Nutrition Issues in Developing Countries: Part I: Diarrheal Diseases, Part II: Diet and Activity During Pregnancy and Lactation (1992)

To provide the reader with a fuller understanding of the forces that women in developing countries are subject to during prepregnancy, pregnancy, and lactation, the subcommittee judged it useful to present a summary of the socioeconomic and health conditions of women living in developing countries. This chapter is not a detailed sociological review, but it highlights those areas that affect women most during those times.

While developed nations managed to recover from the 1981–1982 world recession, with the exception of East and Southeast Asian countries, most developing countries have continued to experience a decrease in their per capita gross domestic product (GDP). At the end of 1984, average per capita GDP in Latin America and the Caribbean was as low as the 1976 level; in Africa, per capita GDP experienced a decline of more than 3 percent per year in the 1981–1983 period. Because of the increase in interest rates around the world, the cost of repaying past debts has increased, becoming a heavy burden on the balance of payments and impairing a reallocation of the scarcer resources. This, in turn, has led many countries to diminish their real per capita expenditures in the social sectors. By 1985, austerity measures were being implemented in half of the developing countries; and the effects of these measures, combined with those of a lower GDP, included a serious deterioration of already low living standards for large sectors of the

population. The situation in many of these countries has been compared with the Great Depression of the 1930s (UNICEF, 1987; 1988; WHO, 1987).

By 1980, a number of countries were producing less food per capita then was produced in the mid-1970s. Since then, per capita energy availability has been increasing in Latin America, the Middle East, North Africa, and East Asia; it has been decreasing in Sub-Saharan Africa and in Southern Asia. In many developing countries, an increasing proportion of the population is falling below the level of poverty. This has serious implications for food consumption levels and nutritional status. By 1985, starvation affected large portions of the population in Africa, and hunger was still prevalent in Asian and Latin American countries (WHO, 1987). Depending on the indicators and cutoff points used, estimates of the malnourished population vary considerably. It is likely that by 1985, at least 430 million of the world's people suffered from malnutrition, and by 1987, severe malnutrition had become very prevalent in Sub-Saharan Africa (WHO, 1987).

It is estimated that 25 to 35 percent of households in the developing world are headed de facto by women because of divorce, separation, desertion, or long-term migration of husbands or because women had children out of wedlock (Tinker, 1979a). These women, who are the poorest in every country, typically are responsible for earning income to support their families.

By the end of the 1976–1985 ''Decade for Women,'' the conditions for the majority of women in developing countries had changed only marginally. Despite the fact that women work nearly two-thirds of the total hours worked, they constitute only one-third of the world's official labor force, receive only one-tenth of the world's income, and own less than 1 percent of its property (WHO, 1985a). However, some of the conditions for the advancement of women are being met in some developing countries. Although wide regional variations still exist among countries, access of young females to education is improving, as is their access to health care services, including family planning (WHO, 1987).

While the proportion of adult literacy increased in the 1970–1980 period from 52 to 60 percent due to population growth, the absolute number of illiterates increased from 731 million to 800 million. The mentioned

increase in the adult literacy rate has taken place without altering the gender gap that favors the male's literacy rate over the female's. While in 1970, 60 percent of males and 43 percent of females aged 15 and over were literate, by 1980 this proportions had risen to 68 and 51 percent, respectively, keeping the same 17 percentage points of difference.

As a consequence, almost two-thirds of illiterate adults in developing countries are women. According to 1980 estimates, the general level of adult illiteracy was 60 percent in Africa, 40 percent in Asia and the Pacific, and 20 percent in Latin America (WHO, 1987).

The total fertility rate, as estimated for the 1975–1980 and 1980 –1985 periods, is following a downward trend worldwide, being 3.9 and 3.5 respectively (live births/women) (WHO, 1987).

Except for African countries, where total fertility rates are the highest in the world and have kept stable at 6.6 for the periods mentioned, the total fertility rate in all other regions is lower and falling, albeit at a faster rate in Southeast Asia and the western Pacific than in the Americas and the eastern Mediterranean countries (Table 2-1).

It has been estimated that 45 percent of married women of reproductive age worldwide used a contraceptive method in 1980–1981. However, when the People's Republic of China was excluded, this proportion fell to 38 percent. This proportion varied from very low levels in Africa (11 percent) to low levels in South Asia (24 percent), intermediate levels in Latin

America (43 percent), and rather high levels in East Asia (69 percent) (WHO, 1987). Even though most countries provide some form of public support for family planning programs, generally within the maternal and child health care programs, reduced geographical and economic access to family planning services is a limiting factor for millions of couples who do not desire additional children, but who are not using any effective method of family planning (WHO, 1987).

During 1975–1985, the infant mortality rate decreased in approximately 150 countries. However, in more than 25 percent of the world's countries, which represent 29 percent of the world's population, infant mortality rates still are higher than 100 per 1,000 live births (Table 2-2).

