1. Objectives

To examine the effects of daily iron supplementation in menstruating women on anaemia and iron status, and on physical, psychological and neurocognitive health

2. How studies were identified

The following databases were searched in November/December 2015:

• CENTRAL (The Cochrane Library 2015, Issue 10)
• MEDLINE
• EMBASE
• LILACS
• CINAHL
• Conference Proceedings Citation Index - Science
• Science Citation Index
• POPLINE
• WHO Global Health Library
• African Index Medicus
• Western Pacific Region Index Medicus
• Index Medicus for the Eastern Mediterranean Region
• Index Medicus for the South East Asia Region
• WorldCat
• DART-Europe E-theses Portal
• Australasian Digital Theses Program
• Proquest Dissertations and Theses Global
• WHO International Clinical Trials Registry Platform (ICTRP)

Reference lists were also handsearched

3. Criteria for including studies in the review

3.1 Study type

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

3.2 Study participants

Menstruating women, defined as non-pregnant, non-lactating women who had passed menarche and had not yet reached menopause and had no conditions preventing menstruation, irrespective of their baseline iron or anaemia status

(Studies including results disaggregated for women aged 12 to 50 years were included, as were studies in which more than half of all subjects met this age criterion)

(Studies were excluded if they targeted hospitalized or ill people, purely evaluated iron supplement pharmacology, or if they targeted populations with conditions affecting iron metabolism, such as those with inflammatory bowel disease)

3.3 Interventions

Daily oral iron supplements, with or without folic acid and vitamin C, compared with no supplemental iron

(Co-interventions were acceptable providing they were not haematopoietic agents and if both intervention and control arms received the co-intervention)

(Iron compounds included ferrous sulphate, ferrous fumarate, ferrous gluconate, carbonyl or colloidal iron, and may have been delivered as a tablet, capsule, or liquid, on at least five days per week)

3.4 Primary outcomes

• Anaemia (haemoglobin below a study-defined concentration)
• Haemoglobin (g/L)
• Iron deficiency (defined by trialists)
• Iron-deficiency anaemia (defined by anaemia plus iron deficiency by trialists)
• All-cause mortality
• Adverse side-effects (including abdominal pain, vomiting, nausea, heartburn, diarrhoea, constipation)
• Cognitive function (including school grades/performance, intelligence testing, memory, attention, other cognitive domains)

Secondary outcomes included iron status (as reported by trialists), physical exercise performance (VO₂ max/peak, heart rate, percentage VO₂ max, energy consumption, endurance), psychological health, adherence, anthropometry, serum zinc (μmol/L), vitamin A status (serum retinol mmol/L or retinol binding protein mmol/L), red cell folate (mmol/L), and malaria incidence and severity

4. Main results

4.1 Included studies

Sixty-seven trials, enrolling 8506 women, were included in this review

• Most studies recruited women between 13 and 45 years of age, although ages ranged from 10 to 60 years
• Iron formulations included: ferrous sulphate (33 trials), ferrous fumarate (5 trials), ferrous gluconate (4 trials), carbonyl iron (3 trials), ferrous carbonate (2 trials), amino acid chelate (2 trials), ferric pyrophosphate, ferric ammonium citrate, ferrous succinate, niferex ferrous glycine sulphate, ferrous sodium citrate, iron polystyrene sulfonate; 12 trials did not specify the iron formulation used
• The daily elemental iron dose ranged from one to 300 mg, and duration of supplementation ranged from one to 24 weeks

4.2 Study settings

• Australia (5 trials), Bolivia, Brazil, Canada (2 trials), Chile, China (2 trials), Finland (2 trials), India (2 trials), Iran (3 trials), Israel (2 trials), Japan (2 trials), Mexico, New Zealand (2 trials), Nepal, Norway, Peru, the Philippines, the Republic of Korea, Serbia, Sri Lanka (3 trials), Sweden (3 trials), Switzerland (2 trials), Thailand, the United Kingdom of Great Britain and Northern Ireland (4 trials), the United Republic of Tanzania, and the United States of America (21 trials)
• Nine studies were specified as being conducted in urban settings, four in rural settings, one in both rural and urban settings, and the remainder of trials did not specify whether they were carried out in urban or rural settings
• Two studies were reported as being conducted in low socioeconomic settings, and one reported targeting middle-class participants
• Two trials were reported as being conducted in malaria endemic areas

