1. Objectives

To assess the effects of daily oral iron supplements, either alone or in conjunction with folic acid or other micronutrients, as an antenatal public health intervention in pregnant women

2. How studies were identified

The following databases were searched to January 2015:

• CENTRAL (The Cochrane Library)
• EMBASE
• CINAHL
• MEDLINE

Relevant journals and conference proceedings were handsearched, and relevant organizations were also contacted to identify any ongoing and unpublished studies

3. Criteria for including studies in the review

3.1 Study type

Randomized and quasi-randomized trials, including cluster-randomized trials

3.2 Study participants

Pregnant women of any gestational age and parity

3.3 Interventions

Daily oral iron supplementation, alone or in combination folic acid or other micronutrients, in comparison to no treatment or placebo, or versus the same micronutrients without iron

(Studies using iron supplied from food-based interventions were excluded, as were studies using iron as a medical treatment, and studies using intermittent, i.e., not daily, supplementation regimens)

3.4 Primary outcomes

Infant

• Low birth weight (<2500 g)
• Birth weight (g)
• Preterm birth (<37 weeks’ gestation)
• Neonatal death (≤28 days post-delivery)
• Congenital anomalies, including neural tube defects

Maternal

• Anaemia at term (haemoglobin <110 g/L at ≥37 weeks’ gestation)
• Iron deficiency at term (as defined by trialists, based on any indicator of iron status at ≥37 weeks’ gestation)
• Iron deficiency anaemia at term (as defined by trialists, at ≥37 weeks’ gestation)
• Death (death while pregnant or within 42 days of termination of pregnancy)
• Side effects
• Severe anaemia at any time during second or third trimesters (haemoglobin <70 g/L)
• Clinical malaria (as defined by trialists)
• Infection during pregnancy

Secondary outcomes for the infant included very low birth weight (<1500 g), very premature birth (<34 weeks’ gestation), haemoglobin concentration in the first six months of life, ferritin concentration in the first six months of life, development and motor skills, and admission to a special care unit. Secondary outcomes for the mother included anaemia at or near term (hemoglobin <110 g/L at ≥34 weeks’ gestation), iron deficiency at or near term, iron deficiency anaemia at or near term (haemoglobin <110 g/L and at least one additional laboratory indicator at ≥34 weeks’ gestation), haemoglobin concentration at or near term, haemoglobin concentration within six weeks postpartum, high haemoglobin concentrations at any time during second or third trimester (haemoglobin >130 g/L), high haemoglobin concentrations at or near term (haemoglobin >130 g/L at ≥34 weeks’ gestation), moderate postpartum anaemia (haemoglobin between 80 and 109 g/L), severe anaemia at or near term (haemoglobin <70 g/L at ≥34 weeks’ gestation), severe postpartum anaemia (haemoglobin <80 g/L), puerperal infection, antepartum haemorrhage, postpartum haemorrhage, transfusion given, diarrhoea, constipation, nausea, heartburn, vomiting, wellbeing/satisfaction, placental abruption, premature rupture of membranes, and pre-eclampsia

4. Main results

4.1 Included studies

Sixty-one trials were included in this review, with 44 trials involving 43,274 women contributing data to quantitative analyses:

• Twenty-four trials specifically recruited non-anaemic women; five studies recruited women with mild and moderate anaemia, but these did not contribute data to the review; the remaining trials included mixed samples, with eight of these studies excluding women with severe anaemia
• In 13 trials supplementation was initiated after 20 weeks’ gestation; in 16 trials supplementation was initiated before or after 20 weeks’ gestation, or was not specified; and in the remaining trials supplementation began before 20 weeks’ gestation
• Doses of elemental iron ranged from 9 mg to 900 mg per day. Ferrous sulphate was used in 36 trials, ferrous fumarate was used in six trials, ferrous gluconate was used in six trials, two trials used ferrous betainate hydrochloride, and one trial each used haeme iron from porcine blood, ferritin, chelated iron aminoates, ferrous iron, and iron EDTA
• Twenty-three trials included folic acid supplementation; doses ranged from 0.01 mg to 5 mg of folic acid per day; some studies used co-interventions in addition to iron alone or iron plus folic acid, such as other micronutrients, deworming agents, and chloroquine

