Contents - Previous - Next

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

Urban nutrition in developing countries: some lessons to learn

Rainer Gross and Carlos Augusto Monteiro

There is no doubt that in recent years concern about nutrition and health in urban areas has been increasing. Several important publications have appeared, such as that of Schürch and Favre [1] and the URBIS newsletter. Two workshops have been held in Great Britain [2; 3], and an article on this subject appeared in the Food and Nutrition Bulletin [4].

Definition of the urban environment

The nutritional condition of a population and the causes of problems depend on among other things the environmental characteristics of the community. Therefore it is necessary to distinguish precisely between rural, urban, and metropolitan areas. Until now, what constitutes an urban area has been defined structurally by the number of inhabitants living in the conglomeration: communities with more than 20,000 inhabitants have generally been called urban. However, there are many conglomerations with more than 20,000 inhabitants that are just large villages and should be designated rural. Therefore, the urban area needs to be defined from a functional, not a structural, point of view.

A rural society lives mainly from agriculture, whereas urban society depends on a higher level of economic diversification. This can have important consequences with respect to nutrition. Whereas the nutritional status of a rural community often suffers from seasonal climatic fluctuations [5-7], particularly when there is a latent lack of food, that of an urban population is less influenced by such changes [8] because of its economic diversity. In the case of drought, for example, urban populations are affected less be cause of their differentiated possible sources of income; the income of the rural population, which depends mainly on agricultural production, is affected severely by lack of rainfall.

The appearance of the metropolis in this century makes a further functional definition necessary to distinguish it from urban areas. The metropolis is more complex, not only from the economic but also from the cultural and environmental points of view. This may influence the etiology of health and nutritional problems in different ways.

The importance of family income

Several studies have shown that the nutritional status of a population is determined by its economic situation. In general, the poorer the population, the higher the prevalence of malnutrition. This can be demonstrated by comparing the gross domestic product of countries and the nutritional status of their populations [9], and also the nutritional status of populations classified by their wages within a country or a smaller area of a country [10], or even within a city [11]. For this reason international development agencies, such as the World Bank, use nutritional indicators to assess the level of development achieved.

Within urban communities such as shanty towns, slums, and residential areas, however, the nutritional status of children as stratified by family income may not necessarily show significant differences [12]. This could be due to the smaller economic differences in the communities, but there may be other reasons for it. Figure 1 shows the growth retardation of preschool children from different socio-geographic areas of Sao Paulo classified by their families' income level [11]. It can be seen that an increase in family income does not result in a linear decrease in growth retardation. In very poor communities, improvement in family income first leads to increased body growth, but further economic improvement has very little biological effect in pre-school children.

Furthermore, children in the slums and shanty towns are shorter than those from families of the same income group in the residential areas. These two facts suggest that, besides family income, other factors such as environment (water supply, sewerage facilities, health services, etc.) or even cultural background could limit growth. It has been shown in other studies that socio-cultural factors within a community such as origin and education of parents, or environmental factors such as sanitation facilities, can be of much more importance than the economic conditions of households [12; 13].

The metropolis shows several characteristics of a biological system, such as hierarchical organization of its elements, cybernetic control of subsystems, structural and temporary oscillations and instability, a partially open and partially closed state, and self organization of the structure and its function [14]. Although the metropolis is subject to a process of breakneck development, it has a high level of inherent stability due to a variety of complex interrelationships. Greater diversity and complexity in a biological system provide a better buffer against environmental disturbances. This is valid also for a complex system such as a metropolis.

To take the analogy further, what the intake of energy and nutrients is to a micro-organism, income is to the community and the family. On the basis of a survey undertaken in Rio de Janeiro, Brazil, between 1980 and 1983 (during an economic recession), it has been hypothesized that a deterioration in the family wage does not necessarily lead to a decrease in the nutritional status of children [15]. Families try to buffer the decrease in the wages by purchasing cheaper foodstuffs, reducing waste, and pursuing diversified forms of employment.

These facts lead to the conclusion that an increase in family income alone does not necessarily lead to an automatic improvement in nutritional conditions.

A new pattern of feeding practice

It has been observed worldwide that urbanization reduces the period of breast-feeding and causes earlier weaning. However, studies in Sao Paulo and Rio de Janeiro show a new trend. More and more mothers tend to breast-feed their children longer [15; 16]. Those who are establishing this new feeding pattern are better-educated mothers from higher-income families.

