SCOPE OF THE PROBLEM
Physiologically, ID occurs when the body's requirements for iron exceed the amount of iron absorbed from the diet. Requirements are raised by rapid increases in body mass (such as in pregnant women and young children) or by high losses of iron (as through menstruation or hookworm infection). Iron absorption is low (∼5%) from plant-based diets (which are common in developing countries) because of factors that inhibit iron uptake (such as phytates and polyphenols), and iron absorption is higher (∼15%) from diets containing more meat and fish (which are more common in developed countries).3 In populations, ID is related to factors such as socioeconomic status, food insecurity and food quality, genetic background, and infectious disease burden. Data from prevalence studies on anemia or ID are useful forassessing the extent of the problem and can provide valuable information on risk factors, associated conditions, and relative risks of vulnerable population groups. Determination of the underlying risk factors is useful for identifying opportunities and developing strategies to address ID. An important limitation is that developing countries often only report anemia data, not indicators of iron status. Although more people are affected by ID than by IDA, many cases of anemia are not due to ID; rather, they are due to other causes such as nutritional deficiencies other than iron (vitamin B12, folic acid, vitamin A, selenium), genetic traits (hemoglobinopathies such as sickle-cell or thalassemia), renal insufficiency, or chronic inflammation. Therefore, anemia is not a good indicator of iron status, especially in some populations, e.g., with endemic malaria or high prevalence of hemoglobinopathies. Reporting only anemia data hampers the identification of risk factors for ID and the assessment of interventions.
Based on currently available prevalence data, the highest burdens of anemia and ID are found in Africa, the Middle East, Asia, and the Western Pacific.1 In Africa, the prevalence of anemia in WRA is estimated to be approximately 47%, and in pregnant women it is around 57%,1 although prevalence rates differ widely from country to country.12 Fortunately, several studies have recorded prevalence rates for both anemia and ID, making the data much more informative. In Mali, for example, anemia is present in 47% of pregnant women (hemoglobin [Hb] < 110 g/L), while only 13% of the women have ID (serum ferritin <12 µg/L).13 In this country, infectious diseases are the major contributors to anemia. Among pregnant women in northern Nigeria, 30% are classified as anemic (Hb < 105 g/L) and the major contributing factor to anemia was ID (ferritin <10 µg/L).14 Besides iron, vitamin B12 and folic acid deficiencies were probably also prevalent. In Ghana, anemia (Hb < 110 g/L) was observed in 34% of pregnant women from urban areas, with 16% of the women having ID (ferritin ≤16 µg/L) and only 7.5% having IDA.15 Malaria is a greater risk factor than ID for being anemic; hence, in sub-Saharan Africa, while ID is prevalent, it is not strongly related to anemia because other causes of anemia, such as malaria infection, are also prevalent in vulnerable groups. Therefore, ID should be considered separately from anemia in intervention strategies.
In India, the prevalence of anemia is among the highest in the world, with >50% of pregnant and non-pregnant women being anemic,1 and anemia is estimated to be responsible for 40% of maternal deaths.16 The incidence of low-birth-weight deliveries and perinatal mortality double with Hb concentrations <80 g/L during pregnancy. In rural Bangladesh, more than 30% of nulliparous married women are anemic when entering pregnancy, with 15% being iron deficient and 11% having IDA.17 More than 80% have inadequate iron stores, defined as <500 mg of iron stores. Anemia prevalence in Pakistan is even higher. In a large, prospective, observational study of 1,369 pregnant women, 90.5% were anemic (Hb < 110 g/L).18 Unfortunately, this study did not provide data on iron status. In South Asia it appears that anemia prevalence is very high, and more closely related to ID than in the African studies. However, when using anemia as an outcome, deficiencies of nutrients besides iron should not be forgotten, especially as nutritional status in general is poor in these populations.
