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

Public health nutrition

Female nutritional status across the life-span in sub-Saharan Africa. 1. Prevalence patterns
Female nutritional status across the life-span in sub-Saharan Africa. 2. Causes and consequences
Blood cholesterol and triglycerides in adolescent Egyptian girls: Relation to anthropometric measurements
Daily versus weekly iron supplementation: Programmatic and economic implications for Indonesia

Female nutritional status across the life-span in sub-Saharan Africa. 1. Prevalence patterns

Abstract
Introduction
Macronutrient disorders
Micronutrient deficiencies
Gender differences in malnutrition in sub-Saharan Africa
Conclusion
Acknowledgements
References
Appendix 1. Selected findings concerning information on female nutritional status in sub-Saharan Africa

Joanne Leslie, Elizabeth Ciemins, and Suzanne Bibi Essama

Joanne Leslie is with the University of California, Los Angeles, School of Public Health, Department of Community Health Sciences, and The Pacific Institute for Women’s Health, in Los Angeles, California, USA. Elizabeth Ciemins is with the Los Angeles County Department of Health Services, STD Program, in Los Angeles. Suzanne Bibi Essama is with the Tulane University School of Public Health in New Orleans, Louisiana, USA.

Abstract

This article reviews and synthesizes existing nutritional studies that provide gender-disaggregated data from sub-Saharan Africa. The analytic focus is on female nutritional status across the life-span. However, it was found that available data are biased towards preschool children and women of reproductive age. As in other economically disadvantaged parts of the world, the two most prevalent nutritional deficiencies among females in sub-Saharan Africa are iron-deficiency anaemia and protein-energy malnutrition. In comparison with other regions of the world, sub-Saharan African females seem to be nutritionally better off than females in South Asia, but as malnourished as, or more malnourished than, females elsewhere. Indirect indicators of nutritional status, such as birthweight and maternal mortality, suggest that the nutritional situation of women in Western Africa is poorer than that of women in Eastern and Southern Africa. In comparison with males in sub-Saharan Africa, however, no consistent pattern of female nutritional disadvantage was found.

Introduction

Increasingly, planners recognize that the health and nutrition needs of females differ from those of males not only because of physiological sex differences, but also because of gender differences. Culturally defined roles and opportunities differ between males and females, leading to significant differences in their knowledge of health and nutrition, their exposure to health and nutrition risks, their access to care, and the social consequences they experience as a result of poor health and nutrition. Therefore, a broad analysis of the life circumstances of girls and women in Africa, not just a narrow focus on biological aspects of nutritional status, is required in order to design interventions that will be effective in improving the nutritional situation of sub-Saharan African females [1, 2].

As in most developing countries, the nutritional status of girls and women in sub-Saharan Africa is compromised by the cumulative and synergistic effects of many risk factors. These factors include limited availability of, or access to, food resources caused by natural and human-made disasters; lack of control over inputs and resource allocation at the household level; traditional feeding practices and other customs that limit women’s consumption of certain energy- or nutrient-rich foods; the energy demands of heavy physical labour; the nutritional demands of frequent cycles of pregnancy and lactation; and a high burden of infections with limited access to preventive or curative care.

This overview of the nutritional implications of the multiple time- and energy-demanding roles of sub-Saharan African women would lead the reader to anticipate a high prevalence of malnutrition among sub-Saharan African females. As this article makes clear, girls and women in this region are, indeed, severely malnourished. However, given their limited economic resources and their physically arduous lives, it is perhaps surprising that their nutritional status is not even worse than it is.

Two caveats are worth noting at the outset. Available data on nutrition in sub-Saharan Africa are biased towards pre-school children and women of reproductive age. Thus, our analysis is limited by the scarcity of reliable data on the female population outside these age groups. In addition, only rarely are the available data nationally representative, or are there comparable data over time.

Macronutrient disorders

Protein-energy malnutrition

As part of its Second Report on the World Nutrition Situation, the United Nations Administrative Committee on Coordination/Subcommittee on Nutrition (ACC/SCN) has published the most thorough and up-to-date global review of the nutritional situation of women of reproductive age. Thus, the ACC/SCN report provides the starting point for our assessment of the nutritional status of females in sub-Saharan Africa. (The ACC/SCN database was compiled from about 340 small- and medium-scale studies of the nutritional status of women 15 to 49 years of age carried out since the late 1970s [3]).

The four anthropometric measures of nutritional status that the ACC/SCN was able to use for their regional estimates of the prevalence of protein-energy malnutrition among women were height, weight, body mass index (BMI = weight in kilograms divided by height in meters squared), and arm circumference. Adult women in sub-Saharan Africa were found to be surprisingly tall. The average height is 158 cm, only 3 cm less than the average height of 161 cm for European women, whereas the mean height of women in South America and most of Asia is only about 151 cm. Although these differences might be assumed to be attributable to differences in genetic potential, the fact that the mean height of women in China was found to be exactly the same as that of sub-Saharan African women suggests that factors other than genetic potential must be important determinants [3]. Short stature or stunting among adults is usually taken as an indicator of cumulative malnutrition during childhood and adolescence, and it is associated with a range of negative functional outcomes, including reduced work capacity and poorer reproductive outcomes [4, 5]. The cut-off point for the definition of stunting used by the ACC/SCN is 145 cm, which is quite conservative. (As is discussed in our companion paper in this issue [6], increased obstetrical risk has been associated with short stature even at heights well above 145 cm). When this cut-off point was used, fewer than 5% of women in sub-Saharan Africa were classified as stunted, compared with more than 15% in Middle America and Asia (excluding China), and about 12% in South America.

The three other anthropometric measures of nutritional status used in the ACC/SCN report are all more indicative of current nutritional status. By these measures, women in sub-Saharan Africa also do fairly well in comparison with women in other regions, but the differences are not so striking as for height. Absolute weight, of course, is highly correlated with height. Using 45 kg as the cut-off point (again quite conservatively), the ACC/SCN found 20% of African women to be underweight. This is about the same percentage as in Middle America and China, more than in South America, but significantly less than in Asia (excluding China), where 45% of women in South-East Asia and a shocking 60% of women in South Asia are underweight, in large part because of stunting. The percentage of women with arm circumferences below 22.5 cm could only be calculated for sub-Saharan Africa, South Asia, and South-East Asia, and these percentages followed very closely those for weight below 45 kg. The relative position of women in sub-Saharan Africa is the worst when BMI, a measure of relative thinness, is considered. When BMI below 18.5 is used as the cut-off point, fewer than 20% of women in Middle America, South America, and China, slightly more than 20% of women in sub-Saharan Africa, and about 40% of women in South and South-East Asia are classified as excessively thin.

It is instructive to compare the regional information on child nutritional status from the Second Report on the World Nutrition Situation with the information on women’s nutritional status (although, unfortunately, the ACC/SCN does not report child nutritional status disaggregated by sex). Anthropometric indicators of child nutrition have been more routinely collected than information on adult nutritional status, so it is possible to estimate regional trends in child nutrition over the past 20 years. It is when trends in child nutrition are examined that the basis for the current concern about the nutritional situation in sub-Saharan Africa becomes clearer. In all other regions of the developing world, there was a marked decline between 1975 and 1990 in the prevalence of underweight pre-school children (defined as the percentage of children below two standard deviations from the mean weight-for-age in the age range zero to five years), ranging from a 50% decline in South America to a 10% decline in South Asia. However, the prevalence of child malnutrition in sub-Saharan Africa appears to have remained essentially unchanged. Because of the continuing rapid rates of population growth in sub-Saharan Africa, the absolute number of undernourished pre-school children increased from 18.5 million in 1975 to 28.2 million in 1990. This means that the present number of undernourished pre-school girls in sub-Saharan Africa is almost as great as the total number of undernourished pre-school children in 1975. It also appears that the relative position of children in sub-Saharan Africa is somewhat less favourable than the relative position of women when compared with other regions of the developing world. Perhaps, because of their smaller size and less mature immune systems, children have been less able than adults to withstand the particularly harsh conditions that have been endured by most sub-Saharan African countries over the last decade.

It would be speculative to try to assess the nutritional situation of women and girls in individual countries, or to compare one part of the region with another, because only aggregate regional estimates of protein-energy malnutrition in sub-Saharan Africa are available from the Second Report on the World Nutrition Situation. Although a substantial amount of research has been carried out on the nutritional situation in sub-Saharan Africa during the post-colonial period, the number, scope, and quality of the studies vary substantially from one country to another, and many otherwise excellent studies (particularly of child malnutrition) do not report sex-disaggregated results. Two approaches have been taken in this paper to begin to disaggregate the extent and diversity of the nutritional problems of females in sub-Saharan Africa.

The first approach is to examine three indirect indicators related to women’s nutritional status that are available for essentially all countries in the region. Table 1 presents the daily per capita energy supply, the percentage of infants with low birthweight, and the maternal mortality rate for all countries in sub-Saharan Africa for which such data could be found. Table 1 and other tables are organized alphabetically by subregion, and then alphabetically by country within each subregion. The four sub-regions - Eastern, Middle, Southern, and Western Africa - and the countries within them follow World Health Organization (WHO) usage [7]. For table 1, two World Bank sources were consulted [8, 9], and where conflicting numbers were encountered, both are presented in table 1. In most cases the numbers were close, but in a few cases they were so different as to make one or both suspect.

Daily per capita energy supply is not specific to individual households, much less to females within those households, but countries with a daily per capita energy availability below 2,100 kcal are usually designated as food insecure and believed to be at risk of having a substantial number of food-deficient households [10]. We also know that in sub-Saharan Africa, women who live in the most food-deficient households are usually the same women who do the most physically demanding work and who give birth to and breastfeed the largest number of infants. Therefore, without the ability to estimate a specific number, it seems safe to assume that in any country with a daily per capita energy supply below 2,100 kcal, there will be a reasonably high prevalence of female malnutrition. Of the 37 countries in table 1 for which kilocalorie estimates of daily per capita energy supply in 1989 were available, 12, or about one-third, fell below 2,100 kcal.

Several studies in sub-Saharan Africa (as well as many studies from other parts of the world) have demonstrated a relationship between women’s nutritional status - both current and past - and the birthweight of their infants [11-13]. Although low birthweight can occur in the absence of maternal malnutrition, and moderately malnourished women can give birth to infants of adequate birthweight, the correlation between poor maternal nutritional status and low birthweight at the aggregate level is sufficiently strong that it is appropriate to use the percentage of low-birthweight infants as an indirect indicator of female nutritional status. In general, we would expect to find a substantial degree of malnutrition among women of reproductive age (both stunting and thinness) in countries reporting a rate of more than 10% low-birthweight infants. Twenty-five of the 37 countries in table 1 for which information was available, or approximately two-thirds, reported more than 10% low-birthweight infants in the mid-1980s.

Maternal mortality rate is also a reasonable indicator of maternal nutritional status, particularly in the absence of adequate, accessible prenatal and childbirth services. Maternal mortality is clearly a major problem in the region; Africa has 20% of the world’s births but 40% of the world’s maternal deaths [14]. Stunting is associated with a greater risk of obstructed labour, and both obstructed labour and anaemia are among the major causes of maternal mortality in sub-Saharan Africa [5]. Therefore, although the maternal mortality rate does not differentiate among the different kinds of nutritional problems that may affect girls and women, a high maternal mortality rate is strongly suggestive of a high prevalence of female malnutrition. Virtually all of the countries in table 1 had maternal mortality rates in 1980 of over 100 maternal deaths per 100,000 live births, and 13 of the 36 countries for which data were available had maternal mortality rates above 500.

Comparisons among countries and even subregions of sub-Saharan Africa based on the data in table 1 must be made quite cautiously because of missing data and some lack of comparability among the data (e.g., maternal mortality rates for some countries are based entirely on hospital data). In addition, there are some inconsistencies in data between the two sources used to compile this table. Nonetheless, with these caveats in mind, the data in table 1 do present some slightly surprising findings when the four subregions are compared. In terms of food availability, table 1 suggests that Eastern Africa and Middle Africa are the two most food-insecure parts of the region. In both subregions, more than half the countries have a daily per capita energy supply below 2,100 kcal, and for most of the remainder of the countries, per capita energy availability is only slightly above the 2,100 kcal level. The food availability situation appears to be distinctly better in Southern and Western Africa.

TABLE 1. Selected health And agricultural indicators related to female nutritional status in sub-Saharan Africa

Country	Daily per capita energy supply, 1989a	% low-birthweight babies, 1985b	Maternal mortality rate per 100,000 live births, 19806
Eastern Africa
Burundi	1,932	14/18	800
Comoros	(89%)	7	460
Ethiopia	1,667	13	360
Kenya	2,163	13/18	170 c
Madagascar	2,158	10	300
Malawi	2,139	10	250
Mauritius	2,887	9/8	99
Mozambique	1,680	15/11	479/300 d
Rwanda	1,971	17	210
Somalia	1,906	-	1,100
Uganda	2,153	10	300/500
United Republic of Tanzania	2,206	4/13	185
Zambia	2,077	14	110
Zimbabwe	2,299	15/6	90
Middle Africa
Angola	1,807	17/21	-
Cameroon	2,217	13	303
Central African Republic	2,036	15	600
Chad	1,743	11	700/1,000
Congo	2,590	12	200
Gabon	2,383	16/8	130
Zaire	1,991	16	800
Southern Africa
Botswana	2,375	8	300/200
Lesotho	2,299	10	370
Namibia	1,946	-	-
South Africa	3,122	12	550 e
Swaziland	(110%)	7	-
Western Africa
Benin	2,305	10	1,680 d
Burkina Faso	2,288	18/11	600
Cape Verde	(112%)	-	107
Côte d’Ivoire	2,577	14/15	-
Gambia	(97%)	-	1,034
Ghana	2,248	17	1,070 d
Guinea	2,132	18
Guinea-Bissau	(105%)	14	400
Liberia	2,382	-	173
Mali	2,314	17/13	-
Mauritania	2,685	10	119/1,100
Niger	2,308	20	420
Nigeria	2,312	25	1,500
Senegal	2,369	10	530 c,d
Sierra Leone	1,799	14	450
Togo	2,214	20	418

Sources: World Bank: World Development Report 1992 [8] and Better Health in Africa [9].

a. Entries in kilocalories are from the 1992 World Development Report. Those printed in bold are below 2,100 kilocalories per person per day, which a recent FAD/WHO document labels as indicative of household food insecurity based on a very low level of average food consumption [10]. For countries with populations under one million, no data were available from the World Development Report, and the estimated daily per capita energy supply in 1985 as a percentage of requirement from Better Health in Africa [9] is given. Although figures vary from country to country according to the age and sex distribution of the population, the daily per capita energy supply needs to be about 2,300 to be equivalent to 100% of average requirements.

b. Where two numbers are given, the first is from the World Development Report [8] and the second from Better Health in Africa [9]. Where only one number is given, the two sources agreed, or only one source had an entry for that indicator and country.

c. Before 1980.

d. Hospital data only.

e. Rural data only.

Unlike daily per capita energy availability, the percentage of infants born with low birthweights and maternal mortality rates are specific to females. Therefore, table 1 appears to suggest that Western and Middle Africa are the two subregions where female malnutrition is the most prevalent, despite overall better food security. A question that may warrant further investigation is whether there are any dietary or behavioural factors in Eastern Africa that contribute to protecting female nutritional status in the presence of extremely low household food availability. Alternatively, it may be that access to health care is better in Eastern and Southern Africa, thus to some extent compensating for the negative effect of food insecurity.

Table 2 shows data that measure the nutritional status of women more directly, using BMI as a measure of chronic energy deficiency. Table 2 also shows BMI data for men from the studies where they were available, but discussion of the gender differences shown in table 2 is reserved for later in this section. Although different cut-off points have been recommended, in general a BMI below 18 or 18.5 is considered excessively thin, and such a person is categorized as suffering from chronic energy deficiency. It should be noted that although 18.5 seems to be an appropriate cut-off in terms of obstetrical risk, a lower cut-off, perhaps as low as 16, is more predictive of increased morbidity risk [1].

Among the countries from Eastern Africa with BMI data available, it is clear that the situation is most severe in Ethiopia, with half the women suffering from chronic energy deficiency. Twelve percent of women in the Zimbabwe study, and probably a similar percentage in the Kenya study, would be considered malnourished. For Middle Africa, data are only available from two studies done in Zaire. Although the data are not presented in terms of percentages below a cut-off, the relatively low range of the mean BMI values (19.7 to 21.7) suggests that the percentage of women suffering from chronic energy deficiency in these Zairian populations would be lower than in Ethiopia, but higher than in Zimbabwe or Kenya. As far as Western Africa is concerned, except for Côte d’Ivoire, where the mean BMI of women is similar to that found in the Kenya study, the BMI data support the conclusion of substantial female malnutrition suggested by the indirect indicators in table 1. The data in table 2 are consistent with the estimated regional average of 21% of women in sub-Saharan Africa who have a BMI below 18.5 [3].

It seems reasonably clear from the information presented in tables 1, 2, and 4 (discussed in more detail in the next section) that the problem of protein-energy deficiency is of considerable magnitude among females in sub-Saharan Africa, with somewhere in the range of 5% to 10% of girls suffering from acute protein-energy malnutrition, and 20% to 40% suffering from chronic protein-energy malnutrition. Among adult women, from 1% to 6% may suffer from severe chronic energy deficiency (chronic energy deficiency based on low BMI), and from 10% to 40% may suffer from mild to moderate chronic energy deficiency. In acute famine situations, of course, the proportion of females suffering from acute protein-energy malnutrition will be much higher.

Obesity

Although inadequate energy intake is certainly the major macronutritional problem among females in sub-Saharan Africa, it is important to assess the prevalence of excess energy intake or obesity as well. Obesity is known to substantially increase the risk of many chronic diseases, and with the decline in the rates of many infectious diseases and increasing life expectancy, a rapid increase in the prevalence of non-communicable diseases can be expected in the near future in sub-Saharan Africa [24]. The prevalence of obesity is a question that has received little attention from researchers concerned with nutrition in sub-Saharan Africa, so relevant data are extremely limited. One of the few studies to examine simultaneously the prevalence of chronic energy deficiency and obesity in the same population groups was a comparative study of adult nutritional status in India, Ethiopia, and Zimbabwe [16]. In the Ethiopian population, there was a high prevalence of chronic energy deficiency (BMI was less than 18.5 in 58% of women and less than 16 in 6%) and essentially no obesity, defined as BMI greater than 25. In contrast, in the Zimbabwean population, 12% of women had a BMI below 18.5 (of whom only 1% were below 16) and 17% were defined as obese, with 2.5% having a BMI over 30. It is clear that a high prevalence of chronic energy deficiency is correlated with a low prevalence of obesity and vice versa, but also that moderate prevalences of both can be found in the same population.

TABLE 2. Body mass index (BMI) of non-pregnant, non-lactating women and of men in selected sub-Saharan African countries

Country	BMI		Comments
	Women	Men
Eastern Africa
Ethiopia [15]	19.5	18.42 a	Significant seasonal variation; lowest BMI in Mar-Apr.
Ethiopia [16]	58% < 18.5 BMI (42% BMI 18.5-24.9)	50% < 18.5 BMI (51% BMI 18.5-24.9)	0.7% of women and no men were classified as obese (BMI ³ 25.0).
Kenya [1]	22.08	-
Tanzania [17]	29.5%< 20 BMI (61.2% BMI 20-25)	-	9.3% of women were classified as obese (BMI ³ 25.0).
Zimbabwe [16]	12% < 18.5 BMI (71% BMI 18.5-24.9)	15% < 18.5 BMI (80% BMI 18.5-24.9)	17.4% of women and 5.6% of men were classified as obese (BMI ³ 25.0).
Middle Africa
Zaire [18]	Lese: 21.7 Efe: 20.2	Lese: 21.6 Efe: 20.2	Significant seasonal variation in weight loss Dec-Jun.
Zaire [19]	Tembo: 19.7 Ntomba: 21.1 Highland: 21.35	-
Western Africa
Benin b [20]	11.5% < 18.0 BMI (67.4% BMI 18.0-23.0)	-	22% of women had BMI >23.0; significant seasonal variation in rural areas; lower BMI in May-Jun.
Côte d’Ivoire [21]	22.50	22.23
Gambia [1]	20.6	-
Ghana [1]	20.38	19.39
Ghana [22]	18.8% < 18.6 BMI (60.6% BMI 18.6-23.8)	40.1% < 19.9 BMI (54.8% BMI 19.9-25.0)	20.6% of women had BMI >23.8; 5% of men had BMI >25.0.
Guinea c	20.52	-	23.5% of women had BMI <18.6; 9.7% had BMI >23.8.
Mali [23]	20.8	20.0	11.9% of women and 12.4% of men were malnourished (BMI <18).

a. BMI of men was significantly lower than that of women.
b. All women in sample were lactating.
c. Personal communication, NB Mock and MK Konde, 1991.