New patterns of nutritional problems


The inhabitants of metropolitan areas (and their nutritional problems) are more heterogeneous than those of rural areas. On the one hand we find a society living under poor socio-economic conditions displaying the classic problems of malnutrition. In epidemiological studies carried out between 1983 and 1986 in the three largest metropolitan areas of Brazil, infants and pre-school children mainly showed a high prevalence of stunting (10%15%), but a low prevalence of wasting (2%-5%) [11; 12; 15]. The reduction in body growth rate and weight was found only in children 10 to 12 months old and older (fig. 2). In contrast to the case in rural areas, this growth retardation started several months after breast-feeding had ceased and dietary supplementation commenced [12]. These data may indicate that malnutrition in this age group is caused less by hunger (or, rather, lack of energy) than by poor sanitation and health facilities.

School-age children show a different picture. The older those from the deprived section of the population become, the higher the prevalence of wasting, without there being a major increase in stunting [17]. Children in some government schools showed a prevalence of wasting of up to 24%. Undernutrition seems to increase in this age group, although all the schools surveyed maintained feeding programmes with food of appropriate quality. There may be different causes (such as low food intake in the family or in the school) that are as yet unknown.

On the other hand, there is a socio-economically better-off group which displays the nutritional problems of populations of the industrialized countries. For instance, children at a private school in the above mentioned study area displayed a prevalence of obesity of 18% [17].

Fig1 The influence of family income on stunting in children up to six yeas old.

FIG. 2. Mean height and weight of infants and pre-school children in urban communities of Sao Paulo (N = 1,003) and Belo Horizonte (N = 253) compared to NCHS growth curves


In the metropolitan areas of Brazil anaemia seems to be more prevalent than acute malnutrition (tables 1 and 2). Two studies [12; 18] found no association between these nutritional problems. This may indicate that in metropolitan areas there are more groups with different nutritional risks.

Neglected urban risk groups

The paucity of data currently available demonstrates that the comparison of social categories such as urban and rural is too broad and non-specific to help us understand the nutritional problems of the metropolitan population, since urban society and the causes of nutritional problems are too heterogeneous to allow generalizations. It seems rather more useful to identify risk groups and search for the causes of their particular problems.

In most cases, pregnant women, lactating mothers, and infants are the main risk groups in rural communities, and, therefore, nutritional surveillance and interventions are concentrated on them. However, it is not clear whether the pattern of risk is the same in urban areas. In Brazil, there are some indications that deterioration of nutritional status in low-income families due to food restriction during economic crises occurs particularly in older children (table 2) and physically highly active adults [19], and less in infants [15].

The elderly

Until now, international nutritional research has mainly been concerned with mothers and children as the most vulnerable groups. The life expectancy of the population in developing countries is increasing [9]. With the slow decrease in birth rate, the proportion of elderly people in these societies is growing. This is particularly so in urban regions, where life expectancy is higher and the birth rate is lower than in rural areas. We are far from knowing the magnitude and gravity of nutritional problems of the elderly in developing countries. There is not even a simple, widely accepted methodology for measuring their nutritional status, such as exists for infants and children (e.g., anthropometry).

The urban homeless

Despite the fact that homelessness is recognized as a serious and growing urban problem, the magnitude and causes of this problem are not known. Since the part of the population that is homeless has no stable physical base, it is difficult to evolve scientifically acceptable methods for estimating their number and composition. Estimates of the number of homeless in the United States vary from about 250,000 to upwards of 3 million [20]. Practically no data are available from urban areas of developing countries. Unaffiliated persons living in extreme poverty suffer from an extremely high prevalence of physical and mental disability.

Within the homeless population, street children need special consideration. In contrast to the homeless adult, who often lives in social isolation without permanent contact with any other person, street children respond to the challenge of their life problems by interacting with other children, from informal groups to highly structured gangs.

The prevalence of malnutrition in these children varies considerably, depending on the city [21]. In Bogota, for example, vitamin deficiency and malnutrition seem not to be serious problems because most restaurants allow street children to scrape leftovers from the plates. In contrast, in Karachi and Calcutta many of these children are malnourished. The lack of hygiene, early sexual contacts and prostitution, the wide use of drugs, and the inaccessibility of health care facilities are responsible for high morbidity, which in turn is directly or indirectly connected to the nutritional status of this marginal group.

