Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Changes in food and essential amino acid availability from 1961 to 1992

Group and country data: 1961 versus 1992

The availabilities of food energy (kilocalories per day), total protein, animal protein, cereal protein, and pulse-soya protein (grams per day) are shown in table 13 (see TABLE 13. Comparison of food energy, protein, and essential amino acids (mg/day) between 1961 and 1992 for selected regions and countries). Also shown are the availabilities of essential amino acids (milligrams per day) for developed countries, developing countries, and least developed countries as well as for the three major countries of the Indian subcontinent -Bangladesh, India, and Pakistan. These are tabulated for 1961 and 1992, the first and last years of the current FAO/Agrostat [17] data set.

Significant changes took place between 1961 and 1992 in food availability for both the developed and developing country groups. For the developing countries, increases can be seen for food energy (26%), total protein (26%), and animal protein (76%), although pulse-soya protein declined by 35%. The availabilities of the essential amino acids (milligrams per day) all increased, with lysine increasing by 25% from a mean value that was below the estimated need in 1961 to one slightly above the need in 1992. Developed countries also showed increases of 15% to 30% for food energy, total protein, animal protein, and lysine although these all took place from a considerably higher baseline level.

In contrast, however, these increases between 1961 and 1992 did not occur for the poorest countries. For the least developed countries and for Bangladesh, over the same period, changes are almost non-existent, and some nutritional indicators, including lysine even declined. For example, for the least developed countries, animal protein availability in 1961 (9.8 g/day) was greater than in 1992 (8.7 g/day), as was animal protein availability for Bangladesh (7.1 versus 4.8 g/day). In contrast, India and Pakistan both showed increases over this period, especially for food energy, total protein, and animal protein. Legume availability declined by as much as 47% in India, however.

When lysine values are considered, the developed countries showed increases from 5,413 mg/day in 1961 to 6,666 mg/day in 1992, and the developing countries from 2,401 to 3,014 mg/day, while the least developed countries declined from 2,433 to 2,278 ma/day during this same period. Similarly, in Bangladesh lysine availability declined from 2,193 mg/day in 1961 to 1,883 mg/day in 1992, while in India and Pakistan there were modest increases.

In table 14 (see TABLE 14. Comparison of food energy, protein, and essential amino acids (mg/g protein) between 1961 and 1992 for selected regions and countries) these same groups and countries are compared, with the animal, cereal, and pulse-soya protein tabulated as percentages of total protein, and the essential amino acids shown in terms of protein quality as milligrams per gram of protein. Animal protein percentage increased in both developed and developing countries, as well as in India and Pakistan, but declined in the least developed countries and Bangladesh. The increases were uneven but were as high as 40% for developing countries as a group. In contrast, animal protein percentage declined by 8% in the least developed countries and by 27% in Bangladesh.

Pulse-soya protein percentage showed significant declines in almost all groups, with a nearly 50% decline in the developing countries as a group, a 51 % decline in India, and a 63% decline in Pakistan. The proportion of protein from cereals increased in all countries except the developed countries. These changes in dietary patterns are reflections of both wealth and the generally greater availability of cereals.

For lysine expressed as milligrams per gram of protein, only the developed countries were above the FAD/WHO 1991 requirement figures in both 1961 and 1992 [5]. The other countries or groups changed little over this period, except for Bangladesh, where lysine fell from 48 to 44 mg/g protein as a result of the lower availability of animal protein.


Countries of South-East Asia: 1961 to 1992

Although for reasons of space, the full tabular data are not presented, the changes in food energy, fat, protein, and lysine (both milligrams per day and milligrams per gram of protein) were evaluated for Bangladesh, India, Japan, South Korea, Pakistan, Sri Lanka, and Thailand for the period 1961 to 1992 using the FAO/Agrostat database [17]. Protein (grams per day) was categorized as total protein, animal protein, cereal protein, and pulse-soya protein. Lysine values were calculated from these data using the regression equation shown above.

Several patterns of change with time emerge from these data and are illustrated in figures 3 (see FIG. 3 Six-country comparison of total protein availability 1961 1992 [17]) to 5 for all the above countries except Sri Lanka. Values for all variables can be deduced for Sri Lanka, since they were generally between those for Bangladesh and those for India. The key determinant of dietary modification appears to be changes in economic status of the countries over the years. Wealth (per capita gross national product) significantly correlates with protein availability (quantity and quality) in the diet. Bangladesh and Sri Lanka changed least, and India, Pakistan, and Thailand showed modest increases. In South Korea and Japan the changes were marked. In Japan protein availability was already at a reasonable level in 1961. In South Korea, however, the values in 1961 were quite low, especially for animal protein, and major changes did not really start until the 1970s and then proceeded rapidly.

