Anemia among Adolescent and Young Women in Low-and-Middle-Income Countries

Objective: Anemia is a global public health problem that affects maternal and infant mortality as well as human capital development. Yet there is not much research on anemia among young women in low-and-middle-income countries with nationally representative samples. The aim of the current research is to assess the extent of anemia in a critical age group: adolescents and young adults ages 15 to 24. Methods: The data are from 34 Demographic and Health Surveys and are used to describe the prevalence of anemia among pregnant and non-pregnant women by age, rural/urban residence, and household wealth. Anemia was assessed using the HemoCue® blood hemoglobin testing system. Findings: The prevalence of anemia among young women ranges from 15% to over 50%. This is substantially higher than 5%, which is the cutoff to identify a population where anemia is a public health problem. African countries show the highest prevalence of anemia; Benin, Ghana and Mali have over 60% anemia prevalence. Moreover, the prevalence of moderate to severe anemia is particularly high in African countries, over 20% in Ghana and Guinea. Our results show that anemia is a public health concern for adolescents and young adult females in all 34 countries we analyzed. Conclusion: The high prevalence of anemia among youth is alarming. Considering the importance of the adolescent and young adult years, when human capital development is consolidated and family formation begins, these findings call for interventions to redress the problem of anemia.


INTRODUCTION
Anemia is a condition in which the number of red blood cells or their oxygen-carrying capacity is insufficient to meet physiologic needs [1].The most common cause is iron deficiency; other causes are vitamin B12 deficiency, chronic inflammation, parasitic infections, and genetic hemoglobin disorders [1].Asia and, particularly, Africa are regions with higher prevalence of anemia [2].Fifty-six million pregnant women and 468 million non-pregnant women are affected globally [2].Determinants of the prevalence and distribution of anemia in a population involve a complex interplay of political, ecological, social, and biological factors [3].Nutritional iron deficiency is due to a low-iron diet [4].Gastrointestinal parasites are also common in many developing countries [4].Hookworm causes intestinal blood loss and is an important cause of iron-deficiency anemia [4].HIV and malaria are other factors associated with anemia.One Ghana-based study found that pregnant women with malaria and HIV co-infection are at higher (twice) risk of anemia compared to pregnant women with one infection or none [5].
*Address correspondence to this author at the Johns Hopkins University at 533 Mergenthaler Hall, 3400 N. Charles Street Baltimore, MD 21218, USA; Tel: 410-516-7626; Fax: 410-516-7590; E-mail: suzumi@jhu.eduAnemia is a major factor in both maternal and infant mortality.Anemic mothers are more likely to experience fetal loss or stillbirth, preterm birth, and low birth weight [6,7].Additionally, maternal anemia is a strong predictor of anemia in children [3].Anemia is much more than a health problem, however.Without sufficient iron, brain development is significantly impaired, adversely affecting learning and cognition [8].Maternal anemia also deters a mother's interactions with her children; mothers are less responsive to their infants [9].Anemia reduces workers' productivity [10,11], particularly in developing countries where there is a predominance of physical labor [12].Because of these negative consequences the WHO classifies populations into four groups with respect to anemia: 1) no public health concern (prevalence <=4.9%); 2) a mild public health concern (5-19.9%);3) a moderate public health concern (20.0% to 39.9%); and 4) a severe public health concern (prevalence >=40.0 %) [13].
The adolescent period requires higher iron intake because of the body's rapid growth [14].Although growth slows down after menarche, iron loss from menstruation must be countered by further high iron intake for young women [14].Thus, the risk of anemia increases during adolescent years with the onset of menstruation and pregnancy [4].The negative impacts of anemia on maternal and infant health and on worker productivity also brings adolescence and young adulthood to the foreground, since these are the years when young women complete schooling, begin working, and start their families.Existing research on anemia has focused on infants, children, or pregnant women, however; or, in some studies, all women of reproductive age [15][16][17][18][19]. Yet women's biology and life experiences differ dramatically between ages 15 and 49.This research gap is particularly troubling considering the current emphasis on the importance of preconception care for women to improve health outcomes of mother and children.For all these reasons, a study of anemia in the adolescent and young adult years is particularly important.Despite its significance, there has yet to be specific analyses examining the prevalence and severity of anemia by age and region among adolescent and young women.In the present analyses we describe the prevalence of anemia among young women 15-24 years of age from 34 Demographic and Health Surveys (DHS) countries where the data were collected between 2003 and 2011.In addition, we explore health inequalities and anemia by age, residence, and household wealth.