Region	Pregnant women		Non-pregnant women		All women
	%	Number (000s)	%	Number (000s)	%	Number (000s)
Eastern Africa	47	3,380	41	13,540	42	16,920
Middle Africa	54	1,290	43	5,330	45	6,620
Southern Africa	35	380	30	2,500	30	2,880
Western Africa	56	4,170	47	15,120	48	19,290

Source: ref. 7, table 2 (p. 10).

Micronutrient deficiencies

Iron-deficiency, anaemia

Iron-deficiency anaemia is the most common nutritional deficiency in the world, and given that it particularly affects pre-school children and women of reproductive age, it is undoubtedly the most wide spread nutritional problem affecting girls and women in sub-Saharan Africa. It is generally accepted that about half the anaemia worldwide is due to iron deficiency, and there is emerging evidence that low iron stores, even in the absence of anaemia, can have negative functional consequences [3,7]. Therefore, the prevalence of iron-deficiency anaemia can be taken as a minimum estimate of the problems of both anaemia and iron deficiency.

Other important causes of anaemia, in addition to diets that are deficient in iron, folate, or vitamin B12, are haemolysis due to malaria and haemorrhage due to hookworm or schistosomiasis. In many African countries, genetic diseases, such as sickle-cell anaemia, and human immunodeficiency virus (HIV) infection can also lead to severe anaemia.

A publication prepared jointly by the Maternal Health and Safe Motherhood Programme and the Nutrition Programme of WHO gives the most recent estimates of the prevalence of nutritional anaemias in women in the world based on studies carried out since 1970 [7]. Table 3 presents data for the four sub-Saharan African subregions on the number and percentage of women with haemoglobin levels below normal, most of which is attributable to iron deficiency. The cut-offs used by WHO were less than 120 g/L haemoglobin for non-pregnant adult women and less than 110 g/L for pregnant women; almost all studies in the review by WHO included only women of reproductive age.

The relative nutritional situation of women by subregion in terms of nutritional anaemia is quite similar to the situation presented in table 1. Table 3 suggests that although women in Western Africa have the highest prevalence of nutritional anaemia, the prevalences in Western Africa, Middle Africa, and Eastern Africa are quite similar, while women in Southern Africa are definitely less anaemic. The better situation of women in Southern Africa has been attributed by some to the widespread use of iron cooking pots in this region [7].

One disturbing finding of the Second Report on the World Nutrition Situation is the apparent increase in iron deficiency in sub-Saharan Africa and many other regions of the developing world except for the Near East, North Africa, and South America. Although the ACC/SCN cautions that estimates of the trends over time in the prevalence of anaemia should be considered quite tentative, their data suggest that the prevalence of anaemia among non-pregnant adult women of reproductive age in sub-Saharan Africa was around 37% from the mid-1970s to the mid-1980s, and had increased to about 46% by the late 1980s. Given the population increase over this time, an unavoidable conclusion is that the absolute number of anaemic women in sub-Saharan Africa has probably increased quite dramatically in the past decade. Part of the reason for the increase in iron-deficiency anaemia is that the iron density in the diet appears to be decreasing rather than increasing in sub-Saharan Africa and in most other parts of the developing world. Although dietary iron density in sub-Saharan Africa appears relatively good compared with other parts of the developing world (7 to 8 mg per 1,000 kcal), the percentage of dietary iron from animal sources is lower than for any region except South Asia, and the general bioavailability of the dietary iron must be quite low.

Anaemia during pregnancy is widely recognized as one of the major health and nutritional problems among pregnant women in sub-Saharan Africa [5]. Factors that contribute to the high incidence of anaemia among pregnant women in the region include poor dietary practices during pregnancy due to sociocultural food taboos, infection, malabsorption, malaria, and increased foetal demand. A case-control study of 122 pregnant anaemic women in Nigeria [25] found that women were aware that they were at high risk for anaemia, but their traditional preventive practices against anaemia proved ineffective.

Iodine-deficiency disorders

Iodine-deficiency disorders exist in most regions of the world, although usually in pockets, rather than throughout a country. Iodine-deficient environments are those in which iodine, which is normally supplied from soil and water, has been leached from the topsoil by rain, flooding, glaciation, or snow. These environments tend to be either mountainous inland regions or floodplains. Iodine deficiency is less strongly correlated with food insecurity than are protein-energy malnutrition and iron-deficiency anaemia. It is rare to encounter iodine deficiencies in populations living near the sea or where the soil has adequate iodine, regardless of how impoverished or subject to seasonal food shortages they may be.

The main manifestations of iodine deficiency are goitre, impaired mental function, and increased rates of foetal wastage, stillbirths, and infant deaths. Severe mental and neurological impairment, known as cretinism, occurs among infants born to mothers who are seriously iodine deficient.

The extent of iodine-deficiency disorders is usually assessed by the prevalence of goitre in affected populations, although this understates the number of people affected by iodine-deficiency disorders, particularly if those suffering from reversible lethargy or mild mental impairment associated with iodine deficiency are included. The ACC/SCN estimates that the prevalence of goitre in Africa is about 8%. In Africa there are 39 million people with goitre, half a million with overt cretinism, and another 227 million who are estimated to be at risk for iodine-deficiency disorders [3, 26]. The ratio of those at risk for iodine-deficiency disorders to those with goitre is extremely high in Africa, reflecting, in part, the lack of control programmes, and suggesting that iodine-deficiency disorders will continue to be a serious public health problem in this region for many years to come.

Iodine-deficiency disorders are of particular concern among women for two reasons. First, the range of functional consequences of iodine deficiency is broader for women than for men, since it includes severe negative reproductive outcomes for both mothers and infants [26]. In addition to the broader range of functional consequences, the prevalence of goitre appears to be significantly higher among females than among males in virtually all studies with sex-disaggregated data [27]. In one Africa-specific study that reported sex-disaggregated data from Zaire, Thilly et al. [28] found a significantly higher prevalence of goitre among females than males at all ages from 10 years upwards. The prevalence of visible and voluminous goitres in the age range of peak prevalence (20 to 30 years) was almost 50% among females and about 20% among males.

Vitamin A deficiency

Vitamin A deficiency, as defined by eye damage (ranging from reversible night blindness through ulceration of the cornea to permanent scarring and blindness), has been identified as a widespread public health problem in at least 37 countries [3]. Each year it is estimated that between 250,000 and 500,000 pre-school-age children go blind from vitamin A deficiency, and that within months of going blind, two-thirds of these children die [3]. In addition, there is growing evidence that even children who do not necessarily have eye signs may have subclinical vitamin A deficiency that puts them at greater risk for morbidity and mortality from infectious diseases. Although all children older than six months, as well as pregnant and lactating women, are at risk for vitamin A deficiency, the peak prevalence seems to fall in the age range of two to four years [29]. The literature supports the general finding of a higher prevalence of eye damage due to vitamin A deficiency among pre-school-age boys than girls [30]. It is not well established whether adult men are similarly at greater risk compared with adult women, because few studies have been done on vitamin A deficiency among adults. However, in a small number of somewhat older country-specific studies from South Africa, Ethiopia, and Rwanda, higher deficiency rates were found among adult men than adult women [31-33].

In Africa about 7.2% of pre-school-age children (1.3 million) are estimated to have eye damage due to vitamin A deficiency, and another 7.2 million suffer from a mild to moderate deficiency [3, 30]. The proportion of the pre-school-age population affected in Africa is similar to the proportion in most other parts of the developing world. Within the sub-Saharan African region, the areas most affected are Eastern Africa, Southern Africa, and the Sahelian parts of Western Africa. Populations living where red palm oil is produced or distributed, that is, along the coastal parts of Western Africa and in some parts of Central Africa, are reasonably well protected against vitamin A deficiency.

Gender differences in malnutrition in sub-Saharan Africa

Throughout the world, differences between males and females in the prevalence of micronutrient deficiencies appear to be substantially attributable to biological differences between the sexes. The higher prevalence of iron-deficiency anaemia found among adolescent girls and adult women, for example, is due primarily to the increased iron losses associated with menstruation and the increased iron demands of pregnancy and lactation, although this biological risk can be exacerbated by female diets that are lower in animal protein or, in some cases, a higher prevalence of hookworm or malaria among females [34]. Similarly, higher prevalences of iodine-deficiency disorders among adult women and of vitamin A deficiency among pre-school-age boys (and perhaps among older males) are documented by studies carried out in many different cultural settings and appear to be primarily physiological, although the specific mechanisms are less well understood than in the case of iron-deficiency anaemia. In addition, there may be local dietary practices that enhance or reduce the biological gender differences.

In contrast, as far as macronutrient disorders are concerned, there are no underlying physiological reasons to expect differences between males and females in the prevalence of thinness or obesity. Where such a pattern does emerge, behavioural and cultural factors must provide the explanation. Although there are significant differences in the roles and opportunities of males and females in sub-Saharan Africa, no widespread pattern of gender differences in protein-energy malnutrition has emerged from studies to date [35, 36]. On the basis of recent sex-disaggregated Demographic and Health Survey data for children (see table 4), BMI data for adult men and women (see table 2), a secondary analysis of somewhat older height and weight data from Eveleth and Tanner’s Worldwide Variation in Human Growth [37] undertaken by Svedberg [36], and other country-specific studies [see, for example, refs. 22, 38, and 39], it seems clear that in contrast to the situation in South Asia, in sub-Saharan Africa there is no significant pattern of female disadvantage according to anthropometric measures of nutritional status.

In his extensive analysis of gender bias in undernutrition in sub-Saharan Africa, Svedberg actually proposed that the slight anthropometric advantage shown by girls, women, or both in many countries may suggest a historical pattern of preferential treatment of females due to the high value placed on women’s agricultural labour [36]. On the basis of a study of gender biases among the Mukogodo of Kenya, Cronk [40] suggested that favouritism towards daughters occurred as a result of lowered socio-economic status. In a case study in south-central Ethiopia, Vesti [15] found that females scored significantly better than males on weight-for age measurements, but only within the lowest income bracket. However, given that there are also a number of studies in sub-Saharan Africa that report dietary discrimination against females [see, for example, refs. 41-43], any firm conclusion of a nutritionally advantaged position of females in the region seems premature.

One particularly interesting report of gender differences in dietary intake comes from a study of child-feeding practices in Zinder, Niger. Field research conducted by CARE on traditional knowledge and practices related to child care and feeding revealed that in some villages girls were weaned one month later than boys. It was traditionally believed that an excess of breastmilk would make a child stupid, and accordingly boys were weaned earlier “so that they would be intelligent” and have a better chance of success in school [44]. Although this practice may be beneficial to girls’ nutrition in the short run, it is actually motivated by a lower value placed on education for females, which may be detrimental to female health and nutritional status in the longer run. A similar gender difference in breastfeeding patterns was reported in a study of women in Dakar, Senegal, and the surrounding area, where girls were breastfed up to 24 months and boys only to 18 months. However, no link with intelligence was made [45]. It is quite likely that the aggregate finding of little gender difference in anthropometric measures of protein-energy malnutrition in sub-Saharan Africa reflects the cumulative effect of a number of specific behaviours and practices, some of which may favour females and some of which may favour males. In specific settings or during certain seasons, significant gender differences have been reported. (See appendix 1 for further examples).

TABLE 4. Child health and nutrition status indicators by gender for selected sub-Saharan African countries (late 1980s) ^a

Country	Child mortality per 1,000 b		Height-for-age c		Weight-for-height c		Weight-for-age c		% with diarrhoea d		% taken to health facility with diarrhoea
	M	F	M	F	M	F	M	F	M	F	M	F
Eastern Africa
Burundi	101.0	113.8	48.3	47.8	6.2	5.1	37.5	39.0	17.7	17.0	12.8	11.7
Uganda	97.3	86.0	47.3	41.6	1.8	1.9	23.1	23.4	25.4	23.2	13.8	15.8
Zimbabwe	30.2	32.5	29.9	28.0	1.4	1.3	11.3	11.9	20.5	19.0	33.5	33.2
Western Africa
Ghana	78.3	79.4	30.2	29.8	9.0	6.9	30.3	31.1	27.0	26.1	40.8	45.6
Mali	166.0	174.0	23.8	24.9	12.0	9.8	30.0	32.2	35.7	133.0	68.4	68.3
Nigeria	93.7	89.1	43.4	42.7	9.8	8.3	35.8	35.7	19.4	16.4	23.6	26.8
Senegal	131.0	129.7	24.8	20.6	7.6	4.1	22.5	20.6	38.9	36.9	19.4	19.4
Togo	74.9	90.1	32.2	26.9	6.2	4.4	25.0	23.8	29.6	29.2	24.2	26.7

a. Data were kindly provided by Hill of the World sank They are assembled from Demographic and Health survey Reports and are for all sub-Saharan African countries for which data on these indicators were available.

b. Child mortality is usually considered most indicative of child nutritional status.

c. For children 3-36 months old, percentage below 2 SD.

d. Percentage of children under age five with diarrhoea during the two weeks preceding the survey.

Conclusion

As in other parts of the developing world, the two most prevalent nutritional deficiencies among females in sub-Saharan Africa are iron-deficiency anaemia and protein-energy malnutrition. Evidence for other micronutrient deficiencies among girls and women in the region is quite limited, but undoubtedly iodine-deficiency disorders are a major problem in many inland parts of sub-Saharan Africa, and vitamin A deficiency is probably quite prevalent in rural Sahelian communities where there are significant seasonal fluctuations in quantity and quality of the diet.

In comparison with other regions of the world, sub-Saharan African females seem to be nutritionally better off than females in South Asia, but as malnourished as, or more malnourished than, females in most other parts of the developing world. The nutritional situation of females in Western Africa seems to be poorer than in other parts of the region. In contrast with South Asia, there is no consistent pattern of a higher prevalence of protein-energy malnutrition among females than males, despite a generally higher work burden among adult women than among men in sub-Saharan Africa. However, small-scale studies from a few countries have found evidence of discrimination against or disadvantage experienced by females in breastfeeding and dietary patterns, and the lack of attention to gender differences in much of the work that has been done on the nutritional problems of sub-Saharan Africa means that no firm conclusion can be reached about the relative nutritional status of males and females. The companion article [6] explores evidence concerning the determinants and consequences of the nutritional status of girls and women in sub-Saharan Africa.

Acknowledgements

The review on which this and its companion article are based was originally undertaken as part of the work of the Institute of Medicine (IOM) Committee to Study Female Morbidity and Mortality in sub-Saharan Africa, of which the first author was a member. We would like to acknowledge the extremely useful comments and input of the other committee members (Maureen Law [Chair], Uche Amazigo, Judith Fortney, Philip L. Graitcer, Françoise F. Hamers, H. Kristian Heggenhougen, Karungari Kiragu, Walinjom F. T. Muna, Jonathan E. Myers, Benjamin O. Osuntokun, Patience W. Stephens, Judith N. Wasserheit, and Belmont E. O. Williams) and of the IOM staff who worked with the committee (folly F. Harrison, Dana Hotra, Delores Sutton, and most especially, Christopher P. Howson, Project Director).

References

1. Kennedy E, Garcia M. Effects of selected policies and programs on women’s health and nutritional status. Washington, DC: International Food Policy Research Institute, 1993.

2. Leslie J. Women’s lives and women’s health: using social science research to promote better health for women. J Wom Health 1992; 1(4): 307-18.

3. United Nations Administrative Committee on Coordination/Subcommittee on Nutrition. Second report on the world nutrition situation. Vol. I. Global and regional results. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, 1992.

4. Buzina R. Bates CJ, van der Beek J. Brubacher G. Chandra RK, Hallberg L, Heseker J. Mertz W. Pretrazik K, Pollitt E, Pradilla A, Suboticanec K, Sandstend HH, Schalch W. Spurr GB, Westennofer J. Workshop on functional significance of mild-to-moderate malnutrition. Am J Clin Nutr 1989; 50: 172-6.

5. Royston E, Armstrong S. eds. Preventing maternal deaths. Geneva: World Health Organization, 1989.

6. Leslie J. Essama SB, Ciemins E. Female nutritional status across the life-span in sub-Saharan Africa. 2. Causes and consequences. Food Nutr Bull 1997; 18: 44-55.

7. World Health Organization. The prevalence of anaemia in women: a tabulation of available information. 2nd ed. WHO-MCH-MSM/92.2. Geneva: World Health Organization, 1992.

8. World Bank. World development report 1992: development and the environment. Washington, DC: World Bank, 1992.

9. World Bank Africa Technical Department. Better health in Africa. Washington, DC: World Bank, August 1992.

10. Food and Agriculture Organization/World Health Organization. Improving household food security. A theme paper prepared for the International Conference on Nutrition. PREPCOM/ICN/92/INF/6. Rome: Food and Agriculture Organization, 1992.

11. Harrison KA, Rossiter CE, Chong H. Relations between maternal height, fetal birthweight and cephalo-pelvic disproportion suggest that young Nigerian primi-gravidae grow during pregnancy. Br J Obstet Gynaecol 1985(suppl 5): 40-8.

12. Prentice AM, Cole TJ, Foord FA, Lamb WH, White-head RG. Increased birthweight after prenatal dietary supplementation of rural African women. Am J Clin Nutr 1987; 46: 912-25.

13. Neumann C, Bwibo NO, Sigman M. Functional implications of malnutrition: Kenya Project Final Report. Human nutrition collaborative research support program. Los Angeles, Calif, USA: University of California, Los Angeles, School of Public Health, 1992.

14. WHO says maternal mortality figures substantially underestimated. Nation’s Health, March 1996.

15. Vesti SA, Witcover J. Gender differences in levels, fluctuations and determinants of nutritional status: evidence from South Central Ethiopia. In: Kennedy E, Garcia M, eds. Effects of selected policies and programs on women’s health and nutritional status. Washington, DC: International Food Policy Research Institute, 1993: 1-33.

16. Ferro-Luzzi A, Sette S. Franklin M, James WPT. A simplified approach to assessing adult chronic energy deficiency. Eur J Clin Nutr 1992; 46: 173-86.

17. Kavishe FP. Analysis of the nutrition situation and trends in Tanzania during the 1980s. In: Review of nutrition relevant actions in Tanzania. Report for the United Nations ACC/SCN. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, February 1992: 27-50.

18. Bailey RC, Jenike MR, Ellison PT, Bentley GR, Harrigan AM, Peacock NR. The ecology of birth seasonality among agriculturalists in Central Africa. J Biosoc Sci 1992; 24: 393-412.

19. Caraël M. Relations between birth intervals and nutrition in three central Africa populations (Zaire). In: Mosley WH, ed. Nutrition and human reproduction. New York: Plenum Press, 1978: 365-84.

20. Fakambi LK. Factors affecting the nutritional status of mothers. The food and nutrition program of the Ouando Horticulture and Nutrition Center in the People’s Republic of Benin. Washington, DC: International Center for Research on Women, June 1990.

21. Thomas D, Lavy V, Strauss J. Public policy and anthropometric outcomes in Côte d’Ivoire. Living Standards Measurement Study Working Paper No. 89. Washington DC: World Bank, 1992.

22. Alderman H. Nutritional status in Ghana and its determinants. Social dimensions of adjustment in sub-Saharan Africa. Working Paper No. 3. Washington, DC: World Bank, 1990.

23. Dettwyler KA. Nutritional status of adults in rural Mali. Am J Phys Anthropol 1992; 88: 309-21.

24. Feachem RG, Jamison DT, Bos ER. Changing patterns of disease and mortality in sub-Saharan Africa. In: Feachem RG, Jamison DT, eds. Disease and mortality in sub-Saharan Africa. New York: Oxford University Press for the World Bank, 1991: 3-27.

25. Elegbe I, Ojofeitimi EO, Elegbe IA. Traditional treatment of pregnancy and anemia in Nigeria. Trop Doct 1984; 14: 4, 175-7.

26. Hetzel BS. The prevention and control of iodine deficiency disorders. ACC/SCN State-of-the-Art Series, Nutrition Policy Discussion Paper No. 3. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, 1988.

