Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu



Continue - 2

These large differences show that income changes affect low- and high-income families in significantly different ways. Low-income families tend to increase fat intake and % fat more than high-income families. The increase is usually attributed to consumption of high-fat foods. These differences demonstrate two perspectives: as income improves, diet quality increases, which helps to eliminate undernutrition; and at the same time, as income increases, there are some negative effects in that rapid increases in fat intake are associated with increased likelihood of obesity and related chronic diseases. Programmes to alleviate poverty or develop the economy should take into account the issues of dietary deficit and excess.

TABLE 8. Signs of the coefficients from the two regimens

Low- income regimen High-income regimen
Fat Calories % fat
+ education smoking moderate activity alcohol use income male smoking alcohol use northern moderate activity heavy activity income education smoking urban age moderate activity income
- urban heavy activity family size northern age male age education family size urban   male alcohol use family size northern heavy activity

TABLE 9. The simulated effects of changes in selected variables on nutrient intake

  Change Fat Calories % fat
One regimen Low- income regimen High - income regimen One regimen Low- income regimen High - income regimen One regimen Low- income regimen High income regimen
Income + Y 1.000 0.800 3.40 -0.200 -5.1 2.8 -0.3 0.300 1.000 - 0.100
Age +1 year -0.071 -0.110 -0.065 -5.9 -6.9 0.3 0.022 0.015 -0.007
Gender female to male -6.461 - 7.042 -1.779 260.1 259.7 270.8 -3.963 -4.039 - 3.138
Alcohol use   2.038 2.724 -1.370 112.6 127.7 24.2 -0.044 -0.008 -0.141
Urban rural to urban 1.020 -0.484 3.714 -143.2 -127.9 -243.6 2.001 1.338 4.14 1
Physical activity light to moderate 4.295 5.002 0.834 83.2 92.4 53.7 0.449 0.583 -0.476
  light to heavy -6.392 -5.671 1.847 291.6 301.4 262.2 -4.250 -4.053 -1.972

Other factors affecting nutrient intake

The differences in nutrient consumption behaviour between the two groups are also present for age, gender, alcohol use, urban or rural residence, and physical activity effects. For example, when age increases one year, fat intake decreases significantly 0.11 g for low-income families yet decreases only insignificantly 0.065 g for high-income families. Caloric intake decreases significantly 6.9 kcal for low-income families yet increases insignificantly 0.3 kcal for highincome families. The % fat from energy increases 0.015 percentage point and decreases 0.007 percentage point for the two respectively. Both are insignificant. When rural residents move to urban areas, the fat intake of low-income families decreases 0.48 g, while that of high-income families increases 3.7 g, although both are insignificant. Caloric intake significantly decreases for both types of families; the decrease in high-income families is twice that in low-income families. In high-income families, the % fat increase is almost four times that of low-income families. This shows that the diet of low-income families may be impaired when they move to urban areas, whereas the diet of high-income families could be improved.

Similarly, physical activity is associated differently with fat and caloric intakes according to income. Moderate physical activity considerably increases the absolute fat intake of the low-income regimen. Heavy physical activity considerably decreases the fat intake of the low-income regimen but increases that of the high-income regimen. There is no clear interpretation of these differences.

Although the effects of some of the variables on certain nutrients are not statistically significant, they are theoretically or practically associated with food or nutrient consumption and may potentially help to interpret consumption behaviour. Thus, they are still included in the equation and the discussion.

Other considerations

The results suggest that when income and other factors vary, the propensity to increase or decrease some nutrients is substantially different for low-income and high-income families. The effects on the former families are stronger.

The estimates of coefficients a and are similar for some variables in terms of their magnitude and direction. It is possible that the effects of some components of a and are equal. Therefore, it is worth while to perform some additional tests to verify whether Hi: a i= i. Computed statistics are based on the t statistic [20]. The results (table 10) indicate no significant difference among the influence of education, family size, smoking status, and northern residence on fat intake in low- and high-income households.

TABLE 10. t test of equality of coefficients of two regimens

Explanatory variable Fat Calories % fat
Education 0.92 6.37* 1.52
Smoking 1.77 4.03* 3.61 *
Family size 1.02 0.58 1.10
Northern residence 1.64 4.83 1.28

Values are t statistics.
Significant at 1% level.

