Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Tiefu Shen and Jean-Pierre Habicht
Introduction
Nutrition surveillance means watching over nutrition in order to make decisions that will lead to improvements in population nutrition according to the internationally agreed upon definition [1]. This definition has two critical implications: first, that decision making should be relative to interventions or activities at the population level, in contrast to decisions made at the level of the individual on a case-by-case basis; and, second, that the information issuing from nutrition surveillance should actually be used in decision-making. The ultimate purpose is to provide pertinent and timely information to enable policy makers better to take nutrition considerations into account and programme planners and managers better to design and implement interventions to protect nutrition [2].
It has been recognized that nutrition surveillance can be classified operationally into four different types: a timely warning information system, a means to manage and evaluate programmes, a means to establish policy and planning [2], and, more recently, a means to identify problems and provide advocacy. Although the underlying principles and general goals for all these subtypes are basically the same, their differentiation in terms of implementation and of the specific policy served by the information is situation dependent. In this paper, we will discuss only one of these: nutrition surveillance for policy and planning.
We chose this topic because China is now considering establishing nutrition surveillance, and pilot data collection is already proceeding. This effort is intended ultimately to cover all of China, not just a few provinces. Those involved want information for policy formulation and programme planning for decisions about economic development, health, and food security. At some future time it may be necessary to be concerned about improving the efficiency and efficacy of programmes. One may hope that never again will a timely warning nutrition surveillance system be needed to institute short-term prevention measures against famines; however, other needs for rapid interventions may become apparent as China's food and nutrition circumstances become better understood. Thus we are not belittling the future importance of these other kinds of surveillance but are only concentrating first on what is most immediate.
The role of the information user in designing nutrition surveillance
The major goal of nutrition surveillance for policy and planning is to make nutrition and nutrition-related information available to decision makers. However, the nutrition of populations cannot be protected adequately if it has not been considered in the policies that affect populations. Analogously, nutrition cannot be considered if the information needs of decision makers are not taken into account during the process of information generation. The information should not only enable policy makers to foresee nutrition problems but also provide insights into how to prevent them. Because of the heterogeneous nature of the causes of these problems, all government and non-government sectors that affect economic and physical well-being also affect the nutrition of the population. Decision makers in all these sectors could use information from surveillance to be sure that the impact of their decisions on nutrition is taken into account, even if the action to be taken has non nutrition objectives. However, in order to use surveillance information, the potential users must know that it is available, have it available at the moment they need it, and understand its implications.
FIG. 1. Sectors and agencies whose policies affect nutrition and food security
The best way to be sure that users have the right information at the right time is to involve them in the design of the surveillance system to be sure that the data coming from the system meet their needs. The Chinese Academy of Preventive Medicine is undertaking user surveys as a first step to involve these decision makers and their advisers. The purposes of the surveys are (1) to make decision makers and their advisers aware of the proposed nutrition surveillance system, (2) to identify major users of the information, in particular, some of each kind of user with whom the surveillance unit can work more closely to ensure that all users of that kind receive relevant and timely information, and (3) to ascertain the users' concerns about specific policy issues so as to incorporate their needs into deciding the timing and format of the information presentation.
A pilot survey was conducted in 1989 to ascertain what kinds of influential persons would use the data and for what purposes. Twenty-three potential users, representing a wide range of agencies and backgrounds, who were known to be interested in food and nutrition were chosen by the senior staff of the Chinese Academy of Preventive Medicine. They came from the State Council, the State Scientific Commission, the ministries of Agriculture, Public Health, Planning, Commerce, and Light Industry, several national academies and universities, and the mass media. The survey was done by a questionnaire, followed by personal interviews. Most of those interviewed responded that they had used some sort of food and nutrition information from various sources before. All of them agreed strongly that such information was useful and requested that it be provided to them. Neither of these answers was surprising in view of the selection process.
We grouped the potential users into four categories relative to policy-making: seven administrators. Five policy consultants, eight researchers and scientists, and three people working in mass media. As expected, only the administrators said they had direct influence on setting policies and planning programmes, but the high-level consultants and researchers had direct access to decision makers and influenced their decisions. This contrasted with those from the mass media, who concentrated on more general information diffusion and affected decisions in this way. However, the consultants and researchers said that they also wrote articles and reports to advocate improvements.
The major findings were that researchers and consultants prefer regular reports, whereas mass media and administrators want information when they need it, often at irregular intervals. All preferred long and comprehensive reports, but most also wanted short summaries. One administrator requested summaries only.
