normal “healthy” children and adults have varied considerably with 2007 FAO/WHO protein requirement estimates for children lower, but dietary essential AA requirements for adults more than doubled. Requirement estimates as presented do not account for common living conditions, which are prevalent in developing countries such as energy deficit, infection burden and added functional demands for protein and AAs. This study examined the effect of adjusting total dietary protein for quality and digestibility (PDCAAS) and of correcting current protein and AA requirements for the effect of infection and a mild energy deficit to estimate utilizable protein (total protein corrected for biological value and digestibility) and the risk/prevalence of protein inadequacy. The relationship between utilizable protein/prevalence of protein inadequacy and stunting across regions and countries was examined. Data sources (n = 116 countries) included FAO FBS (food supply), UNICEF (stunting prevalence), UNDP (GDP) and UNSTATS (IMR) and USDA nutrient tables. Statistical analyses included Pearson correlations, paired-sample/non-parametric t-tests and linear regression. Statistically significant differences were observed in risk/prevalence estimates of protein inadequacy using total protein and the current protein requirements versus utilizable protein and the adjusted protein requirements for all regions (p < 0·05). Total protein, utilizable protein, GDP per capita and total energy were each highly correlated with the prevalence of stunting. Energy, protein and utilizable protein availability were independently and negatively associated with stunting (p < 0·001), explaining 41 %, 34 % and 40 % of variation respectively. Controlling for energy, total protein was not a statistically significant factor but utilizable protein remained significant explaining~45 % of the variance (p = 0·017). Dietary utilizable protein provides a better index of population impact of risk/prevalence of protein inadequacy than crude protein intake. We conclude that the increased demand for protein due to infections and mild to moderate energy deficits, should be appropriately considered in assessing needs of populations where those conditions still prevail.

Type
Full Papers
Copyright
Copyright © The Authors 2012

Introduction

Protein and amino acid requirement recommendations for “healthy” infants, children and adults have varied considerably over the years(1). The most recent FAO/WHO report provides estimates of protein requirements that are lower than previously established for adults and children. However the relative importance of protein quality is greater since estimates of the dietary essential amino acid requirements are twice the previous recommendations with lysine requirement estimates having increased 2·5 times from 12 mg/kg body weight to 30 mg/kg body weight in adults(1). In children, the estimated dietary essential amino acid requirements are only slightly lower (94 % of previous estimate for lysine).

In environments where individuals have persistent or repeated infections and impaired intestinal absorptive capacity, there is an increased demand for protein, despite the absence of overt clinical symptoms(1, Reference Dewey, Beaton and Fjeld2). Thus in vulnerable populations such as women and children commonly affected by acute and chronic infections, protein and amino acid needs are likely to be greater. Growth rates (linear and ponderal) are likely to be affected by repeated infections (bacterial and parasitic) with long-term functional deficits compromising learning and adult productivity(Reference Dewey, Beaton and Fjeld2, Reference Scrimshaw, Taylor and Gordon3).

Interactions between energy deficit and protein needs have been documented by determining nitrogen balance and these studies have been previously reported(4). Energy imbalance below and above energy needs affects body nitrogen equilibrium. At a protein intake of 0·57 g/kg body weight, N equilibrium is achieved if energy intake is ~10 % above that required for balance(Reference Garza, Scrimshaw and Young5, Reference Garza, Scrimshaw and Young6). Conversely people in energy deficit will need additional protein, considering that even a modest energy deficit of 5 % increases protein needs by about 10 %(Reference Kishi, Miyatani and Inoue7). Thus, beyond the role of protein in the maintenance of nitrogen equilibrium, commonly accepted as the criterion for sufficiency, mild energy deficit and infections place additional demands on body nitrogen.

The operational definition of what is a sufficient amount of dietary protein has up to now been based mostly on body N balance; thus this is the method used to define requirements and establish recommendations. However there are multiple other roles that amino acids and protein play which impact health and wellbeing. Other possible roles such as the role of amino acids in up-regulating growth hormone and IGF-1 secretion thus driving anabolism and linear growth could lead to estimates of protein requirements that differ from those established based on the need for N equilibrium and the fractional daily N accretion related to body growth. There is now evidence indicating that both protein quantity and quality affect growth hormone release and thus have the potential for modulating linear growth. Protein intake in infancy has been shown to stimulate early growth and is also associated with body weight and length at later stages of life(Reference Hoppe, Molgaard and Thomsen8). Protein intake at 9 months of age has also been positively associated with height and weight but not percentage of body fat at 10 years of age(Reference Hoppe, Molgaard and Thomsen8). Currently, there is no consideration given to the potential role of specific dietary essential and non-essential amino acids in defining hormonal responses associated with linear growth, nor are other organ-specific functional needs related to protein and amino acids, reflected in the estimation of protein requirements.