In Vietnam, among the 901 women (non-pregnant, pregnant, and post-partum) surveyed in 2001, more than half were anemic (mean Hb value, 111 g/L). The highest rate (62%) of anemia was found in post-partum women.19 In Thailand, of 590 pregnant women surveyed, 14.1% were anemic (Hb < 110 g/L) and 6% had IDA.20 Thalassemia prevalence was high (∼25%) and a significant risk factor for anemia. Hence, in Southeast Asia, anemia and ID are less prevalent than in South Asia, and although ID is the most important nutritional cause of anemia, only around half of the anemia prevalence can be attributed to ID, and genetic traits are an important contributor to anemia in this region.21
ID and anemia are also common in developed countries. For example, in a small study in the United Kingdom around 40% of women of reproductive age were iron deficient (ferritin <10 µg/L) and 70% had ferritin concentrations <20 µg/L.22 In France, ID and IDA were found in almost 30% and 4% of children <2 years of age respectively,23 while in the United States, ID was found in 14.4% of children between 1 and 2 years of age, and in 9.2% of women of reproductive age.24 Worryingly, there appears to be no decrease in the prevalence of ID in children over the last decades, with especially overweight children at risk for ID.25 In these countries with a relatively low prevalence of anemia, ID without anemia is a more important public health problem than anemia.
Given the high prevalence of anemia and ID in the world, the significant contribution of ID to anemia, and the negative effects of anemia and ID on maternal and child health, large-scale interventions are warranted.
WHAT WORKS AND WHAT MIGHT WORK
Many studies have looked at the efficacy of iron supplementation with or without folic acid in pregnancy in terms of anemia, indicators of iron status, maternal health, pregnancy outcome, and neonatal health. Indeed, already in the 1970s, studies were carried out to determine the optimal doses of iron and folic acid for pregnant women.29 From these earlier studies, it was concluded that pregnant women need at least 120 mg iron per day, and that a supplement containing folic acid and vitamin B12 as well as iron increases Hb concentrations more than iron alone. But even in the group of women receiving 240 mg iron/day (in addition to folic acid and vitamin B12), >50% of the women had Hb concentrations <110 g/L at the end of pregnancy. Moreover, there were no effects on birth weight or infant Hb concentrations at 3 months of age.29 Since then, recommended daily iron doses during pregnancy were progressively decreased to be, most often, about 60 mg/day for at least 6 months or 120 mg/day in case of shorter duration of supplementation or when ID or anemia is prevalent.30
A detailed meta-analysis on iron and folic acid supplementation in pregnancy was recently conducted by Pena-Rosas and Viteri.31 Their meta-analyses were comprised of 49 trials with >23,000 women. Data on anemia and IDA were available for 1,108 women from six trials. In women taking iron supplements, 30.7% were still anemic at term, whereas only 4.9% had IDA; for women not receiving iron, these figures were 54.8% and 15.5%, respectively, demonstrating again that anemia is multifactorial and not only due to ID. Interestingly, despite the association of daily or weekly iron supplementation with increased Hb levels in maternal blood before delivery (+8.8 g/L, confidence interval [CI] 6.6–11.1) and reduced risk at term for anemia (relative risk [RR] 0.27, CI 0.17–0.42), ID (RR 0.44, CI 0.27–0.70) and IDA (RR 0.33, CI 0.16–0.69), no significant effects were found on important health outcomes for infants, such as premature delivery, low birth weight (RR 0.79, CI 0.61–1.03), birth weight (+36.1 g, CI -4.8–77.0), perinatal death, or infant Hb concentrations at 6 months of age, and there was no evidence of significant reductions in substantive maternal and neonatal adverse clinical outcomes.
It is important to consider these conclusions. Currently, blanket programs for the supplementation of pregnant women with iron and folic acid are in place in many countries, with the expectation that this will substantially improve maternal and infant health.8 But, apparently, the benefits of supplementation with only iron and folate during pregnancy are not that clear-cut. In their commentary on this in a Cochrane review, Kongnyuy and Van den Broek state that pregnancy outcomes such as preterm delivery, birth weight, and low birth-weight-related perinatal mortality are more likely to be dependent on factors other than anemia and iron status per se and that it may be unrealistic to expect that these outcomes will be affected by iron supplementation.32
Another explanation for the lack of clear effects could be that the number of studies providing data on outcomes such as perinatal mortality was too low to draw consistent findings. Titaley et al.33 favor this explanation through their analysis of demographic data from Indonesia with >40,000 pregnancies, showing that the iron and folic acid supplementation program protected against neonatal death in the first week after birth (RR 0.53, 95% CI 0.36–0.77). Interestingly, this study shows a significant effect of the Indonesian program on infant health (effectiveness), whereas many efficacy trials did not find significant effects. Indeed, they also show that any visit to a health clinic is beneficial, in addition to receiving iron and folic acid. This beneficial effect also holds true in countries with endemic malaria, provided that intermittent malaria treatment is provided as well.34 These findings are in line with an earlier report from China in which the same group of researchers indicated that iron + folic acid supplementation during pregnancy reduced early neonatal mortality by 54% as compared to supplementation with folic acid alone.35 In this study, iron + folic acid had no significant effect on birth weight (+24 g, P > 0.1) as compared to folic acid alone, whereas multiple-micronutrient supplementation had a significant effect on birth weight (+42 g, P < 0.05). Surprisingly, none of the reports from these studies noted the effects of iron and folic acid on maternal mortality. Yet, in the Lancet series on Maternal and Child Undernutrition, iron supplementation during pregnancy is estimated to result in a 23% reduction in the risk of maternal mortality.8 Given the large sample size of the recent meta-analyses, such an effect should have been apparent. It could be that the effect of iron supplementation during pregnancy on reducing maternal mortality is overestimated.