Based on the results of this first survey, a second survey is being designed that will go beyond identifying potential users of nutrition surveillance information to identify the kinds of policy and Programme planning needs that must be met. Special attention is being given to the sectors named in figure I and the agencies in these sectors. This information should help to ensure that policies in education, development, welfare, health, food marketing, and food supply will result in eliminating conditions of want (e.g. starvation, nutrient deficiencies), in controlling diseases of affluence (e.g. diabetes, cardiovascular disease, cancer), and in fostering food security. As indicated above, these surveys are but a first step in what must be a continuing dialogue with the users of nutrition surveillance information so that the system can better serve them.
FIG. 2. Height of seven-year-old boys in China. 1920-1989
The relationship of nutrition surveillance to policies and programmes
Rural-urban discrepancies in socio-economic development are characteristic for many developing countries, and this is true also for China. Rural development policies address this issue by identifying possible reasons for this discrepancy and correcting them. For instance, in China one reason may be that the government bought food at much lower prices than farmers could sell it on the free market by exacting quotas from the farmers. This food was then sold to urban dwellers at subsidized prices. Thus farmers earned less and urban dwellers ate cheaper foods than if the food had been bought and sold on the free market. For the farmers. the difference between the price of the food on the free market and that paid by the government is a tax; and the difference between the price paid for the food by the urban dweller and the price in the urban market is a subsidy. The result of this policy may be one reason why rural children did not eat enough to grow properly.
Figure 2 shows the height of seven-year-old boys in the lao-shao-bian-qiong (old, ethnic, border, poor) areas, which are thought to be the poorest rural areas, all rural areas, suburban areas, all urban areas, and large cities. It also shows the evolution of these heights over time. Some of the data include all children and others only healthy children; the 2-cm difference in height between the two is taken into account in the figure. Dotted lines join data points from different surveys in similar populations to make it possible to visualize the secular trends more easily. These data capture the deterioration in growth during World War II but do not show the effects of the famines of 1958-1962, which took some 30 million lives-more deaths than in any other famine in history [7]. The stunting of the famine survivors is undocumented.
In spite of all these reversals and uncertainties. it is clear that the growth of urban Chinese children has improved markedly since the 1940s. The same is true for suburban children in the 1970s and for many rural children in the 1980s. However, the growth of urban children is levelling off well below the likely growth potential of Chinese children as indicated by international growth standards [8] based on a well-nourished population. The mean growth of well-fed children in all populations is within 1 or 2 cm of these two standards [9, 10]. The reason for the levelling off in urban children in the 1980s is not clear. but it may indicate that some subgroup is not adequately nourished and that the proportion of these children is not becoming smaller. This is worrisome.
Even more worrisome is the growth of levelling off in suburban children at double the degree of mean stunting (as measured by the standards) of urban children, which bodes ill for a continued rise in the growth of rural children above these stunted levels. Of greatest concern is the fate of children in the lao-shao-bian-qiong areas, who are three times as stunted as the urban children and have shown no improvement over the past decade.
The major differences in the growth of these young boys are not due to any racial factors but to the rural urban discrepancy in socio-economic status and an inadequate diet [9]. Endemic diseases such as diarrhoea cannot produce so much stunting when an adequate diet is available [11].
This kind of information about the rural-urban discrepancy in growth and some ideas about its causes are typical of what one should expect from nutrition surveillance. It can be used to advocate policies and programmes to improve rural well-being. It can help to identify reasons for inadequate nutrition that can be remedied by policies and programmes. It can also target development programmes to areas of greatest need for these interventions, such as the government allocation of poverty-relief aid to lao-shao-bian-qiong areas, which attempts to redistribute certain resources at the national level to underdeveloped rural areas and regions at risk of famine. Finally, it can monitor the improvement of nutrition status brought about by policies and programmes. Thus, for instance, the Chinese economic reforms are known to have improved the income of some farmers [12]. Is this a general phenomenon, or are there some who have not benefited or who are even worse off and who therefore need policies and programmes to protect their nutrition even while the benefits of economic reform are retained? Nutrition surveillance, especially of the growth of children, can monitor this evolution.
It appears in China that the size of two- and three year-old children is the best indicator, because growth is stunted most severely between 6 and 24 months of age [6]. For nutrition surveillance purposes, growth is a more easily interpreted indicator of nutritional wellbeing than is dietary intake [13], although food data are essential to understand what has to be done.