27. Simon PA, Jamison DT, Manning MA. Gender differences in goiter prevalence: a review. Los Angeles, Calif, USA: University of California, Los Angeles, Graduate School of Education, 1990.

28. Thilly CH, Delange F. Ramioul L, Lagasse R. Luvivila K, Ermans AM. Strategy of goiter and cretinism control in Central Africa. Int J Epidemiol 1977; 6: 43-54.

29. Eastman SJ. Vitamin A deficiency and xerophthalmia: recent findings and some programme implications. In: Mandl P-E, ed. Assignment children. New York: UNICEF, 1987.

30. Levin HM, Pollitt E, Galloway R. McGuire J. Micro-nutrient deficiency disorders. In: Jamison DT, Mosley WH, Meashan A, Bobadilla JL, eds. Disease control priorities in developing countries. New York: Oxford University Press for the World Bank, 1993: 421-51.

31. May JM, McLellan DL. The ecology of malnutrition in seven countries of Southern Africa and in Portuguese Guinea. Stud Med Geog 1971; 10: 3-414.

32. May JM, McLellan DL. The ecology of malnutrition in Eastern Africa and four countries of Western Africa. Stud Med Geog 1970; 9: 1-637.

33. May JM, McLellan DL. The ecology of malnutrition in Middle Africa. Stud Med Geog 1965; 8: 5-241.

34. Gillespie S. Kevany J. Mason J. Controlling iron deficiency. ACC/SCN State-of-the-Art Series, Nutrition Policy Discussion Paper No. 9. Geneva: World Health Organization, 1991.

35. Kennedy E, Bentley M. Women’s health and nutrition in sub-Saharan Africa: a review and case study from Kenya. In: Sorkin A, Siralgedin I, eds. Research in human capital and development: nutrition, food policy and development. Greenwich, Conn, USA: JAI Press, 1994: 243-76.

36. Svedberg P. Undernutrition in sub-Saharan Africa: Is there a gender bias? J Dev Stud 1990; 26: 469-86.

37. Eveleth PB, Tanner JM. Worldwide variation in human growth. Cambridge, UK: Cambridge University Press, 1976.

38. Thomas D. Gender differences in household resource allocations. Living Standards Measurement Study Working Paper No. 79. Washington DC: World Bank, 1991.

39. Strauss J. Households, communities, and preschool children’s nutrition outcomes: evidence from rural Côte d’Ivoire. Econ Dev Cult Change 1990; 36: 231-62.

40. Cronk L. Low socioeconomic status and female-biased parental investment: the Mukogodo example. Am Anthropol 1989; 91: 414-29.

41. Caplan P. Perceptions of gender stratification. Africa 1989; 59: 196-208.

42. Zumrawi FY. Local influences and customs on diets in Sudan village community. Ahfad J 1988; 5(2): 27-32.

43. Trueblood LW. Preface to May JM, McLellan DL, eds. The ecology of malnutrition in Eastern Africa and four countries of Western Africa. Stud Med Geog 1970; 9: vii-viii.

44. Swimmer K. Field research on traditional knowledge and practices related to child care and child feeding. Final report. The Zinder Child Health Project. Zinder, Niger: CARE International, 1990.

45. Faladé S. Women of Dakar and the surrounding urban area. In: Paulme D, ed. Women of tropical Africa. Berkeley and Los Angeles, Calif, USA: University of California Press, 1963: 217-29.

46. James WPT, Ferro-Luzzi A, Watenlow JC. Definition of chronic energy deficiency in adults. Eur J Clin Nutr 1988; 42: 969-81.

47. Gebre-Medhin M, Gobezie A. Dietary intake in the third trimester of pregnancy and birthweight of offspring among nonprivileged and privileged women. Am J Clin Nutr 1975; 28: 1322-9.

48. UNICEF. The wellbeing of children. In: Situation analysis of children and women in Kenya. Nairobi, Kenya: Central Bureau of Statistics, Ministry of Finance and Planning, 1984: sect. 4: 1-135.

49. Ferguson A. Women’s health in a marginal area of Kenya. Soc Sci Med 1986; 23: 17-29.

50. UNICEF. Third rural child nutrition survey 1982. Nairobi, Kenya: Central Bureau of Statistics. Ministry of Finance and Planning, 1993.

51. Kulin HE, Bwibo N. Mutie D, Santner SJ. The effect of chronic childhood malnutrition on pubertal growth and development. Am J Clin Nutr 1982; 36: 527-36.

52. Jansen AAJ, Kusin JA, Thiuri B. Lakhani SA, Mannetje W. Anthropometric results in pregnancy and lactation. In: Van Ginneken JK, Muller AS, eds. Maternal and child health in rural Kenya. London and Sydney: Croom Helm, 1984: 143-52.

53. Jansen AAJ, Gemert W, Thiuri B, Lakhani SA. Growth of infants of low and normal birth weight. In: Van Ginneken JK, Muller AS, eds. London and Sydney: Croom Helm 1984: 197-206.

54. Liljestrand J. Maternal morbidity in Mozambique. Uppsala, Sweden: Uppsala University, 1985.

55. Bantje HF. Female stress and birth seasonality in Tanzania. J Biosoc Sci 1988; 20: 195-202.

56. Carswell FL, Hughes AO, Palmer RI, Higginson J, Harland PS, Meakins RH. Nutritional status, globulin titers, and parasitic infections of two populations of Tanzanian school children. Am J Clin Nutr 1981; 34: 1292-9.

57. Kwofie K, Brew-Graves E, Adika GH. Malnutrition and pregnancy wastage in Zambia. Soc Sci Med 1983; 17: 539-43.

58. Tsu VD. Maternal height and age: risk factors for cephalopelvic disproportion in Zimbabwe. Int J Epidemiol 1992; 21: 941-6.

59. Fleisher K, Schuon R. Nutritional status of 6,867 primary school children in the Masvingo District. Cent Afr J Med 1988; 34(8): 185-9.

60. Chandiwana SK, Kambaza A, Mutetwa SM. A study of nutritional status, parasitic infections and haematology in a farmworker community in Zimbabwe. Cent Afr J Med 1984; 30(9): 172-5.

61. May JM, McLellan DL. The ecology of malnutrition in the French speaking countries of West Africa and Madagascar. Stud Med Geog 1968; 8: 5-241.

62. United States Agency for International Development. Cameroon national nutrition survey. Washington, DC: USAID, 1978.

63. Schoept BG. Social structure, women’s status and sex differential nutrition in the Zairian copperbelt. Urban Anthropol 1987; 16: 73-102.

64. Vis HL, Bossuyt M, Hennart P. Caral M. The health of mother and child in rural central Africa. Stud Fam Plann 1975; 6: 437-41.

65. Serdula MK, Akphane JM, Kunene PF, Gama DM, Staehling N. Peck R. Seward J. Sullivan B. Trowbridge FL. Acute and chronic undernutrition in Swaziland. J Trop Pediatr 1987; 33: 35-42.

66. Bleiberg FM, Brun TA, Goiham S. Gouba E. Duration of activities and energy expenditure of female farmers in dry and rainy seasons in Upper-Volta. Br J Nutr 1980; 43: 71-82.

67. Thomas D, Lavy V, Strauss J. Public policy and anthropometric outcomes in Côte d’Ivoire. Living Standards Measurement Study Working Paper No. 89. Washington, DC: World Bank, 1992.

68. Bates CJ, Prentice AM, Prentice A, Paul AA, White-head RG. Seasonal variations in ascorbic acid status and breast milk ascorbic acid levels in rural Gambian women in relation to dietary intake. Trans R Soc Trop Med Hyg 1982; 76: 341-7.

69. Lawrence M, Whitehead RG. Physical activity and total energy expenditure of child-bearing Gambian village women. Eur J Clin Nutr 1988; 42: 145-60.

70. Prentice AM, Whitehead RG, Roberts SB, Paul AA. Long term energy balance in child-bearing Gambian women. Am J Clin Nutr 1981; 34: 2790-99.

71. Tripp RB. Farmers and traders: some economic determinants of nutritional status in northern Ghana. J Trop Pediatr 1981; 27: 15-22.

72. Ofosu-Amaah, Neumann A. Danfa project final report: the Danfa comprehensive rural health and family planning project. Accra, Ghana, and Los Angeles, Calif, USA: University of Ghana Medical School, Korie Bu Hospital, and University of California, Los Angeles, School of Public Health, 1979.

73. Loutan L, Lamotte JM. Seasonal variations in nutrition among a group of nomadic pastoralists in Niger. Lancet 1984; 1(8383): 945-7.

74. Fleming AF. Tropical obstetrics and gynaecology. 1. Anaemia in pregnancy in tropical Africa. Trans R Soc Trop Med Hyg 1989; 83: 441-8.

75. United States Agency for International Development. Sierra Leone national nutrition survey. Washington, DC: USAID, October 1978.

Appendix 1. Selected findings concerning information on female nutritional status in sub-Saharan Africa

Country	Study description		Findings
	Year	Sample
Eastern Africa
Ethiopia [15]	1993	Four rural peasant associations in Sike Awraja, Southern Shewa Province	Females were consistently better off nutritionally than males. Seasonality variations affected both males and females, regardless of age. Household structure, marital status, and household sharing arrangements had a detrimental effect on adult female nutritional status, but had no effect on males. Higher household income levels meant higher nutritional status for women, but not for men. Timing and sources of income affected men’s nutritional status less than women’s, especially seasonally.
Ethiopia [16]	1992	1,087 study population	Mean BMI was 18.5 for women and 18.6 for men. Less than 1% of women and men were obese. 57% of women and 50% of men were classified as having CED. Average energy intake for normal women was 2,067 kcal/d. Average energy intake for women with CED was 1,979 kcal/d.
Ethiopia/Somalia [46]	1988	255 Ethiopian peasant women aged 36.1 ± 12.1 yr; 482 male Ethiopian peasants and Somalian army recruits aged 33.4 ± 13.1 yr	3% of Ethiopian women had grade III CED (BMI <16.0). 24% of 182 Ethiopian and Somalian men had grade III CED. In these same groups, 38% of the women and 63% of the men were classified as having less than a normal BMI of 18.5 (grades I and II CED). Among women particularly, subjects from the third world had a lower proportion of body fat than subjects of equivalent BMI in developed countries.
Ethiopia [32]	1970		Protein-deficiency syndromes were found in 8% of females and 2% of males. Anaemia was found in 19% of females and 6% of males. Enlargement of the thyroid gland was found in 9% of females and 2% of males. Vitamin A deficiency was found in 3% of females and 10% of males. Bitot’s spots were found in 1.8% of girls and 3.3% of boys in Addis Ababa.
Ethiopia [47]	1970	20 non-privileged (income <US$100/mo) and 10 privileged (income >US$650/mo) primigravidae	Non-privileged women consumed a diet deficient in all nutrients except iron and thiamine, with an average daily protein and energy intake less than 60% of FAO/WHO recommendations. The diet of privileged women met the recommendations for all nutrients except calcium (45% of recommendations) and vitamin A (95% of recommendations). Babies of non-privileged women had lower mean birthweights than babies of privileged women. Percentage of dietary energy from cereals was 41.2% for privileged and 65.5% for non-privileged women. Percentage of dietary energy from dairy products, meat, and fish was 22.0% for privileged women and 4.3% for non-privileged women.
Kenya [13]	1984-86	290 rural Embu households	Women progressively lowered their food intake during pregnancy, not only during times of food shortage, but also during normal years. Mean weight gain during pregnancy was only 50% of the recommended weight gain for US or UK women. Mean birth length was below normal (20th per-centile), and 16% of infants were considered intrauterine growth-retarded, indicative of maternal malnutrition.
Kenya [48]	1984		Average height of Kenyan children (girls and boys) was that of the lowest 3% in the reference population. Extent of chronic malnutrition may be partly due to cultural practices, such as serving food first to the father and other men, next to the first son and other children according to age, and last to the mother.
Kenya [49]	1983	524 households in Kibwezi	Women had a higher level of stress accompanied by ill health than men. Women were thinner and shorter than women in other regions of the district. Prevalence of chronic disability was higher among women than among men, especially women of child-bearing age.
Kenya [50]	1982	Survey 1: 1,400 rural children aged 1-4 yr Survey 2: 4,000 urban and rural children aged 6-60 mo	24% of all children were stunted (height-for-age <90% of standard); 3% were wasted (weight-for-height <80% of standard); 7% were severely malnourished (weight-for-age <60% of standard).
Kenya [51]	1982	342 privileged urban children aged 10-13 yr; 347 impoverished rural adolescents aged 10-18 yr	Rural children were nutritionally worse off than urban children by anthropometric measures. Among malnourished rural adolescents, sexual maturity was delayed by 2.1 yr in girls and 3 yr in boys compared with WHO standards.
Kenya [52]	1978-80	1,739 mothers, mean age 26.3 yr	In early pregnancy, mean weight of women was higher than standard; towards the end of pregnancy, it was slightly less. Average weight gain between 3.1 and 7.6 mo was 57.8% of the reference value for women of the same height. Mean upper arm circumference remained below 90% of the standard for non-pregnant women. At all times, mothers of LBW babies had lower values for anthropometric measurements than mothers of NBW babies.
Kenya [53]	1978-80	Children aged 0-2+ yr: 113 LBW girls, 80 LBW boys, 97 NBW girls, 62 NBW boys	Mean weights of NBW girls were persistently less than those of NBW boys. Mean weights of LBW girls were also less than those of LBW boys, but the mean ratios for boys are less than those for girls; towards the end of the first year of life the mean ratios converge. The same applies to length. Curves for weight and length of NBW and LBW children deviate from the standard at 18 wk for girls and 12 wk for boys. Weight curves for LBW children show catch-up growth during the first months of life, more pronounced in girls than in boys. Later both curves run a course below the 80% line. Mean birth length curves for LBW children follow the 90% line for boys and a slightly higher level for girls. Mean weights of both LBW and NBW boys lag slightly behind those of girls at 52 wk. There was a gradual fall in mean weight-for-height with advancing duration of pregnancy.
Malawi [32]	1970		Protein-energy malnutrition was very common among pre-school children, and a striking feature was the general lack of well-being of women, especially mothers (Dr. A. Burgess, WHO, 1964).
Mozambique [54]	1985	1,060 pregnant women (68% rural) from 10 sites	Women from the North were shorter than women from the South (average height, 152 and 159 cm, respectively). Nulliparous women who needed Caesarean section because of cephalopelvic disproportion were, on average, significantly shorter than controls; 36% of women who had Caesarean section and 9% of those who delivered vaginally were <150 cm tall. Maternal height was also significantly correlated with previous perinatal loss. Percentage of women in a site with haemoglobin levels below 90 g/L varied from 5% to 15%; 1% of all women had haemoglobin levels below 70 g/L. Main causes of anaemia were found to be iron deficiency and malaria. Low haematocrit values were associated with small upper arm circumference and low skinfold thickness, indicating that anaemia was correlated with undernutrition. Occasional cases of folic acid deficiency were found among severely anaemic women.
Rwanda [34]	1965		Prevalence of Bitot’s spots per 1,000 was 2.6 for women, 13.3 for men, 3.6 for girls, and 15.4 for boys. Goitre was more frequent among females than among males in Rwanda and Burundi.
Sudan [32]	1970		Kwashiorkor and marasmus were both common, perhaps because of the widespread custom of allowing fathers and older boys to eat first, leaving women and children to fend for themselves.
United Republic of Tanzania [55]	1988	Data obtained from routine delivery records of two towns, populations 17,000 and 40,000, registering 800 and 1,800 deliveries per year, respectively.	Births were markedly seasonal in areas with holendemic malaria, especially among older women, as a result of physical exhaustion due to food shortage, heavy work, and anemia, aggravated by a high rate of infection with malaria.
United Republic of Tanzania [55]	1985		Birthweights were higher in urban areas and higher in years of lower rainfall. Seasonal and yearly variations were in the range of 100-200 g. The magnitude of the variation is comparable to that recorded in famine conditions, so it is probably significant. The variation is affected by differences in the food supply, which can cause variations in birthweight of 50 g. Seasonal variation was primarily due to the coincidence of seasonal food shortage and heavy agricultural work.
United Republic of Tanzania [56]	1976	244 pupils, 51% male, mean age 12.5 yr	43 of 117 girls (37%) had weight-for-age between 70% and 80% of standard, and 31 (26%) had weight-for-age below 70% of standard; 31 of 121 boys (26%) had weight-for-age between 70% and 80% of standard, and 62 (51%) had weight-for-age below 70% of standard.
United Republic of Tanzania [32]	1970		In Kisarawe district, 50% of women and 27% of men had clinical signs of anaemia, which was thought to be a result more of malaria and hookworm infestation than of nutritional deficiency. In the same district, 39% of women, 20% of men, and 24% of children were infested with hookworm.
Zambia [57]	1983	Subjects from recent nutritional status and dietary surveys	Of 572 pregnant and lactating mothers examined, 95.3% had low or deficient serum albumin levels. Of the same group, 90.5% had high serum globulin levels. 12% of examined women were deficient and 10.2% had low levels of haemoglobin. Nearly 2% of the women had deficient retinal or vitamin A levels. 82% of the women had low retinal levels.
Zambia [32]	1970	48 girls, average age 14.5 yr	84.4% had signs of malnutrition, including skin rashes (48.4%), corneal vascularization (46.8%), glossitis (78%), and dental caries (50%). Anthropometric measures, particularly height, showed these girls to be behind European children of the same age groups.
Zimbabwe [58]	1992	203 women with operative deliveries due to cephalopelvic disproportion; 299 women with normal, unassisted deliveries	Maternal height <160 cm was associated with a twofold increased risk for cephalopelvic disproportion.
Zimbabwe/Ethiopia [16]	1992	4,528 individuals; 1,288 adults aged >18 yr	Adults in this population were taller and heavier than Indians. Mean BMI was 22.0 for women and 20.7 for men. BMI for women significantly exceeded that for men. 18% of women and 6% of men had different degrees of obesity (grade I). 11% of women and 14% of men were classified with CED.
Zimbabwe [59]	1983-84	6,867 primary-school children aged 6-17 yr	High prevalence of stunting (15.8%) was found for all children; boys were more affected than girls.
Zimbabwe [60]	1984	200 people of farmworker community; 40% female and 60% male, aged 4-40 yr	Percentage of wasted females was greater than the percentage of wasted males. 16.9% of females and 0.0% of males were severely malnourished; 11.9% of females and 6.6% of males were moderately malnourished; 28.9% of females and 22.0% of males were mildly malnourished.
Zimbabwe [32]	1970		56 of 71 females and 26 of 40 males had goitrous thyroid enlargement. In a survey of 2 villages (Dewe and Tendenguwo), the total village prevalence was 74%. 186 of 243 children had goitre (80% of girls and 59% of boys). In another region (Omay Tribal Trust Land), the highest prevalence of goitre (77%) occurred among adolescent girls; the prevalence for girls aged 6-12 yr was 67.9%. Among 341 patients at Harare Hospital, anaemia was noted in 13.5% of females and 26.5% of males, due to iron deficiency in 8.3% of females and 2.7% of males.
Middle Africa
Cameroon [61]	1968		More than 10% of women and slightly more than 5% of men were underweight.
Cameroon [62]	1978	900 children aged 3-59 mo	No consistent relation between sex and the prevalence of undernutrition was found.
Zaire [63]	1987	Data from various studies in Lemba village, Shaba region, 1975-78	Because of a distinctive form of matrilineal social organization, emphasizing the social value of women, the Lemba of south-eastern Shaba in Zaire maintain greater nutritional equality between women and men than do neighbouring groups with different social organizations and lower status of women.
Zaire [18]	1980-85	118 adult men and 91 adult women, both Efe (foragers) and Lese (farmers) living in tropical rain forest	Mean height of Lese men was 161.8 cm: among Efe (sometimes known as pygmies), mean height was 137.1 cm for women and 144.8 cm for men. Both groups were thin and showed substantial seasonal weight gain and loss (greater among Lese than Efe): mean BMI of Lese women ranged from 21.9 to 19.8, whereas BMI of Lese men showed a broader range from 22.3 to 19.4
Zaire [64]	1975	Several hundred reproductive-age women and their children followed at 4 Maternal and Child Health clinics in Kiva Province	Women suffered from general endemic malnutrition, causing delayed puberty (50% reached menarche at 15-16 yr), low levels of milk production (600 ml/d), and prolonged post-partum amenorrhoea (2 yr after delivery, only 1%-15% of women were menstruating). Infants were breastfed for a long time (60% were still breastfeeding at 24 mo), but supplements to breastmilk were introduced early (average age, 3 mo) because of maternal malnutrition.
Zaire [19]	1974-75	Comparison of 433 women from the equatorial forest and 316 women from the highlands (Kivu region)	Highland women (average height, 153.0 cm) were much shorter than equatorial forest women (159.0 cm). Highland women and Ntomba women of the equatorial forest had similar BMIs, but Tembo women of the equatorial forest had lower BMI (19.7). Seasonal deficiencies of protein and lipids lead to extended post-partum amenorrhoea among rural highland women: BMI of amenorrhoeic lactating women was lower than that of menstruating lactating women.
Southern Africa
Lesotho [31]	1971		On the basis of comparisons of their height with Institute of Nutrition of Central America and Panama (INCAP) standards, girls and boys were similar, with the majority being at least 1 yr growth-retarded (83% of girls and 85.7% of boys). 71.9% of girls and 74.8% of boys were classified as having grade I malnutrition or worse according to weight. Skinfold measurements revealed 66.4% of girls and 63.9% of boys were below Canadian Standards. No important differences between boys and girls of any age group were found for prevalence of goitre, an unusual finding.
South Africa [31]	1971		Of 1,116 children with kwashiorkor, 43 girls and 65 boys had xerophthalmia. Of 45,350 pregnant women, 10.1% had <10 g haemoglobin/100 ml blood, 86% had low iron levels, and 21% had low folic acid levels.
Swaziland [65]	1983-84	4,698 rural children and 772 peri-urban children aged 0-59 mo	Overall rates of chronic undernutrition were 28.7% for girls and 32.0% for boys in the rural sample, and 20.0% for girls and 25.7% for boys in the peri-urban sample. Rates of underweight children did not consistently differ by age in either sample. Overall, Swazi children appear slightly better off nutritionally than children in other African countries.
Western Africa
Burkina Faso [66]	1980	15 women aged 18-47 yr	Triceps skinfold values were more than 60% below standard value; other anthropometric measurements were normal.
Côte d’Ivoire [39]	1986	More than 500 rural children aged 6 yr or younger from national probability sample	10.5% of children were below 90% height-for-age, and 4.4% were below 80% weight-for-age. No significant gender differences in anthropometric measurements were found. Percentage of children with low weight-for-age was highest in the savanna (15%) and lowest in the rural south-eastern region (7%).
Côte d’Ivoire [67]	1987-88	1,678 males and 2,032 females	No significant difference was found between BMI of females (22.5) and males (22.2). Data suggest that women’s nutritional status is more likely than men’s to be substantially influenced by increases or decreases in household income and per capita expenditure.
Gambia [68]	1989	60 healthy adolescents aged 12-17 yr, and 60 elderly people aged 57-69 yr	Before supplementation, elderly women were more deficient in riboflavin than elderly men or adolescents; there were no differences between male and female adolescents in the prevalence of riboflavin deficiency. Riboflavin supplements were given for 5 wk during the rainy season when indicators of riboflavin deficiency increase (and mean weights decrease). Riboflavin status improved progressively and reached normal values for both age groups and sexes.
Gambia [12]	1987	197 supplemented and 182 control rural women	Before intervention, women were in positive energy balance during the dry harvest season, with pregnancy weight gain >1,200 g/mo. In the wet season, women were in negative energy balance, with pregnancy weight gain <500 g/mo. Birthweight was correlated with women’s energy balance and averaged 2,944 and 2,808 g in the dry and wet seasons, respectively. Supplementation was ineffective during the dry season but highly effective during the wet season. Girl babies were on average 189 g lighter than boy babies. Girls had a shorter mean gestation time than boys (37.8 vs 38.3 wk).
Gambia [69]	1982-85	32 pregnant women aged 20-35 yr (primiparous women excluded)	Levels of total energy expenditure ranged from a minimum of 2,300 kcal/d (1.7 x BMR) in Jan-Mar to a maximum of 2,700 kcal/d (2 x BMR) during the agricultural season (Jul-Oct). Reduction in physical activity by pregnant and early-lactating women reduced total energy expenditure by 140 ± 18 kcal/d (0.59 ± 0.08 MJ/d) between 28 wk of gestation and 4 wk post-partum.
Gambia [70]	1976-80	All pregnant women in Keneba	Pre-supplemented dietary energy intake of pregnant women varied from an average of only 1,480 kcal/d in the dry season to a minimum of 1,300 kcal/d in the wet season. In the dry season, women were in positive energy balance (satisfactory weight gain and deposition of subcutaneous fat). In the wet season, women were in marked negative energy balance because of a reduction in energy intake and high energy expenditure. Maternal protein intakes were not grossly deficient.
Gambia [70]	1978-79	196 women of child-bearing age	Levels of dietary energy intake were low; during the optimum dry season months, they were only 62% and 64% of the international recommended intake for pregnant and lactating women, respectively. Average energy intakes during the dry and wet seasons, respectively, were 1,483 and 1,417 kcal/d during pregnancy, 1,773 and 1,474 kcal/d during the 1st trimester of lactation, and 1,662 and 1,413 kcal/d during subsequent trimesters of lactation. Pregnant women gained 1.4 kg/mo body weight in the dry season and lost weight in the wet season. Lactating women gained weight in the dry season and lost weight in the wet season. Non-pregnant women lost 1.0 kg/mo in the wet season. Maternal nutritional status did not deteriorate with increasing parity. Wet season energy intakes were clearly inadequate. No maternal depletion in iron, haemoglobin, riboflavin, or vitamins A and C was noted.
Ghana [22]	1990		Females do not appear to be nutritionally at a disadvantage in Africa generally or in Ghana specifically. Levels of acute malnutrition at 6-24 mo of age were appreciably higher among boys than among girls, for most severe cases as well as more moderate cases. Adult men were leaner than women overall, although more women were observed in the lowest categories. There was a surprisingly large number of females with BMIs in the higher brackets, indicating overweight or obesity.
Ghana [71]	1976-77	196 children aged 4-60 mo	There was a slight relationship between sex and nutritional status; boys tended to be better nourished, even though no preferred treatment for boys was articulated or observed.
Ghana [72]	1973	3,700 male and female subjects of all ages	There was a drop in mean weight and height compared with standards in girls at 18-35 mo and in boys at 12-35 mo. Mean weight-for-age and mean height-for-age as a percentage of standard were higher in females than in males; the differences between females and males were greater for weight-for-age; however, all differences between females and males were very slight. Mean weight-for-height as a percentage of standard was also higher in females than in males, except for age 11-16 yr, where values for males were higher.
Guinea (personal communication, NB Mock and MK Konde, 1991)	1991	913 households	Of 780 women, 11.3% were at a health risk and 12.2% were abnormal according to BMI. Rural mothers were worse off than urban mothers. Maternal demographic factors and household activities may have a significant impact on maternal nutritional status.
Mali [23]	1990	441 adults, 320 women and 121 men, aged 19 yr and older	When compared with National Center for Health Statistics (NCHS) references, women were less nutritionally disadvantaged than men. Mean height for women was 160.4 cm (Z score, -0.5) compared with 171.3 cm (Z score, -0.84) for men. Mean BMI was 20.8 for women and 20.0 for men. No sexual dimorphism was found in arm circumference, due to heavy physical work done by women starting at an earlier age than men.
Niger [73]	1980-81	54 nomadic pastoralist families: 30 non-pregnant women and 32 men over 18 yr of age	Marked seasonal variation in weight was seen in both men and women. In February, at the end of the rainy season, women were on average 87% of weight-for-height; by May they had lost on average 4.6% of body weight, declining to 84% of weight-for-height. Men were only 82% of weight-for-height in February and dropped to 78% by May.
Nigeria [11]	1985	2,281 primigravidae with singleton infants (hospital births); 1,140 aged 10-16 yr	The youngest and shortest women had the highest prevalence of contracted pelvis and were at greatest risk of developing cephalopelvic disproportion and of requiring delivery by Caesarean section or embryotomy. With increasing maternal age and height, rates of these three pregnancy complications fell. The tallest women, irrespective of status, rarely had contracted pelvis or cephalopelvic disproportion.
Nigeria [11]	1985	69 Hausa primigravidae	Early teenagers were enrolled in combinations of antimalarial, folic acid, and iron supplementation treatment during pregnancy; height was measured once during pregnancy and once during puerperium (1-60 after delivery). Height growth ³ 2 cm occurred among 29% of those in placebo and antimalarial groups and in 71% of those receiving folate and/or iron supplementation. Study concluded that protecting young teenage pregnant girls from malaria and adding nutritional supplements to their diets enhanced maternal and foetal growth and led to a statistically significant reduction in the proportion requiring abdominal deliveries (see also ref. 74).
Senegal [62]	1968		More than 65% of children were found to show some sign of nutritional deficiency, regardless of ethnic group or sex. Anaemia was common, except among adult men.
Sierra Leone [75]	1978	480 children aged 0-59 mo, 3,724 mothers, 1,965 households	8.2% of mothers were less than 150 cm tall, and 6.1% were undernourished according to measurements of arm circumference. 17.2% of mothers had triceps fatfold thickness below 7.5 mm, indicating low caloric reserve. Almost twice as many pregnant as non-pregnant women were classified as undernourished. More pregnant than non-pregnant women had low fatfold measures (20.8% vs 16.6%). Child nutritional status data were not disaggregated by sex.