Perhaps, providing only iron (and folic acid) is not enough. Are multiple-micronutrient supplements during pregnancy more effective? It is known that multiple micronutrient deficiencies often coexist. Indeed, deficiency of multiple micronutrients in one individual is more common than deficiency of a single micronutrient36 and, given the multifactorial etiology of anemia, a greater benefit might be expected from multiple-micronutrient supplementation than from iron + folic acid supplementation alone.
In the last decade, results from several large studies have been published on the efficacy of multiple-micronutrient supplementation during pregnancy. However, the results have been conflicting or confusing, partly because studies used different combinations of micronutrients, different amounts of micronutrients, or evaluated different outcomes. In a 2006 Cochrane review of nine trials with >15,000 women, multiple-micronutrient supplementation decreased the prevalence of low-birth-weight (RR 0.83, CI 0.76–0.91) and small-for-gestational-age (RR 0.92, CI 0.86–0.99) babies, as well as maternal anemia (RR 0.61, CI 0.52–0.71).37 However, the effect of multiple-micronutrient supplementation was not different from that of iron + folic acid supplementation, and the authors found no effects on preterm births or perinatal mortality. Hence, there appears to be no additional benefit of multiple-micronutrient supplementation over supplementation with iron + folic acid during pregnancy. However, another more recent meta-analysis using different criteria and including two recently published large trials concluded that prenatal multi-micronutrient supplementation was associated with a significantly reduced risk of low birth weight and with improved birth weight when compared with iron + folic acid supplementation.38
The Maternal Micronutrient Supplementation Study (MMSS) Group also recently conducted a meta-analysis of 12 studies, using raw data from each study, and thus allowing much greater power for detecting differences.39 Another advantage of this last meta-analysis is that most studies (9 of 12) used the same multiple-micronutrient supplement.39 This supplement, known as the United Nations International Multiple Micronutrient Preparation (UNIMMAP), was formulated by a group of experts and contains 15 vitamins and minerals, including 30 mg of iron and 400 µg of folic acid. The sample size of this meta-analysis was much larger than that of the earlier meta-analyses, namely >52,000 pregnancies. However, data on over half of these pregnancies was provided by one study from Indonesia (SUMMIT trial). This meta-analysis shows a small but significant increase in birth weight (+22.4 g, CI 8.3–36.4) and an 11% decrease in the prevalence of low-birth-weight infants (RR 0.89, CI 0.81–0.97). However, subgroup analyses show that this effect occurs especially in women with a body mass index score of >20 kg/m2, with no effect in women with a lower score. Moreover, the meta-analysis shows no beneficial effects on early or late neonatal death (RR 1.23 and CI 0.95–1.59 versus RR 0.94 and CI 0.73–1.23, respectively).40 This latter finding is surprising, as the largest trial included in the analysis (the SUMMIT trial, performed in Lombok, Indonesia), shows a significant, beneficial effect on early infant survival (RR 0.82, CI 0.70–0.95) and post-neonatal mortality (RR 0.70, CI 0.55–0.89).41 In contrast, in the meta-analysis, there was a significant increase in the prevalence of large-for-gestational age babies (RR 1.13, CI 1.00–1.28), raising the question as to whether this could increase the number of obstructed deliveries.42
An introductory article in a recent supplement of the Food and Nutrition Bulletin reporting the findings of the MMSS Group meta-analysis states “. . . The World Health Organization will hopefully make a global recommendation that governments provide multiple micronutrient supplements instead of iron–folic acid . . . ”.43 But how strong is the evidence for such a recommendation at the moment? Given the large sample size of this meta-analysis and the use of raw pooled data, it comes as no surprise that the effect on birth size was statistically significant, even though the effect was only a meager 22 g. Of course, it is possible that tail-effects result in a much more pronounced effect on the prevalence of low-birth-weight than on average-birth-weight deliveries. Indeed, some studies suggest that different combinations of micronutrients have different effects on the birth weight distribution curve, with some combinations (iron and folic acid) especially affecting the low end of the distribution curve, whereas other combinations (multiple micronutrients) shift the whole distribution curve.44 Indeed, in the meta-analysis by the MMSS group, the effect of multiple-micronutrient supplementation during pregnancy appears to shift the whole birth weight distribution curve upwards, rather than having distinct tail effects.42 Compliance in these 12 studies was high, and most studies (8 of 12) started providing supplements before month 4 of pregnancy.