When food intake is marginal or inadequate, disease affects nutrition status adversely [11|. Monitoring the effects of the Chinese restructuring of health services on nutrition and the growth of children is therefore particularly important in poorer areas with stunted children. With the dissolution of rural agricultural communes, local collectively financed support for basic health workers has disappeared in many areas [14]. With the management of health services returned to the private sector, it is estimated that nearly half of more than 700,000 village-level health stations require substantial payment for services and medicines [14]. This cost of medical care may deter the use of this service by the poor. Backed by the new commitment of the government to restructure the rural primary health care system [15] to attain "Health for all by the year 2000," nutrition surveillance can play an important role in quickly identifying any detrimental effects of these changes in health services and in assisting policy makers in directing limited resources to the areas and populations that most need them. Again, the growth of children is the most satisfactory indicator.
In these poorer populations breast-feeding and good weaning practices are important. This is particularly true where environmental sanitation is poor [5, 16]. Thus nutrition surveillance must not only collect data about breastfeeding and weaning but also deliver information about the adequacy of water and sanitation so as to target appropriate interventions to the most needy populations.
Another area of health concern is a marked rise in the consumption of animal foods. In the decade 19781988 pork consumption doubled in rural communities and beef consumption almost doubled in urban areas [17]. These animal foods are unhealthy when eaten in excess because they are rich in fats associated with cardiovascular disease [18]. Figure 3 shows how the proportion of energy from fat increases with income in both rural and urban areas. There is a large increase in this proportion among those with incomes above 800 yuan per year in urban areas, which is surprising because one expects a smooth progression of increasing energy from fat with increase in income. Is this because these urban households benefit from hidden subsidies so that their real income is greater than what they earn? In any case, this rise in fat consumption is unhealthy. It is recommended that no more than 30% of calories in the diet should come from fat.
FIG. 3. Percentage of energy from fat by income level (calculated from refs. 19 and 20)
For the wealthier urban population, the mean level of energy in the diet from fat is already more than 30%, which means that many individuals are consuming much more fat, mostly from pork and beef [19]. Fat from mammalian sources and eggs is considered less healthy than that from fish, poultry, and vegetables. Unfortunately, intake of mammalian and egg fat will continue to go up with increased income in groups with higher fat consumption [19].
On the other hand, when a diet has too little fat, it is not rich enough in energy for young children. This is probably the case in much of rural China, where the mean intake of fat is low (15%). It is even lower in the poorest quintile, among whom fat provides only 12.6% of calories [20], meaning that many consume even lower percentages. This is a probable explanation of the poor growth in these children. Fortunately, small increments in income among those poor will go preferentially to increasing the fat content in their diet [20], which may be a sensitive indicator of the effectiveness of programmes and policies to increase their income.
An increase of energy intake is another characteristic of increasing income in China [19, 20]. This is beneficial among the rural poor, whose intakes relative to energy expenditures are probably marginal. Among the urban rich, who tend to expend less energy in work, it leads to obesity. Obesity is the main determinant of diabetes, which rises to high proportions among the overweight [21] and also contributes to hypertension [18]. Monitoring and characterizing the obese through nutrition surveillance will indicate ways to reduce obesity and where to target these interventions. Such interventions will have to go beyond price policies in the context of food security, and must include nutrition education and exercise. If activities directed toward food security take the unhealthy effects of a high mammalian and egg fat intake into account, they will also include food pricing and other marketing interventions to reduce consumption of these items.
Here we discuss only national food security, which is concerned with ensuring that enough food is available to meet demand. National food security is a prerequisite for, although not a guarantee of, the security of market food (i.e. assurance of local supply), household food, and individual food. Ensuring national food security requires monitoring food production and supply, which is usually done by agricultural and food data systems that lie outside the purview of nutrition surveillance. These systems generally rely on historical information about consumption to estimate future needs. The estimates are satisfactory so long as food consumption behaviour does not change. However, they lead to serious errors when behaviours change, as when famine occurred in Indonesia even though the food supply was adequate. Extremes in seasonal loss of income meant that people could not buy food in the traditional way, and thus many died, not for lack of food in the marketplace but for lack of access or entitlement to that food [22].
China faces an analogous problem at the national level, but the cause will be increases rather than decreases in purchasing power with continued improvement in socio-economic development, especially if it is equitably distributed among all Chinese. The first finding from the new Chinese Nutritional Surveillance System is that eating behaviour will change markedly as people have more money to spend on food [17, 19, 20]. Families will turn more and more to animal foods, most of which will be from grain-consuming animals, and this will increasingly be the case as family income rises. The Chinese agricultural potential for grain may not be able to support a major shift to higher-quality grain-fed livestock, which consume three to six times more grain to produce a given quantity of calories for human consumption than if people ate the grain themselves [23]. In this sense, grain-fed animal food is inefficient.