Abbreviations: BMI, body mass index; BMR, basal metabolic rate; CED, chronic energy deficiency; LBW, low birthweight; NBW, normal birthweight.

Female nutritional status across the life-span in sub-Saharan Africa. 2. Causes and consequences

Abstract
Major determinants of female nutritional status
Functional consequences: Taking a life-cycle perspective
Conclusion
Acknowledgements
References

Joanne Leslie, Suzanne Bibi Essama, and Elizabeth Ciemins

Joanne Leslie is with the University of California, Los Angeles, School of Public Health, Department of Community Health Sciences, and The Pacific Institute for Women’s Health, in Los Angeles, California, USA. Suzanne Bibi Essama is with the Tulane University School of Public Health in New Orleans, Louisiana, USA. Elizabeth Ciemins is with the Los Angeles County Department of Health Services, STD Program, in Los Angeles.

Abstract

This article reviews existing data concerning the causes and consequences of female malnutrition in sub-Saharan Africa. As in most parts of the world, the primary cause of female malnutrition is household food insecurity compounded by low household and individual incomes. Gender-specific factors that further undermine women’s nutritional status are the severe physiological burden of frequent child-bearing and the continuous long hours of energy-intensive work. Negative consequences of malnutrition among females include high rates of mortality and morbidity, impaired learning, low birthweights, and reduced energy for discretionary activities. We question the conclusion of other studies that African women have developed special “adaptive mechanisms” to compensate for nutritional deprivation, and recommend that further research investigate the hidden individual and societal costs of malnutrition among women.

Major determinants of female nutritional status

Crucial conflicts face poor women in low-income countries as they try to fulfil their economic, biological, and social roles at each stage in the life cycle, particularly during the child-bearing years. Changes in behaviour that enhance their contribution to one area can have crucial negative effects on their other roles and activities. This role conflict relates to the tremendous time, energy, and money-resource constraints facing these women....

Conflicts between the economic, reproductive, and cultural roles of women can have detrimental effects on their nutrition and/or that of their families [1].

Models of determinants of both children’s and women’s nutritional status standardly include quality and quantity of dietary intake, infection, and energy expenditure as proximate determinants, which are themselves seen to be determined by a range of household, community, national, and global variables related to wealth, food production, education, and availability of health services, among others [see refs. 3-5]. In sub-Saharan Africa in particular, women devote enormous amounts of time and energy to both their productive and their reproductive roles. The time- and energy-consuming nature of their tasks has significant implications for their own nutritional status and that of their children.

Women’s role in food production and acquisition

Women’s agricultural labour in sub-Saharan Africa is extremely important both as a percentage of total agricultural labour and as a percentage of women’s total labour force participation. Almost 80% of economically women in sub-Saharan Africa are working in agriculture [6]. Throughout most of the region, women not only put in longer work days overall than men [7, 8], but also spend more hours per week in agricultural work [1, 6]. Women’s responsibility is at least equal to that of men in determining the quantity of food available at the household level, and it is significantly more important in determining the variety and palatability of the household diet [9]. In sub-Saharan Africa, women’s dominant role in both subsistence food production and food preparation may give them more control over their own and their children’s dietary consumption than in some other regions of the world; however, this comes at the cost of extremely long and energy-demanding work days [9-11]. Findings from a study in Malawi of the gender division in agriculture-related decision-making are not atypical of the region as a whole [12]. This study found that husbands made most of the decisions regarding major farm inputs and had almost complete control over decisions concerning cash crop production. Women had significant input into decisions regarding the production of food crops, and full responsibility for decisions relative to the cultivation of selected vegetables, such as pumpkins and beans, that are used in food preparation. With respect to the use of income, men appeared to have control over formal, more regular sources of income (i.e., income from the sale of cash crops or employment), whereas women tended to manage the income from the sale of beer, fruits, or cooked food items.

An increase in demand for male labour during colonialism may help explain women’s current dominance in agricultural labour. The current distribution of gender roles is, however, lopsided and asymmetrical. While women have taken over many of the tasks that were traditionally outside their domain, such as agricultural labour, men have not done the same in regard to women’s roles [11]. The result has been an inequity in the burden of labour between men and women and simultaneous declines in women’s social and economic status. Events occurring during colonialism adversely affected women in several ways. A redistribution of land caused women to lose their limited access to and ownership of land and their subsequent access to credit, loans, and technologies. An increase in cash cropping affected women by increasing demand for their labour. Finally, the imposition of Western notions of women’s inferiority added to their declining status and clashed with African women’s defined independent roles in their societies, which undermined women’s ability to maintain good household nutrition [11,13].

Today, traditional food production and security strategies are rapidly changing in sub-Saharan Africa in response to population increases, deteriorating environmental conditions, and changing market circumstances, resulting in significant urban migration and increased linkages between rural communities and major urban centres. This adaptation process has led to significant modifications in food production patterns, in the distribution and acquisition of food, and in household food consumption. In particular, gender asymmetries in access to productive resources have meant that women are less able to take advantage of agricultural intensification strategies in their role as food producers [11, 14].

Concerns have also been raised about the effects of agricultural intensification strategies on the health and nutritional well-being of women and children [11,15, 16]. Some studies have suggested that an expansion and intensification of commercial agriculture in sub-Saharan Africa has contributed to gradual declines in food production levels and reductions in the amount of food available for household consumption. It has also been argued that women and children may be adversely affected by a shift to cash cropping as a result of increased demands on women’s labour for agricultural activities and reductions in women’s individually earned income, and that children, in particular, may be adversely affected by earlier weaning and a reduced frequency of meals during the peak of the agricultural season. Conversely, it has been argued that a shift to cash cropping has produced higher household incomes that lead to better household diets and, therefore, to improved nutritional status for all members of the household, including women and children.

The most rigorous examination of the effects of cash crop production on child health and nutrition comes from a comparative analysis of six methodologically similar studies carried out by researchers at the International Food Policy Research Institute (IFPRI). Four of these studies were conducted in sub-Saharan Africa (Gambia, Kenya, Malawi, and Rwanda [4]), and the other two in Guatemala and the Philippines. In the African countries, a comparison of child outcomes in households that participated in a cash cropping scheme and those that did not found no evidence of a negative effect of participation in cash cropping on the nutritional status of children, but only weak evidence of a positive effect. The comparison with the findings from the Guatemala study, focusing on a vegetable cooperative, is illuminating. Guatemala was the only one of the six countries in which household participation in cash cropping was significantly associated with better child health and nutrition outcomes. This was attributed to the fact that the vegetable cooperative directly invested some of its profits in community health and social services.

Other studies from Africa have reported a negative association between increased cash cropping and both the quantity and quality of foodstuffs available for household consumption. Spring [17] presents numerous examples of the negative effects of increased production of non-food crops and sales of foodstuffs outside the household. In Zambia, Malawi, and Kenya, cash cropping resulted in decreased amounts of food available for household consumption, but whether this resulted in poorer nutritional status was not determined. One classic example is the introduction of cocoa production in Ghana, which resulted in a shift of responsibility for the labour-intensive production of yams from men to women. Unable to cope with their increased workloads, women eventually switched to producing cassava, which, while requiring less labour, is also less nutritious and less agriculturally advantageous than yams [18].

Another study from the Gambia, however, reported a situation in which household nutritional status was positively affected by investments in technology, despite the fact that women lost control over a crop [19]. Rice, a crop traditionally controlled by women, was the focus of introduction of new modes of production. As technology increased, the percentage of rice fields for which women were responsible declined, which would normally result in reduced household energy consumption. In this case, however, overall household income increased, resulting in a significant improvement in the nutritional status of children and women, especially in the most nutritionally vulnerable households. The new rice technology also helped to level out seasonal fluctuations of women’s weight, which was particularly beneficial for nutritional status during the rainy season, when food stores are low and energy expenditure is high.

A separate analysis of data from a study in Kenya by IFPRI examined the effects of the commercialization of agriculture on the allocation of time and the patterns of food consumption by women, as well as on their nutritional status [20]. The central findings were that women from households that produced sugar cane spent no more time away from home than women from households that did not produce sugar cane; that there were no significant differences in the amount of time spent on the various household activities (as the amount of time women spent on sugar cane production was negligible); and that the mean weights of women were similar in households that did or did not produce sugar cane.

The seasonal patterns of the agricultural cycle in sub-Saharan Africa impose different demands on women’s energy expenditure throughout the year and have a significant influence on household food availability, women’s energy intake, and women’s nutritional status. The periods of greatest nutritional stress for rural women usually occur during the pre-harvest period (generally known as the “famine” or “lean” months), when household food stocks are low, the energy demands of agricultural work are highest, and energy intake is low [9, 10, 12, 21, 22]. In urban areas, periods of nutritional stress occur when the market prices of basic food commodities are highest.

The effect of seasonality on women’s nutritional status in sub-Saharan Africa is particularly well illustrated by a study in southern Benin that compared the effects of seasonal changes in food availability on women’s nutritional status in rural and peri-urban areas [23]. Findings were reported for a sample of 567 non-pregnant (but lactating) women, of whom 366 lived in rural areas and 201 lived in a pert-urban setting. The findings concerning seasonality were reported in terms of nutritional status changes between pre- and post-harvest seasons. Twenty-five percent of the rural women gained more than 2 kg between the pre- and post-harvest seasons, and the average body mass index (BMI) also increased significantly during this period. In contrast, 25% of the pert-urban women lost more than 2 kg during the same time period and their BMI decreased, although the decrease was not statistically significant. The Benin study illustrates that the food security and nutritional consequences of seasonal changes in food availability can differ significantly between rural women (who depend mostly on food from their own production) and pert-urban women (who obtain a large proportion of their food through purchase).

Pregnancy and lactation

With an average regional fertility rate of 6.5% in 1990, women in sub-Saharan Africa had significantly higher fertility than women in any other region of the world [24]. Studies of food consumption during pregnancy and lactation in sub-Saharan African countries indicate that macronutrient intakes are low, in the range of 1,400 to 2,000 kcal and 25 to 50 g protein per day, while vitamin and mineral intakes are often extremely low [1, 25]. Although this low dietary intake clearly carries risks for both mothers and infants, several detailed studies suggest that when the energy cost of activity, reproduction, and lactation can be partially met by mobilization of maternal tissue stores, the impact of low energy intake on foetal growth and lactation performance is less than might be anticipated [26]. Other studies, however, emphasize that since lactation has even higher energy requirements than pregnancy, it is essential to increase energy intake, reduce energy expenditure, or both during breastfeeding to protect women’s long-term nutritional status [27].