It may be argued that a number of changes may have occurred concurrently with growth in income. For example, education level improved, and changes in agricultural policy resulted in changes in the nature of the food supply by expanding possible choices or by reducing market prices because of a more stable supply. Indeed, the market price increased considerably in past years, which affects the lower-income group more than the high-income group. Compared with full income growth and its effect on diet and all other aspects in China over the past 10 years, most of the other changes occurring during the time would scarcely bias our results.

 


Discussion

There is substantial structural difference in consumption behaviour between groups whose household income is below Y7,000 per year and those whose income is greater. The results indicate that income increases operate differently for very high-income families than for other families. Among high-income families in the multivariate analysis, income increases are associated with decreased fat and calorie intake, whereas the opposite occurs among lower-income families. In particular, a Y1,000 increase would result in a 3.4-g increase in fat intake in lower-income families. If this same relationship were examined with one model, then a Y1,000 increase in income would result in a 0.8-g increase. It is clear that a very different sense of the magnitude of the income and fat relationship would be derived from these two models. The results also show very different effects of area of residence, physical activity patterns, and several other factors on diet.

At present in China the focus is on the nutrition transition. The government organized the National Commission for Food Reform and Development to address many of the problems noted in this paper.

This represents a path-breaking effort for a low-income country to try to address problems of under-and over-nutrition concurrently. The size of relationships such as income and fat, particularly as it relates to income and price increases, is of particular importance to this and a second commission in China. Ignoring this relationship can mask important income and dietary intake relationships and lead to misleading conclusions.

This study also fits into a larger set of changes that are affecting many lower-income countries as they develop. A large transition in diet is occurring in these countries, and its implications for each income group should be understood [22].

References

1. Popkin BM. Time allocation of the mother and child nutrition. Ecol Food Nutr 1980;9:114.

2. Tucker K, Sanjur D. Maternal employment and child nutrition in Panama. Soc Sci Med 1988;26:605-12.

3. Nicklas TA, Webber LS, Thompson B. Berenson GS. A multivariate model for assessing eating patterns and their relationship to cardiovascular risk factors: the Bogalusa Heart Study. Am J Clin Nutr 1989;46:1320-7.

4. Monteiro C, Mondini L, Medeiros de Souza AL, Popkin B. The nutrition transition in Brazil. So Paulo: University of So Paulo, 1994.

5. Johnson RK, Smicklas-Wright H. Crouter AC, Willits FK. Maternal employment and the quality of young children's diets: empirical evidence based on the 19871988 nationwide food consumption survey. Pediatrics 1992;90(2,pt.1):245-9.

6. Doran HE. Applied regression analysis in econometrics. New York: Marcel Dekker, 1989.

7. Akin JS, Guilkey DK, Popkin BM. The school lunch program and nutrient intake: a switching regression. Am J Agric Econ 1983;65:478-85.

8. Food and Agriculture Organization. Income effect on the structure of diet. In: Provisional indicative world plan for agricultural development. Vol 2. Rome: FAO, 1970:50005.

9. Chen C. Dietary guidelines for food and agricultural planning in China. In: Proceedings of the International Symposium on Food, Nutrition and Social Economic Development. Beijing: Chinese Academy of Preventive Medicine, 1991:40-8.

10. Timmer CP, Falcon WP, Pearson SR. Food policy analysis. World Bank Publications. Baltimore: Johns Hopkins University Press, 1983.

11. Alderman H. New research on poverty and malnutrition: what are the implications for research and policy? In: Lipton M, Van der Gaag J. eds. Including the poor. Washington, DC: World Bank, 1992:115-31.

12. Chaudhri R. Timmer CP. The impact of changing affluence on diet and demand patterns for agricultural commodities. World Bank staff working paper. Washington, DC: World Bank, 1986.

13. Mincer J. Market prices, opportunity costs, and income effects. In: Christ CF, Friedman M, Goodman LA et al., eds. Measurement in economics. Stanford, Calif, USA: Stanford University Press, 1963:67-82.

14. Pinstrup-Andersen P. Yang D, Xian Z. Yang Y. Changes in incomes, expenditures, and food consumption among rural and urban households in China during the period 1978-88. In: Chen C, ed. Proceedings of the International Symposium on Food, Nutrition and Social Economic Development. Beijing: Chinese Academy of Preventive Medicine, 1990:447-58.