Just as important as changes in behaviour in response to improved income are the changes caused by differences in prices of food. Policies and programmes that increase the price or decrease the availability of inefficient foods and decrease the prices of more efficient foods can reduce the growth in demand for agricultural production. The changes in food purchasing behaviour with increases in income and changes in food prices can be estimated from the nutrition surveillance data collected by the State Statistical Bureau [19, 20].
As indicated above, modern knowledge of nutrition shows that many of the most inefficient foods are unhealthy, and their consumption should be discouraged not only for food security reasons but also for health reasons [18]. As we begin nutrition surveillance in China, it is essential to give special attention to these inefficient foods and to identify healthier and more efficient foods in order to predict how their consumption will change with improved socioeconomic development and how changes in relative price and availability between the healthy and unhealthy foods can decrease consumption of the latter. With this knowledge from nutrition surveillance, policy makers and planners can devise and implement ways to encourage greater production and sale of efficient, healthy foods relative to inefficient, unhealthy foods.
For instance, as the price of chicken goes down, Chinese will buy 12% more chicken per 10% drop in price. They will only buy 6% more pork for a similar percentage fall in pork prices [17]. Thus, for an equal proportional drop in prices they will increase their consumption twice as fast for chicken as for pork. The same information can be elicited less quantitatively by asking people whether they would prefer to eat chicken or pork if both had the same price; the answer overwhelmingly favours chicken. Incidentally, this finding is similar to that in most populations around the world. Thus the only reason more pork is eaten in China right now is that chicken is so much more expensive.
The implications of this finding are important. It means that by investing to improve efficiency in chicken production and marketing so as to bring the price of chicken below that of pork, as it is already in much of the world, and not investing in pork production and marketing so that pork prices do not fall commensurably, it would be possible to foster a shift in the consumption of the richer populace from less healthy, inefficient pork to healthier, more efficient chicken. Lowering the price of chicken, which can more easily be accomplished through appropriate investments than could similarly lowering the price of pork, would also benefit the poorer Chinese, who could then more easily afford this healthy meat. Equally important, improving the availability and decreasing the costs of healthy, efficient foods not only improves the health of the population but also assures food security for the nation.
Finally, nutrition surveillance can monitor the success of programmes and policies to improve nutrition, health, and food security. This monitoring is essential because the programmes and policies will inevitably be less than perfect at the beginning. Good surveillance data will be necessary to permit fine-tuning them efficiently.
These examples touch on only some of the issues of concern to users of nutrition surveillance information, but they give a good idea of the relationship between such information and its use in policy-making and planning.
Issues relating to data and their analysis
Various data sources in China are already part of the nutrition surveillance system (e.g. the State Statistical System), and others probably could be. To be useful, the sources must be able to deliver valid, relevant, and timely information to decision makers on a continuous basis.
Some examples of data sources appear useful for nutrition surveillance. The latest national nutrition survey in China was done in 1982, and the national-level aggregated results were published in 1986 [24]. Further disaggregated analyses would have been useful, but significant changes can reasonably be expected to have occurred in the nutrition status in the population since 1982 because of economic reform, and thus some of this information may now be irrelevant. The problems are twofold. We need more up-to-date information, which would require more frequent surveys. But more frequent surveys are not useful unless the data can be analysed in a timely fashion. This latter is the real bottleneck, which the Chinese Academy of Preventive Medicine and the State Statistical Bureau have addressed so that information from more frequent surveys can be generated quickly enough that increasing their frequency will be justified.
Monitoring over time requires that the data be comparable from period to period. This has not historically been the case for anthropometry, because different kinds of children have been sampled, as indicated above. Even national data from 1985 [25] and 1987 [26] are not comparable for this reason. Nutrition surveillance requires stability in survey design and data collection to ensure data comparability over time.
For this reason, data from monitoring the growth of children in the context of primary health care have never been successfully used for national policymaking and planning anywhere in the world. Such data are not collected to help in making decisions for policies and planning but rather for use in patient care. Therefore, the samples are poorly described and usually inadequate, with the result that one cannot be sure who and what kind of persons were measured. Caution with this approach is therefore urged because of the bad experiences of many countries which have spent a great deal on growth-monitoring data that then could not be put to practical use. Other approaches for the local use of growth-monitoring data are more promising [27] but still are not relevant for making decisions at the national level.