Kosin et al. [28] analysed cross-sectional data on food consumption in Machakos area, Kenya, by pregnant, non-pregnant, and lactating women from October 1977 to December 1979, and found that the diets of pregnant women, and to a lesser extent those of lactating women, were inadequate both when compared with the intakes recommended by the World Health Organization (WHO), and when compared with the diets of non-pregnant, non-lactating women in the same population. Compared with the recommended daily intakes, pregnant women received adequate amounts of protein, thiamine, and ascorbic acid. However, their energy intakes were low, with median values ranging from 70% of recommended intake during the first trimester to 62% during the third trimester. An even larger deficit was recorded in the median calcium, iron, retinol equivalents, and riboflavin intakes of pregnant women. Kosin could not explain clearly why food intake was reduced in the last trimester of pregnancy and suggested that cultural factors could be major determinants, as food availability was not a constraint in that region. The dietary intake of lactating women was found to be inadequate in energy and in all nutrients except protein and ascorbic acid. However, the deficits noted were lower than the deficits found among pregnant women. The study found that mean weights remained the same during the first and second trimesters of pregnancy and were only 2 kg higher in the third trimester. The mean weight-for-height of lactating women at 15 to 24 months was slightly lower than the mean for women during the first year of lactation. The authors concluded that the nutritional status of Kenyan women deteriorated as pregnancy and lactation progressed.

Studies of European and American women report significant increases in skinfold thickness at triceps and subscapular sites of the body between 10 and 20 weeks of pregnancy. Comparable analyses of changes in skinfold thickness at various stages of pregnancy in Africa are rare. A study of pregnant Nigerian women is one of the few published studies on the subject [29]. It provides an excellent analysis of the magnitude and patterns of subcutaneous fat deposition at triceps and subscapular sites in a group of “normal” pregnant Yoruba women living under low socioeconomic conditions in rural areas. All the women in the study had had multiple pregnancies. Their mean age was 27 years, and on average each woman had completed four pregnancies at the time of the study. Other anthropometric measurements collected on the women included arm circumference, weight, height, age, and parity number. Like the Kenyan study described above and other research from sub-Saharan Africa, this study found low weight gains during pregnancy. The total mean weight gain between 20 and 30 weeks of pregnancy was 3.8 kg, which was about half of that reported for elite Nigerian women in Ibadan. The study also reported gradual declines in arm circumference and in triceps and subscapular skinfold thickness throughout pregnancy; the average total decline for the whole group was 4.1 mm. Thus, instead of showing a normal increase between 20 and 30 weeks of pregnancy, both individual and combined skinfold thickness at the triceps and subscapular sites of low-income pregnant Nigerian women declined. The authors interpret these declines as indicating a continuous depletion of the energy stores during the course of pregnancy to compensate for inadequate dietary intake. In addition, the study found a negative correlation between parity and subscapular skinfold thickness, which was interpreted as indicating that the ability of pregnant women to store body fat at the subscapular site decreased as parity increased.

There is mixed evidence concerning the extent to which women are actually able to reduce energy expenditure to compensate for the increased energy demands of pregnancy or lactation. Most research has found little evidence of a change in activity patterns or energy expenditure during pregnancy or lactation by women in sub-Saharan Africa or elsewhere in the developing world [12, 30]. However, a recent detailed study of the functional consequences of malnutrition among the Embu in Kenya found that reduction in energy expenditure during the third trimester was a major mechanism by which pregnant women were able to achieve reasonable infant birthweights in the face of inadequate dietary intake [31]. Evidence of the contribution of energy-sparing mechanisms to partially meet the additional energy demands of pregnancy also emerges from a series of studies of women in three rural Gambian villages. The energy expenditure of pregnant women on activities with relatively low energy demands did not appear to change over the course of pregnancy, whereas activities with higher energy demands were reduced during the second and third trimesters [3234]. Peacock and colleagues [35] reported a similar finding of a reduction in energy-intensive activities in one, but not both, of the Zairian tribal groups they studied. Among Efe women (semi-nomadic foragers), the proportion of time spent in the most energy-intensive activities was reduced during pregnancy and even more during lactation. However, no compensatory reduction in energy expenditure was found among pregnant or lactating Lese women, who work as swidden cultivators.

The general problem of meeting the nutritional demands of pregnancy and lactation on top of the already substantial energy demands of a long and physically demanding workday is particularly acute during seasonal periods of food shortage. This is well illustrated by the Gambian research referred to above. Fifty women subsistence farmers were followed through pregnancy, during which time some were provided with food supplements, and seasonal changes in basal metabolic rate, body fat, activity patterns, and total energy expenditure were assessed [36]. Seasonal variations in body-fat content were found to occur in all women, whether pregnant or not, and fat gain during pregnancy in individual women was found to be dependent upon the times of the year through which the pregnancy progressed. Seasonal fluctuations in body-fat content of rural Gambian women were as large as, or larger than, the changes resulting from pregnancy. Weight loss during the rainy season among unsupplemented non-pregnant, non-lactating women averaged 5 kg, most of which was adipose tissue. Among pregnant women, unsupplemented women who gave birth at the end of the rains (when agricultural activity was intense and food supplies were very low) lost 4.7 kg of body fat, whereas those who were pregnant during the dry season (when little agricultural work was done and food supplies increased) gained as much as 3 kg of body fat. The interaction between seasonality and supplementation was highly significant. In supplemented women, neither weight nor fat gain during pregnancy varied as much with season as in the unsupplemented group. Overall, supplementation increased fat gain during pregnancy by about 2 kg and gave some protection against seasonal weight loss. The authors interpreted their combined findings concerning changes in basal metabolic rate, fat deposition, and energy expenditure as demonstrating that maternal nutritional status in rural Gambian women is significantly compromised by pregnancy during the rainy season.

Given the marginal food availability in most of rural sub-Saharan Africa, and the need for women to continue with heavy physical work throughout most, if not all, of the time they are pregnant and lactating, the importance of ensuring an adequate interval after the end of lactation and before the next pregnancy to replenish maternal reserves of fat and other nutrients cannot be overemphasized. It is estimated, for example, that even when food intake is adequate, it may take two years to replenish body iron stores after a pregnancy [37]. In addition, efforts to reduce the energy demands on women during the preconception period, as well as during pregnancy and lactation, through easier access to needed resources, labour-saving devices, or both would be extremely beneficial in protecting the health of both women and children.

Functional consequences: Taking a life-cycle perspective

Malnutrition is multifactorial in its aetiology and cumulative in its manifestations. Merchant and Kurz [ref. 3, p. 73] note succinctly that “A nutritional problem is generally the consequence of earlier problems and the cause of later problems.” Some of the most important functional consequences of female malnutrition (for example, the obstetrical risks associated with short stature and iron-deficiency anaemia) have been studied directly in sub-Saharan African populations. Many other functional consequences of nutritional status (both positive and negative) have not been studied to any great extent among sub-Saharan African females directly, but reasonable extrapolations can be made based on studies of men or studies from other parts of the world.

Mortality and morbidity

The ultimate consequence of severe malnutrition is death. Malnutrition is a particularly significant contributing cause of infant and child mortality and of maternal mortality. On the basis of estimates made by UNICEF and others, it seems likely that at least a third of infant and child deaths in sub-Saharan Africa are partially attributable to protein-energy malnutrition [38]. And, in times of famine, both the rates of infant and child mortality and the proportion of total deaths attributable to malnutrition increase dramatically [39].

Low birthweight, which can be due to either prematurity or intrauterine growth retardation, is the most significant nutritional risk factor for subsequent infant and child mortality. It has been estimated that maternal nutritional factors account for approximately half the influence of established determinants of intrauterine growth retardation in developing countries [40]. Current or past maternal malnutrition, as evidenced by short stature, low weight-for-height, poor-quality dietary intake, or excessive energy expenditure, is a significant risk factor for bearing infants of low birthweight, showing a direct intergenerational transmission of malnutrition.

Stunting and wasting among pre-school-age children, whether attributable to low birthweight or to poor diet and disease, significantly increase the risk of death. A study in Iringa, Tanzania, for example, found a sharp increase in mortality risk at weight-for-age below 60% of the median, weight-for-height below 70% of the median, and height-forage below 85% of the median [41]. Extensive work by Pelletier and his colleagues at Cornell University has clearly established that both severe and mild-to-moderate malnutrition contribute more significantly to child mortality than previously recognized. For example, from eight community-based prospective studies of the relationship between anthropometry and child mortality (including studies from Malawi and Tanzania), they conclude that the relative risk is 8.4 for severe malnutrition, 4.6 for moderate malnutrition, and 2.5 for mild malnutrition [42]. The finding of an increased mortality risk even among mildly to moderately malnourished children has important programmatic and policy implications, given the much larger number of such children compared with those who are severely malnourished.

After infants and pre-school-age children, those most at risk of mortality associated with malnutrition are women during pregnancy and childbirth. Given the extremely high rates of maternal mortality in sub-Saharan Africa [see ref. 2, table 1], assessing and reducing as many of the major causes of maternal deaths as possible will be a particularly important component of improving women’s health in this region.

Obstructed labour and its sequelae are the most important causes of maternal death in sub-Saharan Africa [43]. The risk of obstructed labour, in which the birth canal is too small or too deformed to allow passage of the baby, is directly related to maternal age, developmental stage, and stature. The growth of the birth canal is not complete until about three years after height growth ceases, and protein-energy malnutrition both slows down the rate at which girls mature and, in many cases, permanently stunts their growth. Thus, protein-energy malnutrition directly increases the risk of obstructed labour, particularly among adolescent mothers. A study in Nigeria found that among a group of primigravidae who received prenatal care, the proportion who required operative delivery because of a small pelvis ranged from 40% among women under 145 cm tall, to 14% among those at least 150 cm tall, to less than 1% among those at least 160 cm tall [44]. A population-based case-control study in Harare, Zimbabwe, reported similar findings. When other factors were controlled, women of short stature (less than 160 cm) were twice as likely as taller women to have an operative delivery (Caesarean section, vacuum extraction, or forceps) because of cephalopelvic disproportion [45]. Given the relatively low rate of stunting among women in sub-Saharan Africa compared with other regions of the developing world, the importance of obstructed labour as a cause of maternal mortality may seem somewhat surprising. However, the proportion of births in sub-Saharan Africa occurring among young mothers who are not yet fully physically mature is part of the explanation. Even more significant is the widespread lack of access to timely medical intervention when obstructed labour does occur.

The other nutritional deficiency that significantly increases the risk of maternal mortality is anaemia. When anaemia is acute, it can cause death directly from heart failure or shock. Fortunately, even among malnourished women, anaemia this severe is quite rare. However, although less severe anaemia may not be a direct cause of maternal death, it is a significant contributory cause. In particular, anaemic women are less able to tolerate haemorrhage (both antepartum and post-partum), which is one of the four leading causes of maternal death in sub-Saharan Africa [43]. Anaemia is estimated to account for a fifth to a tenth of all maternal deaths in many countries of the region, and in the extreme circumstances of two refugee camps in Somalia, more than 90% of maternal deaths were associated with anaemia [37]. Again, however, it is probably more accurate to say that it is the combination of pre-existing anaemia, haemorrhage, and lack of access to medical care that causes women to die.

Scientific studies relating malnutrition to both infectious and non-communicable diseases have proliferated over the past decade. A large number of micronutrient deficiencies have been found to impair the function of the immune system, particularly through their effect on cellular immunity [46, 47]. The negative effect of zinc deficiency on the immune system seems to be particularly notable. As far as non-communicable diseases are concerned, the effect of malnutrition is cumulative and primarily manifests itself in disease outcomes during the post-reproductive years. Obesity, and probably excess dietary fat, are risk factors for both diabetes and coronary heart disease, whereas low consumption of the antioxidant vitamins A, E, and C increases the risk of developing most, if not all, cancers [48, 49].

Although the negative effect of specific nutrient deficiencies on the effectiveness of the immune system has been demonstrated, the functional significance of this in terms of increased morbidity is less well established. The largest body of scientific evidence concerns the relationship between protein-energy malnutrition and diarrhoea. Although virtually all studies show a strong association between these two widespread health problems of childhood, the direction of causality has been more difficult to establish. Careful longitudinal studies suggest that pre-existing protein-energy malnutrition has a limited effect on the incidence of diarrhoea but significantly increases the duration [50, 51]. A study of more than 300 children between 6 and 32 months of age in northern Nigeria at the end of the rainy season found that diarrhoea lasted 37% longer in stunted children and 79% longer in wasted children; in this particular study, wasted children also had diarrhoea more frequently [52].

A longitudinal study of the functional consequences of malnutrition in the Embu district of Kenya produced several important findings concerning the relationship between nutritional deficiency and subsequent morbidity, not only among preschool-age children, but also among reproductive age women [31]. In fact, one of the most striking results of the research was the finding that morbidity rates for the study sample as a whole doubled during the drought-related food shortage period in 1984 compared with 1985, when dietary intake had returned to more normal levels.

The same research found that stunting among toddlers (18 to 30 months) and, to a lesser extent, low weight-for-age significantly increased the risk of acute lower respiratory tract infections [31]. There were also significant sex differences. The percentage of time that female infants and toddlers spent ill was somewhat higher than for boys (47% versus 42%), and girls were found to have a duration of severe illness that was, on average, twice as long as that for boys. Girls also had an energy deficit during severe illness that was more than double that of boys, although this was somewhat balanced by a larger food intake during convalescence. Lagged analysis showed improved quality and quantity of food intake to be protective against severe illness among girls and boys, with both incidence and duration being affected. In terms of life-cycle effects of malnutrition, a particularly notable finding of the Embu study was that lower rates of maternal illness and higher maternal fat intakes were both significant predictors of less morbidity among their toddlers. The researchers interpreted both factors as indicative of higher levels of energy among mothers, who would then be better able to prevent or treat their children’s illness.

Morbidity rates among reproductive-age Embu women were significantly higher among pregnant women than among non-pregnant women [31]. In addition to the well-established negative effects of pregnancy on the immune system, the researchers attributed the greater morbidity among pregnant women to their lower food intake. As with toddlers, among both pregnant and non-pregnant women, higher levels of food intake (particularly total energy, fat, and animal protein) were found to be protective against severe illness. Overall, women in this study were found to have higher illness rates than men, a difference that persisted even when pregnant women were excluded from the comparison. However, the authors caution that since women were the main informants, male illness may have been underreported.

Cognitive development and school performance

One of the clearest intergenerational effects of female malnutrition is the significant level of cretinism, deafness, and other congenital abnormalities among infants born to mothers who are severely iodine deficient. Endemic cretinism is estimated to affect up to 10% of the population living in severely iodine-deficient areas [53]. There are pockets of severe iodine deficiency in a majority of sub-Saharan African countries; in the region as a whole, there are estimated to be at least 500,000 overt cretins due to maternal iodine deficiency during pregnancy [48]. In addition to congenital cretinism, children who suffer from iodine deficiency during their pre-school or school years also show delayed mental development, although unlike cretinism, these cognitive impairments can be reduced with appropriate nutritional intervention.

Protein-energy malnutrition in children is also strongly associated with impaired motor and mental development [38, 54]. The effects appear to be both direct and indirect. A child who is malnourished is often apathetic or irritable and thus tends to receive less attention and positive stimulation than a better-nourished child in a similar environment. Protein-energy malnutrition is negatively associated both with the likelihood that children will go to school and with how well they are able to learn in school [55].

The study of functional consequences of malnutrition among the Embu in Kenya discussed above found negative effects of malnutrition on cognitive development among both toddlers and school-age children [31]. In both age groups, stunted children were found to do less well on cognitive tests than children with normal height-for-age when other factors were controlled. Better dietary quality, particularly an increased intake of animal protein, fat, and several micronutrients (including but not limited to iodine), was found to have a significant positive effect on cognitive development. In addition, current activity level was strongly related to concurrent energy intake, and activity level and exploratory behaviour were found to be positively linked to learning among school-age children.

There are very few studies, in sub-Saharan Africa or elsewhere, of the effect of malnutrition on attendance, repetition, or drop-out rates among school-age children, although it is virtually inevitable that a high prevalence of malnutrition or other health problems among school-age children will make them “inefficient” users of the educational resources available to them [55]. A detailed study of health and nutrition problems among school-age Yorubu children in Nigeria documents a high proportion of children going to school without breakfast and a high prevalence of growth retardation and micronutrient deficiencies. The study concludes that these lead to a high drop-out rate, poor intellectual performance, and low educational attainment, representing a serious economic loss to the government of Nigeria, which spends one-fourth of its annual recurrent budget on primary-school education [56].

The life-cycle consequences of malnutrition are also well illustrated by the linkages between malnutrition and schooling. Malnutrition during the preschool and school-age years has negative effects on girls’ (as well as boys’) school participation and performance. Low levels of maternal education are in turn significantly associated with poor child nutritional status, as well as with higher levels of child mortality [57].

Reproductive function

The functional consequences of childhood malnutrition discussed above affect both males and females, although the long-term consequences may be different, and in some cases more severe, among females. However, the detrimental effect of malnutrition on reproductive function is specific to females (with the possible exception of effects of severe malnutrition on male fertility) and has grave life-cycle and intergenerational consequences.

Significant declines in fertility during famine, as well as a predictable return to previous levels of fertility once the famine is over, have been well documented [58]. Frisch [59] has developed a comprehensive model relating female malnutrition to a shorter and less efficient reproductive span through delayed menarche, reduced fecundity, lengthened post-partum amenorrhoea, and perhaps earlier menopause. Nonetheless, there remains considerable debate about the magnitude of any effects of female malnutrition on fertility in the chronically mildly to moderately malnourished populations of sub-Saharan Africa. An analysis of data from unrelated nutrition and fertility surveys in Senegal, for example, found evidence of, at most, a minor negative effect of malnutrition on fertility [60]. Two studies from Zaire, however, have reached conclusions that suggest a much more significant effect of nutritional status on fertility. Caraël’s [61] analysis of data on lactation status and duration of post-partum amenorrhoea among women from two different ecological zones of Zaire found a strong relationship between birth intervals and nutritional patterns. Although the study did not have individual-level dietary intake data, the researchers concluded that when duration of lactation was controlled, severe seasonal inadequacies of protein and lipids prolonged postpartum amenorrhoea seven to nine months among rural highland women. Even more compelling evidence comes from a more recent study in Zaire. Women living in different ecological zones were compared, but in this case the researchers were able to relate longitudinal anthropometric data and salivary measures of steroids (indicating ovarian function) with seasonal variations in conception. They concluded that

Variability in the seasonal pattern of rainfall in the Ituri Forest causes variability in Lese garden size, which translates into significant changes in nutritional status. Declines in female nutritional status result in reduced ovarian function, which produces seasonal reductions in rates of conception and implantation [ref. 22, pp. 404-5].

Some researchers have hypothesized, in addition, an unusual capacity on the part of pregnant and lactating African women to adjust to or compensate for low food intake, but the findings are not consistent. A series of studies in the Gambia, for example, led to the conclusion that women are able to produce adequate-birthweight infants and adequate amounts of breastmilk by mobilizing rather than building up fat stores during pregnancy, particularly during the wet season, and by achieving considerably greater metabolic efficiency than women in industrialized countries [32, 63]. In particular, researchers in the Gambia reported that women raised their basal metabolic rates so little during pregnancy that the net extra cost of basal metabolism was only 1,000 kcal, rather than the usual estimate of 36,000 kcal, over the course of a pregnancy [33]. These findings are not entirely supported by the results of a longitudinal study of reproduction among Embu women in Kenya [31]. The Kenyan research also reports surprisingly good infant birthweight outcomes, in spite of low energy intake during pregnancy and a weight gain during pregnancy only half of what is recommended. However, in contrast with the reported findings from the Gambia, the Kenyan study found no compensatory lowering of resting energy expenditure among pregnant women. Instead, the researchers found a compensatory behavioural adaptation: late in pregnancy, women doubled their inactive time (at the expense of household care, child care, economic and agricultural activities, and food preparation) in order to accommodate to their low energy intakes relative to the energy requirements at this stage of pregnancy. In addition, the study found that as in industrialized countries, pre-pregnancy maternal size, energy intake during pregnancy, and pregnancy weight gain were all important determinants of infant birthweight and of net post-partum maternal weight and fat gain.

Physical work capacity

Both the long-term consequences of childhood malnutrition and current nutritional deficiencies may have significant effects on women’s capacity for physical work. Given the strenuous nature of the major tasks of rural women in sub-Saharan Africa - pounding grain, carrying water and fuel, doing non-mechanized agricultural work, and walking to and from markets - a woman’s physical capacity for work may be one of the most important determinants of her own and her family’s nutritional well-being [5].

In attempting to assess the effects of female malnutrition on physical work capacity, considerable extrapolation must be done from studies of males, since surprisingly little research has been done on women. The long-term effects of childhood malnutrition, acting through short stature and reduced muscle mass, have a reasonably clear negative effect on the productivity of men engaged in strenuous activities such as cutting cane or moving earth [64]. In addition, deficiencies of several micronutrients, particularly iron, but also vitamin C and the B-complex vitamins, have been found to have a negative effect on physical work capacity [46]. Intervention studies in which anaemic female tea pickers in Asia were given supplemental iron showed that those who were supplemented were significantly more productive than unsupplemented controls [5].