The evaluation of protein quality has been discussed extensively in several key FAO/WHO policy documents(1, 9). While different methods exist for evaluating protein quality, the currently accepted method is the protein digestibility-corrected amino acid score (PDCAAS). The PDCAAS method assesses the quality of protein of mixed diets and/or mixed formulations based on the digestibility of the protein sources and the essential amino acid composition thus giving an estimate of “utilizable protein”. Diets that have the right amino acid composition (matching the reference protein) have a high score and if the source(s) are well digested the protein is considered of high quality. Foods of animal origin are likely to have total protein and utilizable protein values higher and closer to each other compared to diets that have a low content of essential amino acids and that are poorly digested, such as cereals and other grains(1, 9). The PDCAAS method can be used for evaluating food products (e.g. cereal-legume mixtures), for evaluating overall protein quality of mixed diets (individual) and evaluating the per capita availability and accessibility of good quality protein at the country/global level. While amino acid scoring methods have been used for evaluating protein quality at country and global level, the use of PDCAAS to determine the effect of digestibility and the amino acid composition of per capita availability of protein quality has not been examined extensively. The evaluation of protein quality and quantity for growth (specifically linear growth) has been examined mainly in developed country populations, but not sufficiently in developing country settings. Thus the aim of our analysis was the evaluation of the impact of applying the PDCAAS method correcting for protein quality to assess adequacy at the level of the national food supply, examining its relationship with the corresponding national data on linear growth. The underlying hypothesis proposed by our analysis was that linear growth retardation (stunting) would be more strongly related with available protein corrected for quality by PDCAAS, and also better related to the estimated risk/prevalence of protein inadequacy if protein requirements were adjusted for mild energy deficit and for the prevalence of infections, factors prevalent in developing countries that are known to increase protein needs. The latest WHO report on protein and amino acid recommendations(1) noted that there was very little work assessing the protein requirements in children and adult populations with high disease burden. Furthermore, given the high levels of stunting and the early onset of stunting, such an analysis would also allow for the development of testable hypotheses on the potential impact of protein quality on linear growth of young children in developing countries.

Experimental methods

Data sources

National level data from 180 countries, used to estimate total protein and utilizable protein supply and to estimate the risk of protein inadequacy, were obtained from food balance sheets (FBS) from the Food and Agriculture Organization of the United Nations (FAO) for the year 2005 using methods previously described(Reference Pellett10Reference Sasaki and Kesteloot12). The FBS provide estimates of per capita supplies of specific food items available for human consumption in a given year and a given country. Per capita supplies of each listed commodity available for food consumption (i.e. supply) are the sum of domestic production, stocks, and imports minus exports and amounts used for feed, seed, processing, and other purposes.

To assess the relationship between protein quality and linear growth, national level data were acquired for moderate to severe stunting (n = 116) from UNICEF(13). The indicator “Moderate to Severe Stunting” (referred to as “stunting”) is defined by UNICEF as the percentage of children under five years of age in a particular country who fall below − 2 and − 3 standard deviations for height-for-age z-score. Gross domestic product (GDP) data were extracted for the same 116 countries from the United Nations Development Programme (UNDP) Human Development Report(14). The regional breakdown included 17 countries in East and Southern Africa, 23 in West and Central Africa, 2 in Oceania, 6 in South East Asia, 7 in South Asia, 23 in Latin America and the Caribbean, 3 in Eastern Asia, 16 in North Africa and the Middle East, 12 in Eastern Europe, 5 in Central Asia and 2 in North America (with prevalence estimates of stunting and wasting). Data on Infant Mortality Rates (IMR) were obtained from the United Nations Statistics division that compiles data on such indicators(15).

Nutrient database development

The nutrient database was developed for FAO FBS food commodity categories using USDA food composition tables(16). Protein digestibility values for each food item were obtained from FAO/WHO guidelines on protein quality evaluation as well as specific digestibility studies(9, Reference Young and Pellett17Reference Gabert, Brunsgaard and Eggum22).

Data calculations and analysis

Utilizable protein and Prevalence of Protein Inadequacy: Based on Current Requirements

Utilizable protein was calculated using the Protein Digestibility Corrected Amino Acid Score (PDCAAS) method described by WHO(1) and compared with WHO (2007) protein requirements to allow estimation of the adequacy of available dietary protein in each country or region. For the FAO FBS data, requirement estimates pertaining to the adult for total energy, total protein and amino acids (total requirement and the amino acid reference pattern) were assumed and related to the entire population. The requirement values used were 0·66 g/kg body weight for utilizable protein, 2525 kcal total energy calculated for a moderately active adult (PAL 1·75, average for moderately active adult male and female) and the amino acid reference pattern for lysine (30 mg/kg body weight/day), tryptophan (4 mg/kg body weight/day), sulphur amino acids (SAA) (15 mg/kg body weight/day) and threonine (15 mg/kg body weight/day)(1, 23). Risk of protein inadequacy at the country level was computed using the protein requirement of a 60 kg adult, considering a 25 % coefficient of variation in “intake” (supply) and assuming that quality protein would be uniformly distributed per person(Reference Wuehler, Peerson and Brown24, Reference Ghosh, Smriga and Vuvor25).

Utilizable Protein and Prevalence of Protein Inadequacy: Corrected Requirements

Given the high rates of disease and infections in most developing countries, we examined the effect of correcting current estimates of protein requirements to account for energy deficit as well as for the increased need for protein for infectious episodes as well as for the period of recovery post-infection. On the basis of data taken from the studies of Garza et al. (Reference Garza, Scrimshaw and Young5) and Kishi et al.(Reference Kishi, Miyatani and Inoue7) in adults, a 10 % increase per day was added for energy deficit (assuming a moderate deficit in energy). This was applied across all countries.

To calculate an additional dietary protein need due to infection, countries were ranked in tertiles of infant mortality rates (IMR), as a proxy for the burden of infection in early life(Reference Black, Allen and Bhutta26Reference Mata29). Infections in early life were assumed to be primarily respiratory and diarrhoeal in nature as these are the main causes of death and disability in developing country children(Reference Victora, Adair and Fall30). Daily protein requirements are augmented a further 10 % during each day of illness(Reference Kurpad, Regan and Nazareth31, Reference Kurpad, Regan and Raj32), assuming 7 days per episode (with 5 days ill and 2 days of recovery) and 8 episodes per year for the highest IMR tertile(Reference Black, Brown and Becker33), 5·3 for the middle and 2·7 for the lowest tertile. To find the increase in daily protein needs due to infection over one year, a weighted average of the protein requirements on total days ill and not ill over one year was calculated. Combined with the 10 % increased protein needs for moderate energy deficit, the final figure for each IMR tertile represented the new daily protein needs accounting for both infection and energy deficit. Estimates of the above calculation and resulting adjusted requirements are presented in Table 1. These new requirement estimates based on adjustments for energy deficit and infection were used in place of the standard adult protein requirement estimate of 0·66 g/kg, in the calculation of prevalence of protein inadequacy for each country.