In a study from China, the effects of micronutrient supplementation during pregnancy on outcomes such as birth weight were only significant when supplementation started before week 12 of gestation.45 Neither iron + folic acid nor multiple-micronutrient supplementation had an effect on birth weight when it began later than gestational week 12. Other studies confirm that Hb concentrations early in pregnancy are related to low birth weight in a U-shaped curve. Women between weeks 4 and 8 of pregnancy, with Hb concentrations between 90 and 99 g/L had a 3.27 (CI 1.09–9.77) higher risk for a low-birth-weight baby than the reference category (110–119 g/L), whereas risks for low-birth-weight and preterm birth also increased with Hb concentrations >130 g/L.46 Anemia early in pregnancy appears to be much more strongly related to low birth weight than anemia in the third trimester.47 It is unlikely that these conditions (high compliance, early start of supplementation) can be met by standard national programs in which women are more likely to report to the health system for the first time at around week 16 of pregnancy. Based on the above observations, it can be expected that the effects of supplementation programs for pregnant women, whether providing iron, iron + folic acid, or multiple micronutrients, will be disappointingly small or absent.
So, if women cannot be reached in time during pregnancy to start providing much-needed micronutrients, why not start before pregnancy? Surprisingly, only a few studies have investigated the effects of pre-conception supplementation with iron or multiple micronutrients on maternal or neonatal health. As for pregnant women, deficiencies of more than one micronutrient are also likely in women of reproductive age.48 To meet iron needs during pregnancy, women need an iron reserve of at least 300–500 mg prior to conception,47,49 so as not to become iron deficient after the first trimester. As iron stores are directly related to ferritin concentrations, and presuming that 1 µg/L serum ferritin reflects 140 µg/kg body weight of stored iron, a woman weighing 40–50 kg should have a serum ferritin concentration above approximately 50 µg/L.50 Many women in developing countries will have ferritin concentrations (and hence iron reserves) below this level, and will thus be at risk of becoming iron deficient during pregnancy. Indeed, even in industrialized countries many women fail to enter pregnancy with adequate iron stores. Based on serum transferrin receptor/serum ferritin ratios, 56% of non-pregnant women in the United States (according to the NHANES III survey) had iron stores below 300 mg,47 and <20% of women in Denmark are estimated to have adequate (>500 mg) iron stores before pregnancy.49 Hence, a promising approach could be to improve the iron and folic acid status of women before they become pregnant – first, because it gives a much larger time window for intervention; second, because improving micronutrient status seems to have the greatest effect early in pregnancy; and third, because nutritional status around the time of conception is important. Especially for folic acid, pre-conceptual increases in status have been shown to have a strong effect on the reduction of neural tube defects, with possibly >70% of neural tube defects being prevented by adequate intakes of folic acid.51 For other micronutrients such as iron, the effect of pre-conception status on maternal and child health is less clear, mainly due to a lack of studies in humans.52 A multi-country trial in Southeast Asia is one of the few examples of research investigating the effects of supplementing women of reproductive age with iron and folic acid, and following the women through pregnancy and delivery.53 Earlier studies had shown that weekly supplementation with iron and folic acid improved iron status in adolescent Malaysian54 and Indonesian girls.55 Therefore, with support of the WHO, effectiveness trials were conducted with unsupervised weekly iron + folic acid supplementation for women of reproductive age in Cambodia, Vietnam, and the Philippines. The supplements provided 3.5 mg of folic acid and 60 mg of elemental iron for non-pregnant women and 120 mg of iron + 3.5 mg folic acid for pregnant women. Supplements were free of charge for pregnant women, but sold to non-pregnant women through a social marketing program. In the non-pregnant women, iron status improved significantly over the intervention period. In Vietnam, anemia prevalence decreased from 45% at baseline to <20% after 9 months to 1 year. Interestingly, prevalence of ID (ferritin concentrations <12 µg/L) was much lower than the prevalence of anemia at baseline (∼9%), but it also decreased significantly over the study period. Indeed, even at the first follow-up after 4½ months, when women had taken, on average, 14 tablets, Hb concentration had already improved significantly. But did this have an impact on iron status during pregnancy? The report on this intervention indicates that longer