Total protein and animal protein are major determinants of lysine availability and are illustrated for six countries in figures 3 and 4 (see FIG. 4. Six-country comparison of animal protein availability 1961-1992 [17]). In both Bangladesh and Sri Lanka (not shown), total daily protein availability remained on the order of 40 to 45 g/day, and animal protein availability averaged only 5 to 10 g/day, with lower values for animal protein in Bangladesh than in Sri Lanka. In South Korea animal protein availability was 6 g/day in 1961, but this had increased to 33 g/day by 1992. Japan, whose total protein availability ranged from 73 to 98 g/day over this period, showed changes in animal protein availability from 24 to 56 g/day. India and Pakistan changed far less, with total protein availability averaging between 50 and 60 g/day for both countries. Animal protein availability ranged between 6 and 10 g/day in India and between 13 and 15 g/day in Pakistan. Although not shown, changes in food energy availability generally were parallel to increases in total protein. Hence Bangladesh and Sri Lanka changed the least and South Korea the most.

For all the countries, the lysine values (milligrams per day and milligrams per gram of protein) parallel the changes in food energy and protein. These changes for lysine (milligrams per day) are graphically illustrated (except for Sri Lanka) in figure 5 (see FIG. 5. Six-country comparison of lysine availability 1961-1992 [17]). Very distinct differences can be observed. Japan differs from all others in that lysine (milligrams per day), already at a moderately high level in 1961 (4,000 mg/day), increased continuously until 1992 (6,500 mg/day). In contrast, South Korea and Thailand had inadequate lysine values in 1961 (2,300 and 2,200 mg/day), but both had increased by 1992. Thailand increased to a modest extent, but the South Korean diet was providing nearly 4,700 ma/day by 1992.

The other three countries remained almost static over the three decades, with Bangladesh at inadequate levels throughout and even declining. Because of the very high proportion of wheat in the Pakistani diet, the values for Pakistan may have been slightly overestimated by the use of the regression equation and are thus closer to the values for India than they appear from figure 5. The mean values for both these countries approximate requirement estimates, and therefore on statistical grounds many would be at risk in both societies. Bangladesh, India, Pakistan, and Sri Lanka all have lysine values that appear to be marginal or below requirements for the whole period, but South Korea and Thailand, which had inadequate lysine values in the earlier years, have shown significant increases since the early 1980s.

When lysine is considered as milligrams per gram of protein, again only the values for Japan can be considered adequate throughout the whole period. Although the South Korean diet has shown significant changes, from 40 mg/g protein in 1961 to 54 mg/g protein in 1992, protein quality must still be considered marginal. Poorer quality is, however, compensated for by adequate protein quantity. Other countries, especially Bangladesh but also Pakistan and India, average below the estimated needs of 58 mg/g protein [5].


Changes for Pakistan: Food balance sheet calculations for selected years

Because of future research plans and hence particular concern for the lysine value (in milligrams per day and milligrams per gram of protein) of diets in Pakistan, calculations made directly from food balance sheet data are shown in tables 15 and 16 (see TABLE 15. Changes in food energy, protein, and essential amino acids (mg/g protein); selected years Pakistan 1961-92 calculated directly from FAO food balance sheets and TABLE 16. Changes in food energy, protein, and essential amino acids (mg/day); selected years Pakistan 1961-92 calculated from FAO food balance sheets) for selected years, at about five-year intervals, between 1961 and 1992. When the lysine values are compared with those shown in figure 5 for Pakistan, i.e., those that were calculated from the regression equation, it will be seen that although the values parallel each other, the values directly calculated from food balance sheets are somewhat lower than those calculated from the regression equation. In practice such a discrepancy will be larger for Pakistan than for other countries of the region because of the greater predominance of wheat in the Pakistani diet.

The regression equation used for prediction uses cereals as a group rather than individual cereal species. Since rice has a higher lysine content than wheat (38 versus 24 mg/g protein), lysine availability will be overestimated for Pakistan because of its predominantly wheat diet.

