1. Objectives

To assess the effects and safety of home (point-of-use) fortification of foods with multiple micronutrient powders on nutritional, health and developmental outcomes in children under two years of age.

2. How studies were identified

The following databases were searched in February 2011:

• Cochrane Infectious Diseases Group Specialized Register
• Cochrane Central Register of Controlled Trials (CENTRAL)
• MEDLINE
• EMBASE
• African Index Medicus
• LILACS
• CINAHL
• POPLINE
• WHO International Clinical Trials Registry Platform (ICTRP)
• metaRegister of Clinical Trials (mRCT)

Reference lists of included studies were handsearched and relevant organizations and authors were also contacted

3. Criteria for including studies in the review

3.1 Study type

Randomized controlled trials and quasi-randomized controlled trials, with individual or cluster randomization

3.2 Study participants

Apparently healthy infants aged six to 23 months at the beginning of the intervention

(Although apparently healthy, some children may have been at risk of conditions such as malaria, diarrhoea or undernutrition)

3.3 Interventions

Micronutrient powders (MNP) including a minimum of iron, zinc and vitamin A, compared with no intervention, placebo or usual supplementation (iron-only; iron plus folic acid; the same micronutrients supplied as a supplement). Co-interventions such as education were included only if they were the same in both the intervention and comparison groups

(Trials of lipid-based supplements, crushable micronutrient tablets and fortified complementary foods or other fortified foods were excluded)

3.4 Primary outcomes

• Anaemia (haemoglobin <110 g/L)
• Iron deficiency (as defined by trial authors)
• Haemoglobin concentrations (g/L)
• Iron status (as defined by trial authors)
• Weight-for-age Z-scores
• All-cause mortality

Secondary outcomes included length-for-age Z-scores, weight-for-height Z-scores, all-cause morbidity, side-effects, diarrhoea, upper respiratory tract infections, ear infections, iron overload, serum retinol (μmol/L) and zinc (g/dL) concentrations, mental and motor skill development, and malaria incidence and severity

4. Main results

4.1 Included studies

Eight trials, enrolling 3748 children, were included in this review:

• Two studies compared the effects of MNP versus iron drops or syrup
• Six studies compared the effects of MNP versus no intervention or placebo
• No trials evaluated the effects of MNP versus iron plus folic acid or against multiple micronutrient supplements
• All trials had intervention arms providing 12.5 mg of elemental iron (ferrous fumarate), 5 mg of elemental zinc (zinc gluconate), 300-400 μg of vitamin A, and folic acid

4.2 Study settings

• Cambodia, Ghana (2 studies), Haiti, India, Kenya, Kyrgyzstan, and Pakistan
• Five studies were conducted in malaria-endemic areas (Cambodia, Ghana, Haiti, Kenya)
• No studies excluded anaemic infants

4.3 Study settings

How the data were analysed
Two comparisons were made: MNP versus no intervention or placebo (6 trials/3182 children) and MNP versus iron drops or syrup (2 studies/565 children). Significant heterogeneity was present in all analyses and could not be attributed to standard sensitivity analyses, thus a random-effects model was used.

The following subgroup analyses were specified: by anaemia status at baseline, by iron status at baseline (iron deficient, not iron deficient, unknown), by age at baseline (six to 11 months, 12 to 17 months, 18 to 23 months), by refugee status, by malaria status of the area, by frequency (daily, weekly, flexible) and duration (< six months, ≥ six months) of the intervention, and by iron (<12.5 mg, ≥12.5 mg) and zinc (<5 mg, ≥5 mg) content of the intervention product.

Results
Home (point-of-use) fortification of foods with MNP versus no intervention or placebo
Anaemia (haemoglobin <110 g/L)
Compared to no treatment or placebo, children receiving MNP had a reduced risk of anaemia at follow-up, with an average risk ratio (RR) of 0.69, 95% confidence interval (CI) [0.60 to 0.78], p<0.00001; (6 trials/3182 children). No differences in the effectiveness of the intervention were apparent across subgroups.