METHODS
We use the Demographic and Health Surveys (DHS), which are "nationally-representative household surveys that provide data for a wide range of indicators in the areas of population, health, and nutrition," [20] usually conducted every five years, for our analysis.The DHS uses a stratified two-stage cluster sampling design.We use 34 DHS surveys conducted since 2000 with information on anemia.The countries represented include 20 from sub-Sahara African, 2 from North Africa and Middle East, 4 from West Asia and Europe, 4 from South and Southeast Asia, and 4 from Latin America.Almost all these surveys have a response rate of at least 90% [21].We use data from women 15 to 24 who were interviewed and tested for anemia.
Our outcome is anemia status.DHS staff tested women for anemia with the HemoCue® blood hemoglobin testing system [22].Testing is voluntary.DHS defines mild anemia among women as between 10.0 and 11.0 grams of hemoglobin per deciliter (g/dl).Women with 7.0 to 9.9 g/dl have moderate anemia; and those with less than 7.0 g/dl of hemoglobin have severe anemia.For pregnant women, a hemoglobin level below 12 g/dl is considered anemic.We created two dichotomous variables where "0" denotes no anemia and "1" any anemia and, secondly, where "0" denotes no anemia and mild anemia and "1" denotes moderate or severe anemia based on the DHS definition.
Age is a categorical variable distinguishing between young women who are 15-19 and 20-24 years of age.Residence is a dichotomous variable distinguishing between urban and rural.DHS provides household wealth quintiles of the household based on the wealth index.The wealth index is calculated using principal components analysis "on a household's ownership of selected assets, such as televisions and bicycles; materials used for housing construction; and types of water access and sanitation facilities" [23].
We report the prevalence of anemia in each of the countries and then the prevalence of moderate to severe anemia.In addition we present the prevalence by world region, age, urban residence, and household wealth.Our estimates are weighted and adjusted for the complex survey design using Stata11.

RESULTS
In Figure 1 and Tables 1a through 1f we present estimates of anemia prevalence for all the countries in our analysis by region.Figure 1 shows the prevalence of anemia among young women from 34 DHS countries.No country has prevalence lower than 5%, so anemia is a public health concern for all 34 countries.
For Sub-Saharan African countries (Table 1a), we have a sufficient number of countries to also present aggregate statistics (Table 1b).In Table 1c to Table 1f we present data on North African and Middle East, West Asian and European, South and Southeast Asian, and Latin American countries.
In Sub-Saharan Africa (Table 1a), the lowest prevalence of anemia observed was in Ethiopia, with a prevalence of lower than 15%.In sub-Saharan Africa, 12 countries in the present analyses had prevalence of anemia at or above 40%.It is striking that in Table 1a the inequalities by age, residence, and wealth are small and not consistent.In North Africa and Middle East and in West Asia and European countries (Tables 1c and  d) the levels of anemia are lower but still rise to the level of public health problem.Just as in Sub-Saharan Africa, wealth inequalities for anemia are not particularly notable.
Table 1e contains the results for four countries in South and Southeast Asia.The table shows that the prevalence of anemia in India and Cambodia is considerably higher than in Nepal and Timor-Leste.While the inequalities by age and residence are not notable, the inequalities by wealth are more substantial than in countries in Sub-Saharan Africa and in West Asia and European countries.For example, the prevalence of anemia among the highest wealth quintile in India is over 46 percent, which while very high is twenty points lower than that in the lowest quintile, where two-thirds of the population is anemic.
In Table 1f we present data for the four Latin American countries.For Bolivia and Guyana the patterns are similar to those in the other regions: very high prevalence with few age and residence inequalities, some inequalities by household wealth, but unacceptably high levels even among young people in the wealthiest households.Honduras is similar to these other two Latin American countries with little variance by wealth.
Haiti is unique among the countries we examined because the inequalities by wealth are not as seen in other countries.Specifically, rather than the poorest quintile with the highest prevalence of anemia, in Haiti it is the wealthiest quintile.And rather than the rural population, in Haiti the urban population of young women has the highest prevalence of anemia.
When we turn to more severe anemia, we see results that parallel those for anemia in general.Tables 2a through 2f are analogous to the tables we just discussed, but they contain information on moderate to severe anemia.Overall, the patterns are similar to those described.The inequalities by wealth, however, in Sub-Saharan Africa are more notable, where the median prevalence is 19.3% for the lowest wealth quintile and 13.05% for the highest wealth quintile.