For the Pakistani diet as a whole, total protein and animal protein availability (grams per day) have increased modestly. Cereal protein percentage has been fairly constant (50%60%). Animal protein percentage has also been fairly constant but has increased slightly since the late 1980s and was 32% in 1992. Although pulse and soya protein were never significant contributors to the protein supply in Pakistan as they were in India (pulse-soya protein percentage in 1961 was 13% in Pakistan and 24% in India), by 1992 pulse-soya protein percentage had declined to 5% in Pakistan compared with 12% in India. Consequent on these several changes the lysine value has been fairly constant (44-48 mg/g protein: table 15) over the years. However, because of the slight increase in total protein, the total lysine increased slightly from 2,374 mg/day in 1961 to 2,621 mg/day in 1992. These values remain below the estimated requirement values (table 4) and give cause for concern as to the lysine availability and the protein quality of the Pakistani diet as a whole.

Thus, increases in national wealth significantly affect the amount of lysine consumed. A more expensive diet containing more legumes or more animal protein will provide more lysine Further elaboration of survey data from Pakistan and India and the lysine value of regional diets will be discussed below.

Wealth was also a key determinant of the lysine value of diets in the past. The lysine values for a number of diets from the older literature have been calculated. These values are not reported here, but calculations from dietary information presented by Hutchinson [22] and Sherman [23] show that the types of diets consumed by workers and peasants in Britain in the later years of the nineteenth century were generally adequate in their protein value. In addition, as in the case of current world diets, lysine was the amino acid that was most deficient in the diet in relation to requirements. The relations between food cost and lysine value have also been confirmed by evaluation of diets reported by the US Department of Agriculture [24] and the Carnegie United Kingdom Trust [25]. In the latter monograph detailed information was presented on a large-scale dietary survey performed during the depression in pre-war Britain.


Survey data from Pakistan and India

After regional and country data originating from food balance sheets have been evaluated, the next level of information comes from survey data showing intra-country variability [26].

Pakistan survey data

In the survey report, food consumption data (grams per day) are tabulated for the food groups—cereals, animal foods, milk, pulses, vegetables, fruit, sugar, and oils—for a number of regions, provinces, and cities of Pakistan. By making assumptions concerning the individual foods comprising the reported food groups, nutritional intake data were calculated from these values together with the lysine value (milligrams per gram of protein) using Massachusetts Nutrient Data Bank values and standard dietary calculation procedures. These data, following adjustment to match reported protein intakes, are shown in table 17 (see TABLE 17. Food energy, protein, animal protein percentage, and essential amino acids (mg/g protein) for Pakistani regions) and illustrate regional variations in intake. That the assumptions made concerning the individual foods in the food groups were reasonable can be seen by comparing the values with the whole country data for the mid-1980s presented in tables 15 and 16, which were calculated from food balance sheet data [17].

The intra-country variations observable from these calculations are important and demonstrate significant differences from the country average. The cities and the surrounding areas have lower food energy availability than the rural areas in general and the north in particular. Total protein and animal protein consumption was, however, often greater in the urban and city areas. When compared with requirement values (milligrams per gram of protein: table 5), lysine is the only amino acid in significant deficit. The next limiting amino acid is threonine. The latter, however, is only a little below requirement levels. Both observations agree with expectations for cereal-based diets.

Food energy, protein, and fat as well as the daily availabilities of essential amino acids (milligrams per day) are shown in table 18 (see TABLE 18. Food energy, protein, fat, and essential amino acids (mg/day) for Pakistani regions). When comparing availability with requirement values (see table 4), only lysine can be considered marginal. In general, lysine availability is roughly proportional to food energy availability. There are, however, a number of exceptions. Both Sind and North-West Frontier Province (NWFP) cities have high lysine values despite moderate food energy and protein intakes. This can be accounted for by the higher availability of animal protein. In the north (where intake data appear questionable), because of the very high intake of protein recorded, the intake of lysine in milligrams per day is high even though the amount of lysine in milligrams per gram of protein is low. For a number of areas, e.g., Baluchistan urban and city, Lahore, Karachi, and Azad K, the overall diets, including lysine values, give cause for concern. Mean values below mean requirements are indicative of a high probability that a significant section of the population is at nutritional risk.

In table 19 (see TABLE 19. Means, standard deviations, and coefficients of variation for food energy, protein, and essential amino acids for 16 Pakistani regions), means, standard deviations, and coefficients of variation (CV%) are shown for data from the 16 regions presented in tables 17 and 18 in order to demonstrate the variability in statistical terms. When food energy, protein, and essential amino acids (milligrams per day) are considered, CV% values are similar throughout and range between 17.5% and 21.4%, with the CV% for fat being somewhat higher at 31.4%. The CV% values are much lower when amino acids are expressed as milligrams per gram of protein. The CV% is 8.9% for lysine, 3.0% for threonine, and ranges between 0.8% and 2.5% for the other essential amino acids. As with other observations (e.g., table 9), the CV% for lysine supply is greater than those for the other essential amino acids.