4.2 Study settings

• Australia (3 trials), Belgium, Canada, China (6 trials), Denmark, Ecuador, Finland, France, the Gambia, Indonesia, Iran (4 trials), Ireland, Italy, Jamaica, Myanmar, Nepal, the Netherlands (2 trials), Niger, Nigeria, Norway (2 trials), the Philippines, South Africa, Republic of Korea, Sweden (2 trials), Thailand, Turkey, the United Kingdom (12 trials), the United States of America (8 trials), and Viet Nam
• In the majority of studies, the intervention was delivered in hospital or community-based antenatal clinics by healthcare professionals. In eight studies the intervention was delivered by community workers, traditional birth attendants or village-based healthcare staff, and supplements were provided during visits to women’s homes or in local community settings
• Twenty-three studies were conducted in settings considered to have some risk of malaria, although only two trials reported on malaria outcomes

4.3 Study settings

How the data were analysed
Eight comparisons were planned: i) any supplements containing iron versus the same supplements without iron or no treatment/placebo; ii) any supplements containing iron plus folic acid versus the same supplements without iron or folic acid; iii) supplementation with iron alone versus no treatment/placebo; iv) supplementation with iron plus folic acid versus no treatment/placebo; v) supplementation with iron plus folic acid versus folic acid alone; vi) supplementation with iron plus other micronutrients versus the same micronutrients without iron; vii) supplementation with iron plus folic acid plus other micronutrients versus folic acid plus the same micronutrients without iron; and viii) supplementation with iron plus folic acid plus other micronutrients versus the same micronutrients without iron plus folic acid. Results were pooled in meta-analysis using random effects models, providing average relative risks (RR) for dichotomous data and average mean differences (MD) for continuous data, with corresponding 95% confidence intervals (CI). The following subgroup analyses were planned to investigate potential sources of heterogeneity:

• By gestational age at initiation of supplementation: <20 weeks’, ≥20 weeks’, mixed/unspecified
• By anaemia status at initiation of supplementation: anaemic, non-anaemic, mixed/unspecified
• By dose of iron: ≤30 mg/day, >30 to <60 mg/day, ≥60 mg/day
• By type of iron formulation: slow release, non-slow release/not specified
• By iron compound bioavailability in comparison to ferrous sulphate: higher (iron EDTA), equal or lower (all other compounds)
• By malaria risk: malaria risk-free settings, settings with some malaria risk

Results
Any supplements containing iron versus the same supplements without iron or no treatment/placebo (44 studies)
Infant primary outcomes
In 11 trials including 17,613 women, the risk of having a low birth weight infant <2500 g was reduced among those receiving iron, although the results were not statistically significant (RR 0.84, 95% CI [0.69 to 1.03], p=0.091). The same was found for birth weight, where infants born to mothers receiving iron were non-statistically significantly 23.8 g heavier than those born to mothers receiving no treatment or placebo (95% CI [-3.0 to 50.5 g], p=0.082; 15 trials/18,590 participants). The risk of preterm birth, neonatal death, and congenital anomalies were lower with iron supplementation, but results did not reach statistical significance (preterm birth: RR 0.93, 95% CI [0.84 to 1.03], p=0.19; 13 trials/19,286 women; neonatal death: RR 0.91, 95% CI [0.71 to 1.18], p=0.48; 4 trials/16,603 women; congenital anomalies: RR 0.88, 95% CI [0.58 to 1.33], p=0.54; 4 trials/14,636 women).

Additional outcomes for the infant
The risk of very premature birth <34 weeks’ gestation was almost halved with maternal iron supplementation (RR 0.51, 95% CI [0.29 to 0.91], 5 trials/3743 women). Infant ferritin concentrations at six months of age were higher with maternal iron supplementation in one study of 197 participants (MD 11.0 μg/L, 95% CI [4.4 to 17.6]). No difference between groups was reported for the outcomes very low birth weight; low Apgar score at five minutes; mean infant haemoglobin levels at three and six months; admission to special care; head circumference at birth; and stunting at long-term follow-up.

Maternal primary outcomes
Women receiving a daily iron supplement were 70% less likely to be anaemic at term (RR 0.30, 95% CI [0.19 to 0.46], p<0.00001; 14 trials/2199 women), although heterogeneity was high (I²=80%). In subgroup analysis, the effect size was greater in areas without malaria risk (RR 0.18, 95% CI [0.10 to 0.34], 11 trials/1669 women). The risks of iron deficiency at term (RR 0.43, 95% CI [0.27 to 0.66], p=0.00015; 7 trials/1256 women) and iron deficiency anaemia at term (RR 0.33, 95% CI [0.16 to 0.69], p=0.003; 6 trials/1088 women) were reduced among women taking iron supplements. No statistically significant difference was found between treatment groups in risk of maternal death (RR 0.33, 95% CI [0.01 to 8.19], p=0.5; 2 trials/12,560 women), any side effects (RR 1.29, 95% CI [0.83 to 2.02], p=0.26; 11 trials/2423 women), or severe anaemia during the second or third trimesters (RR 0.22, 95% CI [0.01 to 3.20], p=0.27; 3 trials/786 women). In addition, there were no differences between groups in the risk of infection during pregnancy (1 trial), or in the risks of placental malaria and parasitaemia (2 trials).