However, this downward trend has slowed or even reversed as a consequence of economic deterioration (UNICEF, 1987).

The principal causes of neonatal mortality in the world are related to the perinatal period. For example, in a study by Haas et al. (1987), they showed that proportionally growth-retarded infants had nearly twice the mortality rate of full-term, appropriate-weight infants, and disproportionally growth-retarded infants had 2.9–5.7 times the mortality rate. Infants delivered before 37 weeks of gestation had 23–100 times the mortality rate of full-term infants born at normal weight. Postnatal mortality in developing countries is due mainly to malnutrition and infection in children under 5 years of age, particularly respiratory and diarrheal diseases. Nutritional

deficiency is the associated cause in 61.9 percent of the deaths from infectious diseases (Puffer and Serrano, 1973).

Mortality rates are lower for female children Compared with those for males. The female advantage is accentuated more during fetal life, infancy, and childhood. However, cultural values and lifestyles can cancel or even reverse this natural advantage. For example, feeding practices favor male children over female children. This discrimination occurs more frequently in societies in which the status of women is low.

In the majority of developing countries, most deaths among women of reproductive age are due to maternal mortality. However, the rate is still considerably underestimated due to a failure to record both deaths and cause of death. It has been estimated that 500,000 women die each year from pregnancy-related causes. The maternal mortality rate in various African countries may be as high as 550 times that in some European countries. By 1985, only 75 of the 117 developing countries that were WHO member states were able to provide information on maternal mortality (WHO, 1987).

In developing countries, the main causes of maternal deaths are hemorrhage (often with anemia as an underlying cause) and sepsis. In some Latin American countries, up to 50 percent of maternal mortality is due to illegal abortion (WHO, 1987).

While nutritional anemia affects members of both sexes and people in all age groups, the problem is more prevalent among women and contributes to maternal morbidity and mortality, as well as to low birth weight. Table 2-3 gives rates of nutritional anemia for pregnant women from selected developing countries. It has been estimated that nutritional anemia affects almost two-thirds of the pregnant and 50 percent of the nonpregnant women in developing countries. The estimated prevalence of nutritional anemia among pregnant women in South Asia is over 65 percent, while in Latin America it is 30 percent, in Oceania it is 25 percent, and in Europe it is 14 percent (WHO, 1987).

Since birth weight is substantially determined by the health and nutritional status of the mother, the prevalence of low birth weight reflects the health and social status of women and of the communities into which children are born (WHO, 1987). The birth weight of the infant is perhaps the single most important predictor of survival and is also a strong predictor of growth and development.

Unfortunately, most of the studies on which the figures in Table 2-4 are based do not distinguish between prematurity (gestational age of fewer than 37 weeks) and intrauterine growth retardation. In most developing countries, the majority of infants are born at home or in other locations outside of hospitals. Women are often unsure of the date of their last menstrual period, and investigators are generally content to collect accurate birthweight measurements for defined populations. Villar and Belizan (1982), however, have analyzed data from 11 different regions in the developed world and 25 areas in developing countries. In developing countries, they found that most low birth weight appears to be due to intrauterine growth retardation, whereas in developed countries (especially those with the lowest rates of low birth weight), most is due to prematurity. The relative rates of intrauterine growth retardation in developing countries compared with those of developed countries (6.6-fold higher) are far greater than the relative rates of prematurity (2.0-fold higher).

According to a review and analysis of 200 studies on breastfeeding and its duration, carried out in 86 countries, the patterns of incidence, duration, and suckling frequency vary widely between and within regions, as well as between rural and urban populations and between social classes (WHO, 1987).

Most infants born in rural African areas are breastfed for a period of I to 2 years, while in urban areas, fewer children are breastfed. In most of South Asia, breastfeeding is the norm, with an average duration of 6 months in urban areas and of 1 to 2 years in rural areas. In East Asia, 80 percent to 95 percent of infants in rural areas are breastfed, as are 80 percent of infants in urban areas; however, in urban areas, supplementary feeding starts at about 3 months and breastfeeding duration is shorter. In Latin America, a higher proportion of newborns are breastfed than in South or East Asia or Africa, but the average duration varies widely. Duration is longer in rural areas, especially in Central America and the Caribbean, but it is still shorter than in Africa and Asia; in urban areas the pattern is one of early supplementary feeding and subsequent weaning (WHO, 1987).

In developing countries, women's roles include both income-producing and household-production activities. Income-producing activities are often essential for the household to function, especially among poor women, whose contribution to the family income is particularly important. In addition to labor force participation, women are responsible for maintaining the household, which includes child rearing, food processing and preparation, and fuel and water gathering. Because of the lack of adequate transportation, running water, and easy access to fuel or electricity, these basic activities often require that heavy burdens be carried for long distances daily. Rather than purchasing preprocessed food at a high cost or paying others to supply water and fuel, women provide the labor. Thus, household-production roles often result in income savings for the family.