Table 1 Calculation of increased protein needs due to infection and moderate energy deficit, by IMR tertile (infant mortality rate per 1000 births) for 115 countries

Statistical analysis

All data calculation and analyses were conducted in Statistical Analysis Systems statistical software package version 9.2 (SAS Institute, Cary, NC, USA), SPSS version 15.0 (SPSS, Inc, Chicago, IL, USA) and Microsoft Excel 2007 (Microsoft Corp, Redmond, WA, USA). Nutrient calculation and estimation of PDCAAS was conducted in SAS while prevalence of inadequacy calculations in MS Excel. All statistical analyses were conducted in SPSS. Statistical analyses and tests included descriptive statistics, frequency analyses, non-parametric Chi-square, independent sample tests and linear regression analyses. Significance was set at the 0·05 level.

Results

Estimates of per capita dietary energy, utilizable protein, percentage of stunting and prevalence of protein inadequacy are presented in Table 2 by region. Total energy supply per capita per day ranged from 2323 ± 398 kcal in East and South Africa to as high as 4017 ± 686 kcal in North America. Total protein values ranged from 60 ± 14 g/capita/day to over 120 ± 23 g/capita/day. While there was a general trend of increasing total energy and total protein, in the case of Eastern Asia despite lower energy availability (2360 ± 279 kcal), total protein levels were 84 ± 17·5 g/capita/day. Stunting prevalence ranged from 36·5 ± 10·6 % in Oceania (countries of Comoros and Sao Tome and Principe) to 9·5 ± 12·0 % in North America while wasting ranged from 12·6 ± 2·4 % in South Asia to 4·0 ± 2·8 % in North America. Protein requirements (g/kg body weight) were adjusted as described above. Figure 1 depicts the extra needs imposed by infection and moderate energy deficit as aggregated by IMR (Infant Mortality rate) tertile (n = 115 countries). The adjusted protein need for an adult in a given country was 0·73 (both low and middle IMR) and 0·74 g/kg (for countries with high IMR), respectively (Fig. 1).

Table 2 Supply per capita per day of energy, protein and utilizable protein, prevalence of stunting and wasting, Gross Domestic Product (GDP) by regions

1 Defined as the percentage of children under five years of age in a particular country who fall below − 2 standard deviations for height-for-age (stunting) or weight-for-height(wasting) z-score

Fig. 1 Estimates of adult daily protein requirement, with added needs for infection and moderate energy deficit, by country-level IMR tertile, for 115 countries.

Risk of Protein Inadequacy

Risk of protein inadequacy was calculated using the estimates of total protein and utilizable protein (Fig. 2) both compared to the current estimates of requirements (Fig. 2 lines a, b) and utilizable protein compared to the revised requirements by country (Fig. 2 line c). Prevalence of inadequacy ranged from 0·7 ± 0·8 % in North America to 37·2 ± 23 % in East and Southern Africa. It was the highest in East and Southern Africa (irrespective of type of protein and requirement used) and the lowest in North America. Correction of prevalence using utilizable protein (line b) increased the inadequacy estimates for East and Southern Africa, West and Central Africa, Oceania, South Asia and South East Asia and to a lesser extent in Latin America and the Caribbean. As can be seen from Fig. 2 (line c), inadequacy estimates based on utilizable protein and requirements adjusted for infection and energy deficit were significantly higher than the estimates based on total protein and utilizable protein that were compared to the current requirements (p < 0·05).

Fig. 2 Differences in risk estimates of protein inadequacy, calculated using total protein, utilizable protein (UP), and UP plus higher requirements for infection and moderate energy deficit, by regions of the world. * Significant difference between risk estimates of protein inadequacy using total protein (line a) and utilizable protein (line b) compared current requirements versus inadequacy estimates using utilizable protein (line c) compared current requirements that have been adjusted for infection and energy deficit. Analysis conducted using paired t-tests or non-parametric tests (p < 0·05).

Examining changes in estimates at the country level, countries with less than 2000 kcal/capita/day of energy supply, show a change in estimates in risk of protein inadequacy from 20–80 % (Eritrea) to 40–90 % (Guinea Bissau) (Fig. 3) (going from total protein and current requirements to utilizable protein and revised requirements). Estimates in the Congo (Democratic Republic) are high to begin with and signify very low levels of total protein in the diet. In countries with 2000–2500 kcal energy supply/capita/day (Fig. 4 a and b), risk estimates of protein inadequacy using total protein and current estimates of requirement range from 60 % (Liberia) to 10 % (Djibouti) in Sub Saharan Africa and 10 % (India) to 5 % (Thailand) in South Asia and South East Asia. On the basis of total protein and the current requirement estimates alone, countries in South Asia and South East Asia have relatively low risk estimates of protein inadequacy. Correcting for quality (utilizable protein) and needs (adjusted requirements) changes the prevalence in Sub Saharan Africa to 90 % (Liberia) and almost 40 % (Djibouti) while in South Asia and South East Asia, risk of protein inadequacy changes to 45 % (India) to about 15 % (Thailand). Examining countries with total energy intakes ranging from 2500–3000 kcal/capita/day, similar changes are observed. In countries like Burkina Faso and Niger, a very low prevalence (less than 2 %) changes to 20 % (Burkina Faso) and about 15 % (Niger) on using the figures for utilizable protein and adjusted requirements (Fig. 5a). This is also observed in South Asia in the case of Nepal, and Sri Lanka, Myanmar and the Philippines (Fig. 5b).