There is less consistency in studies that have tried to assess the association between energy status and physical work capacity, in part because a frequent initial response to inadequate energy intake is to mobilize fat stores rather than to reduce work [38]. One study from Kenya did find a positive association between nutritional status and work capacity among women in sugar cane-farming households [65]. Women of higher BMI were able to spend more time in work-related activities, including home production, and at a given level of BMI, taller women appeared to engage in more energy-intensive work activities. In addition, the frequently cited report from researchers in the Gambia of women who received a dietary supplement and then sang while they worked in the fields (which they had not done previously) supports the importance of looking beyond physical capacity for work or labour productivity when assessing the functional consequences of adult malnutrition [66].

Conclusion

The main cause of malnutrition of females in sub-Saharan Africa is the same as for males: household food insecurity due to unreliable food availability, compounded by extremely low, and for the most part falling, incomes. The individual nutritional status of both males and females in sub-Saharan Africa is further undermined by the continuing high burden of infectious disease in this region, which is particularly significant as a determinant of children’s nutritional status. Additional important causes of poor nutritional status among adult women in sub-Saharan Africa are the high physiological burden of reproduction and the long hours of energy-intensive work of most rural women in this region.

Female malnutrition in sub-Saharan Africa is responsible for a broad range of both short-term and long-term negative consequences. As a result of malnutrition, girls (like boys) suffer high levels of mortality and stunting in early childhood and poor school performance in later childhood. Malnutrition among adult women poses severe risks both to themselves and to their infants. Although stunting is not as widespread among women in most sub-Saharan African countries as in the rest of the developing world, the lack of access to timely medical intervention for cephalopelvic disproportion and prolonged labour puts women with inadequate pelvic development (whether due to size, age, or both) at extremely high risk. In addition, the high proportion of low-birthweight infants in many sub-Saharan African countries is substantially attributable to maternal malnutrition, both before and during pregnancy.

Although some researchers have suggested that sub-Saharan African women seem to “accommodate” remarkably well to inadequate food intakes, multiple pregnancies, long duration of breastfeeding, and long hours of energy-intensive domestic and market work, such a positive conclusion seems unwarranted. First, the extremely high infant, child, and maternal mortality rates and the short life expectancy in sub-Saharan Africa suggest that women and children who are severely malnourished in this region simply fail to survive, undoubtedly in part because of the high prevalence of infectious diseases and the lack of access to medical care. Even among those who survive, however, it seems almost certain that their marginal nutritional status severely restricts the energy that girls and women have for any activities beyond those that are essential for survival. It may turn out that this restriction on discretionary activities is most responsible for the perpetuation of malnutrition from one generation to the next and, therefore, that interventions to reduce female malnutrition could be one of the most effective investments that could be made in social and economic development in sub-Saharan Africa.

Acknowledgements

The review on which this and its companion article are based was originally undertaken as part of the work of the Institute of Medicine (IOM) Committee to Study Female Morbidity and Mortality in sub-Saharan Africa, of which the first author was a member. We would like to acknowledge the extremely useful comments and input of the other committee members (Maureen Law [Chair], Uche Amazigo, Judith Fortney, Philip L. Graitcer, Françoise F. Hamers, H. Kristian Heggenhougen, Karungari Kiragu, Walinjom F. T. Muna, Jonathan E. Myers, Benjamin O. Osuntokun, Patience W. Stephens, Judith N. Wasserheit, and Belmont E. O. Williams) and of the IOM staff who worked with the committee (folly F. Harrison, Dana Hotra, Delores Sutton, and most especially, Christopher P. Howson, Project Director).

References

1. McGuire J. Popkin BM. Beating the zero-sum game: women and nutrition in the third world. Part 1. Food Nutr Bull 1989; 11(4): 38-63.

2. Leslie J. Ciemins E, Essama SB. Female nutritional status across the life-span in sub-Saharan Africa. 1. Prevalence patterns. Food Nutr Bull 1997; 18: 20-43.

3. Merchant KM, Kurz KM. Women’s nutrition through the life cycle: social and biological vulnerabilities. In: Koblinsky M, Timyan J. Gay J. eds. The health of women: a global perspective. Boulder, Col, USA: Westview Press, 1993: 63-90.

4. Kennedy E, Bouis H. von Braun J. Health and nutrition effects of cash crop production in developing countries: a comparative analysis. Soc Sci Med 1992; 35: 689-97.

5. Leslie J. Women’s nutrition: the key to improving family health in developing countries? Health Policy Plann 1991; 6: 1-19.

6. United Nations. The world’s women 1970-1990: trends and statistics. Social Statistics and Indicators, Series K, No. 8. New York: United Nations, 1991.

7. Juster FT, Stafford FP. The allocation of time: empirical findings, behavioral models, and problems of measurement. J Econ Lit 1991; 29: 471-522.

8. Leslie J. Women’s time: a factor in the use of child survival technologies? Health Policy Plann 1989; 4: 1-16.

9. Holmboe-Ottesen G. Mascarenhas O. Wandel M. Women’s role in food chain activities and the implications for nutrition. ACC/SCN State-of-the-Art Series, Nutrition Policy Discussion Paper No. 4. Geneva: World Health Organization, 1989.

10. Bleiberg FM, Brun TA, Goihman S. Duration of activities and energy expenditure of female farmers in dry and rainy seasons in Upper-Volta. Br J Nutr 1980; 43: 71-82.

11. Lado C. Female labour participation in agricultural production and the implications for nutrition and health in rural Africa. Soc Sci Med 1992; 34: 789-807.

12. Lamba C, Tucker KL. Work patterns, prenatal care, and nutritional status of pregnant subsistence farmers in central Malawi. Washington, DC: International Center for Research on Women, 1990.

13. Caplan P. Perceptions of gender stratification. Africa 1989; 59(2): 196-208.

14. Dey J. Gender asymmetries in intra-household resource allocation in sub-Saharan Africa: some policy implications for land and labour productivity. Washington, DC: International Food Policy Research Institute, 1992.

15. Ferguson A. Women’s health in a marginal area of Kenya. Soc Sci Med 1986; 23: 17-29.

16. Raikes A. Women’s health in East Africa. Soc Sci Med 1989; 28: 447-59.

17. Spring A. Women farmers and food in Africa: some considerations and suggested solutions. In: Hansen A, McMillan DE, eds. Food in sub-Saharan Africa. Boulder, Col, USA: Lynne Rienner Publishers, 1986: 332-48.

18. Bukh J. The village woman in Ghana. Uppsala, Sweden: Scandinavian Institute of African Studies, 1979.

19. Von Braun J. Effects of technological change in agriculture on food consumption and nutrition: rice in a West African setting. World Dev 1988; 16: 1083-98.

20. Kennedy ET, Cogill B. Income and nutritional effects of the commercialization of agriculture in southwestern Kenya. Washington, DC: International Food Policy Research Institute, 1987.

21. Loutan L, Lamotte JM. Seasonal variations in nutrition among a group of nomadic pastoralists in Niger. Lancet 1984; 1(8383): 945-7.

22. Bailey RC, Jenike MR, Ellison PT, Bentley GR, Harrigan AM, Peacock NR. The ecology of birth seasonality among agriculturalists in Central Africa. J Biosoc Sci 1992; 24: 393-412.

23. Fakambi LK. Factors affecting the nutritional status of mothers. The food and nutrition program of the Ouando horticulture and nutrition center in the People’s Republic of Benin. Washington, DC: International Center for Research on Women, June 1990.

24. World Bank. World development report 1992: development and the environment. Washington, DC: World Bank, 1992.

25. Prentice AM. Variations in maternal dietary intake, birthweight, and breast-milk output in the Gambia. In: Aebi H. Whitehead R. eds. Maternal nutrition during pregnancy and lactation. Bern, Switzerland: Hans Huber, 1980: 167-83.

26. Lawrence M, Singh J. Lawrence F. Whitehead RG. The energy cost of common daily activities in African women: increased expenditure in pregnancy? Am J Clin Nutr 1985; 42: 753-63.

27. Parker LN, Gupta GR, Kurz KM, Merchant KM. Better health for women: research results from the maternal nutrition and health care program. Washington, DC: International Center for Research on Women, 1990.

28. Kusin JA, van Steenbergen WM, Lakhani SA, Jansen AAJ, Renquist U. Food consumption in pregnancy and lactation. In: van Ginneken JK, Muller AS, eds. Maternal and child health in rural Kenya. London: Croom Helm, 1984: 127-40.

29. Hussain MA, Akinyele IO. Skinfold thickness of a group of Nigerian village women during pregnancy. Nutr Rep Int 1980; 21(1): 87-92.

30. Subcommittee on Diet, Physical Activity, and Pregnancy Outcome. Nutrition issues in developing countries. Washington, DC: National Academy Press, 1992.

31. Neumann C, Bwibo NO, Sigman M. Functional implications of malnutrition: Kenya Project Final Report. Human nutrition collaborative research support program. Los Angeles, Calif, USA: University of California, Los Angeles, School of Public Health, 1992.

32. Lawrence M, Coward WA, Cole TJ, Whitehead RG. Energy requirements of pregnancy in the Gambia. Lancet 1987; 2(8567): 1072-5.

33. Lawrence M, Whitehead RG. Physical activity and total energy expenditure of child-bearing Gambian village women. Eur J Clin Nutr 1988; 42: 145-60.

34. Heini A, Schutz Y. Diaz E, Prentice AM, Whitehead RG, Jeguier E. Free-living energy expenditure measured by two independent techniques in pregnant and non-pregnant Gambian women. J Physiol 1991; 261(1P+1): E9-17.

35. Bailey RC, Jenike MR, Ellison PT, Bentley GR, Haarrigan AM, Peacock NR. The ecology of birth seasonality among agriculturalists in Central Africa. J Biosoc Sci 1992; 24: 393-412.

36. Lawrence M, Coward WA, Lawrence F. Cole TJ, Whitehead RG. Fat gain during pregnancy in rural African women: the effect of season and dietary status. Am J Clin Nutr 1987; 45: 1442-50.

37. World Health Organization. The prevalence of anaemia in women: a tabulation of available information. 2nd ed. WHO/MCH/MSM/92.2. Geneva: World Health Organization, 1992.

38. Pinstrup-Andersen P. Burger S. Habicht JP, Peterson K. Protein-energy malnutrition. In: Jamison DT, Mosley WH, Meashan A, Bobadilla JL, eds. Disease control priorities in developing countries. New York: Oxford University Press for the World Bank, 1993: 391-420.

39. World Health Organization. Nutritional status of Somalia refugees in Eastern Ethiopia, September 1988-May 1989. Wkly Epidemiol Rec 1990; 65: 93-100.

40. Kramer MS. Determinants of low birthweight: methodological assessment and meta-analysis. Bull WHO 1987; 65(5): 663-737.

41. Yambi O. Nutritional status and risk of death: a prospective study of children 6-30 months old. Doctoral thesis, Cornell University, Ithaca, NY, USA, 1988.

42. Pelletier DL, Frongillo EA, Schroeder DG, Habicht JP. The effects of malnutrition on child mortality in developing countries. Bull WHO 1995; 73(4): 443-8.

43. Royston E, Armstrong S. eds. Preventing maternal deaths. Geneva: World Health Organization, 1989.

44. Fortney J. Reproductive mortality in two developing countries. Outlook, June 1986: 6-7.

45. Tsu VD. Maternal height and age: risk factors for cephalopelvic disproportion in Zimbabwe. Int J Epidemiol 1992; 21: 941-6.

46. Buzina R. Bates CJ, van der Beek J. Brubacher G. Chandra RK, Hallberg L, Heseker J. Mertz W. Pretrazik K, Pollitt E, Pradilla A, Suboticanec K, Sandstend HH, Schalch W. Spurr GB, Westennofer J. Workshop on Functional Significance of Mild-to-Moderate Malnutrition. Am J Clin Nutr 1989; 50: 172-6.

47. Rose D, Martorell R. The impact of protein-energy supplementation interventions on child morbidity and mortality. In: Hill K, ed. Child health priorities for the 1990s. Baltimore, Md, USA: The Johns Hopkins University Institute for International Programs, 1992: 191-214.

48. United Nations Administrative Committee on Coordination/Subcommittee on Nutrition. Second report on the world nutrition situation. Vol. I. Global and regional results. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, 1992.

49. Slater TF, Block G, eds. Antioxidant vitamins and b-carotene in disease prevention. Am J Clin Nutr 1991; 53(suppl 1): 189S-396S.

50. Leslie J. Child malnutrition and diarrhea: a longitudinal study from northeast Brazil. Doctor of Science thesis, The Johns Hopkins School of Hygiene and Public Health, Baltimore, Md, USA, 1982.

51. Tomkins A, Watson F. Malnutrition and infection: a review. ACC/SCN State-of-the-Art Series, Nutrition Policy Discussion Paper No. 5. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, 1989.

52. Tomkins A. Nutritional status and severity of diarrhoea among preschool children in rural Nigeria. Lancet 1981; 1: 860-2.

53. Hetzel BS. The prevention and control of iodine deficiency disorders. ACC/SCN State-of-the-Art Series, Nutrition Policy Discussion Paper No. 3. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, 1988.

54. Pollitt E. Malnutrition and infection in the classroom. Paris: Unesco, 1990.

55. Leslie J, Jamison DT. Health and nutrition considerations in education planning. 1. Educational consequences of health problems among school-age children. Food Nutr Bull 1990; 12(3): 191-203.

56. Oduntan SO. The health of Nigerian children of school age. Doctor of Medicine thesis, University of London and World Health Organization, London and Brazzaville, Congo, 1975.

57. Cochrane SH, Leslie J, O’Hara D. Parental education and child health: intracountry evidence. Health Policy Educ 1982; 2: 213-50.

58. Stein Z. Susser M. Famine and fertility. In: Mosley WH, ed. Nutrition and human reproduction. New York: Plenum Press, 1978: 123-45.

59. Frisch R. Nutrition, fatness, and fertility: the effect of food intake on reproductive ability. In: Mosley WH, ed. Nutrition and human reproduction. New York: Plenum Press, 1978.

60. Cantrelle P. Ferry B. The influence of nutrition on fertility: the case of Senegal. In: Mosley WH, ed. Nutrition and human reproduction. New York: Plenum Press, 1978: 353-63.

61. Caraël M. Relations between birth intervals and nutrition in three central African populations (Zaire). In: Mosley WH, ed. Nutrition and human reproduction. New York: Plenum Press, 1978: 365-84.

62. Kennedy E, Bentley M. Women’s health and nutrition in sub-Saharan Africa: a review and case study from Kenya. In: Sorkin A, Siralgedin I, eds. Research in human capital and development: health and nutrition in a changing economic environment. Greenwich, Conn, USA: JAI Press, 1993; 8: 243-76.

63. Prentice AM, Cole TJ, Foord FA, Lamb WH, Whitehead RG. Increased birthweight after prenatal dietary supplementation of rural African women. Am J Clin Nutr 1987; 46: 912-25.

64. Martorell R. Arroyave G. Malnutrition, work output, and energy needs. In: Collins KJ, Roberts DF, eds. Capacity for work in the tropics. Cambridge, UK: Cambridge University Press, 1988: 57-75.

65. Kennedy E, Garcia M. Effects of selected policies and programs on women’s health and nutritional status. Washington, DC: International Food Policy Research Institute, 1993.

66. Beaton GH. Energy in human nutrition: perspectives and problems. Nutr Rev 1983; 41(11): 325-40.

Blood cholesterol and triglycerides in adolescent Egyptian girls: Relation to anthropometric measurements

Abstract
Introduction
Subjects and methods
Results
Discussion
References

Iman A. Hakim, Amina H. Awad, Nagwa H. Mohamed, and Salwa El-Husseiny

The authors are with the Child Health Department, National Research Center, in Dokki, Giza, Egypt. Iman Hakim is currently a Visiting Assistant Professor in the Department of Family and Community Medicine at the College of Medicine, University of Arizona, in Tucson, Arizona, USA.

Abstract

High serum cholesterol is a major risk factor for atherosclerosis. This cross-sectional study (n = 102) investigated the levels of total cholesterol (TC) and triglycerides (TG) in Egyptian girls aged 11 to 16 years attending a middle-class public school. The mean TC level was 194.27 ± 21.97 mg/dl, and 33.33% of girls had TC levels > 200 mg/dl. The mean TG level was 160.07 ± 30.83 mg/dl, with 3.92% of the girls showing TG levels > 200 mg/dl. Most of the girls (65.69%) were overweight, with body mass index (BMI) >25. Univariate analyses revealed an association of TC and TG with all anthropometric measures. Using stepwise regression analyses, the best model for prediction of TC was with BMI and central body fat (explaining 24.76% of TC variance); the final model for TG was with BMI, central body fat, and abdominal skinfold thickness (explaining 47.49% of TG variance). Our data show that these adolescent Egyptian girls were heavier and had higher blood lipid concentrations than subjects in the Bogalusa study and other studies worldwide. Further studies are needed to determine the factors associated with these higher lipid levels and to develop appropriate intervention programmes.

Introduction

Findings of epidemiological, family-genetic, and autopsy studies have demonstrated that the atherosclerotic process starts in childhood and further develops in adolescence and early adulthood [1, 2]. One of the most notable components of the atherosclerotic plaque is cholesterol, both in the cells and within the matrix of the arterial tissue [3]. Many paediatric epidemiological studies have demonstrated that blood lipid levels maintain their relative rank order over time, and therefore high levels in adulthood can be linked to high levels in childhood [4-8].

Epidemiological and clinical data point to a possible relation between moderate obesity and premature atherosclerosis. The relationship is stronger with truncal obesity than with fat deposition in the hips or limbs [9-12]. Adolescence is a period of rapid growth, sexual maturation, and emotional, behavioural, and social changes, all of which can have an effect on blood lipids. Evaluation of body-fat distribution in children and adolescents may help to identify persons most susceptible to cardiovascular risk in adulthood [13].

No representative studies on lipids in adolescent girls have been performed in Egypt. Guidelines of the American Heart Association, the American Academy of Pediatrics [14], and the United States National Cholesterol Education Program [15] give the following cut-off points for classifying children and adolescents (1 to 19 years of age) at risk for high blood cholesterol concentrations: 170 to 185 mg%, slightly elevated risk; >185 mg%, high risk.

We report the findings of our recent study on total cholesterol (TC) and triglyceride (TG) levels in middle-class adolescent Egyptian girls, relating them to anthropometric factors.

Subjects and methods

Study population

The study population was enrolled in a public middle school for girls in a middle-class neighbourhood in Giza, Egypt. It was considered representative of the middle social class. Approximately 550 girls (ages 11 to 16) were enrolled in grades 6 to 8. A 20% sample was drawn from each of the five classes per grade, and these girls were interviewed (n = 124). Consent forms were sent to parents asking permission to draw blood. This report concentrates on cholesterol and TG concentrations in 102 girls with a complete blood picture and biochemistry.

Determination of lipid profile

Venous blood samples were collected from each girl after an overnight fast, and concentrations of serum TC and TG were measured with the Boehringer-Mannheim Diagnostic Analysis Kit. Double quality controls were performed and revealed a coefficient of analytical variation of 1.2% for TC and 3.2% for TG.

Measurement procedures

The anthropometric measurements included height, weight, and sitting height; triceps, biceps, subscapular, suprailiac, abdominal, and femoral skinfold thicknesses; and chest, waist, hip, thigh, and mid upper arm circumferences.

Weight (to the nearest 0.1 kg) and height (to the nearest 0.5 cm) were measured on a balance beam scale with the girls wearing an examining gown over their underclothing and without shoes. All other anthropometric measurements were obtained as described by Steinkamp et al. [16, 17], all on the right side of the body. Measurements were carried out by one graduate student who had been trained for six months. The mean of three readings was calculated and used in the analysis. Derived data were calculated as follows: Body mass index (BMI) = weight/height squared (kg/m²). Body build = chest circumference/height [18]. Body-fat distribution: upper-lower fat pattern = arm circumference/thigh circumference [19]; central-peripheral fat pattern = subscapular skinfold/triceps skinfold [19]; waist-to-hip ratio (WHR) = waist circumference/hip circumference [16, 17]. The arm-to-thigh ratio measures the relative distribution of upper versus lower peripheral body fat. The subscapular-to-triceps ratio measures the relative distribution of central versus peripheral upper body fat.