In developing countries, women are traditionally undercounted in estimates of labor force participation, in part because of the definitions used to describe labor force participants. The international guidelines are adopted from categories that are appropriate only to the developed world. For example, activities in nonformal employment sectors such as street vending or producing foods in the home are not included in definitions of labor force

participation, yet they very often contribute enough income or lower household expenses sufficiently to ensure the family's survival. According to international guidelines, farm wives in industrial nations are often assumed not to be working, yet in developing countries, they may work on agricultural activities as much as or more than their husbands (Tinker, 1979b).

Bearing in mind, therefore, that official figures are likely to be underestimates, the economically active female labor force participation can be examined for different parts of the developing world. Africa (except for Muslim countries) has the highest proportion of women in the active labor force, with up to 47 percent of women in Botswana and 32 percent in Nigeria reported to be active. In Asia, with the exception of the People's Republic of China, female labor force participation is also high; for example, 27 percent of women in India are recorded as being economically active. Caribbean countries with populations that are primarily of African descent, such as Jamaica and Haiti, have female labor force participation rates of 46 and 26 percent, respectively, while Central America (10 percent) and South America (13–18 percent) have lower rates. The Arab states have the lowest rates, with 2–5 percent of women counted as being economically active (UNDP, 1980). A major reason for these different rates is development programs, the extent that women participate in agricultural labor. Table 2-5 gives the estimated percentage of women in the agricultural labor force in different parts of the world.

In addition to these activities, women also bear the children. In developing countries, fertility rates are high, and a woman who is either breastfeeding or pregnant, or both, is not uncommon. Although the importance of their income production or savings activities to family survival is great, pregnancy does change to some extent the amount and type of physical activities in which women are involved.

Jimenez and Newton (1979) have examined cultural practices related to work and pregnancy using data from the Human Relations Area Files, a reference source of anthropological studies. Among the 122 societies studied, 45 percent expected a woman to continue with full duties until the onset of

labor. In fewer than 5 percent of the societies, pregnant women are expected to perform light duties only, but most Societies expected partial or full duties to continue for most if not all of the 9 months of pregnancy.

In developing countries, patterns of energy expenditure during pregnancy are of particular concern since women generally consume diets lower in energy and have lower prepregnant weights than women in developed countries. Weight gains during pregnancy are also lower. The prevalence of chronic malnutrition is reflected in the high prevalence of low-birthweight infants. Low birth weight may also be due as much or more to the effects of work in a hot environment directing blood away from the uterus. (For a summary of mean birth weights and low birthweight rates both in developed and developing countries, see Table 2-4). Energy intakes are at least 500 calories less among women in developing countries compared with those seen among women in developed countries (Tables 2-6a and 2-6b). However, caution must be exercised when interpreting intake data from community-based surveys and studies.

Few studies have specifically examined the activity patterns of pregnant women in developing countries. Of those that have examined variations in activities by pregnancy status, most have shown that pregnant women perform the same types of activities as nonpregnant women. However, they generally also show that during the last trimester, and particularly during the last month, women try to reduce the amount of time spent in the most strenuous activities.

In a year-long study of work output and pregnancy in a Javanese village, frequent and repeated time-use studies of 44 households examined the evolution of individual pregnancies and revealed that the work patterns and income-generating activities of women were only minimally disrupted by pregnancy (White, 1984).

Among women studied in Papua New Guinea (Greenfield and Clark, 1975), as reported by Durnin (1976; 1980), energy expenditures were measured among pregnant, lactating, and nonpregnant women by using 24-hour diaries of activities and the oxygen consumption for typical activities. It was found that the energy expenditures decreased in the last 3 months of pregnancy, when women walked less, worked in their gardens for shorter periods of time, and spent more time sitting.

Similar results have been reported by Roberts et al. (1982) in The Gambia. Also using 24-hour diaries, they noted that the total active time

spent by women declined during pregnancy, especially during the last trimester. During the last month of pregnancy, women lowered their activity level to 75 percent of that of nonpregnant women by reducing the amount of housework and nonessential activities, such as those done for leisure. The stage of pregnancy, however, did not affect the time women spent at working on their farms. Even in the month before delivery, they continued to spend as much time there as the nonpregnant, nonlactating women in the study. During the farming season, women did not reduce the amount of time spent farming, but they did reduce the amount of strenuous farming-related activity.

A recent study in India used a 24-hour recall of rural southern Indian women's activities during pregnancy and lactation (McNeill and Payne, 1985). Pregnant women spent less time working in the fields compared with nonpregnant, nonlactating women (51 and 161 minutes/day, respectively) and more time in personal activities, mainly resting (969 and 678 minutes/day, respectively).

Except for the studies indicated above, data are limited on the activity patterns specific to pregnancy, but because women in developing countries seem to vary little the types of activities they perform while they are pregnant, a review of the literature based on types of activities of an women may give useful estimates for those who are pregnant. However, it must be taken into account that there may be qualitative changes in the way tasks are performed. The following section discusses urban and rural women separately because their daily activities demand different energy expenditure levels.