Fig. 3 Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in countries with less than 2000 kcal/capita/day energy supply.

Fig. 4 (a) Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in Sub Saharan African countries with 2000–2500 kcal/capita/day energy supply. (b) Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in South and South-East Asian countries with 2000–2500 kcal/capita/day energy supply.

Fig. 5 (a) Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in Sub Saharan African countries with 2500–3000 kcal/capita/day energy supply. (b) Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in South and South-East Asian countries with 2500–3000 kcal/capita/day energy supply.

Growth status and Protein Adequacy

As noted in Table 1, most regions with a high prevalence of protein inadequacy also have high rates of stunting (ranging from 34·7 ± 12·1 % to 35·4 ± 12·9 % in Sub-Saharan Africa and South Asia). An examination of the relationship between levels of stunting, utilizable protein (total) and prevalence of protein inadequacy are presented in Tables 3, 4 and 5. Using Pearson correlations, stunting was significantly and negatively correlated with total energy, protein, utilizable protein and GDP per capita, (Table 3) (p < 0·001 for all variables).

Table 3 Correlation coefficients for relationships between country-level nutrient supply, prevalence of stunting and GDP per capita, for 115 countries

1 Defined as the percentage of children under five years of age in a particular country who fall below − 2 standard deviations for height-for-age z-score

2 All coefficients are significant at p <  0·001

Table 4 The association (linear regression) between prevalence of stunting and total and utilizable protein supply (g/capita/day) for 115 countries

1 Defined as the percentage of children under five years of age in a particular country who fall below − 2 standard deviations for height-for-age z-score

Table 5 Associations (linear regression) between prevalence of stunting and prevalence of protein inadequacy (total protein and utilizable protein with adjusted protein requirements) and energy supply, for 115 countries

1 Prevalence of protein inadequacy calculated using estimates of total protein (g/capita/day) and current protein requirements of 0·66 g/kg for 60 kg adult

2 Prevalence of protein inadequacy calculated using estimates of utilizable protein (g/capita/day) and protein requirements of 0·66 g/kg for 60 kg adult adjusted for energy deficit and infection needs

3 Defined as the percentage of children under five years of age in a particular country who fall below − 2 standard deviations for height-for-age z-score

Linear regression analyses show that total energy, total protein and total utilizable protein estimates are all independently and significantly negatively associated with prevalence of stunting (Table 4) with p < 0·001. Total energy explained 41 % of the variation in stunting, (r-square of 0·406), total protein explained 34 % of the variation and utilizable protein explained 40 % of the variation in prevalence of stunting (separately). When total energy and total protein were incorporated into the model together, only total energy was significant explaining about 41 % of the variation in prevalence of stunting (p < 0·001, r-square = 0·407); however when total energy and utilizable protein were incorporated in the model together, both were significant together explaining 43 % of the variation in stunting (p = 0·006 and p = 0·017, respectively, r-square = 0·430). GDP per capita was a significant predictor of stunting (p < 0·001) by itself and remained significant when energy was controlled (p = 0·048) (data not shown). When GDP, utilizable protein and total energy variables were included in the model, GDP remained significant (p < 0·001), the utilizable protein factor was not significant and total energy as a factor remained significant (p = 0·022) (Table 4).

Similar to the estimates using total protein and utilizable protein, prevalence of protein inadequacy that was estimated using utilizable protein and the adjusted requirements (adjusted for infection and moderate energy deficit) was significantly associated with stunting (p = 0·003) when controlling for energy (which was also significant, p < 0·001) (Table 5). Prevalence of protein inadequacy calculated from total protein and the current requirements was not significantly associated with stunting once energy was incorporated into the model (Table 5).

Discussion

The aim of the analysis reported here was to examine the effect of applying PDCAAS to total protein values from national food supply data (FAO FBS) and examining the difference in estimates of risk/prevalence of protein inadequacy in relation to total protein versus utilizable protein. We also aimed to calculate the difference in protein requirement estimates accounting for increased needs from infection and moderate dietary energy deficit. National food supply data for the year 2005 were used to calculate total energy, protein and essential amino acid intakes and these intakes corrected for protein quality and digestibility estimates to obtain utilizable protein values. The requirement estimates for protein (adults, 0·66 g/kg body weight) were corrected for moderate energy deficit (10 % correction) and increased demand due to infections using IMR as a proxy for the burden of infections to estimate an exact additional need by country IMR. Based on methods defined by WHO(34), prevalence estimates for protein inadequacy were calculated using total protein and the current requirement estimates as well as utilizable protein and adjusted requirements. For all analyses an estimated intake CV of 25 % was used.

Our findings on total protein and energy availability as well as levels of lysine in the supply are in line with Pellett(Reference Pellett11) with developing country regions specifically in Sub-Saharan Africa and South Asia being the lowest in total energy and protein availability. Correcting total protein for quality and digestibility further increases the gap between developing country regions and developed regions indicating the differences in protein sources and quality. Prevalence of protein inadequacy after correction for quality and digestibility ranges from 5–50 % irrespective of the energy availability by region and/or by country. Almost all the countries with total energy availability less than 2000 kcal/capita/day had a high prevalence of protein inadequacy which further increased when requirements were adjusted for need due to energy deficit and increased need during infections (during the infection and post recovery). All countries had lysine as the primary limiting amino acid. (data not shown). This is confirmed in other examinations of dietary data that have also found lysine as the first limiting amino acids in developing country diets. Prior work in the area also demonstrates that for diets providing over 50 % of protein from cereal sources, protein quality is relatively poor thus affecting biological value and limiting protein utilization(9, Reference Young and Pellett17Reference Young, Bier and Pellett21).

