Contents - Previous - Next

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

Nutrition and urbanization (part 2)

Urbanization and nutrition in low-income countries

Barry M. Popkin and Eilene Z. Bisgrove


The last half of this century has seen a demographic revolution in low-income countries. Urbanization brought on by migration and natural increase has become a dominant factor in all regions. Unlike urbanization in the higher-income countries of the world, which is associated with major advances in science, technology, and social organization as well as absorption of large populations, urbanization in low-income countries has not been accompanied by the same level of "economic and cultural progress. . . [and] has become a source of major concern" [1]. The accelerated growth of urban populations in low-income countries has tremendous social, economic, nutritional, health, and environmental consequences, as increasing numbers of urban poor live in crowded slums and squatter settlements with limited access to the basic resources necessary for a healthy and productive existence.

The goals of this article are to describe the demographic, economic, and environmental dimensions of urban growth in low-income countries and to develop an understanding of the implications of rapid urban growth for the design of food and nutritional policies and programmes. To accomplish this, we examine the question, Are there unique urban policy and programme needs? This question has three components: (1) Are there unique nutritional needs? (2) Are unique factors associated with these nutritional needs and with the use and success of programmes and policies designed to address them? (3) Are there unique policy and programme issues and options to consider?

Most significant of reasons for this focus is the lack of attention paid to urban nutrition needs. Most of the research on nutrition in low-income countries as well as most of our programme design, implementation, and evaluation are based on rural populations and needs [2]. For instance, of the 1,324 English-language citations in publications from INCAP (the Institute of Nutrition of Central America and Panama) from 1949 to 1985, only 2.6% relate to urban communities and only 0.2% focus exclusively on city-based populations [3].


Demographic, economic, and environmental dimensions of urbanization

Eight issues indicated the enormity of needs of the urban sector: the proportion of people living in cities, rapid population-growth rates, absolute increases of population, concentrated population growth, a shift in the proportion of poor residing in cities, the proportion living in slum and squatter areas, environmental conditions, and basic health conditions.

Proportion /living in cities

The proportion of people living in cities of developing countries has grown tremendously since 1950 and is expected to increase rapidly throughout this century. In 1950 the urban proportion was 16.7%; this figure grew to 21.9% in 1960 and 30.5% in 1980, and it is projected that it will be 43.5% in 2000 (see FIG. 1. Proportion of population in urban areas (Source: Ref. 113) ). Latin America will be more than three-quarters urban by the end of the century. The largest relative percentage growth in the last 20 years of this century is projected to be in Africa, followed closely by East Asia. Both regions were 14-16% urban in 1950, and will grow to be over 42% urban in 2000. South Asia, the most populous region, will lag considerably behind these other regions in urbanization, being only one-third urban by the year 2000. Over 60% of this growth has been the result of natural increases rather than urban rural migration; 40% is the result of in-migration and reclassification. Moreover, because of the much younger age distribution in urban areas and the predominance of persons of childbearing age in migrant streams, this pattern is expected to continue.

Rapid growth rate

The population in cities has doubled or risen even more in low-income countries in the past 20 to 50 years. Since 1920 the urban growth rate has been over 3%. Between 1980 and 1990 it will be 4%, with the largest rate of increase (5%) occurring in Africa. This rapid speed of urbanization makes it very difficult for cities to provide basic social services (facilities, services, goods) at a pace that keeps up with demands.

Absolute increases of population

The numbers of persons living in urban areas in low-income countries will increase from almost 440 million in 1960 to about 1,500 million in 1990, a threefold increase in 30 years. In Africa the urban population has grown during this same period from 50 million to 220 million.

Concentrated population growth

Urban growth has been skewed toward a very few cities. The number of cities with more than one million persons has grown from 31 in 1950 to 74 in 1970, and is projected to be 284 in the year 2000. Africa, which had only four cities with a population of one million in 1960, is projected to have 57 in the year 2000; the same number will be found in Latin America. ln many countries, one or two very large cities, called primate cities, dominate (e.g., Bangkok is 40 times the size of the next largest city in Thailand, and 69% of urban Thais live there). In particular, the number of cities with a population above 10 million will increase disproportionately in Latin America and Asia before the end of this century.