pre-pregnancy supplementation was associated with less anemia and better iron status during the first and second trimesters of pregnancy. Indeed, there was no IDA in the first and second trimesters of pregnancy in women who started taking supplements >3 months before conception.56 Another important observation was that, although anemia prevalence increased in the third trimester of pregnancy, there was almost no severe anemia (Hb < 95 g/L). Severe anemia is directly associated with increased perinatal risk for mothers and newborns. Unfortunately, data on birth weight was available for only 200 infants, but there was a tendency towards higher birth weights (+81 g) in the weekly supplementation group (P = 0.15) and towards a lower prevalence of low birth weight (<2,500 g, 3% and 9%, respectively, P = 0.08).56 In Vietnam, weekly supplementation with iron + folic acid (WIFS) in combination with deworming has been continued as a pilot to improve the iron status of WRA and has been shown to be successful as such. The provision of weekly iron + folic acid supplements for free resulted in significant reductions in the prevalence of anemia (from 38% to 19%) and ID, defined as ferritin <15 µg/L (from 23% to 9%).57 The authors conclude that countries with high rates of anemia in women should urgently consider WIFS and regular deworming if sustainable alternatives (such as fortification and dietary diversification) are not yet feasible. This view is shared by the WHO, as indicated in the following statement from 2009: “. . . in population groups where the prevalence of anemia is above 20% among women of reproductive age and mass fortification programs of staple foods with iron and folic acid are unlikely to be implemented within 1–2 years, WIFS should be considered as a strategy for the prevention of iron deficiency, the improvement of pre-pregnancy iron reserves and the improvement of folate status in some women.”58
Pre-pregnancy improvement of iron status would also mean the iron content of the supplement given during pregnancy can be (much) lower, as an iron deficit would not need to be addressed by the supplement, only the increased requirements. Lower doses of iron will not only benefit the absorption of other minerals, they will also decrease the risk of side effects and intestinal damage.49,59 In Vietnam, weekly supplementation with 60 mg iron before pregnancy and 120 mg iron/week during pregnancy was more effective for improving Hb concentrations during pregnancy than daily supplementation with 60 mg iron/day during pregnancy.56
Unfortunately, at present, there are no published results of studies on the effect of multiple-micronutrient supplementation provided before conception. However, following the same argumentation as above for providing multiple micronutrients instead of iron + folic only to pregnant women, the weekly provision of a multiple micronutrient supplement, such as the UNIMAPP, to women of reproductive age could be a very cost-effective intervention that would improve maternal micronutrient status before pregnancy and allow a better micronutrient status during pregnancy, thereby increasing birth weight, reducing the prevalence of low birth weight, and improving overall maternal and child health.
Besides supplementation, other strategies to improve the micronutrient status of WRA need to be considered. Supplements are often not regarded as a feasible, long-term solution, although in developed countries, many women take vitamin supplements before becoming pregnant. Fortification of (staple) foods is a feasible, long-term solution to improve the micronutrient status of a population, including vulnerable groups. In Vietnam, fortification of fish sauce with iron successfully improved the iron status of WRA and decreased the prevalence of both anemia and ID.60 These results, and the success of other food fortification programs, such as iodized salt or vitamin A-fortified cooking oil, indicate that iron fortification of staple or specific foods, such as condiments, can be important strategies to increase iron stores of the most vulnerable groups of a population such as WRA. Several national iron food fortification programs are ongoing in different countries and their impact on the iron status of WRA will be evaluated in the coming years.
Other strategies might also be promising, and interventions can act synergistically. In a comprehensive review article, Mora and Nestel61 discuss potential policy changes that could improve prenatal nutrition and, thereby, maternal and infant health. They suggest that for adolescent girls essential programs should include access to family planning and reproductive health services, nutritional education, and supplementary foods to improve overall nutritional status. Other strategies are also worth considering; for instance, the impact of a cash-transfer program in Mexico (“Oportunidades”) on birth weight was evaluated and showed a 4.6% reduction in low-birth-weight prevalence among women participating in the study.62