The increase in prevalence of protein inadequacy is incremental following the respective corrections for quality, digestibility, energy deficit and infections. This was as expected since these variables play an important role in defining protein needs within the context of developing countries. We justify the use of adjusted protein requirements above and beyond quality and digestibility on the following basis: firstly, protein requirements are known to be higher in the context of chronic and acute infections(34). For example in the case of acute bacterial infections such as pneumonia and diarrhoea, requirements increase by 20–30 %(1). Requirements (using the indicator amino acid oxidation method (IAAO)) of essential amino acids such as lysine are up to 50 % higher in chronically undernourished adults living in India compared to well-nourished controls (high socio-economic levels and clean environments). Furthermore, following successful treatment for parasites, the requirement for amino acids such as lysine return to their usual level, supporting the interpretation that the increased requirement was attributable to the presence of intestinal parasites(Reference Kurpad, Regan and Nazareth31, Reference Kurpad, Regan and Raj32).

Secondly, interactions between dietary protein and energy are well established(4) and changes in food energy (both below and above energy needs) affect body nitrogen balance. Thirty-three percent of total variation can be explained by nitrogen intake while 36 % can be explained by variation in energy intake with both energy intake and nitrogen intake levels individually effective in improving nitrogen balance. Data from studies at MIT indicate that maintenance of N equilibrium at a low protein intake of 0·57 g/kg body weight requires additional energy ranging from 9–20 %. Furthermore, while other substrates are preferentially utilized when energy intakes are slightly below requirements (protein sparing effect), additional protein metabolized accounts for about 10 % of the energy deficit(Reference Garza, Scrimshaw and Young5, Reference Garza, Scrimshaw and Young6). Work done in Japan also found an impact of energy on maintaining nitrogen equilibrium with a reduction of 20 % in energy availability increasing nitrogen requirement by almost 50 %(Reference Kishi, Miyatani and Inoue7).

An examination of other indicators within the country and region indicate that not surprisingly countries with high prevalence rates of protein inadequacy and/or low utilizable protein levels are also those with high rates of stunting and lower GDP per capita. The linear correlation analysis clearly shows an association between the levels of protein available (when corrected for quality and digestibility and using adjusted requirements) and prevalence of stunting, irrespective of energy supply. Quality of dietary protein will affect linear growth especially at intakes close to maintenance or when energy intake is potentially insufficient. The mechanisms for protein effects on growth are multiple including if protein amounts and digestibility are marginal, net retention will be affected thus potentially compromising growth. In addition, specific dietary essential and non essential amino acids play a role in defining hormonal responses to food including effects on GH release(Reference van Vught, Heitmann and Nieuwenhuizen35).

Evidence indicates that protein restriction leads to low levels of IGF-1 in healthy children(Reference Smith, Underwood and Clemmons36). When older children in the study underwent energy restriction (50 % reduction in intake) or protein restriction (reducing protein from 1·0 to 0·66 g/kg body weight per day) with both forms of restriction led to a significant decrease in nitrogen balance and decline in IGF-1 concentrations as well as concentrations of specific IGF binding proteins. IGFBP-2 was responsive to re-feeding only in the children that were protein restricted. There is also a significant impact of the type and quality of protein on gene expression especially genes associated with insulin like growth factor I and insulin like growth factor binding protein I, both of which play an important role in whole body protein synthesis and growth promotion and body composition(Reference Matsukawa, Inoue and Oishi37Reference Ferreira, Barbosa and Stoppiglia40).

Type of dietary protein seems to have a specific stimulating effect on weight and length gain. Milk intake is positively associated with serum IGF-I concentrations and height suggesting a stimulating effect of milk on insulin like growth factor and subsequently on growth(Reference Van Vught, Nieuwenhuizen and Veldhorst52). Eight year old boys receiving a high milk intake had higher IGF-1 levels compared to boys receiving protein from meat(Reference Hoppe, Molgaard and Juul41, Reference Larnkjaer, Hoppe and Molgaard42). It is postulated that amino acids, peptides specific to milk and/or other milk components (e.g. β-lactoglobulin, α-lactalbumin, immunoglobulins, lactoferrin) are likely to be the active components.

An association between intake of dietary protein and linear growth has been observed in pre-pubertal girls over a 3 year follow up period(Reference van Vught, Heitmann and Nieuwenhuizen35) and a 6 year follow up period(Reference van Vught, Heitmann and Nieuwenhuizen43). High arginine and lysine intakes (ranging from 3·8-4·6 g/day) were inversely associated with fat mass index in pre-pubertal lean girls in both studies(Reference van Vught, Heitmann and Nieuwenhuizen35, Reference van Vught, Heitmann and Nieuwenhuizen43). In an analysis of dietary and anthropometric data collected on Ghanaian children aged 2-13 years, an association has been found between utilizable protein and the risk of being stunted(44). While energy intake was low in the population, it was not significantly associated to stunting levels(44). The finding of the combined effect of improved linear growth and reduced fat mass index is especially interesting in growth promotion practice since stunted children in a setting of rapid diet and nutrition transition are increasingly becoming overweight or obese(Reference Uauy and Kain45).

Oral ingestion of dietary protein, amino acid mixtures to resemble soya protein and an arginine-lysine test drink have been shown to increase growth hormone release in normal women(Reference van Vught, Nieuwenhuizen and Brummer46). Dietary restriction of single essential amino acids including leucine, lysine, methionine and threonine have been shown to decrease plasma IGF-I production but not affect plasma insulin like growth factor binding protein 1 (IGFBP-1 production)(Reference Takenaka, Oki and Takahashi38). However it does decrease IGFBP-I production in hepatocyte cultures(Reference Thissen, Pucilowska and Underwood47Reference Straus, Burke and Marten50).