Shift in proportion of poor to the cities

Concomitant with increased concentrations of population in urban areas will be a dramatic shift in the proportion of poor people living in cities. The residence of the absolute poor in all regions will be increasingly urban. At present one-third of the world's poor in low-income countries live in urban areas. This will increase to 57% by 2000. In Africa the change will be from 23% to 42%.

TABLE 1. Predicted growth in number of households in poverty-urban and rural, by region, 1980-2000

  Number of households ('000s) Growth (%)
1980 1990 2000 1980-1990 1990-2000


East Asia and the Pacific 4,156 5,111 5,744 23 12
South Asia 13,970 21,255 32,555 52 53
Latin America and the Caribbean 14,023 16,798 19,328 20 15
Europe, Middle East, and North Africa 6.250 7,574 8.743 21 15
Eastern Africa 1.369 2,544 4,703 86 85
Western Africa 1,405 2.266 3,227 61 42
Total 41,173 55,548 74,300 35 34


East Asia and the Pacific 12,553 11,719 9,872 -7 -16
South Asia 48,799 41.036 32,709 -16 - 20
Latin America and the Caribbean 4,932 3,028 1,621 -39 -46
Europe, Middle East, and North Africa 3,761 2,333 1,403 -38 -40
Eastern Africa 6,458 7,558 8,625 17 14
Western Africa 2,938 2,488 2,238 -15 -10
Total 79,441 68,162 56,468 - 14 - 17

A sizeable proportion of the urban poor are casual, unskilled workers [4]. Like the rural landless poor, they lack permanent employment security. A large number work in the informal sector in unskilled or semi-skilled occupations.

Proportion living in slum and squatter areas

A majority of the urban population in developing nations lives in deteriorated, dense slums or makeshift, dense squatter settlements [1, 5]. Given the tremendous difficulty of defining and measuring housing conditions within and across countries and cities, one must be very cautious in using statistics estimating the precise numbers. Grimes [6] estimated for the World Bank the percentage of urban populations living in squatter and slum areas; other estimates have come from the 1974 World Housing Survey [7]. Overall, most knowledgeable observers consider that in the 1970s dwellers in these areas represented 30-60% of the urban population. Older and larger cities tend to have more slum and squatter households.

Physically, shelters in urban and rural areas do not vary greatly. Often the rural forms are simply transposed to urban settings. There are, however, several unique differences in urban and rural shelter issues. In rural areas people rely on individual initiative in providing shelter; in urban areas the concentration of people and unavailability of low-cost building materials and land require collective, public solutions. The absolute need for urban shelter is growing so fast that it has outstripped all policies and programmes trying to remedy the problem; moreover, investments required to deal with the shelter problem are enormous [8]. Private housing supply has clearly not adjusted quickly to changes in demand, and the poor appear to be most affected by the breakdown in the market system [9].

Environmental conditions

In general, urban households have greater access to safe and adequate water supplies and appropriate sanitation facilities than rural households [10]. For the urban poor, however, lack of adequate water and sanitation facilities appears to be the rule [4, 9, 11-14]. Poor water and sanitation may represent "a more serious problem in cities, because of the [disease! risks it brings. But it is also a more intractable one in many cases because the high population densities of urban areas often render rustic [simple, low-cost] forms of technology [such as pit latrines] useless or even dangerous" [4].

Inadequate water supply presents an additional economic dilemma for the poor in many cities. White et al. [14] showed that the cash price of water for the urban poor was much greater than that for higher-income groups in Nairobi, and many others have found the same relationship. Conversely the time costs of obtaining limited water from scarce public water taps or wells may be considerable.

Another urban problem is environmental pollution. It affects the poor "most severely, since most of them live at the periphery where manufacturing, processing and distilling plants are often built, and where environmental protection is frequently weakest" [7].