TABLE 1. Prevalence of malnutntion in children of various age groups living in different types of dwelling area in cites in Brazil

Location and date of study Type of dwelling areaa Age group (years) Number of children

Malnutrition (% of children)

Stuntedb Wastedc Obesed
São Paulo
1984e FS 0-5 208 9.4 1.2 1.0
1984e R 0-5 736 3.9 3.1 5.4
Belo Honzonte
1986f F 0-6 254 12.0 2.9 2.1
Rio de Janeiro
1983g F 0-2 1,205 6.7 0.7 -
1985h F 7-14 484 10.3 12.3 2.8
1986h R 7-14 272 2.3 3.5 18.0

a. F = shanty town (favelas); S = slum; R = residential area.
b. <90% of normal NCHS reference height for age.
c. <80% of normal NCHS reference weight for height.
d. >120% of normal NCHS reference weight for height
e. Ref. 11.
f Ref. 12.
g. Ref. 15.
h. Ref. 17.

TABLE 2. Prevalence of anaemia in children living in different types of dwelling types of dwelling area in cities of Brazil

Location and date of study Type of dwelling areaa Age group (years) Number of children

Prevalence (% of children)

Anaemiab Severe anaemiac
São Paulo
1984d FS 0-5 209 45.4 21.9
1984d R 0-5 740 32.6 12.8
Belo Horizonte
1986e F 0-6 244 29.9 10.7

a. F = shanty town; S = slum; R = residential area.
b. Haemoglobin < 11 ml/dl.
c. Haemoglobin < 9.5 ml/ dl.
d. Ref. 18.
e. Ref. 12.

Individuals with psycho-social disorders

For a long time it has been observed that the prevalence of psycho-social disorders, such as depression, is higher in urban than in rural areas [22]. On the basis of a theory that, with the social process of urbanization, community relationships (gemeinschaftlich) are replaced by secondary networks (gesellschaftlich)

[23], and with the support of the findings of other researchers [e.g. 24], the conclusion has been drawn that the higher-risk psycho-social disorders are caused by the greater social isolation of the urban population [22]. It has been reported that rural habitation appears to be a buffer against major psycho-social disorders [25]. According to one study, though, the traditional urban-rural dichotomy may be inappropriate for socio-psychiatric research, since differences were found to be concentrated in two minorities, namely, unemployed men and unpartnered women [26]. All these findings, however, were made in developed countries with Western-style societies. Little is known about the magnitude or causes of psycho-social disorders in the metropolises of developing countries. If the observations of Kovness et al. [26] are also to be found there, the magnitude of the problem is likely to be much greater.

Individuals with psycho-social disorders often show significant weight loss or gain due to poor or increased appetite. However, we do not know to what extent the disorders lead directly or indirectly to the various forms of malnutrition. The whole syndrome, including loss of interest in activities and lack of energy, diminished ability to concentrate, indecisiveness, and decreased effectiveness and work productivity, may lead to neglect of the rest of the family. This culminates in poorer nutritional status of the most vulnerable group, the children.


Because of the differences in ecology in metropolitan dwelling areas, the higher population densities, and differences in social behaviour, the metropolitan risk profile for infectious diseases differs from that in rural areas. For example, urban populations without sanitation facilities show higher prevalence rates for diarrhoeal diseases than do those in rural areas [27].

The prevalence of human immune virus (HIV) infections in Africa varies significantly between rural and urban populations [28]. In Rwanda, 1.3% of the rural population are infected, compared with nearly 18% in the urban conglomerations. Although it is unclear whether nutritional status influences susceptibility to HIV infection or affects the development of manifest acquired immunodeficiency syndrome (AIDS), the drastic reduction of weight that occurs during the development of the disease means that its prevalence must be considered in future, at least in the context of interpreting anthropometry in nutritional monitoring and surveillance.


As shown by highly industrialized countries, precipitate industrialization and development may burden the environment with incalculable amounts of substances, many of them synthetics. Lack of financial resources has forced many developing countries to dispense with expensive measures for reducing this contamination. The high concentration of industry combined with a high population density creates a particular public health problem for urban populations. Cities such as Alexandria [29], Jakarta [30], Mexico City [31], Bombay [32], and Seoul [33] suffer from high air pollution. Many of the people exposed to this contamination, unlike those in more industrially developed countries, also show signs of malnutrition.

TABLE 3. Rates of diarrhoea and acute respiratory infections (ARI) in pre-school children from different socioeconomic strata of Sao Paulo

  Number of children Diarrhoeaa ARIb
Family income
high 106 2.8 25.5
middle high 163 6.1 24.5
middle low 279 7.9 2x.3
low 264 11.7 30.1
Housing area
residential 739 6.1 26.2
slums and shanty towns 209 12.9 35.4

Sources: Refs. 34 and 35.