While we did not examine individual amino acid effects in this study, evidence exists for the specific roles of amino acids (including arginine and lysine) within the context of growth hormone release mechanisms via an effect on the somatotropic axis(Reference van Vught, Nieuwenhuizen and Brummer46). It is postulated that arginine and lysine could have individual or combined effects within this mechanism via the somatotropic axis(Reference van Vught, Nieuwenhuizen and Brummer46). Soya protein ingestion has also been found to increase growth hormone secretion (an effect similar to direct arginine supplementation) however this effect is reduced when soya protein is ingested as part of a meal(Reference van Vught, Nieuwenhuizen and Brummer51, Reference Van Vught, Nieuwenhuizen and Veldhorst52). Ingestion of soya proteins with a carbohydrate or fat alone increased secretion at the same level as soya and/or arginine alone however when soya was combined with both carbohydrate and fat, the effect was reduced. Peak plasma arginine concentrations were higher indicating a role for arginine in the somatotropic activity of protein(Reference Van Vught, Nieuwenhuizen and Veldhorst52).

In conclusion, the findings of this analysis indicate the need to reconsider the adequacy of dietary protein intakes in relation to protein requirement estimates after adjusting for quality, digestibility, burden of infection and energy deficit. There is clearly an association between the quality of protein available at the national level and the prevalence of stunting. Whether this association exists at the individual level needs to be explored by examining cross sectional and cohort data specifically in infants 6 months and older to determine if there is a true effect of protein quality on linear growth pattern. Such a finding would have significant public health and policy implications from the perspective of targeting linear growth and the prevention of stunting, a major cause of disease and disability in the developing world(Reference Zulfiqar, Tahmeed and Robert53).

Acknowledgements

Shibani Ghosh conceptualized, led the analysis, writing and interpretation of the data analysis. Devika Suri was responsible for statistical analysis and preparation of the manuscript. Ricardo Uauy conceptualized the study and contributed to the writing and interpretation of results. The authors declare no conflicts of interest. This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. The authors would like to thank Dr. Nilupa Gunaratna for SAS coding used in food balance sheet data analysis.