15. Popkin BM, Paeratakul S. Zhai F. Ge K. Body weight distribution pattern in China: the results from 1989 and 1991 China health and nutrition survey. Am J Health 1995;85:690-4.

16. Popkin BM, Ge K, Zhai F et al. The nutrition transition in China: a cross-sectional analysis. Eur J Clin Nutr 1993;47:333-46.

17. National Research Council, Commission on Life Science, Food and Nutrition Board. What is America eating? Washington, DC: National Academy Press, 1986.

18. Beaton CH, Milner J. Corey P et al. Sources of variance in 24-hour dietary recall data: implications for nutrition study design and interpretation. Am J Clin Nutr 1979;32:254659.

19. Beaton GH, Milner J. McGuire V, Feather TE, Little JA. Source of variance in 24-hour dietary recall data: implications for nutrition study design and interpretation-carbohydrate sources, vitamins, and minerals. Am J Clin Nutr 1983;37:98695.

20. Kleinbaum D&, Kupper LL, Muller KE. Applied regression analysis and other multivariable methods. 2nd ed. Boston, Mass, USA: PWS-Kent, 1988.

21. Goldfeld SM, Quandt RE, eds. Studies in nonlinear estimation. Cambridge, Mass, USA: Ballinger, 1976.

22. Popkin BM. The nutrition transition in low income countries: an emerging crisis. Nutr Rev 1994;52(9): 285-98.

 


World balance of dietary essential amino acids relative to the 1989 FAO/WHO protein scoring pattern


Hoshiai Kazuo

Abstract

The world imbalance of dietary essential amino acids was studied using the latest available protein-supply data (1987-89) and the revised 1989 FAO/WHO protein scoring pattern in comparison with the 1973 FAO/WHO pattern, the 1985 FAO/WHO/UNU pattern, and a pattern proposed by Young et al. in 1989. The results obtained using the 1989 FAO/WHO scoring pattern indicate that the first limiting amino acid for developed countries is usually tryptophan, and that for developing countries is mainly lysine. Similar findings resulted with the Young pattern, but results using the 1973 and 1985 patterns differed substantially. On the basis of the 1989 FAO/WHO pattern, lysine was found to be the first limiting amino acid in the dietary protein supplies of 121 of the 164 countries studied worldwide; it is estimated that the total lysine deficiency in these 121 countries, the amount that would be needed to bring it to the level of the second limiting amino acid, was 1.15 million metric tons per year for 1987-89. In addition, same global correlations of protein and amino acid supplies with gross domestic product were recalculated in US dollars at 1985 prices.

 

Introduction

In an earlier paper [1], I reported on the imbalance between supplies of and requirements for essential amino acids (EAA) by country, region, and economic system and for the world, with averages for four three-year periods, 1972-74, 1975-77, 1979-81, and 198486, calculated in accordance with the 1973 FAO/WHO scoring pattern [2], the 1985 FAO/ WHO/UNU scoring pattern [3], and a scoring pattern proposed in 1989 by Young et al. [4, 5].

In 1989, on the basis of new evidence, FAO/WHO published a revised EAA scoring pattern [6] to correct the 1985 pattern. The 1989 report recommended that the amino acid composition of human milk should continue to be the basis of the scoring pattern to evaluate protein quality in foods for infants under 1 year of age, but that the amino acid scoring pattern proposed in 1985 by FAO/WHO/ UNU for children of preschool age (2-5 years) should be used to evaluate dietary protein quality for all age groups above infancy (table 1). The 1989 FAO/WHO pattern has EAA requirements similar to those of the Young pattern but with a higher requirement for lysine, a slightly higher requirement for tryptophan, and a slightly lower one for leucine.

In addition, the statistical data for average annual protein supply and gross domestic product (GDP) by country are now available for 1987-89.

In view of the worldwide significance of those developments, the change in the calculated nutritional imbalance of EAA is here re-evaluated using the 1989 FAO/WHO scoring pattern in comparison with the previous patterns to revise the conclusions of my earlier paper [1]. (Part of this study has been presented elsewhere [7].) Furthermore, the correlations of GDP with total protein supplies and animal protein ratios (APR) are recalculated in US dollars at 1985 prices to update similar data in the previous paper.


Contents - Previous - Next