Additional outcomes for the mother
Women receiving iron had higher haemoglobin concentrations at term (MD 8.88 g/L, 95% CI [6.96 to 10.80], 19 trials/3704 women) and within six weeks postpartum (MD 7.61 g/L, 95% CI [5.50 to 9.72], 7 studies/956 women), but were at increased risk of having high haemoglobin concentrations (>130 g/L) during the second and third trimester (RR 2.37, 95% CI [1.34 to 4.21], 9 studies/2188 women) and at term (RR 3.07, 95% CI [1.18 to 8.02], 8 trials/2156 women). Severe anaemia at or near term was not statistically significantly different between groups in two trials of 494 women, while in two trials of 553 women the risk of severe postpartum anaemia was reduced by 96% (RR 0.04, 95% CI [0.01 to 0.28]). There was also a reduced risk of puerperal infection with iron supplementation (RR 0.68, 95% CI [0.50 to 0.92], 4 trials/4374 individuals). No differences between treatment groups were observed for the following outcomes: provision of transfusion, ante- or postpartum haemorrhage, individual side effects, maternal wellbeing, placental abruption, preterm rupture of the membranes, pre-eclampsia, and moderate anaemia in the postpartum period.

Any supplements containing iron plus folic acid versus the same supplements without iron or folic acid (8 studies)
Infant primary outcomes
There was no significant difference in the risk of low birth weight between infants born to women taking iron plus folic acid versus those who received no treatment or placebo in two trials including 1311 women (RR 1.07, 95% CI [0.31 to 3.74]); however, infant birth weight was greater by 57.7 g (95% CI [7.7 to 107.8 g]). No difference between treatment groups was evident for the risk of preterm birth (RR 1.55, 95% CI [0.40 to 6.00], 3 trials/1497 women), neonatal death (RR 0.81, 95% CI [0.51 to 1.30], 3 trials/1793 participants), or congenital anomalies (RR 0.70, 95% CI [0.35 to 1.40], 1 trial/1652 women).

Additional outcomes for the infant
No statistically significant differences between treatment groups were found for the outcomes very low birth weight <1500 g, very premature delivery, or admission to a special care unit.

Maternal primary outcomes
Women receiving iron plus folic acid were less likely to be anaemic at term (RR 0.34, 95% CI [0.21 to 0.54], 3 trials/346 women), less likely to have iron deficiency at term (RR 0.24, 95% CI [0.06 to 0.99], 1 trial/131 women), and less likely to have severe anaemia during the second or third trimester (RR 0.12, 95% CI [0.02 to 0.63], 4 trials/506 women) than women who did not receive iron plus folic acid. No statistically significant difference in maternal iron deficiency anaemia at term was found between treatment groups in one trial including 131 women (RR 0.43; 95% CI [0.17 to 1.09]), and no differences were reported for the outcomes infection in pregnancy and maternal deaths. The risk of any side effects was higher among women receiving iron plus folic acid supplementation in one trial of 456 women, with no side effects being reported in the control group (RR 44.32, 95% CI [2.77 to 709.09]).

Additional outcomes for the mother
Maternal haemoglobin concentrations were greater at term among those receiving iron plus folic acid in comparison to those receiving no treatment or placebo (MD 16.13 g/L, 95% CI [12.74 to 19.52], 3 trials/140 women), and remained greater within six weeks postpartum (MD 10.07 g/L, 95% CI [7.33 to 12.81], 2 trials/459 women). No statistically significant differences were found between treatment groups for the outcomes maternal high haemoglobin at term or during the second or third trimesters and maternal severe anaemia at or near term.

Supplementation with iron alone versus no treatment or placebo (33 studies)
Infant primary outcomes
The risk of delivering a low birth weight infant was not significantly different between women receiving iron alone and those receiving no treatment or placebo (RR 0.63, 95% CI [0.30 to 1.32], 6 trials/1136 women). In addition, birth weight was not different between treatment groups (MD -1.0 g, 95% CI [-78.8 to 76.7], 9 trials/1331 women). In six trials including 1713 women, the risk of preterm birth was not different between treatment groups (RR 0.82, 95% CI [0.58 to 1.14]), and in two trials of 2402 women, the risk of congenital anomalies did not differ significantly between groups (RR 0.86, 95% CI [0.55 to 1.35]).