A typical Zambian woman's day during the planting season is described in Table 2-7, showing the amount of physical activity performed each day by a woman in rural Africa. Much of this activity involves bending, walking, and carrying loads; all of these activities are more difficult for pregnant women. To estimate the weight of loads women carried, it was assumed that a woman commonly carries a child (up to age 3 years), firewood or other fuel source, water, and often, agricultural products.

The following estimates can be applied: The average age of a carried child is 18 months, and the mean weight of an 18-month old in a developing country is 9 kg. The average daily consumption of water in a rural area where piped water is not available is about 20 liters per person. Thus, for a family of six, at least 120 liters of water (or about 30 gallons) weighing more than 250 pounds must be transported daily. Although the mother may not be the only person carrying the water from the source to the home, one person can carry up to 50 pounds per trip, which means a minimum of five trips a day.

The weight of the fuel that must be carried varies by type of fuel. Firewood is one of the heaviest fuels and is commonly used in many parts of Latin America and Africa. Other common fuels are cow dung, plant residues, and charcoal. Not only are sufficient amounts of these fuels often heavy, but also collecting them involves bending, stooping, and walking over rough terrain, all of which are more difficult during pregnancy.

The amount of time women in developing countries spent in gathering water and fuel and in processing food has been shown in several studies to exceed 2–3 hours each day. These activities have also been classified in several studies as moderate to heavy in terms of energy expenditure (Bleiberg et al., 1980; Roberts et al., 1984). Many women in developing countries are involved in agricultural activities that have higher energy costs. The overall energy expenditure among women in Burkina Faso (formerly Upper Volta) increased by more than 500 kcal/day when they worked in agricultural activities, bringing their energy expenditure to 2,900 kcal/day during the busiest farming season (Bleiberg et at., 1980). Among women in The Gambia, the amount of time spent performing moderate or heavy activities varied between 45 and 65 percent of the working day.

A study of peasant households in the Sierra region of Peru found that mothers are responsible for 93 percent of the conversion of agricultural products to food and meal preparation, 78 percent of water collection, 58 percent of fuel gathering, and 62 percent of animal care (Deere, 1983). In a study in Tanzania, 91 percent of the weeding and harvesting, 64 percent of the ploughing, and 79 percent of the slashing of grasslands was by women (Tobisson, 1980). Direct observations of villagers in Burkina Faso illustrated that women carried out 64 percent of production and supply tasks, including food and cash crop production, food storage and processing, and water and fuel supply maintenance. They also performed 23 percent of crafts activities, 97 percent of household tasks (child care, cleaning, and cooking), and 23 percent of community obligations. Overall, women performed 56 percent of all work compared with 44 percent for men (McSweeney, 1979).

Women spend 1–3 hours daily processing agricultural products (e.g., threshing and grinding grain), 1–2 hours preparing food, 2–5 hours farming, up to 1 hour maintaining the fuel supply, and up to 4 hours caring for children. In total, rural women work 10–12 hours per day on both home and market production.

The amount of time spent in child care may seem surprisingly low. However, this figure reflects a pattern of child care in which small children and infants are rarely separated from their mothers, so that it is often not feasible to distinguish between the time spent doing child care and the time spent marketing or food processing. Typically, a child is with the mother and receives attention while she is engaged in other tasks.

Census figures show that women's participation in agricultural activities is much lower in Latin America (excluding the English-speaking Caribbean) than in Africa or Asia. Although it is well known that national census efforts consistently, and sometimes grossly, underestimate the female agricultural labor force (Rechinni de Lattes and Weinerman, 1979), it appears that women are not the principal source of agricultural field labor in rural Latin America, in contrast to the situation in rural Africa.

Buvinic (1982) conducted a case study of the amount of pay women receive for agricultural work in the coffee and tobacco plantations of Honduras and compared these figures with those of a national census conducted 3 years earlier for the same geographical region. Based on plantation production and salary records, Buvinic estimated the extent of female participation in the paid agricultural labor force for the previous harvest season. When compared with the case study figures, it was found

that census figures had underestimated women's participation in the agricultural labor force by between 30 and 40 percent. Women in Honduras make up almost 90 percent of the paid labor force in coffee production. Similar evidence of census under-reporting is available for Chile, Paraguay, Bolivia, and the Dominican Republic (Rechinni de Lattes and Weinerman, 1979).

Guatemalan women also work in large plantations, either seasonally or year round, in the field and at tasks ancillary to agricultural activities, such as cleaning and classifying produce, cooking and washing clothes for the seasonal work force, and tending livestock (Bossen, 1984). An in-depth study of a sugarcane plantation in southeastern Guatemala found that among the 80 households permanently attached to the plantation (there was also a seasonal work force of about 400) 50 percent of the women performed subsistence agricultural work and service work, compared with 88 percent of the men. Bossen described their circumstances as follows:

Another study conducted in the same coastal plantation belt region provides additional insights and measures of women's physical activities. McGuire (1979) conducted a longitudinal prospective study of energy intake, energy expenditure, and work. The women in this sample worked as unpaid family workers in subsistence crop production and as coffee pickers for a local cooperative. This work varied seasonally.