Principles for nutrition surveillance for policy and planning
From the above examples, certain principles can be summarized. Nutrition surveillance for policy and planning should meet the needs of the decision makers. Linking the information directly to the policymaking process is the single most crucial step and also the most obvious difference between this and other data-generating processes. At present, China is beginning surveillance at the national level for national decision makers. In the future, as more and more autonomy is given to the provinces, expansion may be necessary to ensure adequate sample sizes for more local information.
The timing of information for policy and planning is also important. Policy-making and Programme planning are usually long-range endeavours with predictable cycles. Therefore, the necessary linkages to the policy-making process must be built even before the information is generated so as to be sure the timing is correct. Establishing such links is also a remedy for a common difference in perception between scientists and decision makers about the utility of the information.
The information necessary to guide policies and programmes almost always requires data that go beyond those usually considered related to nutrition such as dietary intake and anthropometry, because most decisions have to take into account the economic, sociocultural, and biological determinants of nutrition. Thus the information required often has to cover not only dietary intake and outcome data but also the other important determinants of nutrition status, including income, food prices, sanitation, education, mortality, and morbidity.
Certain prerequisites for a national nutrition surveillance system require careful consideration. First, designing the data collection, implementing it to adequate levels of data and sampling quality, and analysing the data collected require highly skilled, stable, and well-motivated personnel. Second, the skills for all these activities across all the disciplines are rarely found in one institution. This is why presently both the Chinese Academy of Preventive Medicine and the State Statistical Bureau are involved. Even then, more information can always be obtained from the data than even the largest consortium of data-collection agencies can deliver. This argues for methods to release data to other analysts quickly and efficiently so that more information can be generated. This is difficult to implement, less for technical than for bureaucratic reasons, as we know from experience [28]. A clear policy should be enunciated that methods and means will be found to release data quickly to other agencies with qualified analysts (e.g. universities and research centres). This policy has to be made now at this early stage before bureaucratic tradition makes changes difficult.
Another reason why it is important that data from the nutrition surveillance system be analysed widely is that relevant information will come from various sources, which will make it more likely that relevant issues will be considered at the time decisions are made, especially those decisions that have non nutrition objectives. Different information sources have access to decision makers in different sectors and know how to present the information in the language and format that all of these individuals understand. Releasing the data to a wide range of qualified information sources will be facilitated by the food and nutrition advisory committee that is presently being formed.
Finally, the system itself should be put under surveillance to be sure that it is associated with a policymaking process and is available for use in policy considerations. The quality of the information delivered by the system and its availability to decision makers should be differentiated from the specific decisions made on the basis of the information. Wise decisions from a nutrition point of view may be frustrated because nutrition is often not a major consideration in the development of policies that have major impact on nutrition, such as income generation, taxation and subsidies, and even food security. Nutrition surveillance information only makes it more likely that these issues will be taken into account.
Acknowledgements
This paper was presented in part at the International Symposium on Diet, Nutrition, and Socioeconomic Development, 5-8 June 1990, Beijing, China. The work was supported by the Chinese government and the Interagency Food and Nutrition Surveillance Programme of the United Nations Subcommittee on Nutrition. Special thanks are given to Dr. Robert Parker (UNICEF, China), Dr. Alberto Pradilla (WHO, Geneva), Dr. William Kean (WHO, China), and Dr. Elizabeth Morris Hughes (FAO, Rome).
References
1. Habicht J-P, Mason JB, Kevany J. Report of a Joint FAO/UNICEF/WHO Expert Committee on Methodology of Nutritional Surveillance. World Health Organization Technical Report Series, no. 593. Geneva: WHO, 1976.
2. Mason JB, Habicht J-P, Tabatabai H. Valverde V. Nutritional surveillance. Geneva: WHO, 1984.
3. Piazza A. Food consumption and nutritional status in the PRC. Boulder. Colo, USA: Westview Press, 1986.
4. UNICEF. Children and women of China. Beijing: UNICEF, 1989.
5. Ge KY et al. Further analysis of 1987 National Growth Survey. In: Proceedings of the International Symposium on Diet, Nutrition, and Socioeconomic Development. Beijing: Chinese Academies of Agriculture and of Preventive Medicine, 1991.