Data analysis

Means, standard deviations, and ranges were calculated for each study variable. Pearson correlation coefficients were calculated among all independent variables and between each independent variable and each of the serum lipids (TC and TG).

To test the relative contribution of selected study variables to the variability of each lipid level, a multiple linear regression was performed followed by a stepwise variable selection procedure. In this analysis, we first seek the linear combination of anthropometric factors and lipid levels (TC and TG) that has the maximum correlation.

The independent variables in the regression procedure were potential correlates of each lipid level and were selected to minimize multicolinearity, or interdependence, so that two variables did not describe the same phenomenon. That is, if a high degree of association was found between two variables describing a similar phenomenon, only one would be entered in the stepwise model. Only variables selected as most potent from the preceding correlation analysis were treated as independent variables in a stepwise multiple regression. This method allowed us to determine the order of importance of each anthropometric variable for describing or predicting TC and TG. Only subjects with data available on every study variable were used in the stepwise regression analysis.

Statistical analyses were performed using the statistical package Stata [20] at the University of Arizona, Tucson. Two-sided p values <.05 were considered significant.

Results

Table 1 shows the distribution of the anthropometric measurements and lipid profiles for the study population. Most of the girls (66%) had BMI >25.

The distribution of the girls’ TC and TG levels is shown in figures 1 and 2. The mean TC and TG values of all girls were 194 ± 22 mg/dl and 160 ± 31, respectively. The percentages of TC and TG levels in various ranges are shown in table 2. Seventy-seven percent of the girls had TC levels greater than 185 mg/dl, and 22% had TG levels greater than 185 mg/dl. Thirty-four percent had TC levels greater than 200 mg/dl, and 4% had TG levels greater than 200 mg/dl.

A worldwide review [21-26] of TC and TG levels in adolescent girls in the last 20 years (table 3) allowed us to compare our results with those of other studies in different countries. The mean serum TC level in the adolescent Egyptian girls exceeded the levels in the Bogalusa study and other American reports by more than 15 mg/dl.

Univariate analysis

Pearson correlation coefficients between the various study variables and serum TC and TG are given in table 4. The findings show that neither TC nor TG was associated with age or pubertal stage. On the other hand, both TC and TG were significantly associated with all anthropometric measures except for height, span, and upper fat pattern. In addition, TC was not associated with either biceps skinfold thickness or WHR, and sitting height was not associated with TG.

TABLE 1. Physical characteristics of the study population of adolescent girls in Giza. Egypt (n = 102)

Variable		Mean	SD	Range
Age (yr)		14.2	1.3	11-16
Pubertal stage (Tanner)		3.7	0.8	1-4
Total cholesterol (mg/dl)		194.3	22.0	107-285
Triglycerides (mg/dl)		160.1	30.8	79-224
Body measurements
	weight (kg)	65.8	14.6	33-120
	height (cm)	155.9	8.0	131-176
	sitting height (cm)	80.8	5.6	59-93
	BMI (kg/m2)	27.1	5.3	15.6-44
	body build (chest circumference/height)	0.6	0.1	0.5-0.8
Skinfolds (mm)
	triceps	22.7	6.4	6-38
	subscapular	23.2	8.3	6-38
	suprailiac	21.7	7.5	5-42
	femoral	34.3	7.1	6-53
	abdominal	30.9	8.5	9-45
	biceps	14.1	5.2	4-29
Circumferences (cm)
	mid upper arm	27.9	3.7	19-39
	chest	97.3	10.9	70-127
	waist	88.7	10.7	63-120
	hip	110.0	11.1	83-154
	thigh	54.7	7.2	38-77
Fat pattern
	upper a	0.5	0.0	0.38-0.68
	central b	1.0	0.3	0.49-1.92
	WHR c	0.8	0.1	0.71-0.94

a. Arm circumference/thigh circumference.
b. Subscapular skinfold/triceps skinfold.
c. Waist circumference/hip circumference.

Regression analyses were then performed to evaluate the effect of several factors simultaneously and to quantitate the relative contribution of each variable to serum lipids when examined in the presence of all other study variables. Only those variables considered significant at the univariate level were considered as potential variables for entry, i.e., BMI; suprailiac, femoral, and abdominal skinfold thicknesses; central fat pattern (triceps skinfold/subscapular skinfold); waist and thigh circumferences; and WHR.

The results of the final model are shown in table 5. Two variables entered the model as significant predictors of TC level: BMI and central body-fat pattern. These two explained 25% of the variance in TC level. TG level was best predicted by BMI, central body-fat pattern, and abdominal skinfold thickness, which explained 48% of the variance.

Discussion

Hypercholesterolaemia is well known as a major risk factor for coronary atherosclerosis. There is interest in measuring blood cholesterol levels in the young because there is increasing evidence that atherosclerosis has its beginning in childhood [1, 2]. Tracking of high TC and TG levels from childhood to adolescence and on into early adulthood is well documented in Western populations [4-8]. The need for cholesterol determination on a population basis and in the paediatric age group has been well described. One strategy of population comparison is to select a specific age and sex group for comparison across several populations. In our study, although most girls were overweight, they represented a wide spectrum of body weights and lipid levels. Our data revealed that the mean TC (194.27 mg/dl) and TG (160.07 mg/dl) serum levels of normal adolescent Egyptian girls were high in comparison with levels in other countries [21-26]. Unfortunately, no qualifying survey was found from the Eastern Mediterranean region. The cause of these high levels requires further investigation, but it may be related to nutritional habits, considering the sedentary lifestyle of adolescent girls in Egypt. Gliksman et al. [27] studied 5,211 schoolchildren aged 10 to 15 years and reported that BMI was the highest, and aerobic fitness was the lowest, in children from Mediterranean and Middle Eastern countries.

FIG. 1. Distribution of TC levels of 102 girls.

FIG. 2. Distribution of TG levels of 102 girls.

TABLE 2. Percentage of total cholesterol (TC) and triglyceride (TG) levels in various ranges selected according to the US cut-off points for TC¹⁵ m children and adolescents

Level (mg/dl)	TC		TG
	Frequency	%	Frequency	%
<170	4	3.9	53	52.0
170-185	19	18.6	27	26.4
186-200	45	44.1	18	17.7
>200	34	33.3	4	3.9
Total	102	100.0	102	100.0

TABLE 3. Mean and standard deviation (SD) of total cholesterol (TC) and triglyceride (TG) levels (to nearest mg/dl) in adolescent girls from some selected epidemiologic studies

Location of Study		TC				TG
		Age (yr)	n	Mean	SD	n	Mean	SD	Year	Ref.
Bogalusa
	white	6-14	1,030	164	27	976	77	38	1976	21
	black	6-14	605	171	30	565	64	24	1976	21
Princeton
	white	12-17	1,144	157	24	1,144	80	31	1977	22
	black	12-17	411	165	28	411	68	25	1977	22
Japan		15	217	182	28	213	79	25	1982	23
Austria		<20	484 a	167	26	484 a	91	47	1992	24
Italy		school-children	-	178	2	-	82	3	1992	25
Tanzania		15-19	924	142	-	924	42	-	1993	26
Egypt		11-16	102	194	22	102	160	31	1996	this study

a. Includes troth boys and girls.

There is evidence that body composition in children and adolescents may be more closely correlated with TC levels in certain ethnic groups [31]. Lipidaemia and total cholesterolaemia increase with BMI [32]. In women, BMI and waist circumference by themselves did as well as body-fat distribution indices in explaining variation in TC and TG levels, suggesting the involvement of visceral fat in the relation between disease and body fat or body-fat distribution [33]. WHR has been used as an index of body-fat distribution in several studies [34-36]. Other studies [18,19] used the ratios arm circumference/thigh circumference and subscapular skinfold thickness/triceps skinfold thickness to distinguish upper and central fat patterns.

TABLE 4. Correlation coefficients for study variables with serum cholesterol (TC) and triglycerides (TG) in adolescent girls in Giza, Egypt

Parameter		TC	TG	Age
Age (yr)		0.11	0.11	1.00
Pubertal stage (Tanner)		-0.08	-0.04	0.17
Weight (kg)		0.40***	0.47***	0.33***
Height (cm)		0.01	-0.16	0.43***
Sitting height (cm)		0.27**	0.18	0.40***
BMI (kg/m2)		0.45***	0.59***	0.18
Body build (chest circumference/height)		0.42***	0.60***	0.15
Skinfolds (mm)
	triceps	0.22*	0.38***	0.13
	subscapular	0.36***	0.56***	0.19
	suprailiac	0.43***	0.57***	0.06
	femoral	0.29**	0.49***	0.02
	abdominal	0.38***	0.63***	-0.00
	biceps	0.18	0.28**	0.06
Circumferences (cm)
	mid upper arm	0.38***	0.55***	0.17
	chest	0.41***	0.51***	0.32**
	waist	0.46***	0.55***	0.21*
	hip	0.47***	0.52***	0.22*
	thigh	0.35***	051***	0.15
Fat pattern
	upper a	0.05	0.11	0.06
	central b	0.30**	0.42***	0.16
	WHR c	0.17	0.28**	0.06

Significance levels: *p <.05; **p <.01; ***p <.001.

a. Arm circumference/thigh circumference.
b. Subscapular skinfold/triceps skinfold.
c Waist circumference/hip circumference.

TABLE 5. Results from the final model to predict the variability of TC and TG ^a

Variables	TC partial r2(%)	TG partial r2 (%)
BMI	19.8	35.3
Central fat pattern	5.0	8.9
Abdominal skinfold thickness		3.3
total r2 =	24.8%	47.5%
F =	14.3	25.9
P =	0.0	0.0

a. Only girls having all study variables were included in the analysis. The variables tested were BMI; suprailiac, femoral, abdominal, and triceps/subscapular (central fat pattern) skinfold thickness; and waist, thigh, and WHR circumferences.

In conclusion, our findings demonstrate the high prevalence of elevated levels of TC and TG in middle-class adolescent Egyptian girls and their association with overweight and central fat pattern, thus providing the basis for appropriate intervention by establishing school health education programmes that include both vigorous physical activity and lipid screening. Preventing or slowing the atherosclerotic process in childhood and adolescence could mean years of healthy life for many people.

References

1. McNamara JJ, Molot MA, Stremple JF, Cutting RT. Coronary artery disease in combat causalities in Vietnam. JAMA 1971: 216: 1185-7.

2. Newman WP, Freedman, DS, Voors AW, Gard PD, Srinivasan SR, Cresanta JL, Williamson GD, Webber LS, Berenson GS. Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis: the Bogalusa Heart Study. N Engl J Med 1986; 314: 138-44.

3. McMillen GC. Nature and definitions of atherosclerosis. Ann NY Acad Sci 1985; 454: 1-4.

4. Clarke WR, Schrott HG, Leaverton PE, Connor WE, Lauer RM. Tracking of blood lipids and blood pressures in school age children: the Muscatine Study. Circulation 1978; 58: 626-34.

5. Webber LS, Cresanta JL, Voors AW, Berenson GS. Tracking of cardiovascular disease risk factor variables in school age children. J Chron Dis 1983; 36: 647-60.

6. Laskarzewski P. Morrison JA, deGroot I, Kelly KA, Mellies MJ, Khoury P. Glueck CJ. Lipid and lipoprotein tracking in 108 children over a four-year period. Pediatrics 1979; 64: 584-91.

7. Stuhldreher WB, Orchard TJ, Donahue RP, Kuller LH, Gloninger MF, Drash AL. Cholesterol screening in childhood: sixteen-year Beaver County Lipid Study experience. J Pediatr 1991; 119: 551-6.

8. Porkka KV, Viikari JS, Akerblom HK. Tracking of serum HDL-cholesterol and other lipids in children and adolescents: the Cardiovascular Risk in Young Finns Study. Prev Med 1991; 20: 713-24.

9. Freedman DS, Srinivasan SR, Harsha DW, Weber LS, Berenson GS. Relation of body fat patterning to lipid and lipoprotein concentrations in children and adolescents: the Bogalusa Heart Study. Am J Clin Nutr 1989; 50: 930-9.

10. Baumgartner RN, Roche AF, Guo S. Chumlea WC, Ryan AS. Fat patterning and centralized obesity in Mexican-American children in the Hispanic health and nutrition examination survey (HHANES 1982-1984). Am J Clin Nutr 1990; 51: 936S-43S.

11. Zwiauer K, Widhalm K, Kerbl B. Relationship between body fat distribution and blood lipids in obese adolescents. Int J Obesity 1990; 14: 217-27.

12. de Ridder CM, Bruning PF, Zonderland ML, Thijssen JHH, Bonfrer JMG, Blankenstein MA, Huisveld IA, Erich WBM. Body fat mass, body fat distribution and plasma hormones in early puberty in females. J Clin Endocrinol Metab 1990; 70: 888-93.

13. Zwiauer KF, Pakosta R. Mueller T. Wildhalm K. Cardiovascular risk factors in obese children in relation to weight and body fat distribution. J Am Coll Nutr 1992; 11(suppl): 41S-50S.

14. Committee on Nutrition: American Academy of Pediatrics indications for cholesterol testing in children. Pediatrics 1989; 83: 141-4.

15. National Cholesterol Education Program: Report of the expert panel on population strategies for blood cholesterol reduction. Bethesda, Md, USA: NIH Publication 3046, 1990.

16. Steinkamp RC, Cohen NL, Siri WE, Sargent TW, Walsh HE. Measures of body fat and related factors in normal adults-I. J Chron Dis 1965; 18: 1279-89.

17. Steinkamp RC, Cohen NL, Gaffey WR, McKey T. Bron G. Siri WE, Sargent TW, Issacs E. Measures of body fat and related factors in normal adults-II. J Chron Dis 1965; 18: 1291-1307.

18. Foster CJ, Weinsier RL, Birch R. Norris DJ, Bernstein RS, Wang J. Pierson RN, Vanitallie TB. Obesity and serum lipids: an evaluation of the relative contribution of body fat and fat distribution to lipid levels. Int J Obesity 1987; 11: 151-61.

19. Weinsier RW, Norris DJ, Birch R. Bernstein RS, Wang J. Yang MU, Pierson RN Jr, Van Itallie TB. The relative contribution of body fat pattern to blood pressure level. Hypertension 1985; 7: 578-85.

20. Stata Statistical package. Release 3.0. Computing Resource Center, 1640 Fifth Street, Santa Monica, CA 90401, USA.

21. Frerichs RR, Srinivasan SR, Webber LS, Berenson GR. Serum cholesterol and triglyceride levels in 3446 children from a biracial community. Circulation 1976; 54: 302-9.

22. Morrison JA, deGroot I, Edwards BK, Kelly KA, Raugh JL, Mellies M, Glueck CJ. Plasma cholesterol and triglyceride levels in 6,775 school children, ages 617. Metab Clin Exp 1977; 26: 1199-211.

23. Ueshima H. Kitada M, Iida M, Tanigaki M, Shimamoto T. Konishi M, Nagano E, Nakanishi N. Takayama Y. Ozawa H. Komachi Y. Serum total cholesterol, triglyceride level, and dietary intake in Japanese students aged 15 years. Am J Epidemiol 1982; 116: 343-52.

24. Widhalm K, Koch S. Pakosta R. Schurz M, Brendinger M. Serum lipids, lipoproteins and apolipoproteins in children with and without familial history of premature coronary heart disease. J Am Coll Nutr 1992; 11(suppl): 32S-5S.

25. Giovannini M, Bellu R, Ortisi MT, Incerti P, Riva E. Cholesterol and lipoprotein levels in Milanese children: relation to nutritional and familial factors. J Am Coll Nutr 1992; 11(suppl): 28S-31S.

26. Kitange HM, Swai AB, Masuki G, Kilima PM, Alberti KG, McLarty DG. Coronary heart disease risk factors in sub-Saharan Africa: studies in Tanzanian adolescents. J Epidemiol Commun Health 1993; 47: 303-7.

27. Gliksman MD, Lazarus R, Wilson A. Differences in serum lipids in Australian children: Is diet responsible? Int J Epidemiol 1993; 22: 247-54.

28. Omura T, Takizawa Y, Kojima S, Funaki S, Sawabe K, Takakuwa K, Ono Y, Kishi M, Yamazaki T, Iida M, Komachi Y, Wakamatsu W. A five-year follow-up study on blood pressure and serum cholesterol in junior high school children. Jpn J Public Health 1991; 38: 417-24.

29. Morrison JA, Laskarzewski PM, Rauh JL, Brookman R, Mellies M, Frazer M, Khoury P, deGroot I, Kelly K, Glueck CJ. Lipids, lipoprotein and sexual maturation during adolescence: the Princeton Maturation Study. Metab Clin Exp 1979; 28: 641-9.

30. McColl P, Amador M, Diaz M. Blood cholesterol and triglycerides in adolescents during sexual development. Rev Chil Pediatr 1991; 62(1): 14-7.

31. Wong ND, Bassin SL, Deitrick R. Relationship of blood lipids to anthropometric measures and family medical history in an ethnically diverse school-aged population. Ethn Dis 1991; 1: 351-63.

32. Guzzaloni G. Moreni G. Ardizzi A, Moro D, Grugni G. Morabito F. Correlation between lipid pattern and body mass index in obesity. Minerva Med 1991; 82: 339-44.

33. Mueller WH, Wear ML, Hanis CL, Emerson JB, Barton SA, Hewett-Emmett D, Schull WJ. Which measure of body fat distribution is best for epidemiologic research? Am J Epidemiol 1991; 33: 859-69.

34. Kalkhoff RK, Hartz AH, Rupley D, Kissebah AH, Kelber SL. Relationship of body fat distribution to blood pressure, carbohydrate tolerance, and plasma lipids in healthy obese women. J Lab Clin Med 1983; 102: 621-7.

35. Hartz AJ, Rupley DC, Kakhoff RD, Rimm AA. Relationship of obesity to diabetes: influence of obesity level and body fat distribution. Prev Med 1983; 12: 351-7.

36. Lapidus L, Bengtsson C, Larsson B. Pennert K, Rybo E, Sjostrom L. Distribution of adipose tissue and risk of cardiovascular disease and health: a 12 year follow up of participants in the population study of women in Gothenburg, Sweden. BMJ 1984; 289: 1257-61.

37. Gillum RF. The association of the ratio of waist to hip girth with blood pressure, serum cholesterol and serum uric acid in children and youths aged 6-17 years. J Chron Dis 1987; 40: 413-20.

Daily versus weekly iron supplementation: Programmatic and economic implications for Indonesia

Abstract
Introduction
Prevalence of anaemia in different population groups
Current iron supplementation programmes in Indonesia
Weekly dosing in iron supplementation
Costs of iron supplementation
References

Rainer Gross, Imelda Angeles-Agdeppa, Werner J. Schultink, Drupadi Dillon, and Soemilah Sastroamidjojo

Rainer Gross and Werner Schultink are advisors of the Deutsche Gesellschaft fur Technische Zusammenarbeit (GTZ) at the SEAMO-TROPMED Regional Center for Community Nutrition in Jakarta, Indonesia. Imelda Angeles-Agdeppa is a doctoral student. Soemilah Sastroamidjojo and Drupadi Dillon are staff members of the Center.

Abstract

Available information on iron deficiency and anaemia among Indonesian population groups was analysed to identify the at-risk groups in Indonesia and to suggest more efficient intervention programmes to reduce its high prevalence. The results showed that the groups with the highest prevalence of anaemia were pregnant women (52.3%), working adult women (27.9%), pre-schoolers (27.1%), adolescent girls (21.1%), the elderly (10.9%), and primary-school children (6.8%). Pre-school children and adolescents need iron supplementation to provide enough iron for growth and cognitive functioning during childhood. Adolescent girls need iron supplementation to develop sufficient iron reserves before pregnancy and to improve working performance. Pregnant and lactating women also need iron supplementation. Strategies to control anaemia include improvement in dietary habits, food fortification, and supplementation. Unless feeding behaviour is changed or food fortification is adapted nationwide, oral iron supplementation remains the mainstay of the prevention and treatment of anaemia. Weekly supplementation has been shown to be an effective and economic method of supplementation. With the current approach of daily supplementation recommended by the World Health Organization, Indonesia would have to spend US$360 million annually, but weekly dosing would require only US$15 million to cover the same target population. Weekly dosing offers a practical and economic means of improving iron status in developing countries.