Daily intake of energy, protein, and iron, by physiological status, for these women are presented in Table 2-8. McGuire (1979) found that dietary intake was low but within the range of recommended values and that there were no modifications in intake by physiological status or season. Energy expenditure was measured for different activities by using a ''Bolfi bag,'' a modification of the Douglas bag, and a respirometer.

Daily energy expenditure averaged about 2,000 kcal/24-hour period or 41 kcal/kg/day. McGuire found no seasonal effects on energy expenditure. Pregnant women expended significantly more energy per day than lactating women—2,044 ± 158 kcal/24 hours and 1,821 ± 120 kcal/24 hours, respectively; (P < 0.01)—but these differences disappeared when data were corrected for body weight. The energy losses in breast milk are not included in these figures. When women worked outside the home, however, they expended more energy than when they worked at home—2,085 ± 273 kcal/24 hours compared with 1,970 ± 238 kcal/24 hours, respectively. This difference

was significant at the P < 0.01 level. Women work on their feet markedly more on days when they work outside the home. Understandably, they also spend more time and energy walking. When the proportion of workday time spent off their feet was compared between pregnant and lactating women, no discernible differences were found. However, when women worked outside the home, they spent significantly less time (P < 0.05) off their feet. These data are summarized in Table 2-9.

Weight gains were recorded during the second and third trimesters for seven of the pregnant women in the study. During the second trimester, the average weight gain was 0.04 ± 0.02 kg/day; or 3.6 kg/trimester. Identical figures were obtained for the same group of women in the last trimester.

In summary, the available evidence seems to indicate that women increase their participation as agricultural field workers in Latin America as the transition to cash cropping occurs. Nonetheless, women's involvement in agriculture in the region probably will not approximate that of African women.

Results of the above studies indicate the large amount of time spent on home production activities as well as on agricultural production. Both types of activities often necessitate substantial energy expenditures. Using the data from The Gambia, the amount of time pregnant women spend in light, moderate, and heavy work can be estimated, including the amount of energy expended on these activities. During the beginning of the rainy season, farming takes up a major part of women's overall activities, and pregnant women spend the largest proportion of their time throughout the year in heavy activities. Taking the month of June as an example, pregnant women

spent 738 minutes out of a 15-hour workday in activities other than resting and unaccounted light activities, with about 25 percent of the day spent in hard work, 35 percent spent in moderate work, and 20 percent spent in light work. During the least active time (April), only 436 minutes out of a 15-hour workday were spent in activities other than resting or not otherwise specified light activities, with about 3 percent spent in hard work, 32 percent in moderate work, and 15 percent spent in light work, with the rest of the time spent resting. Using data from the same studies, an estimate of energy expenditure can be made (Table 2-10).

The estimates used in Table 2-10 are those observed for women in the second and third trimesters of pregnancy. While the data on energy expenditure for light and moderate work were similar for tasks categorized as light and moderate, expenditure for hard activities varied, thus the three estimates were used. Light activities with an estimated energy expenditure rate of 1.25 kcal/minute included sitting, most food preparation, and standing with a child. Moderate activities were estimated at levels of 2.09 kcal/minute for harvesting rice and 3.28 kcal/minute for walking while carrying a load such as fuel or items from the market place at 4.4 kph. Other activities within this range were beating groundnuts, washing clothes, and drawing water. Heavy activities included harvesting groundnuts (3.74 kcal/minute), digging while standing (4.59 kcal/minute), and pounding grain (4.81 kcal/minute). Carrying heavy loads, for example, chopping wood, clearing land, planting seeds, and digging and weeding groundnuts were also classified as heavy activities.

These estimates were based on 24-hour recall and observations of activity patterns and energy expenditures by using oxygen consumption for specific activities. The total estimated calories expended are substantially greater than the intakes recorded based on food weights. This may reflect an error in measurement of intake data similar to the Guatemalan study. The data on activities probably do not include the amount of time spent not active in each of the activities, such as the time spent resting while chopping wood or carrying loads. However, Table 2-10 can be used for illustrative purposes to indicate the large seasonal differences that are evident, the associated increases in energy expenditure related to field work, and the other home maintenance duties that women are responsible for on a daily basis.