6. Chang Y, Zhai FY. The nutritional status of preschool children in some poor areas of China. In: Proceedings of the International Symposium on Diet, Nutrition. and Socioeconomic Development. Beijing: Chinese Academies of Agriculture and of Preventive Medicine, 1991.
7. Ashton B, Hill K, Piazza A, Zeitz R. Famine in China. Pop Dev Rev 1984; 10:613-45.
8. WHO. Measuring change in nutritional status. Geneva: WHO, 1983.
9. Habicht J-P, Martorell R, Yarbrough C, Malina RM, Klein RE. Height and weight standards for preschool children: how relevant are ethnic differences in growth potential? Lancet 1974,1(7858):611-15.
10. Martorell R, Habicht J-P. Growth in early children in developing countries. In: Falkner F, Tanner JM, eds. Human growth: a comprehensive treatise. 2nd ed. Vol. 3: Methodology: ecological, genetic and nutritional effects on growth. New York: Plenum Press, 1986:24159.
11. Lutter CK, Mora JO, Habicht J-P, Rasmussen KM, Robson DS, Sellers SG, Super CM, Herrera MG. Nutritional supplementation: effects on child stunting because of diarrhoea. Am J Clin Nutr 1989;50: 1-8.
12. Tiang JY. China's current agricultural situation and policy. Beijing Rev 1990; 2:7-8.
13. Filer LJ, Habicht J-P, Beaton GH et al. Nutrient adequacy: assessment using food consumption surveys. Report of the Subcommittee on Criteria for Dietary Evaluation, Coordinating Committee on Evaluation of Food Consumption Surveys, Food and Nutrition Board. Washington, DC: National Academy Press, 1986.
14. Aldis WL. Privatization of care in China. Lancet 1989; 16:1456.
15. Chen MZ. Promotion of health in China: consistency and stability. Outlook Weekly 1990; 2:7-8.
16. Habicht J-P, DaVanzo J, Butz WP. Mother's milk and sewage: their interactive effects on infant mortality. Paediatric 1988; 81(3):456-61.
17. Pinstrup-Andersen P, Yang DN, Xian ZD, Yang Y. Changes in incomes, expenditures, and food consumption among rural and urban households in China during the period 1978-88. In: Proceedings of the International Symposium on Diet, Nutrition, and Socioeconomic Development. Beijing: Chinese Academies of Agriculture and of Preventive Medicine, 1991.
18. National Research Council. Diet and health implications for reducing chronic disease risk. Report of the Committee on Diet and Health. Food and Nutrition Board, Commission on Life Sciences. Washington DC: National Academy Press, 1989.
19. Yang DN, Yang Y, Pinstrup-Andersen P. Food consumption and food acquisition behaviour among urban households in China, 1987 and 1988. In: Proceedings of the International Symposium on Diet, Nutrition, and Socioeconomic Development. Beijing: Chinese Academies of Agriculture and of Preventive Medicine, 1991.
20. Zhu XD, Xian ZD, Pinstrup-Andersen P. Food consumption and food acquisition behaviour among rural households in China, 1988. In: Proceedings of the International Symposium on Diet, Nutrition, and Socioeconomic Development. Beijing: Chinese Academies of Agriculture and of Preventive Medicine, 1991.
21. National Institutes of Health. Health implications of obesity. Ann Int Med 1985;103(6):979-1077.
22. Brooks RM, Abunain D, Karyadi D et al. A timely warning and intervention system for preventing food crises in Indonesia: applying guidelines for nutrition surveillance. Food Nutr 1985;11(2):37-43.
23. Liu CH. Learn to feed one billion people. Proceedings of Chinese Association of Agricultural Students and Scholars, June 24-25,1989, Cornell University, Ithaca, NY, 1989.
24. Chen CM. The national nutritional survey in China, 1982: summary results. Food Nutr 1986;12(1):59-60.
25. Capital Institute of Pediatrics. Children growth survey, 1985. Beijing: Capital Institute of Paediatrics, 1986.
26. State Statistics Bureau of China. Statistics of 1987 survey on the situations of children in China. Beijing: China Statistical Publishing House, 1989.
27. Pelletier D. The role of information in the planning, management and evaluation of community based nutrition programs. In: Pinstrup-Andersen P, Pelletier D, Alderman H, eds. Beyond child survival: enhancing child growth and nutrition in developing countries. Ithaca, NY, USA: Cornell University Press, 1991.
28. National Academy of Public Administration. Improving the health and nutrition examination survey. Washington, DC:NAPA, 1980.