References

1WHO (2007) Protein and amino acid requirements in human nutrition. Report of a Joint WHO/FAO/UNU Expert Consultation. Geneva: World Health Organization (WHO Technical Report Series, No. 935); 2007 Contract No.: Document Number.Google Scholar
2Dewey, KG, Beaton, G, Fjeld, C, et al. (1996) Protein requirements of infants and children. Eur J Clin Nutr 50, S119S150.Google ScholarPubMed
3Scrimshaw, NS, Taylor, CE & Gordon, JE (1959) Interactions of nutrition and infection. Am J Med Sci 237, 367403.Google ScholarPubMed
4Pellett PL & Young VR (editors) (1991) The effects of different levels of energy intake on protein metabolism and of different levels of protein intake on energy metabolism: A statistical evaluation from the published literature. International Dietary Energy Consultancy Group. NH, USA: Waterville Valley.Google Scholar
5Garza, C, Scrimshaw, N & Young, V (1976) Human protein requirements: the effect of variations in energy intake within the maintenance range. Am J Clin Nutr 29, 3, 280287.CrossRefGoogle ScholarPubMed
6Garza, C, Scrimshaw, NS & Young, VR (1978) Human Protein Requirements: Interrelationships between Energy Intake and Nitrogen Balance in Young Men Consuming the 1973 FAO/WHO Safe Level of Egg Protein, with Added Non-Essential Amino Acids. J Nutr 108, 1, 9096.CrossRefGoogle ScholarPubMed
7Kishi, K, Miyatani, S & Inoue, G (1978) Requirement and utilization of egg protein by Japanese young men with marginal intakes of energy. J Nutr 108, 4, 658669.CrossRefGoogle ScholarPubMed
8Hoppe, C, Molgaard, C, Thomsen, BL, et al. (2004) Protein intake at 9 mo of age is associated with body size but not with body fat in 10-y-old Danish children. Am J Clin Nutr 79, 3, 494501.CrossRefGoogle Scholar
9FAO/WHO (1991) Protein quality evaluation. Joint FAO/WHO. 66.Google Scholar
10Pellett, PL (2004) The prediction and tabulation of countries where a significant proportion of the population may be at risk of lysine deficiency: International Nutrition Foundation.Google Scholar
11Pellett, PL (1996) World essential amino acid supply with special attention to South-East Asia. Food Nutr Bull 17, 3, 204234.CrossRefGoogle Scholar
12Sasaki, S & Kesteloot, H (1992) Value of Food and Agriculture Organization data on food-balance sheets as a data source for dietary fat intake in epidemiologic studies. Am J Clin Nutr 56, 4, 716723.CrossRefGoogle ScholarPubMed
13UNICEF (2005) State of the World's Children 2005. New York.Google Scholar
14(UNDP) UNDP, Human Development Report 2009 Overcoming barriers: Human mobility and development 2009: Available from:http://hdr.undp.org/en/reports/global/hdr2009/.Google Scholar
15UNSTATS, 2010 [updated 2010; cited 2010]; Available from:http://unstats.un.org/unsd/mdg/SeriesDetail.aspx?srid = 562.Google Scholar
16USDA Nutrient Data : Home. http://www.ars.usda.gov/main/site_main.htm?modecode = 12-35-45-00. Accessed November 1, 2009. 2009 [updated 2009; cited November 1, 2009]; Available from:http://www.ars.usda.gov/main/site_main.htm?modecode = 12-35-45-00 .Google Scholar
17Young, V & Pellett, P (1994) Plant proteins in relation to human protein and amino acid nutrition. Am J Clin Nutr 59, 5, 1203S12012.CrossRefGoogle ScholarPubMed
18Pellett, PL & Ghosh, S (2004) Lysine fortification: Past, present, and future. Food Nutr Bull 25, 2, 7.CrossRefGoogle ScholarPubMed
19Ghosh, S, Pellett, PL, Aw-Hassan, A, et al. (2008) Impact of lysine-fortified wheat flour on morbidity and immunologic variables among members of rural families in northwest Syria. Food Nutr Bull 29, 3, 163171.CrossRefGoogle ScholarPubMed
20Pellett, PL & Young, VR (1988) The contribution of livestock products to human dietary needs with special reference to West Asia and North Africa. In Increasing Small Ruminant Productivity in Semi-Arid Areas, [, editor]. Dortrecht, The Netherlands: Kluwer Academic Publishers for ICARDA, Aleppo, Syria.Google Scholar
21Young, V, Bier, D & Pellett, P (1989) A theoretical basis for increasing current estimates of the amino acid requirements in adult man, with experimental support. Am J Clin Nutr 50, 1, 8092.CrossRefGoogle ScholarPubMed
22Gabert, VM, Brunsgaard, G, Eggum, BO, et al. (1995) Protein quality and digestibility of new high-lysine barley varieties in growing rats. Plant Foods Hum Nutr 48, 2, 169179.CrossRefGoogle ScholarPubMed
23UNU, WHO, FAO (2004) Human Energy Requirements: UNU/WHO/FAO; 2004.Google Scholar
24Wuehler, SE, Peerson, JM & Brown, KH (2005) Use of national food balance data to estimate the adequacy of zinc in national food supplies: methodology and regional estimates. Public Health Nutr 8, 7, 812819.CrossRefGoogle ScholarPubMed
25Ghosh, S, Smriga, M, Vuvor, F, et al. (2010) Effect of lysine supplementation on health and morbidity in subjects belonging to poor peri-urban households in Accra, Ghana. Am J Clin Nutr 92, 928939.CrossRefGoogle ScholarPubMed
26Black, RE, Allen, LH, Bhutta, ZA, et al. (2008) Maternal and child undernutrition: global and regional exposures and health consequences. Lancet 371, 9608, 243260.CrossRefGoogle ScholarPubMed
27WHO (2008) The global burden of disease: 2004 update Geneva: World Health Organization. Google Scholar
28Mata, L (1983) Breast-feeding, Health, and Growth. In Diarrha and Malnutrition: Interactions, mechanisms and interventions, pp. 177202 [, editor]. New York: UNU Plenum Press.CrossRefGoogle Scholar
29Mata, L (1995) The Santa Maria Cauque study: health and survival of Mayan Indians under deprivation, Guatemala. In Community-based longitudinal nutrition and health studies: Classical examples from Guatemala, Haiti and Mexico, [, editor]. Boston: International Foundation for Developing Countries.Google Scholar
30Victora, CG, Adair, L, Fall, C, et al. (2008) Maternal and child undernutrition: consequences for adult health and human capital. Lancet 371, 9609, 340351, PMCID: 2258311.CrossRefGoogle ScholarPubMed
31Kurpad, AV, Regan, MM, Nazareth, D, et al. (2003) Intestinal parasites increase the dietary lysine requirement in chronically undernourished Indian men. Am J Clin Nutr 78, 6, 11451151.CrossRefGoogle ScholarPubMed
32Kurpad, AV, Regan, MM, Raj, T, et al. (2003) Lysine requirements of chronically undernourished adult Indian men, measured by a 24-h indicator amino acid oxidation and balance technique. Am J Clin Nutr 77, 1, 101108.CrossRefGoogle ScholarPubMed
33Black, RE, Brown, KH & Becker, S (1984) Malnutrition is a determining factor in diarrheal duration, but not incidence, among young children in a longitudinal study in rural Bangladesh. Am J Clin Nutr 39, 1, 8794.CrossRefGoogle ScholarPubMed
34WHO (2007) Protein and Amino Acid Requirements in Human Nutrition. [WHO/FAO/UNU, editor]. Geneva: World Health Organization.Google Scholar
35van Vught, AJ, Heitmann, BL, Nieuwenhuizen, AG, et al. (2010) Association between intake of dietary protein and 3-year-change in body growth among normal and overweight 6-year-old boys and girls (CoSCIS). Public Health Nutr 13, 647653.CrossRefGoogle ScholarPubMed
36Smith, WJ, Underwood, LE & Clemmons, DR (1995) Effects of caloric or protein restriction on insulin-like growth factor-I (IGF-I) and IGF-binding proteins in children and adults. J Clin Endocrinol Metab 80, 2, 443449.Google ScholarPubMed
37Matsukawa, T, Inoue, Y, Oishi, Y, et al. (2001) Up-Regulation of Upstream Stimulatory Factors by Protein Malnutrition and Its Possible Role in Regulation of the IGF-Binding Protein-1 Gene. Endocrinology 142, 11, 46434651.CrossRefGoogle ScholarPubMed
38Takenaka, A, Oki, N, Takahashi, S-I, et al. (2000) Dietary Restriction of Single Essential Amino Acids Reduces Plasma Insulin-Like Growth Factor-I (IGF-I) but Does Not Affect Plasma IGF-Binding Protein-1 in Rats. J Nutr 130, 12, 29102914.CrossRefGoogle ScholarPubMed
39Katsumata, M, Kawakami, S, Kaji, Y, et al. (2002) Differential Regulation of Porcine Hepatic IGF-I mRNA Expression and Plasma IGF-I Concentration by a Low Lysine Diet. J Nutr 132, 4, 688692.CrossRefGoogle ScholarPubMed
40Ferreira, F, Barbosa, HCL, Stoppiglia, LF, et al. (2004) Decreased Insulin Secretion in Islets from Rats Fed a Low Protein Diet Is Associated with a Reduced PKA{alpha} Expression. J Nutr 134, 1, 6367.CrossRefGoogle Scholar
41Hoppe, C, Molgaard, C, Juul, A, et al. (2004) High intakes of skimmed milk, but not meat, increase serum IGF-I and IGFBP-3 in eight-year-old boys. Eur J Clin Nutr 58, 9, 12111216.CrossRefGoogle Scholar
42Larnkjaer, A, Hoppe, C, Molgaard, C, et al. (2009) The effects of whole milk and infant formula on growth and IGF-I in late infancy. Eur J Clin Nutr 63, 8, 956963.CrossRefGoogle ScholarPubMed
43van Vught, AJAH, Heitmann, BL, Nieuwenhuizen, AG, et al. (2009) Association between dietary protein and change in body composition among children (EYHS). Clin Nutr 28, 6, 684688.CrossRefGoogle ScholarPubMed
44S Ghosh, D Suri and F Vuvoret al.(editors) Dietary protein quality is associated with risk of being stunted in peri-urban children in Greater Accra. 2nd World Public Health Congress on Nutrition; 2010; Porto, Portugal.Google Scholar
45Uauy, R & Kain, J (2002) The epidemiological transition: need to incorporate obesity prevention into nutrition programmes. Public Health Nutr 5, 1a, 223229.CrossRefGoogle ScholarPubMed
46van Vught, AJAH, Nieuwenhuizen, AG, Brummer, R-JM, et al. (2008) Effects of Oral Ingestion of Amino Acids and Proteins on the Somatotropic Axis. J Clin Endocrinol Metab 93, 2, 584590.CrossRefGoogle ScholarPubMed
47Thissen, J, Pucilowska, J & Underwood, L (1994) Differential regulation of insulin-like growth factor I (IGF-I) and IGF binding protein-1 messenger ribonucleic acids by amino acid availability and growth hormone in rat hepatocyte primary culture. Endocrinology 134, 3, 15701576.CrossRefGoogle ScholarPubMed
48Jousse, C, Bruhat, A, Ferrara, M, et al. (1998) Physiological concentration of amino acids regulates insulin-like-growth-factor-binding protein 1 expression. Biochem J 334, 1, 147153.CrossRefGoogle ScholarPubMed
49Pao, C, Farmer, P, Begovic, S, et al. (1993) Regulation of insulin-like growth factor-I (IGF-I) and IGF-binding protein 1 gene transcription by hormones and provision of amino acids in rat hepatocytes. Mol Endocrinol 7, 12, 15611568.Google ScholarPubMed
50Straus, D, Burke, E & Marten, N (1993) Induction of insulin-like growth factor binding protein-1 gene expression in liver of protein-restricted rats and in rat hepatoma cells limited for a single amino acid. Endocrinology 132, 3, 10901100.CrossRefGoogle ScholarPubMed
51van Vught, AJAH, Nieuwenhuizen, AG, Brummer, R-JM, et al. (2008) Somatotropic responses to soya protein alone and as part of a meal. Eur J Endocrinol 159, 1, 1518.CrossRefGoogle Scholar
52Van Vught, AJAH, Nieuwenhuizen, AG, Veldhorst, MAB, et al. (2009) Growth hormone responses to ingestion of soyprotein with or without fat and/or carbohydrate in humans. E Spen Eur E J Clin Nutr Metab 4, 5, e239e244.CrossRefGoogle Scholar
53Zulfiqar, AB, Tahmeed, A, Robert, EB, et al. (2008) What works? Interventions for maternal and child undernutrition and survival. Lancet 371, 9610, 417440.Google Scholar
Figure 0

