1. Objectives

To evaluate the effects of vitamin A supplementation in infants one to six months of age in low- and middle-income countries on morbidity, mortality, and adverse reactions

2. How studies were identified

The following databases were searched in October 2010:

• CENTRAL (The Cochrane Library 2016, Issue 2)
• MEDLINE
• CINAHL
• EMBASE
• ClinicalTrials.gov
• WHO International Trials Registry and Platform
• ISRCTN Registry

Reference lists and bibliographies were also handsearched

3. Criteria for including studies in the review

3.1 Study type

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

3.2 Study participants

Apparently healthy breastfed or non-breastfed infants from low- and middle-income countries

(Trials recruiting neonates or infants over six months of age were excluded, as were trials focusing on maternal supplementation with vitamin A. Studies that exclusively enrolled infants who were sick or hospitalized, infants with very low birth weight (<1500 g), or infants with HIV-infected mothers were also excluded)

3.3 Interventions

Oral synthetic vitamin A supplementation of infants aged one to six months of age, compared to placebo, irrespective of maternal postpartum vitamin A supplementation

3.4 Primary outcomes

All-cause mortality

• During infancy, from initiation of the intervention until the final follow-up, up to one year of age

Secondary outcomes included cause-specific mortality due to diarrhoea, acute respiratory infections, meningitis, and measles; infant morbidity until one year of age due to diarrhoea, acute respiratory infection or respiratory difficultly, and fever; and adverse effects within one week following the intervention including bulging fontanelle, vomiting, irritability, diarrhoea, and fever

4. Main results

4.1 Included studies

Twelve randomized controlled trials, including 24,846 participants, were included in this review

• Samples sizes ranged from 89 to 10,297 infants and the median proportion of boys was 51%
• Two factorial trials included maternal supplementation; for these trials separate comparisons were made by maternal supplementation status. One other trial also included maternal vitamin A supplementation
• One factorial study also included vitamin E supplementation; for this trial data were analysed by vitamin A supplementation status irrespective of vitamin E supplementation
• Vitamin A doses ranged from 25,000 IU to 100,000 IU, and the commonest dosing regimen was a 25,000 IU dose given at six, 10, and 14 weeks of age
• Supplements were in liquid form in all trials, and were given orally. Two studies gave vitamin A independently of vaccination times

4.2 Study settings

• Bangladesh (4 trials), Ghana (2 trials), Indonesia, Kenya, Nepal (2 trials), and Turkey, and one multinational trial was set in Ghana, India, and Peru
• Trials were conducted in both rural and urban areas, and all studies were conducted in low- and middle-income countries with endemic vitamin A deficiency

4.3 Study settings

How the data were analysed
Infant vitamin A supplementation initiated between one and six months of age was compared with placebo. In studies assessing different doses of vitamin A and placebo, the intervention groups were combined for analysis. Risk ratios (RR) were generated for dichotomous outcomes with corresponding 95% confidence intervals (CI). Data from cluster-randomized trials were adjusted for clustering. Fixed-effects meta-analysis was employed to summarize data, and the generic inverse variance method was used to pool studies reporting effect estimates in different formats. Subgroup analyses were planned for the primary outcome infant mortality as follows:

• By maternal vitamin A supplementation: received versus not received
• By cumulative infant dose of vitamin A up to the age of six months: low dose (≤50,000 IU) versus high dose (>50,000 IU)
• By birth weight: low (<2500 g) versus normal (≥2500 g)
• By timing of vitamin A delivery: given with vaccination versus given independent of vaccination

Results
Infant vitamin A supplementation
All-cause mortality
Overall, infant vitamin A supplementation did not reduce the risk of mortality in infants up to one year of age compared to placebo (RR 1.05, 95% CI [0.89 to 1.25], p=0.53; 7 trials/21,339 infants; I²=0%). Results did not differ meaningfully when subgrouped by whether supplementation was given at the time of vaccination or by maternal vitamin A supplementation, and other subgroup analyses could not be performed.

Additional outcomes
In one trial including 200 infants, cause-specific mortality before one year of age due to diarrhoea (RR 0.59, 95%CI [0.20 to 1.70]), acute respiratory infection (RR 2.12, 95% CI [0.40 to 11.33]), or meningitis (RR 0.35, 95% CI [0.01 to 8.58]) was not different between treatment and control groups. No difference was noted between groups in pooled analysis of two trials reporting on the point prevalence of diarrhoea (RR 0.99, 95% CI [0.93 to 1.05]). The point prevalence of lower respiratory tract infection and fever were not different between groups in the single trials reporting on these outcomes. The risk of vitamin A deficiency (serum retinol <0.7 µmol/L) did not differ between vitamin A and control groups in pooled analysis of three trials including 1204 infants (RR 0.86, 95% CI [0.70 to 1.06]).

Adverse effects
In pooled analysis of nine trials, the risk of a bulging fontanelle within 48 to 72 hours of supplementation in the treatment group was more than three times that of the control group (RR 3.10, 95% CI [1.89 to 5.09], p<0.00001; 13,493 infants; I²=9%). Subgroup analysis by whether the supplement was given at the time of vaccination revealed no important differences in estimates of the effect size (p=0.99 for subgroup differences), although the analysis for supplementation independent of vaccination was rendered non-significant (RR 3.13, 95% CI [0.86 to 11.32]; 1 trial/2141 infants). No differences between treatment and control groups were noted for other adverse effects. In addition, no differential effect by vitamin A supplementation was found in antibody response to tetanus toxoid, polio, measles, hepatitis B, or influenza vaccines.

5. Additional author observations*

Using GRADE criteria, evidence for all outcomes except bulging fontanelle was regarded as low to moderate quality, primarily due to imprecision and risk of bias. Evidence for the outcome bulging fontanelle within 48 to 72 hours of supplementation was judged as high quality.

Current evidence does not support infant vitamin A supplementation between one and six months of age for the reduction of mortality or morbidity. In addition, high quality evidence shows an increased risk of a bulging fontanelle within 48 to 72 hours of vitamin A supplementation, although no subsequent complications were associated with this increase in risk.

Given that the prevalence of vitamin A deficiency is decreasing, it is unclear whether further trials would provide any evidence of a beneficial effect of vitamin A supplementation in this age group. A reduced risk of mortality among children who had been supplemented with vitamin A in the neonatal period was previously reported in a single trial, and it remains unknown whether this effect exists for supplementation between one and six months of age.