Introduction

Anaemia is recognized as a major nutritional problem affecting a majority of the women and children in developing countries. The World Health Organization (WHO) estimates that more than two billion people are affected by iron deficiency or anaemia. Most are in the Western Pacific and South-East Asia, but they can be found in all countries [1]. Despite its increasing prevalence in South-East Asia [2], anaemia is the most neglected nutritional deficiency disorder in the region today [3]. The most affected groups, in approximate descending order, are pregnant women, preschool-age children, low-birthweight infants, other women of child-bearing age, the elderly, school-age children, and adult men [1, 4].

Iron-deficiency anaemia has severe nutritional and health consequences, including inadequate growth and mental development in children, high maternal mortality, high incidence of low-birthweight infants, and low productivity in adults [1]. Poor school performance among schoolchildren and adolescents has been associated with iron-deficiency anaemia [5-7]. Low physical performance was also observed in a group of adolescents in London [8].

In Indonesia, although some pilot studies on supplementation trials among pre-school children have been conducted [9], routine oral iron supplementation has been focused only on pregnant women [10]. Despite a national programme for the last 20 years, the prevalence of anaemia decreased only slowly from 70% in 1983 to between 55% and 60% in 1992 [11]. The main reasons for the ineffectiveness of the programme are low compliance and inefficient delivery systems [12, 13]. Another factor may be the late administration of supplements. Pregnancy is the period of greatest demand for iron [14], but improving iron nutrition during this period is difficult. The conventional target approach of providing iron plus folate during the second trimester does not adequately address the problem. Careful evaluation of the critical period when supplementation should take place is required to minimize scarce budgetary resources.

FIG. 1. Prevalence of anaemia in different population groups in Indonesia. Abbreviations: Adoles, adolescents; PS, pre-schoolers; SC, schoolchildren

The present overview, based on the current experience gained in Indonesia, describes population groups that are at high risk of anaemia and analyses the effect of weekly dosing of iron so that future intervention programmes can be initiated for not only curative but also preventive purposes.

Prevalence of anaemia in different population groups

Subjects were randomly chosen in each population group from a series of cross-sectional surveys conducted in selected rural and urban areas in Jakarta and Yogjakarta, Indonesia. The methods and results have been reported [12, 13, 15-18]. Anaemia is defined here as haemoglobin levels <120 g/L among adolescent girls and elderly women, <110 g/L among pre-school children and pregnant women, and <130 g/L among adult and elderly men [19].

Figure 1 summarizes the prevalence of anaemia in different Indonesian population groups. The average prevalence was highest among pregnant women (52.3%). In both rural and urban areas, the prevalence of anaemia in pregnant women was not associated with age (p >.05), but it was associated with months of pregnancy (p <.001; X² test) and whether they lived in a rural or an urban area. As shown in figure 2, anaemia in rural areas appeared at a low rate in the third month of pregnancy (18%) and was highest in the fifth month of pregnancy (75%). It decreased to 33% at nine months. In urban areas, anaemia occurred as early as the second month of pregnancy (50%) and reached a prevalence of 100% in the eighth month. Pregnant women in this study belonged to the low- and middle-income groups; urban pregnant women were more severely anaemic than rural ones (fig. 2). Suharno et al. [20] found a similar prevalence of anaemia (49.4%) in pregnant women in West Java. The most commonly identified reason for this high prevalence was the low availability of iron in the rice-based Indonesian diet [21-23].

The prevalence of anaemia in non-pregnant women was 27.9%, and no men were found to be anaemic. The prevalence of anaemia among adult women in this study was lower than that among non-pregnant Chinese cotton mill workers (34%) [24] but remained within the range of the prevalence of anaemia in Indonesia nine years ago [11]. Iron supplements resulted in improved productivity [9] and production efficiency [25]. Among the elderly, the prevalence was higher in women (13.1%) than in men (8.9%).

FIG. 2. Prevalence of anaemia in pregnant women in urban and rural areas in Indonesia

FIG. 3. Prevalence of anaemia in urban pre-schoolers in Indonesia

The prevalence of anaemia among pre-school-age children was 26.4% for boys and 27.9% for girls, with a combined prevalence of 27.1%. The prevalence of anaemia in pre-schoolers was not associated with sex (p >.05; X² test). The highest prevalence of anaemia (32% to 36%) was at 17 to 35 months of age in both boys and girls (fig. 3). The high prevalence of anaemia (34.9%) in this study as early as 17 to 23 months reflects depleted iron stores [26]. An even higher prevalence is found before 17 months [27, 28]. During this period, the iron requirement is so high that breastmilk alone cannot meet the requirement of iron for rapid growth [29]. Breastmilk contains 0.5 mg iron per litre during the first month post-partum, falling to about 0.3 mg/L at 4 to 6 months. Assuming a mean daily intake of 673 ml milk 1 month after birth and 896 ml at 6 months [30], the calculated daily iron intake falls from 0.34 mg at 1 month to 0.27 mg at 6 months. Other factors, such as repeated attacks of infectious diseases and the low content and bioavailability of iron in the diet, put the young child at higher risk for anaemia. The prevalence of anaemia decreased to 12.9% at 48 to 60 months (fig. 3). A lower prevalence in this age group was also reported among Chinese [27] and Chilean [28] pre-school children. This age-dependent dynamic of anaemia is consistent with the epidemiology of nutritional anaemia [31]. The prevalence was 9.1% among school-age boys but only 4.5% among school-age girls, with a combined prevalence of 6.8%.

The prevalence was 21.1% in adolescent girls but only 2.5% in adolescent boys. The high prevalence of anaemia among adolescent girls may be due to menstrual losses, which average 38.0% ± 1.2% of the total iron losses, and the occurrence of the second growth spurt [32] combined with low dietary iron intake [33].

Current iron supplementation programmes in Indonesia

The current programmes to control anaemia in Indonesia include daily iron supplementation of pregnant women through the integrated health posts and community health centres, nutrition education, and food fortification [10]. The low indication of improvement, as shown in this study, simply reflects the limited effectiveness and sustainability of these programmes. To encourage dietary modification, emphasis must be placed also on the diversity of food produced and on improved access to these foods. Changing people’s behaviour requires long and intensive national and community-based actions, such as mass media campaigns and other formal and informal education in the community, that support each other synergistically. Food fortification, the medium-term strategy, offers benefits faster than dietary modification. However, many operational problems may be encountered [1, 34].

The present high prevalence of anaemia in most age groups justifies a broad intervention programme [35]. In developing countries, where dietary iron is unlikely to be sufficient to overcome an iron deficit, iron supplements are often relied upon as a cure for anaemia. This curative strategy of giving iron supplements when the anaemia already exists is frequently of limited effectiveness and sustainability because of poor compliance due to the negative side-effects of daily dosing [13].

Weekly dosing in iron supplementation

The Indonesian experience showed that the prevalence of iron-deficiency anaemia is highest in preschool children and pregnant women. These two peaks of prevalence have different causes: the high growth spurt in infants, and menarche in women leading to depleted iron storage from blood loss. The need for iron is aggravated during pregnancy. A curative approach alone is not sufficiently effective to reduce markedly the high prevalence of anaemia in pregnancy. Preventive iron supplementation of women before pregnancy is required to prevent iron-deficiency anaemia during pregnancy. Dietary changes and food fortification are important long-term strategies but cannot be expected to have a rapid impact.

Several studies in different population groups have shown that weekly dosing successfully reduces anaemia and is as effective as daily administration in preschool children [15], adolescent girls [36], and non-pregnant [37, 38] and pregnant [18] women.

Costs of iron supplementation

WHO/UNICEF recommend that universal iron-folate supplementation should be implemented for all pregnant and lactating women, for infants and children six months through five years of age, and for pre-adolescent girls and women from 10 to 49 years of age, in populations where the prevalence of anaemia is over 30% [39]. In this recommendation, no specific information is provided about when and for how long supplements should be administered. However, an earlier recommendation [40] stated that in areas with high prevalence, pregnant women should receive 60 mg elemental iron and 250 µg folate twice a day for 180 days, and normal infants should receive 1 mg elemental iron per kilogram body weight per day from six months to five years of age. Supplementation from the age of two months is recommended for low-birthweight infants. For adolescent girls, 60 mg elemental iron twice a day for two to three months is recommended. Table 1 shows the recommended duration of iron supplementation for each intervention group. On the basis of a calculation of the price of supplementation for the different intervention groups, a yearly budget of US$360 million would be needed to follow the 1992 WHO recommendation in Indonesia. Although the economic return would be higher because of increased productivity and learning ability and fewer infections, it would be unrealistic to expect an investment of this size for reducing iron-deficiency anaemia alone.

TABLE 1. Duration of iron supplementation of different target populations in Indonesia with the current WHO-recommended daily dosing approach versus the suggested new weekly dosing approach

Target population	Current WHO approach a	Suggested new approach b
Pregnant women	180 d	26 wk
Lactating women	120 d	17 wk
Infants (0.5-1 yr)	180 d	26 wk
Low-birthweight infants	300 d	40 wk
Pre-schoolers (>1 <5 yr)	60 d/yr	16 wk c
Adolescent girls (10 -<19 yr)	60 d/yr	9 wk/yr

a. Two tablets/day for pregnant and lactating women and adolescent girls (one tablet = 60 mg elemental iron and 250 µg folate); 30 mg elemental iron per day for infants.

b. Two tablets/week for pregnant and lactating women; one tablet/week for adolescent girls (one tablet = 60 mg elemental iron and 250 µg folate); 30 mg elemental iron per week for infants.

c. Two eight-week periods in four years.

Target population	Millions of people (% of total) a	Cost in millions of US dollars
		Current WHO approach	Suggested new approach
Pregnant and lactating women	5.5 (2.9)	3.9	0.5 b
Infants (0.5-1 yr)	1.9 (1.0)	11.4	1.6 c
Low-birthweight infants	0.5 (0.3) d	5.4	0.7 c
Pre-schoolers (>1 <5 yr)	19.7 (10.3)	315.2	10.5 c
Adolescent girls (10 -<19 yr)	29.9 (11.6)	23.6	1.6 b
Total	57.5 (25.8) e	359.5	14.9

a. Based on estimated 1996 population of Indonesia (total, 191 million) Source: Directorate of Nutrition, Department of Health, Indonesia.

b. US$1 per 1,000 iron pills [41].

c. US$1 per bottle (150 ml) of iron syrup containing 6 mg elemental Fe per millilitre (local manufacturer’s price).

d. 15% of infants have low birthweight [42].

e. Does not include low-birthweight infants.

First, on the basis of the promising results in Indonesia, weekly doses are suggested for all age groups. This will reduce costs to one-seventh of the daily administration.

Second, it has been shown in pre-school children that iron supplementation for eight weeks has an effect lasting for many months, until the age when iron requirements decline and can be met by common food sources [43].

Weekly supplementation not only is more efficient but also is far less demanding of organizational and administrative efforts, because the reduced number of tablets and syrup bottles allows an increase in the coverage of the programme. In Indonesia traditional birth attendants have access to pregnant and lactating mothers and low-birthweight infants. Schoolteachers have contact with schoolchildren and adolescent girls, and factories employ many non-pregnant women. All of these could be enlisted in support of weekly supplementation. Weekly iron supplementation is cheaper and also offers opportunities for wider coverage.

References

1. Food and Agriculture Organization/World Health Organization. International conference on nutrition. Nutrition and development: a global assessment. In: Micronutrient deficiencies and preventing specific micronutrient deficiencies. Rome: Food and Agriculture Organization, 1992.

2. Garcia M, Mason J. Second report of the world nutrition situation. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, 1992.

3. Gopalan C. Strategies for combating undernutrition: lessons learned for the future. In: Nutrition in developmental transition in Southeast Asia. New Delhi: World Health Organization, 1992: 109-111.

4. Food and Agriculture Organization/World Health Organization. International conference on nutrition. In: Major issues for nutrition strategies. Rome: Food and Agriculture Organization, 1992.

5. Pollitt E, Hathirat P. Kotchabhakdi N. Missell L, Valyasevi A. Iron deficiency and educational achievement in Thailand. Am J Clin Nutr 1989; 50: 687-97.

6. Soemantri AG. Preliminary findings on iron supplementation and learning achievement of rural Indonesian children. Am J Clin Nutr 1989; 50: 698-702.

7. Seshadri S. Gopaldas T. Impact of iron supplementation on cognitive functions in preschool and school-aged children: the Indian experience. Am J Clin Nutr 1989; 50: 675-86.

8. Nelson M, Bakaliou NM, Trivedi A. Iron deficiency anaemia and physical performance in adolescent girls from different ethnic backgrounds. Br J Nutr 1994; 72: 427-33.

9. Hussaini MA, Karyadi D, Soewondo S. Suhardjo, Djojosoebagio S. Pollitt E, Scrimshaw NS. Effect of iron deficiency on physical growth, cognitive process, morbidity, and work productivity. In: Hin CY, Wai TNK, Siong TE, Noor MI, eds. Proceedings of the 6th Asian Congress of Nutrition. Kuala Lumpur, Malaysia: Nutrition Society of Malaysia, 1991: 302-16.

10. Kodyat B. Djokomoelyanto, Karyadi D, Tarwotjo, Muhilal, Hussaini MA, Sukaton A. Nutritional anaemia (iron deficiency). In: Micronutrients malnutrition intervention program: an Indonesian experience. Jakarta, Indonesia: Directorate of Community Nutrition, Directorate General of Community Health, Ministry of Health, 1991.

11. Muhilal, Herman S. Karyadi D. The current national prevalence of anaemia among pregnant women and its preventive measures in Indonesia. In: Tharavaniji S. Fungladda W. Khusmith S. Pruekwatana, eds. 13th International Congress for Tropical Medicine and Malaria. Pattaya, Thailand, 1992 (Abstract).

12. Schultink W. van der Ree M, Matullessi P. Gross R. Low compliance with an iron supplementation program: a study among pregnant women in Jakarta, Indonesia. Am J Clin Nutr 1993; 57: 135-9.

13. Thorand B. Schultink W. Gross R. Sastroamidjojo S. Wentzel S. Efficiency of iron supplementation programme for pregnant women in Jeneponto, Sulawesi, Indonesia. Asia Pacific J Clin Nutr 1994; 3: 211-5.

14. International Nutritional Anaemia Consultative Group. Iron deficiency in women. Washington, DC: INACG, 1981.

15. Schultink WJ, Gross R. Gliwitzki M, Karyadi D, Matulessi P. Effects of weekly iron supplementation in Indonesian preschool children with a low iron status. Am J Clin Nutr 1995; 61: 111-5.

16. Angeles IT, Schultink WJ, Matullessi P. Gross R. Sastroamidjojo S. Decreased rate of stunting through iron supplementation. Am J Clin Nutr 1993; 58: 339-42.

17. Scholz B. Gross R. Shultink W. Sastroamidjojo S. Anaemia is associated with reduced productivity of women workers even in less physically strenuous tasks. Br J Nutr 1997; 7: (in press).

18. Ridwan E, Schultink WJ, Dillon D, Gross R. Effects of weekly supplementation of Indonesian pregnant women are similar to those of daily supplementation. Am J Clin Nutr 1996; 63: 884-90.

19. World Health Organization. The prevalence of anaemia in women: a tabulation of available information. Geneva: World Health Organization, 1992.

20. Suharno D, West C, Muhilal, Logman MHG, de Waart FG, Karyadi D, Hautvast JGAJ. Cross sectional study on the iron and vitamin A status of pregnant women in West Java, Indonesia. Am J Clin Nutr 1992; 56: 988-93.

21. Martoatmodjo S, Abunain D, Muhilal, Enoch M, Hussaini MA, Sastroamidjojo S. Nutritional anaemia among pregnant women and food consumption pattern. Jakarta, Indonesia: Penelitian Gizi den Makanan 1973; 3: 22-41.

22. Kusin J, Kardjati S, Suryohudoyo P, de With C. Anaemia and hypovitaminosis A among rural women in East Java, Indonesia. Trop Geogr Med 1980; 32: 30-9.

23. Hussaini MA, Lamid A, Hussaini YK, Mulyati S, Herno, Suharno D. Study on the implementation and delivery strategies of iron preparate for policy and program formulation to reduce nutritional anaemia among pregnant women. Bogor, Indonesia: Nutrition Research and Development Center and Directorate of Nutrition, Ministry of Health, 1989.

24. Li R, Chen X-C, Yan H-C, Deurenberg P, Garby L, Hautvast J. Prevalence and type of anaemia in female cotton mill workers in Beijing, China. Br J Nutr 1993; 30: 787-96.

25. Li R. Chen X, Yan H. Deurenberg P. Garby L, Hautvast J. Functional consequences of iron supplementation to iron deficient female cotton mill workers in Beijing, China. Am J Clin Nutr 1994; 59: 908-13.

26. Morton RE, Nysenbaum A, Price K. Iron status in the first year of life. J Pediatr Gastroenterol Nutr 1988; 7: 707-12.

27. Chen XS, Ge KY. Nutrition transition in China: the growth of affluent diseases with the alleviation of undernutrition. In: Wahlqvist M, Truswell SA, Smith R. Nestel P. eds. Nutrition in a sustainable environment. Proceedings of the 15th International Congress of Nutrition. Adelaide, Australia: IUNS 1994: 189-98.

28. Ruz M. Trace element intake and nutriture in Latin America. In: Wahlqvist M, Truswell SA, Smith R. Nestel P. eds. Nutrition in a sustainable environment. Proceedings of the 15th International Congress of Nutrition. Adelaide, Australia: IUNS, 1994: 296-300.

29. National Research Council. Recommended dietary allowances. Commission of Life Sciences. 10th ed. Washington, DC: National Academy Press, 1989: 195203.

30. Lonnerdal B. Iron and breast milk. In: Stekel A, ed. Iron nutrition in infancy and childhood. Nestle Nutrition Workshop Series. New York: Raven Press, 1984: 95-117.

31. Dallman PR. Changing iron needs from birth through adolescence. In: Fomon SJ, Zlotkin S. eds. Nutritional anemias. Nestle Nutrition Workshop Series, Nestle Ltd. Vevey, Switzerland: Raven Press, 1992: 29-35.

32. Hallberg L. Iron balance in menstruating women. Eur J Clin Nutr 1995; 49: 200-7.

33. Krijger E, Schultink W. Gross R. van Raaij JMA. Nutritional status of adolescent girls in East Jakarta in relation to their socio-economic status. Master’s thesis, SEAMO-TROPMED Center for Community Nutrition, University of Indonesia, Jakarta, 1995.

34. Mejia L. Fortification of foods: historical development and current practices. Food Nutr Bull 1994; 15: 278-81.

35. Hercberg S. Iron and folate deficiency anemias. In: International child health: a digest of current information. International Pediatrics Association, 1991; 2: 44-60.

36. Kätelhut A, Schultink W. Angeles I, Gross R. Pietrzik K. The effects of weekly iron supplementation with folic acid, vitamin A, vitamin C on iron status of Indonesian adolescents. Asia Pacific J Clin Nutr 1996; 5(3): 181-5.

37. Tee ES, Cavalli-Sforza LT, Kandiah M, Narimah A, Chong SM, Satgunasingan N. Kamarudin L. A study of the effectiveness of iron supplementation in adolescent secondary school girls in Malaysia: preliminary findings. In: Towards achieving nutritional goals of the Asian region. Abstracts of the 7th Asian Congress of Nutrition. Beijing: Chinese Nutrition Society, 1995: 127.

38. Gross R. Schultink W. Juliawati. Treatment of anaemia with weekly iron supplementation (letter). Lancet 1994; 344: 821.

39. UNICEF/WHO World Summit for Children. Strategic approach to operationalizing selected end-decade goals: reduction of iron deficiency anaemia. Geneva: World Health Organization, 1994.

40. Food and Agriculture Organization. Major issues for nutrition strategies. Rome: Food and Agriculture Organization, 1992.

41. United Nations Administrative Committee on Coordination/Subcommittee on Nutrition. Controlling iron deficiency anemia. State of the art series, Nutrition policy discussion paper. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, 1991.

42. United Nations Administrative Committee on Coordination/Subcommittee on Nutrition. Second report of the world nutrition situation. Global and regional results. Geneva: United Nations Administrative Committee on Coordination/Subcommittee on Nutrition, 1992.

43. Angeles IT, Gross R. Schultink WJ, Soemilah S. Is there a long-term effect of iron supplementation on anaemia alleviation? (letter) Am J Clin Nutr 1995; 62: 440-1.