DISCUSSION
Our study has limitations.The DHS does not include an anemia assessment in all countries, thus leaving large gaps in our understanding of prevalence worldwide.Use of population-based data is not especially revealing about the mechanisms through which anemia is produced, which undoubtedly vary in different settings.For example, the relatively low prevalence of anemia in Ethiopia might be due to the widespread consumption of "teff", an iron-rich food [24].In addition, the countries in our analysis differ in withinhousehold patterns of food allocation; in some countries women may eat last [25,26].Such customs could lead to anemia among young women even if the conventional diet is adequate.Nevertheless, the external generalizability of the data we use and its geographic scope are the compelling strengths of our analysis, as is our focus on a population that is most of interest for preconception care and which has been neglected in past research.
The prevalence of anemia among young women between ages 15 and 24 is a public health concern in all 34 countries we examined.Moreover, anemia is a severe public health concern in sixteen countries: the highest prevalence rate is 61.52% of young women in Ghana.While there are age and residence inequalities in many countries, the differences are not large.Household wealth inequalities are somewhat larger but still indicate high levels of anemia among the most advantaged households.Interventions, therefore, should be at the population level and not targeted at any specific group.
Iron supplementation is the most common mechanism for preventing and treating iron-deficiency anemia [22].It is a cost-effective intervention, which makes it attractive [14], although a barrier to this intervention is compliance problems [4] due to side effects such as nausea [14].Pregnant women and infants and children are often the target of iron supplementation [22], but supplementation to female adolescents generally has been shown to be effective in reducing anemia [27].
Iron supplementation might not be the most effective intervention everywhere, however.In areas with high HIV and malaria prevalence, controlling malaria and HIV is also important to reduce anemia [5].Similarly, in countries where intestinal worms are common, it may be necessary to reduce infection since it exacerbates malnutrition and worsens anemia [28].For these countries, deworming is "an often overlooked intervention for improving nutritional status" [30].
Iron fortification is another intervention.One research found that iron fortification is more cost effective than iron supplementation [29].The infrastructure such as large rural population and the technology for food fortification could be an issue.Moreover, fortified food might not be affordable for all.
Our findings that anemia is not particularly concentrated among the most disadvantaged young women makes a population-level intervention especially attractive.Our findings show that the prevalence of anemia is high enough to be an alarming public health concern for young women among all 34 countries, and this has serious negative repercussions for human capital development and population health.They indicate that the focus of past research-on women of the full range of reproductive age and on children-is insufficient, and the unique circumstances of women at different stages of reproductive age ought not to be ignored.As Stoltzfus and colleagues said: "The control of iron deficiency in women and children requires a comprehensive strategy that is based on a lifecycle approach to the problem" [30].