Basic health conditions

The combination of poverty, overcrowding, environmental problems, and stress related to social instability and insecurity is associated with a number of urban health conditions. In particular, the high frequency of tuberculosis, venereal disease, and alcoholism appears quite severe [4, 7, 15]. Environmental conditions such as massive overcrowding and poor hygiene lead to gastrointestinal and infectious diseases and epidemics [16].


Dimensions of food consumption needs

Our focus is on evidence of inadequate diets in urban areas, in particular, evidence linked to key nutrition related problems. Research seems to indicate that urban infants suffer growth retardation at an early age and urban children are very susceptible to rickets [7]. These conditions are related to a great extent to inadequate consumption or inappropriate dietary patterns. Although we can evaluate infant-feeding practices and consumption patterns of urban and rural households, because of inadequate data we are not able to examine consumption for pre-schoolers and women, two other high-risk groups


Infant feeding

Modernization is thought to be leading to earlier supplementation and discontinuation of breast-feeding in low-income countries? particularly in urban and pert-urban areas. Reliable measurement of breast-feeding patterns has been extremely difficult, largely because most studies have been based on small and non-representative samples, usually from hospitals, clinics, or small communities. Our review of urban-rural patterns and trends in breast-feeding is based primarily on nationally representative surveys from low-income countries. For ease of presentation, countries are divided into three geographic regions: Asia and the Pacific, Latin America and the Caribbean, and Africa and the Near East (see FIG. 2. Percentage of mothers breast-feeding, by region, urban or rural residence, and age of child (last two children, or only child, in the four years preceding the World Fertility Survey) (Sources: Refs. 17,114)).

Asia and the Pacific

In general, breast-feeding is practiced less in urban than in rural areas in Asia. Where declines have occurred, however, changes over time are similar in the two areas. Only in the Philippines, which has been subjected to the most complete analysis of trends, is urbanization associated with reduction of breast-feeding.

Nearly all women in the countries reviewed except urban women in the Philippines, Taiwan, Fiji, and Singapore initiate breast-feeding [17-23]. Among both rural and urban women in Korea, Nepal, Bangladesh, and India, 90% or more breast-feed beyond the child's first six months. After six months, the numbers still breast-feeding fall off more steeply among urban than rural women. In Sri Lanka and Indonesia, the differential widens somewhat earlier; fewer than 80% of urban women still breast-feed by the sixth month and fewer than 50% by one year. More extreme cases are the Philippines and Fiji, where fewer urban women initiate breast-feeding, and the drop-off is dramatic. At three months, only 56.6% of urban Filipino mothers and 50.9% of Fijian mothers breast-feed.

Trend data based on findings from four national surveys conducted in Thailand between 1969 and 1979 showed a moderate pervasive decline in breast-feeding duration among urban and rural mothers [24]. Dramatic declines also were noted in urban and rural areas of Taiwan from 1968 to 1980. Haaga [21] reported a decline in percentages of urban Malaysian infants breast-fed in birth cohorts from 1969 to 1974, followed by a slight increase for infants born from 1975 to 1977 compared with those born from 1970 to 1974 (76% and 67% respectively).

In the Philippines the proportion ever breast-fed declined by five percentage points between 1973 and 1983 [19]. Overall, the duration of breast-feeding increased; however, complex multivariate analysis showed that increased urbanization, shifts in women's occupations, and residence in Mindanao (the centre of extensive social unrest during this period) were associated with decreased duration [25].

Latin America and the Caribbean

Of the 10 countries reviewed, the number of women ever breast-feeding was somewhat lower in all areas of Latin America than in Asia. Urban-rural differences were also much greater. Only in Guatemala, Jamaica, and Chile was the proportion of urban women ever breast-feeding above 90%. In rural areas the practice was still the norm up to at least six months, while in most counties 60% or more of urban babies were weaned by age six months. Urban rates fell more quickly than rural rates; approximately 40% of urban women still breast-fed by six months.

Another pattern emerged in the Dominican Republic, Colombia, Peru, and (Guyana. Approximately 80-90% of urban women initiate breast-feeding, but the percentages continuing the practice drop steeply, reaching 50-60% by the third month and 30-50% by the sixth month. Panama and Costa Rica are the most extreme cases, with about 70% of urban mothers breast-feeding in the first month but with the proportion falling to approximately 35% by the third month and 20% by the sixth month.