. Attack rate (number of episodes per too children-months). b. Point prevalence (percentage).

There are only minor differences between different socio-economic and socio-geographic groups in Sao Paulo in the prevalence of acute respiratory infections [34], in contrast to the case for diarrhoea [35] (table 3). This may be due to the widespread air pollution in the city. Although there are some physiological indications that the stress of pollution may impair the immune system and increase the requirement for some particular nutrients in individuals, we do not know the extent of interplay among three epidemiological factors: stress from pollution, infectious diseases, and nutritional status.

Nutritional interventions in formal health care facilities

Significant variation occurs in the effectiveness of nutritional intervention through formal health care facilities in the metropolitan areas of Brazil. It is much more difficult to improve nutritional status by feeding programmes in the metropolitan areas because of the complex origins of nutritional problems and risk groups. Therefore it is not practicable merely to copy feeding programmes from the rural areas. For instance, the free distribution of a cup of milk to each child from low-income families (programa de leite) deals with protein-energy malnutrition, which is of low magnitude, but not with the more urgent problem of anaemia. Moreover, school feeding programmes do not solve the problem of wasting in school children.


Although we are only beginning to gain experience in the field of public health nutrition, some conclusions can already be drawn.

First, we need more basic epidemiological knowledge concerning risk groups, their nutritional problems, and their ecology in the metropolitan areas in the subtropics and tropics. The causal complexities will force science to take a holistic, interdisciplinary approach to research. It must be stressed, however, that interdisciplinary research needs more and better co-ordination than currently exists, as well as human and financial resources.

Second, to date, interventions in urban areas have been based on rural experience and not on the different patterns of causation of urban nutritional problems and different health facilities. Therefore, research into appropriate intervention measures and strategies based on the results of epidemiological experience needs to be carried out. We can already conclude that the high interrelationship of causes of malnutrition will make it necessary for both individuals and public institutions to take responsibility for intervention.

Third, epidemiological results for either the entire urban system at the macro level or for isolated socio geographic sub-systems at the micro level cannot completely describe the complex nutritional and health situation of the metropolitan population. It is therefore necessary to study both the whole system and the different sub-systems.

Finally, it should be stressed that the causes of nutritional problems of urban populations have yet to be studied in depth.


1. Schürch B, Favre AM. Urbanization and nutrition in the third world. Lausanne: Nestle Foundation, 1985.

2. Harpham T, Vaughan P, Riffkin S. Health and the urban poor in developing countries. EPC Publication no. 5. London: London School of Hygiene and Tropical Medicine, 1985.

3. Gross R, Solomons NW. Tropical urban nutrition. GTZ-Sonderschriftenreihe, no. 196. Eschborn, FRG: Deutsche Gesellschaft für Technische Zusammenarbeit, 1987.

4. Solomons NW, Gross R. Urban nutrition in the tropics: a call for increased attention to metropolitan population in the developing world. Food Nutr Bull 1987;9(2):4344.

5. Chen L, Chowdhury AKMA, Huffman SL. Seasonal dimensions of energy protein malnutrition in rural Bangladesh: the role of agriculture, dietary practices and infection. Ecol Food Nutr 1987;8:175-87.

6. Valverde V, Delgado H, Martorell R, Belizán JM, Mejía-Pivaral V, Klein RE. Seasonality and nutritional status. Arch Latinoam Nutr 1982;32:521-40.

7. Dugdale AK, Payne PR. Modelling seasonal changes in energy balance (Africa). In: Taylor TG, Jenkins NK, eds. Proceedings of the XII International Congress of Nutrition. London: John Libbey, 1986:141-44.

8. Pickering H, Lamb WH. Urban and rural differences in seasonal stress in the Gambia. In: Taylor TG, Jenkins NK, eds. Proceedings of the XII International Congress of Nutrition. London: John Libbey. 1986: 148-50.

9. UNICEF. Situaçao mundial de infancia 1987. Brasilia, DF. 1987.

10. Victora C, Vaughan JP, Kirkwood BR, Martines JC, Barcelos LB. Risk factors for malnutrition in Brazilian children: the role of social and environmental variables. Bull WHO 1986;64:299-309.

11. Monteiro CA, Zuniga HPP, Benicio MHD'A, Szarfare SC. Estudo das condições de sauce das crianças do municipio de Sao Paulo, SP (Brasil), 1984-1985. II. Antropometría nutricional. Rev Saúde Públ 1986;20:446-53.