Low-income women living in urban areas in developing countries are often as active as those living in rural areas, although the activities they

perform are probably not as energy intensive. Urban women are also doubly burdened by home production and market work, and the poorer the household in which a woman lives, the more important her economic contribution becomes (Safilios-Rothschild, 1980). In cities where electricity and potable water systems are widely available, low-income women may not necessarily be relieved of the tasks described above for rural women. In general, poor urban women do not participate in the three most strenuous physical activities: gathering fuel, drawing water, and doing agricultural work. However, in most shanty towns and slums in large cities in developing countries, piped water is not available to the entire urban population, so carrying water is a job performed by women and children. Transportation to and from work, to buy and carry essential household goods, including food, or to attend health care facilities (for the children or herself) also require considerable amounts of energy, especially when public transport is expensive and unreliable. With increased rural-urban migration in Latin America, an increasing pattern of seasonal work in the agricultural sector for urban women has been described. In fact, in most Central American countries, seasonal employment of urban and rural women in the coffee-picking activities has long been known. This seasonal involvement of urban dwellers in cash crop agricultural activities is increasing in Brazil (coffee harvest) and Chile (fruit harvesting and packing) (UN-ECLAC, 1988).

The physical activity involved in market work (income-generating work integrated into the market, money-oriented economy) may or may not make up for the differences between rural and urban women noted above, depending on whether the work is sedentary or whether it involves standing, walking, or lifting objects. The opportunities for income-generating work are diverse in urban settings and range from petty trading on the streets and domestic service to formal employment in the modern industrial and service sectors. Overall, women in cities in developing countries are mostly represented in the service sector. In the mid-1960s, approximately 90 percent of domestic employees in Chile were women; in Colombia the figure was 80 percent, and in Mexico it was 68 percent (ICRW, 1980). The participation of women in manufacturing is mostly circumscribed to lighter industries, such as electronic assembly plants, textile or garment factories, and pharmaceutical industries. Table 2-11 presents the proportion of the economically active population engaged in industrial and service work in selected countries by sex. It should be noted, however, that the figures for the Middle East may not be comparable since Muslim women (in purdah) presumably perform tasks very different from those of women in other areas.

Most developing countries subscribe to International Labor Organization conventions that are designed to protect women from heavy work or work that may endanger a woman or her fetus. This protective legislation sometimes prohibits women from working in the mining sector or working at night. Although few of these labor codes are ever enforced, they have contributed to the maintenance of a sex-stratified labor force in developing countries in both rural and urban areas (ICRW, 1980b). There are several exceptions, however, especially in some cities. These changes in women's work roles may not have reached large numbers of women, but may be indicative of future trends. For example, in Jamaica, women are moving into truck driving and the operation of other medium and large vehicles in the transportation and industrial sectors (Powell and Olafson, 1982). In Bahia, Brazil, metalwork is an activity that is opening up to women as vocational schools introduce nontraditional training for young adult females (Crandon and Shepard, 1984). In Chile, the government accepted the participation of women in public road maintenance and street-cleaning programs as part of its efforts to lower unemployment. A survey of 10,000 workers in these programs showed that 52.4 percent were women working up to 8 hours a day. Seventy percent of these women were 18–40 years of age (Buvinic and Mellencamp, 1983). It is not known how this change in women's work roles will affect the outcome of pregnancy.

By and large, women are not represented in great numbers in the industrial or manufacturing sector. They tend to concentrate in the lower levels of the tertiary or service sector and in the informal sector of the labor market (ICRW, 1980b). In Latin America and Asia, women make up more than one-third of the labor force in the service sector, most of it in urban areas. These figures include middle-and upper-class women engaged in professional work.

Informal is the term used to refer to that growing sector of the economy characterized by microenterprises that are usually family based and employ one to five workers, have little capital and less than optimal technology, and buy work that is piecemeal. It is often performed at home or in the street and is unregulated, untaxed, and sometimes illegal. It is generally outside the realm of what labor codes define as adequate both in terms of remuneration and in working conditions. In cities like Bombay, Jakarta, and Lima, the informal sector absorbs 53–69 percent of all urban workers. The informal sector of the labor market receives the oversupply of workers from the formal service sector. In the informal sector of urban economies, women workers proliferate (Buvinic et al., 1983). Market vending, street food vending, and domestic service are examples of informal sector work dominated by women in Latin America, the Caribbean, Africa, and parts of Asia. In Belo Horizonte, Brazil, in the mid-1970s, for example, 54 percent of the informal sector workers were women. When domestic servants were excluded from this population, 40 percent of informal sector workers were still women. Sixty percent of the self-employed within the informal sector were women (Merrick, 1976). In urban India in the mid-1970s, 40–50 percent of women workers were in the informal urban labor market, whereas only 15–17 percent of the male labor force was engaged in informal work (ICRW, 1980). There is no reason to expect that this trend has reversed in a decade. On the contrary, evidence from the current economic crisis indicates that it has become more marked (UNICEF, 1987).

Case studies of Indian women provide evidence that they are working in heavy informal work: collecting garbage, cutting stone, carrying bricks and other construction materials, and doing construction work. Although no figures are available on the involvement of women in construction work, in New Delhi and Bombay, there are now mobile creches that provide care for approximately 5,000 of these women's children (Mahavedan, 1977; Huffman, 1985).