Additional outcomes for the infant
Maternal treatment with iron alone led to an increase in infant ferritin concentrations in the first six months of life relative to controls (MD 11.0 μg/L, 95% CI [4.4 to 17.6], 1 trial/197 women). No clear evidence of a difference between treatment groups was observed for the outcomes very premature birth <34 weeks’ gestation, very low birth weight <1500 g, infant haemoglobin concentrations in the first six months, or admission to a special care unit.

Maternal primary outcomes
The risks of maternal anaemia at term (RR 0.29, 95% CI [0.19 to 0.47], 14 trials/2136 women), iron deficiency at term (RR 0.43, 95% CI [0.27 to 0.66], 7 trials/1256 women), and iron deficiency anaemia at term (RR 0.33, 95% CI [0.16 to 0.69], 6 trials/1088 women) were all statistically significantly reduced among those taking iron supplements in comparison to no treatment or placebo. However, the risk of severe anaemia during the second or third trimester was not reduced (RR 0.75, 95% CI [0.02 to 29.10], 2 trials/466 women), and the risk of any side effects was greater, although this did not reach statistical significance (RR 1.59, 95% CI [1.00 to 2.52], 9 trials/1677 women).

Additional outcomes for the mother
Maternal iron supplementation in comparison to no treatment or placebo resulted in greater haemoglobin concentrations at or near term (MD 9.0 g/L, 95% CI [6.4 to 11.5], 16 trials/1851 women), and within six weeks postpartum (MD 7.3 g/L, 95% CI [4.8 to 9.7], 6 trials/659 women). However, the risk of having high haemoglobin concentrations at any time during the second or third trimester (RR 1.90, 95% CI [1.07 to 3.35], 7 studies/1146 women) and at or near term (RR 3.80, 95% CI [1.74 to 8.28], 7 studies/1189 women) was increased. No significant differences between treatment groups were found for the outcomes provision of transfusion, maternal wellbeing, diarrhoea, placental abruption, pre-eclampsia, moderate anaemia at postpartum, maternal severe anaemia at postpartum, puerperal infection, ante- and postpartum haemorrhage, constipation, nausea, heartburn, or vomiting.

Supplementation with iron plus folic acid versus no treatment or placebo (8 studies)
Infant outcomes
Infants born to women who received iron plus folic acid supplements were not at a reduced risk of low birth weight <2500 g in comparison to controls (RR 1.07, 95% CI [0.31 to 3.74], 2 trials/1311 participants), but were 57.7 g heavier at birth (95% CI [7.7 to 107.8 g], 2 trials/1311 participants). Preterm birth was not significantly different between treatment groups (RR 1.55, 95% CI [0.40 to 6.00], 3 trials/1410 participants), and nor was the risk of neonatal death (RR 0.81, 95% CI [0.51 to 1.30], 3 trials/1793 women) or congenital anomalies (RR 0.70, 95% CI [0.35 to 1.40], 1 trials/1652 participants). There was no evidence of a significant difference between treatment groups for the following secondary outcomes: very low birth weight, very premature birth, and admission to special care unit.

Maternal outcomes
Maternal anaemia (RR 0.34, 95% CI [0.21 to 0.54], 3 studies/346 women) and iron deficiency at term (RR 0.24, 95% CI [0.06 to 0.99], 1 trial/131 women) were statistically significantly less likely with iron plus folic acid supplementation, but iron deficiency anaemia at term was not (RR 0.43, 95% CI [0.17 to 1.09], 1 trial/131 women). The risk of severe anaemia during the second or third trimester was reduced among the treatment group (RR 0.12, 95% CI [0.02 to 0.63], 2 trials/365 women). A greater number of side effects were reported among the treatment group in comparison to controls in one trial of 456 women (RR 44.32, 95% CI [2.77 to 709.09]), with no side effects reported in the control group. Maternal anaemia at or near term (RR 0.34, 95% CI [0.21 to 0.54], 3 trials/346 women), iron deficiency at or near term (RR 0.24, 95% CI [0.06 to 0.99], 1 trial/131 women), haemoglobin concentrations at or near term (MD 16.13 g/L, 95% CI [12.74 to 19.52], 3 trials/140 women), and haemoglobin concentrations within six weeks postpartum (MD 10.07 g/L, 95% CI [7.33 to 12.81], 2 studies/459 women) were all improved with iron plus folic acid supplementation. No evidence of a significant difference was found between women who received iron plus folic acid supplements and those receiving no treatment or placebo for other secondary outcomes.