Indian survey data

Protein and lysine values were calculated from dietary survey data from 18 states in India. These data were obtained from Dr. A. V. Kurpad, Project Coordinator of the Nutrition Research Centre, Bangalore, and were preliminary information on the Indian National Food Intake Survey from the National Nutrition Monitoring Bureau. Since the data available were not original raw food intake data, a number of assumptions were required to allow for processing. The data for food groups consumed (cereals, pulses, vegetables, etc.) were mainly in the form of graphs. Protein (grams per day) and food energy (kilocalories per day) were also shown. Rural data were expressed for individual states, whereas the urban data were shown for different income levels ranging from high-income families to slum dwellers.

Dietary patterns in India differ considerably from state to state, especially in the staple cereal, and hence in the proportions of rice, wheat, and millet. In general, more wheat is consumed in the north and more rice in the south. Assumptions on foods and the proportions of cereals consumed within the various states were made with the assistance of Ms. Shibani Ghosh, an Indian graduate student who was familiar with local dietary patterns. Pulse protein was assumed to be derived equally from lentils, chickpeas, and pigeon peas. No meat protein was described as being in the diets, and thus all animal protein was assumed to be in the form of milk and milk products. This assumption is reasonable because a large proportion of Indians are vegetarian. Further, as will be demonstrated later, the amino acid compositions of most animal foods show broad similarities.

For urban areas the food consumption data were shown for different income levels (high, middle, low, industrial labour, and slum dwellers) without state origin being given. Thus it was not possible to account for the regional dietary differences. For these calculations approximately 60% of the cereal protein was considered to originate from wheat protein and 40% from rice.

Amino acid calculations were performed on the food availability data using the Massachusetts Nutrient Data Bank values for the content of essential amino acids. The results for the Indian states are shown in tables 20 and 21 (see TABLE 20. Food energy, protein, and essential amino acids (mg/day) for Indian states). As with the data for Pakistan, there is considerable intra-country variation, but the availability of essential amino acids (milligrams per day) parallels the availability of food energy and protein. Punjab and Madhya Pradesh have the highest availabilities of food energy and protein, whereas Tamil Nadu and Tripura have the lowest. Punjab also differs from the others in the much higher level of animal protein and hence also has the highest level of lysine At 3,576 mg/day this was in sharp contrast to West Bengal and Tripura, where the lysine supply was less than 1,900 mg/day.

Protein quality (essential amino acids in milligrams per gram of protein) and the percentage sources of dietary protein are shown in table 21 (see TABLE 21. Food energy, protein percentages, and essential amino acids (mg/g protein) for Indian states). The very high dependence on cereal protein, with six states having more than 80% of their protein from cereals, and the general low dependence on animal protein (except for Punjab) is shown clearly. Pulse protein showed varied availability. As would be expected from these distributions and from comparison with earlier diet analyses, there are low values throughout for lysine Indeed, 11 of the 18 states have lysine values below 40 mg/g protein.

Table 22 (see TABLE 22. Means, standard deviations, and coefficients of variation for food energy, protein, and essential amino acids for Indian states) shows means, standard deviations, and CV% for the states. This parallels the summary data for Pakistan shown in table 19 and again illustrates intra-country variability. When food energy, protein, and essential amino acids (milligrams per day) are considered, CV% values are similar throughout and range between 16% and 19%. The CV% values are much lower when amino acids are expressed in milligrams per gram of protein. The CV% for lysine is 6.2%, whereas other essential amino acids generally range between 2% and 4%.

Urban area values, separated into income categories, are shown in table 23 (see TABLE 23. Food energy, protein, and essential amino acids for Indian urban areas in terms of income levels). The influence of income on all variables is strong, with that on lysine being especially noticeable both in milligrams per day and in milligrams per gram of protein. As much as 3,508 mg lysine per day was available in the high-income group, compared with 1,992 mg/day in the slum area group. For quality considerations (milligrams of lysine per gram of protein), the range over the same income groups was from 52 to 38 mg/g protein. Animal (milk) and legume protein were much higher, in both absolute and percentage terms, in the high-income group than in the slum group. These animal and legume protein differences were certainly causal, together with the greater dependence on cereal in the poorer areas, in affecting lysine availability. When CV% values were considered, the values for lysine were far higher, both in milligrams per day and in milligrams per gram of protein, than for the other essential amino acids.

Contents - Previous - Next