Market and street vending is another urban occupation in which women predominate or are well represented in all but the Muslim regions, where purdah is still observed. Market and street vending requires women to be on the streets or in the market for long hours during the day, sometimes even during the night. There is a lack of studies of time use or activity patterns of market vendors and domestic workers, but several intensive life histories of some of these women recently have been published as composite profiles (Bunster and Chaney, 1985). Market and street vendors have, on average, 3.4 children each, and nearly half of these women are the main economic support for their households. The workday may begin as early as 3 or 4 a.m., and some women work from 8 a.m. to 7 p.m. and then perform their own housework, including washing clothes by hand. The total workday may be up to 18 hours long, 6 or 7 days a week. Commuting to the workplace is a major effort, because the women have to carry produce and, sometimes, children, and public transportation systems are often inadequate.

Traders and peddlers alternate street or market work with employment as domestic service workers. Domestic service work often requires that women work 14–16 hours a day, most of the time while standing or bending, and is considered highly demanding work by the women themselves (Bunster and Chaney, 1985). Washing clothes by hand is part of domestic service. This activity requires women to adopt uncomfortable positions and to exert themselves as they pound or rub the wet clothes against a hard, rough surface, such as river stones or cement washboards, and then rinse and force the water out, squeezing the pieces one by one. Furthermore, it is a domestic task that must be undertaken frequently. It may take from 1 to 3 hours a day twice a week for a family of five (two adults and three children); when there are diapers to be washed, it can take up to 4 hours a day (Nieves, 1986). A study of domestics and petty traders revealed that washing and ironing clothes was considered by domestic servants to be the hardest and most unpleasant work (Bunster and Chaney, 1985).

In the Lima study mentioned earlier, the median amount of time domestic servants worked was 11.5 hours a day. Some women in a sample of 50 servants worked up to 17 or 18 hours a day. They reported that most of the time they were standing, walking, or climbing stairs (Bunster and Chaney, 1985). Domestics work more hours per day than do their male counterparts in blue-collar jobs, and they earn about 60 percent of what other informal sector workers do (ICRW, 1980a).

While domestic servants begin their working lives when they are still very young (12–15 years old in Latin America), they work intermittently as domestic servants throughout their adult lives. The work histories collected by Bunster and Chaney (1985) showed that the first pregnancy usually occurs while the woman is employed. If she is not dismissed immediately, she will

continue working as a servant, with no diminished responsibilities, until she is dismissed later in the course of the pregnancy or leaves voluntarily. No quantitative data are available on how many women work in domestic service through the second or third trimesters of pregnancy.

The involvement of women in informal sector work in developing countries is a subject that social scientists are studying intensively. As more data become available, types of work that poor urban women perform will be better identified. More attention needs to be given to documenting the conditions of work and the physical activity involved, as well as the economics and the sociology of work. Physical mobility and the ability to transport themselves and their children from one point in the city to another already have been identified as key aspects of poor women's survival needs in urban areas in Latin America. The identification of other physical demands of urban life and urban work for low-income women, and low-income mothers specifically, deserves more attention.

Bleiberg, F.M., T.A. Brun, S. Goihman, and E. Gouba. 1980. Duration of activities and energy expenditure of female farmers in dry and rainy seasons in Upper-Volta. Br. J. Nutr. 43:71–82.

Bossen, L. 1984. The Redivision of Labor: Women and Economic Choice in Four Guatemalan Communities. State University of New York Press, Albany.

Bunster, X.B., and E.M. Chaney. 1985. Sellers and Servants. Working Women in Lima, Peru. Praeger, New York.

Buvinic, M. 1982. La productora invisible en el agro centroamericano: un estudio de caso en Honduras. Pp. 103–114 in M. Leon, ed. Las Trabojadoras del Agro. Associacion Colombiana para el Estudio de la Poblacion, Bogota.

Buvinic, M., and A. Mellencamp. 1993. Research on and by women in Chile. Report prepared for the Inter American Foundation. International Center for Research on Women, Washington, D.C.

Crandon, L., and B. Shepard. 1984. Women, enterprise and development: The Pathfinder Fund's women in development: projects, evaluation, and documentation program. Funded by AID/PPC/PDPR/IPD under grant number AID/otr-G-1867. Chestnut Hill, Mass.

Deere, C.D. 1983. The allocation of familial labor and the formation of peasant household income in the Peruvian Sierra. In M. Buvinic, M.A. Lycette, and W.P. McGreevey, eds. Women and Poverty in the Third World. The Johns Hopkins University Press, Baltimore, Md.

Durnin, J.V.G.A. 1976. Sex differences in energy intake and expenditure. Proc. Nutr. Soc. 35:145–154.

Durnin, J.V.G.A. 1980. Food consumption and energy balance during pregnancy and lactation in New Guinea. Pp. 86–95 in H. Aebi and R. Whitehead, eds. Maternal Nutrition during Pregnancy and Lactation. Hans Huber Publishers, Bern.