Supplementation with iron plus folic acid versus folic acid alone (5 studies)
Infant outcomes
In four trials including 16,146 participants, the risk of low birth weight was reduced by 12% with iron plus folic acid supplementation versus folic acid alone, although this did not reach statistical significance (RR 0.88, 95% CI [0.78 to 1.00]), and there was no difference in mean birth weight (MD 19.5 g, 95% CI [-6.9 to 45.9]). No evidence of a difference between treatment groups was observed for preterm birth (RR 0.97, 95% CI [0.87 to 1.08], 4 trials/16,146 participants), neonatal death (RR 0.91, 95% CI [0.71 to 1.18], 4 trials/16,603 participants), or congenital anomalies (RR 0.78, 95% CI [0.44 to 1.39], 2 studies/13,586 women). No differences were found between groups for the secondary outcomes very premature birth and very low birth weight.

Maternal outcomes
Anaemia at term was less likely among women receiving iron plus folic acid in comparison to those receiving folic acid alone (RR 0.34, 95% CI [0.21 to 0.55], 2 studies/303 women), as was severe anaemia during the second and third trimesters (RR 0.06, 95% CI [0.01 to 0.47], 1 trial/74 women). No difference between groups was found for risk of side effects (RR 1.10, 95% CI [0.55 to 2.23]) or infection during pregnancy (RR 1.21, 95% CI [0.33 to 4.46]) in one trial of 727 women. Maternal haemoglobin concentrations were greater at or near term with iron supplementation (MD 12.44 g/L, 95% CI [0.95 to 23.93], 2 trials/771 women). Although the risk of high haemoglobin concentrations at or near term was not significantly increased (RR 1.87, 95% CI [0.32 to 10.84], 2 trials/967 women), the risk during pregnancy was more than four times greater (RR 4.33, 95% CI [2.26 to 8.30], 2 trials/1042 women). No evidence of a difference between treatment groups was found for other secondary outcomes.

Supplementation with iron plus other micronutrients versus same micronutrients without iron (3 studies)
Infant outcomes
No evidence of a statistically significant difference between groups was found for low birth weight (RR 0.51, 95% CI [0.22 to 1.15], 1 trial/334 infants) or preterm birth (RR 0.66, 95% CI [0.41 to 1.04], 2 trials/1127 infants). However, mean birth weight was greater among infants born to women receiving iron (MD 55.7 g, 95% CI [3.4 to 108.0], 2 trials/1116 infants).

Maternal outcomes
Haemoglobin concentrations were higher among iron-supplemented women iron at term (MD 10.85 g/L, 95% CI [7.29 to 14.42], 2 trials/809 women), and within six weeks postpartum (MD 14.00 g/L, 95% CI [3.56 to 24.44], 1 trial/27 women). Although there was no difference between groups in the risk of constipation, vomiting or heartburn, less women receiving iron experienced diarrhoea in one trial of 188 women (RR 0.53, 95% CI [0.29 to 0.96]).

Supplementation with iron plus folic acid plus other micronutrients versus folic acid plus the same micronutrients without iron
None of the included trials reported on this comparison.

Supplementation with iron plus folic acid plus other micronutrients versus the same micronutrients without iron plus folic acid
None of the included studies reported on this comparison.

5. Additional author observations*

In general, the methodological quality of the included trials was mixed, with more than half of studies failing to report allocation concealment. Lack of blinding in more than a third of trials and attrition are also potential sources of bias in this review. However, the trials were conducted in a wide variety of settings and included both anaemic and non-anaemic women, increasing the applicability of the findings.

Overall, maternal iron supplementation reduced the risk of having a low birth weight infant or a preterm birth and increased the average birth weight of infants, although these findings did not reach statistical significance. Very preterm birth before 34 weeks’ gestation was lower with iron supplementation, and women taking iron were less likely to be anaemic, iron deficient or have iron-deficiency anaemia at term. The risk of having high haemoglobin concentrations during the second and third trimesters was increased with iron supplementation, and although not statistically significant, the risk of any side effects was also greater. Iron supplementation during pregnancy may be used as a preventive strategy to improve maternal and infant outcomes, particularly in malaria-endemic areas, although the effect may vary depending on the background risk for low birth weight and anaemia in the population.

Future research investigating the effects of providing other micronutrients in conjunction with iron and folic acid on maternal and infant outcomes is warranted, as is the assessment of novel iron compounds for use in prenatal public health programmes.