4.3 Study settings

How the data were analysed
One main comparison was made: daily iron supplementation versus no iron supplementation. Random effects meta-analysis was used to generate summary risk ratios (RR) for dichotomous data and mean differences (MD) for continuous data, with corresponding 95% confidence intervals (CI). For crossover trials, only the first period of randomization was analyzed. For studies with multiple arms, relevant arms were combined to provide a single pair-wise comparison, and, if the control group was shared by more than two arms, it was divided for comparison to avoid double-counting of participants. Sensitivity analyses were performed including only those studies considered to be at low risk of bias. Potential sources of heterogeneity were explored in the following subgroup analyses:

• By age: adolescents (12 to 18 years), older adults (50 to 55 years)
• By nutrient: iron alone or iron plus co-intervention versus co-intervention alone, iron plus vitamin C versus vitamin C alone
• By baseline anaemia status: anaemic, non-anaemic, mixed or unknown
• By baseline iron-deficiency anaemia status: iron deficient with anaemia, iron deficient without anaemia, non-iron deficient or unknown
• By daily dose of elemental iron: <30 mg, 30 to 60 mg, 61 to 100 mg, ≥101 mg
• By duration of iron supplementation: ≤1 month, >1 to ≤3 months, >3 months
• By malaria endemicity: endemic, non-endemic, not reported/unknown

Results
Anaemia
Compared with women who did not receive iron supplements, those who did were 61% less likely to be anaemic at the end of the intervention (RR 0.39, 95% CI [0.25 to 0.60], p=0.000017; 10 trials/3273 women), although heterogeneity was high (I²=93%). Only one trial was considered to be at low risk of bias in which no difference between groups was observed. In subgroup analysis, studies comparing iron alone with control produced a smaller reduction in the risk of anaemia (RR 0.57, 95% CI [0.45 to 0.74], 8 trials/2775 women) in comparison to studies comparing iron plus vitamin C with vitamin C alone (RR 0.10, 95% CI [0.06 to 0.15], 2 trials/498 women). Ferrous sulphate also appeared to have a greater effect (RR 0.20, 95% CI [0.47 to 0.90], 1 trial/69 women).

Haemoglobin
In 51 trials including 6861 women, daily iron supplementation resulted in a higher haemoglobin concentration at the end of the intervention compared with no iron supplementation (MD 5.30 g/L, 95% CI [4.14 to 6.45], p<0.00001; I²=86%). Results did not differ meaningfully when restricted to studies at low risk of bias. A greater difference was seen among women with baseline anaemia (MD 8.67 g/L, 95% CI [5.16 to 12.18], 8 trials/558 women). Iron-replete women did not show an improvement in haemoglobin with iron supplementation (MD 0.84 g/L, 95% CI [-2.26 to 3.95], 5 trials/421 women). Duration of supplementation affected the difference in haemoglobin, with the greatest improvement in trials lasting one to three months (MD 6.14 g/L, 95% CI [4.70 to 7.58], 37 studies/4171 women).

Iron deficiency
Women who received iron supplements were 38% less likely to be iron deficient at the end of the intervention (RR 0.62, 95% CI [0.50 to 0.76], p<0.00001; 7 trials/1088 women). Results did not differ meaningfully when restricted to studies at low risk of bias.

Iron-deficiency anaemia
In one study including 55 women, no cases of iron-deficiency anaemia were reported in either treatment group. In another trial reporting on microcytic anaemia, iron treatment halved the risk of iron-deficiency anaemia (RR 0.51, 95% CI [0.33 to 0.77], 378 women).

All-cause mortality
No trials reported on this primary outcome.

Adverse side effects
The risk of any side effects was not significantly different between treatment groups (RR 2.14, 95% CI [0.94 to 4.86], 7 trials/901 women). For trials with higher doses, the effect of iron on the risk of any side effects became statistically significant (61 to 100 mg/day: RR 2.61, 95% CI [1.44 to 4.75], 2 trials/157 women; >100 mg/day: RR 2.15, 95% CI [1.24 to 3.73], 3 trials/439 women). Gastrointestinal side effects were more frequent among iron-supplemented women (RR 1.99, 95% CI [1.26 to 3.12], 5 trials/521 women), with higher doses again having a greater effect (61 to 100 mg/day: RR 3.00, 95% CI [1.45 to 6.20], 1 trial/145 women; >100 mg/day: RR 2.42, 95% CI [1.45 to 4.05], 2 trials/83 women). Loose stools/diarrhoea were more common with iron supplementation (RR 2.13, 95% CI [1.10 to 4.11], 6 trials/604 women), as were hard stools/constipation (RR 2.07, 95% CI [1.35 to 3.17], 8 trials/1036 women). Abdominal pain, nausea, and headaches were not different between treatment groups, and only one trial reported on reflux/heartburn, in which four cases occurred in the iron group compared with none in the control group. In four trials including 359 women, stool colour was increased with iron supplementation (RR 6.92, 95% CI [3.83 to 12.52]).