12. Bisi Molina MC, Gross R, Schell B, Cuelho Leao MA, Strack U, Brunken B. The nutritional status of children of two urban low-income communities of Belo Horizonte, Brazil. Rev Saúde Publ (in press).

13. Walter J. Two poverties equal one hunger. J Nutr Educ 1973;5: 129-33.

14. Gross R. Thermodynamic and cybernetic viewpoint of the city. In: Gross R, Solomons NW, eds. Tropical urban nutrition. GTZ-Sonderschriftenreihe, no. 196. Eschborn, FRG: Deutsche Gesellschaft fur Technische Zusammenarbeit, 1987:31-35.

15. Gross R, Stange M, Solomons NW, Oltersdorf U, Rios Esquivel I. The influence of economic deterioration in Brazil on the nutritional status of children in Rio de Janeiro, Brazil. Ecol Food Nutr 1987;19:265-79.

16. Monteiro CA, Zuniga HPP, Benicio MHD'A, Rea MF, Tudisco ES, Sigulem DM. The recent revival of breast feeding in the city of Sao Paulo, Brazil. Am J Public Health 1987;77:964-66.

17. Gross R, Dias de Lima F, de Freitas CJ, Gross U. Urban nutritional risk group: schoolchildren of different social strata in Rio de Janeiro. Rev Saúde Públ (in preparation).

18. Monteiro CA, Szarfaro SC. Estudo das condições de saúde das crianças do municipio de Sao Paulo, SP (Brasil), 1984-1985. V. Anemia. Rev Saúde Públ 1987;21 :255-60.

19. Pinheiro A. As mamitas de pobreza. Thesis, Escola de Saúde Publica, Fiocruz, Rio de Janeiro, 1987.

20. Rossi PH, Wright JD, Fisher GA, Wills G. The urban homeless: estimating composition and size. Science 1987;235: 1336-41.

21. Connolly M. Meeting the health needs of street children: a global challenge. New Orleans, La, USA: Tulane School of Public Health and Tropical Medicine, 1985.

22. Mueller D. The current status of urban/rural differences in psychosocial disorder: an emerging trend for depression. J Nerv Ment Dis 1981;169:18-27.

23. Wirth L. Urbanism as a way of life. Am J Sociol 1938;44:3-24.

24. Brown GW, Harris T. Social origins of depression: a study of psychiatric disorder in women. London: Tavis

25. Crowell BA, George LK, Blazer D, Landermann R. Psychosocial risk factors and urban/rural differences in the prevalence of major depression. Br J Psychiatry 1986;149:307-14.

26. Kovaess V, Murphy HBM, Tousignant M. Urban-rural comparisons of depressive disorders in French Canada J Nerv Ment Dis 1987;175:457-66.

27. Bangladesh Bureau of Statistics. Report of the child nutrition status module. Dhaka, 1987.

28. Mann J. Global AIDS: epidemiology, impact, projections and the global strategy. World Summit of Ministers of Health on Programmes for AIDS Prevention, London, 26-28 Jan 1988.

29. Abdel Kader HM. Dust deposits and some trace elements in air-borne dust in the city of Alexandria. Bull High Inst Public Health (Alexandria, Egypt) 1978;8: 137-49.

30. Achmadi UF. Carbon monoxide pollution in Jakarta, Indonesia. Southeast Asian J Trop Med Public Health 1979;10:610-13.

31. Echeverria E. Evaluation of the effect of atmospheric pollution on the respiratory tract in Mexico: preliminary results. Bull Int Union Tuberc 1976;51:461-67.

32. Kamat SR. Correlation of health-morbidity to air-bome pollutant levels in Bombay City: results of prospective 3year survey at one year. J Postgrad Med 1980;26:4562.

33. Myung CY. Present status of air pollution and its effects on health. J Catholic Med College 1971;20:215-32.

34. Monteiro CA, Benicio MHD'A. Estudo das condições de sauce das crianças do municipio de Sao Paulo, SP (Brasil), 1984-1985. Vl. Doença respiratória. Rev Saúde Públ 1987;21:380 86.

35. Benicio MHD'A. Monteiro CA. Estudo das condições de sauce das crianças do município de Sao Paulo, SP (Brasil), 1984-1985. IV. Doença diarréica. Rev Saúde Publ 1987;21:23-28.

Contents - Previous - Next