Greenfield, H., J.A. Clark, and I. Ring. 1974. Proceedings: Changes in body size relative to age and to childbearing in Papua. New Guinea women: A comparison of Highlands women and coastal women. Proc. Nutr. Soc. 33:30A.

Haas, J., H. Baleazar, and L. Caulfred. 1987. Variation in early neonatal mortality for different types of fetal growth retardation. Am. J. Physiol. Anthropol. 73:467–473.

Huffman, S.L., M. Wolff, and S. Lowell. 1985a. Atencion infantil en la communidad. Madres y Niños. 4:4–5.

Huffman, S.L., M. Wolff, and S. Lowell. 1985b. Nutrition and fertility in Bangladesh: Nutritional status of non-pregnant women. Am. J. Clin. Nutr. 42:725–738.

ICRW (International Center for Research on Women). 1980. Keeping women out: a structural analysis of women's employment in developing countries. ICRW, Washington, D.C.

Jimenez, M.H., and N. Newton. 1979. Activity and work during pregnancy and the postpartum period: A cross-cultural study of 202 societies. Am. J. Obstetr. Gynecol. 135:171–176.

Mahadevan, M. 1977. Mobile creches in India. Assignment Child 40:68–86.

McGuire, J.S. 1979. Seasonal Changes in Energy Expenditure and Work Patterns of Rural Guatemalan Women. Ph.D. Dissertation. MIT, Cambridge, Mass.

McNeill, G., and P.R. Payne. 1985. Energy expenditure of pregnant and lactating women. Lancet 2(8466):1237–1238.

McSweeney, B.G. 1979. Collection and analyses of data on rural women's time use. Stud. in Fam. Plan. 10:379–382.

Merrick, T. 1976. Employment and earnings in the informal sector in Brazil: The case of Belo Horizonte. Paper presented at the joint meetings of the LASA and the ASA, Houston.

Nieves, I. 1986. Snowpeas, Maidens and Millions: An In-depth Study of Intra-household Resource Allocation in a Cash-cropping Scheme. International Center for Research on Women, Washington, D.C.

Powell, D., and F. Olafson. 1982. Women in Development and Nontraditional Income—generating Activities in Jamaica. The Pathfinder Fund. Chestnut Hill, Mass.

Puffer, R.R., and C.V. Serrano. 1973. Patterns of Moriality in Childhood. Report of the Inter-American Investigation of Mortality in Childhood. Scientific Publication No. 262. Pan American Sanitary Bureau, Regional Office of the World Health Organization. Pan American Health Organization, Washington, D.C. 470 pp.

Rechinni de Lattes, Z., and C.H. Weinerman. 1979. Informacion de censos y encuestas de hogares pare el analisis de la mano de obra femenina en America Latina y el Caribe: Evaluacion de deficiencias y recomendaciones para superarlas. E/CEPAL/L.2067, UNESCOL, ECLA.

Roberts, S.B., A.A. Paul, T.J. Cole, and R.G. Whitehead. 1982. Seasonal changes in activity, birthweight and lactational performance in rural Gambian women. Trans. R. Soc. Trop. Med. Hyg. 76:668–678.

Safilios-Rothschild, C. 1980. The role of the family. A neglected aspect of poverty. Pp. 311–372 in P.T. Knight, ed. Implementing Programs of Rural Development. World Bank Staff Working Paper No. 403. World Bank, Washington, D.C.

Tinker, I. 1979a. New Technologies for Food Chain Activities: The Imperative of Equity for Women. Office of Women in Development, Agency for International Development, Washington, D.C.

Tinker, I. 1979b. Women and Development. American Association for the Advancement Science, Washington, D.C.

Tobisson, E. 1980. Women, work, food and nutrition in Nyamwigura Village, Mara Region, Tanzania. Tanzania Food and Nutrition Centre Report No. 548 (July).

UNDP (United Nations Development Program). 1980. Rural Women's Participation in Development. Evaluation Study No.3. United Nations, New York.

UNICEF (United Nations Institute of Children Emergency Fund). 1987. G.A. Cornia, R. Jolly, and F. Stewart, eds. Adjustment with a Human Face. I and II. United Nations, New York.

Villar, J., and J.M. Belizan. 1982. The timing factor in the pathophysiology of the intrauterine growth retardation syndrome. Obstet. Gynecol. 37:499–506.

White, B. 1984. Measuring time allocation, decision-making and agrarian changes affecting rural women: Examples from recent research in Indonesia. IDS Bulletin 15(1):18–33.

WHO (World Health Organization). 1985. Women, Health and Development: A Report by the Director-General. WHO Offset Publication No. 90. WHO, Geneva.

WHO (World Health Organization). 1987. Evaluation of the strategy for health for all by the year 2000. Seventh Report on the World Health Situation. Vol. 1. Global Review. WHO, Geneva.