Table 1 Calculation of increased protein needs due to infection and moderate energy deficit, by IMR tertile (infant mortality rate per 1000 births) for 115 countries

Figure 1

Table 2 Supply per capita per day of energy, protein and utilizable protein, prevalence of stunting and wasting, Gross Domestic Product (GDP) by regions

Figure 2

Fig. 1 Estimates of adult daily protein requirement, with added needs for infection and moderate energy deficit, by country-level IMR tertile, for 115 countries.

Figure 3

Fig. 2 Differences in risk estimates of protein inadequacy, calculated using total protein, utilizable protein (UP), and UP plus higher requirements for infection and moderate energy deficit, by regions of the world. * Significant difference between risk estimates of protein inadequacy using total protein (line a) and utilizable protein (line b) compared current requirements versus inadequacy estimates using utilizable protein (line c) compared current requirements that have been adjusted for infection and energy deficit. Analysis conducted using paired t-tests or non-parametric tests (p < 0·05).

Figure 4

Fig. 3 Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in countries with less than 2000 kcal/capita/day energy supply.

Figure 5

Fig. 4 (a) Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in Sub Saharan African countries with 2000–2500 kcal/capita/day energy supply. (b) Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in South and South-East Asian countries with 2000–2500 kcal/capita/day energy supply.

Figure 6

Fig. 5 (a) Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in Sub Saharan African countries with 2500–3000 kcal/capita/day energy supply. (b) Risk of protein inadequacy as determined by protein needs adjusted for infection and moderate energy deficiency, compared with energy supply and risk of protein inadequacy determined by total and utilizable protein, in South and South-East Asian countries with 2500–3000 kcal/capita/day energy supply.

Figure 7

Table 3 Correlation coefficients for relationships between country-level nutrient supply, prevalence of stunting and GDP per capita, for 115 countries

Figure 8

Table 4 The association (linear regression) between prevalence of stunting and total and utilizable protein supply (g/capita/day) for 115 countries

Figure 9

Table 5 Associations (linear regression) between prevalence of stunting and prevalence of protein inadequacy (total protein and utilizable protein with adjusted protein requirements) and energy supply, for 115 countries