Iron deficiency
In four trials enrolling 586 children, those receiving MNP were significantly less likely to be iron deficient at follow-up than those not receiving an intervention or receiving the placebo (RR 0.49, 95% CI [0.35 to 0.67], p=0.000017). No differences were observed between subgroups.

Haemoglobin concentration (g/L)
In six trials enrolling 3182 infants, children receiving MNP had a significantly higher haemoglobin concentration than those in the placebo or no intervention group, with a mean difference (MD) of 5.87 g/L, 95% CI [3.25 to 8.49], p=0.000011. No differences were apparent among subgroups.

Iron status
Children in the MNP treatment group had higher ferritin concentrations at follow-up than those not receiving treatment, MD 20.38 ng/mL, 95% CI [6.27 to 34.49], p=0.0046; 2 trials/264 children.

Weight-for-age Z-scores
In two trials (304 infants) with a duration of six to 12 months, no treatment effect of MNP was found (RR 0.00 weight-for-age Z-score, 95% CI [-0.37 to 0.37], p=0.99).

All-cause mortality
No deaths were judged attributable to the MNP intervention by trial investigators.

In secondary outcome analyses, no treatment effect was detected for length-for-age or weight-for-height Z-scores (2 trials/304 children). Other secondary outcomes were not able to be pooled due to differences in definition (diarrhoea) or were reported in only one trial (recurrent diarrhoea, upper respiratory tract infection, serum zinc concentration, mental and motor skill development, malaria outcomes), or were not reported in any studies (all-cause morbidity, side effects, ear infections, iron overload, serum retinol concentrations, malaria incidence). No statistically significant differences were observed except for in one trial which reported that children receiving MNP were more likely to walk independently at 12 months of age than those receiving no intervention (RR 1.58; 95% CI [1.02 to 2.46]).

Home (point-of-use) fortification of foods with MNP versus iron supplements
Anaemia (haemoglobin <110 g/L)
In one study of 145 infants, no statistically significant differences in anaemia were found between infants treated with a five-micronutrient MNP versus iron drops (RR 0.89, 95% CI [0.58 to 1.39], p=0.62).

Haemoglobin concentration (g/L)
No statistically significant difference in haemoglobin concentrations was observed between those treated with MNP versus iron supplements (RR -2.36, 95%CI [-10.30, 5.58], p=0.56; 2 studies/278 children).

In secondary outcome analyses, children receiving MNP in two trials (395 children) were less likely to suffer stained teeth as a side effect than those receiving iron syrup (RR 0.37, 95% CI [0.16 to 0.82], p=0.015). In one study, children receiving MNP were reported to be less likely to have diarrhoea, recurrent diarrhoea and vomiting than those receiving daily iron supplements (RR 0.52; 95% CI [0.38 to 0.72]), although another study found no differences in diarrhoea episodes between treatment groups. No other secondary outcomes were reported on.

5. Additional author observations*

The overall quality of the evidence for the provision of MNP compared with a placebo was high for iron deficiency, whereas it was moderate for anaemia, haemoglobin concentration and growth, and low for iron status. Of the eight included studies, only one was classified as being at high risk of bias, and its exclusion in sensitivity analyses did not alter the results in a meaningful way. Blinding of caregivers and outcome assessors was not attempted in three-quarters of the trials, which may present a serious source of bias.

Compared with no treatment or placebo, the provision of MNP was effective in reducing the risk of anaemia, and increased haemoglobin and ferritin concentrations at follow-up. These findings were observed in various settings including in populations with a high background prevalence of anaemia and when provided for two months or for six months or more, and to all infants and young children aged six to 23 months. A paucity of comparable data on adverse outcomes such as diarrhoea and malaria and other measures of morbidity made it difficult to assess the overall safety of MNP, although no deaths were judged attributable to this intervention.

In comparison to iron supplements as drops or syrups, MNP has similar benefits on haematological outcomes and was associated with less staining of teeth, however the evidence was limited.

Further research is warranted investigating the effect of home fortification with MNP on additional nutritional outcomes, such as vitamin A deficiency and zinc status.