Cognitive function
Five trials reported on changes in cognitive function, but due to heterogeneity in outcome measures, data could not be pooled in meta-analysis. In one trial of 81 adolescents with non-anaemic iron deficiency, iron treatment resulted in a significant improvement in verbal learning (p<0.02). In another trial of women >20 years of age with anaemia, a reduction in the number of errors made while completing a maze test was observed. A reduction in impulsivity was found in another trial of non-iron deficient women aged 18 to 35 years following daily iron supplementation (p=0.047). No other differences were found between treatment groups.

Additional outcomes
Iron status
In 42 trials including 3881 women, ferritin concentrations were higher with iron supplementation at the end of the intervention (MD 10.27 ng/mL, 95% CI [8.90 to 11.65]), and in 23 studies including 1637 women, transferrin saturation was also increased (MD 5.98%, 95% CI [3.93 to 8.02]). Iron treatment reduced the concentration of soluble transferrin receptor (standardized MD -0.32, 95% CI [-0.49 to -0.16], 11 trials/579 women) and increased the concentration of serum iron (standardized MD 0.47, 95% CI [0.19 to 0.74], 17 trials/902 women), but had no significant effect on total iron binding capacity or erythrocyte protoporphyrin.

Physical exercise performance
In analyses of peak exercise performance outcomes, absolute VO₂ max (MD 0.11 L/min, 95% CI [0.02 to 0.20], 8 trials/276 women) and relative VO₂ max (MD 2.35 mL/kg/min, 95% CI [0.55 to 4.17], 15 trials/407 women) were improved with iron supplementation, but no effect was found on peak expiratory exchange ratio, heart rate, or lactate at longest point of exercise. In five studies including 126 women, a lower proportion of VO₂ max was required in women supplemented with iron to reach a defined submaximal exercise task (MD -3.34%, 95% CI [-6.17 to -0.51]), and in six studies including 212 women, a lower heart rate was required to achieve the same exercise task (MD -4.72 beats/minute, 95% CI [-8.64 to -0.80]). No differences were observed between groups in energy consumption during exercise, submaximal respiratory exchange ratio, achieved workload, or time to exhaustion.

Other secondary outcomes
In two studies examining psychological health, iron supplementation improved self-reported physical condition and vitality, but no other benefits were reported. Although not a pre-specified outcome, in six of the eight studies reporting on fatigue (or outcomes similar to fatigue) improvements were found with iron supplementation, particularly among women who were fatigued at baseline. Overall, iron supplementation did not appear to have an effect on adherence in the 33 studies in which it was reported, although meta-analysis could not be performed due to heterogeneity in outcome measures. While height and weight were not affected by iron supplementation, in six studies including 520 women, body mass index was 0.53 kg/m² greater in women receiving iron at the end of the intervention (95% CI [0.10 to 0.96 kg/m²]). Serum zinc, work productivity and malaria prevalence were not different between treatment groups.

5. Additional author observations*

Overall, only 10 of the 67 included studies were judged to be at low risk of bias, with common concerns being the lack of specification of random sequence generation and allocation concealment methods. The evidence for anaemia, haemoglobin, iron deficiency, and the side effects diarrhoea and constipation was rated as being moderate to high quality, while for any side effects, abdominal pain, and gastrointestinal side effects the evidence was rated as low quality.

Daily iron supplementation in menstruating women increases haemoglobin and iron stores and reduces the risk of anaemia and iron deficiency. In addition, iron supplementation improves both maximal and submaximal exercise performance and potentially improves fatigue. Adverse effects may be reduced with lower doses of iron given for a longer duration.

Further research examining the effects of iron supplementation on cognitive function, psychological health, wellbeing, and productivity are warranted. The effect of iron supplementation on future pregnancies, and, in low- and middle-income countries, infection and micronutrient status, should also be addressed.