Trend data from a few Latin American countries point toward a gradual increase in breast-feeding among urban women. Data from 1978 and 1983 INCAP surveys in Guatemala show slightly increased mean durations in urban and rural sectors [26]. In the Dominican Republic, compared with 1975 data, the proportion breast-fed has increased slightly in urban areas and has remained stable in rural areas. In Jamaica, substantial increases were apparent between 1975 and 1983 in the mean age at weaning-rising from 9.4 to 12.7 months for the urban population and from 10.9 to 13.4 months for the rural population. The proportion ever breast-fed increased slightly in both sectors. Overall, the upward shift in breast-feeding in Jamaica was more pronounced in urban areas [26].

Africa and the Near East

Breast-feeding patterns in Africa appear to be intermediate between those of Asia and Latin America. Except in Kenya, Yemen, and Tunisia, urban-rural differences were smaller than in the other two regions. In Ethiopia, Nigeria, Lesotho, and Zaire, initiation of breast-feeding is nearly universal [18]. From almost 100% of urban and rural women initiating this practice, the decline is gradual, with more than 80% continuing it beyond the sixth month. Egypt, Jordan, and Kenya show slightly more divergence between rural and urban patterns. While about 90% of urban women begin breast-feeding, only approximately 65-75% continue until the sixth month. In Tunisia and Yemen, slightly fewer than 90% of urban women initiate breast-feeding, and the decline is much steeper than for rural women, reaching levels of approximately 60% at three months and 50% at six months.

Trend data from Jordan show a stable mean duration of breast-feeding for the rural population from 1976 to 1983. There was a small increase in the mean duration for the urban population from 9.9 to 10.7 months [26].

Lower proportions of infants ever breast-fed and shorter breast-feeding duration prevailed in urban areas of the countries discussed. Overall, the trends are becoming more variable within geographic regions. In the past, the greatest changes were noted in urban areas. More recently, a decline has also occurred in rural areas of more modernized countries, including Taiwan, Thailand, Malaysia, and Panama [27]. In countries where breast-feeding is already at low levels, future declines will probably take place among the rural population.


Infant supplementation

It is generally felt that early, inadequate supplemental feeding is a problem of urban areas. The dearth of nationally representative data concerning urban-rural patterns and trends does not allow us to assess the issue fully. Moreover, none of the large surveys allow us to evaluate the nutritional quality or processing of these supplements. The following discussion shows important urban-rural differences in the age of introducing supplementary foods, based on the WHO Collaborative Study on Breast-Feeding and studies from Mexico and Malaysia.

TABLE 2. Percentages of urban and rural breast-feeding mothers giving breast milk alone, occasional supplements, or regular supplements. by age of child

  2-3 months 6-7 months 12-13 months    
India (N = 2,165)                  
urban 94 0 6 78 3 19 40 6 54
rural 98 0 2 88 0 12 36 2 62
Philippines (N= 1,601)                  
urban 64 13 23 9 9 82 0 0 100
rural 58 13 29 5 18 77 3 10 87
Chile (N = 739)                  
urban 41 0 59 0 0 100 _ _ _
rural 41 3 56 5 0 95 0 0 100
Guatemala (N = 1,194)                  
urban 48 0 52 13 0 87 0 0 100
rural 86 2 12 34 4 62 0 0 100
Ethiopia (N = 1,185)                  
urban 51 16 33 61 16 23 15 39 46
rural 51 18 31 15 32 53 6 51 43
Nigeria (N = 1,304)                  
urban 37 0 63 8 2 90 4 2 94
rural 65 0 35 14 0 86 2 0 98
Zaire (N= 1,187)                  
urban 64 4 32 4 13 83 9 17 74
rural 51 14 35 24 4 72 2 0 98

BA = breast milk alone; OS = occasional supplements; RS = regular supplements. (The terms "occasional" and "regular" are not defined in the source.)

Source: Ref. 18.


Contents - Previous - Next