1. Objectives

To evaluate the role of vitamin A supplementation in term neonates in low and middle income countries (LMIC) with respect to the prevention of mortality and morbidity

2. How studies were identified

The following databases were searched up to March 2016:

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

Relevant conference proceedings and reference lists of identified trials were handsearched. In addition, relevant organizations and researchers were contacted regarding ongoing trials and unpublished data

3. Criteria for including studies in the review

3.1 Study type

Randomized controlled trials, cluster randomized controlled trials, quasi-randomized controlled trials, and studies using a factorial design

3.2 Study participants

Term neonates (born between 37 to 42 weeks’ gestation) up to 28 days post-birth

3.3 Interventions

Vitamin A supplementation within the first 28 days of life, but not extending beyond the first 28 days, compared to placebo or no supplementation

(Any co-interventions must have been identical in both treatment and control arms)

3.4 Primary outcomes

• All-cause infant mortality at six and 12 months

3.5 Secondary outcomes

• Cause-specific infant mortality associated with acute respiratory infections and diarrhoea at six and 12 months
• Infant morbidity at six months of age, associated with acute respiratory infections and diarrhoea
• Vitamin A deficiency measured biochemically as serum retinol <0.70 μmol/L
• Blindness and signs of xerophthalmia (Bitot’s spots and corneal lesions)
• Mean haemoglobin concentrations or anaemia
• Adverse events due to vitamin A toxicity
• All-cause neonatal mortality between supplementation and 28 days

4. Main results

4.1 Included studies

Twelve trials including 168,460 neonates were included in this review:

• Three trials were cluster-randomized and nine were individually randomized
• Most studies used a single oral 50,000 IU dose of vitamin A in a vegetable oil base
• Sample sizes ranged from 1736 to 44,984 neonates
• One study was in low birthweight infants <2500 g
• Mothers were also supplemented with vitamin A in two studies; one in the postpartum period and one during pregnancy
• Gestational age at birth was not specified in most studies

4.2 Study settings

• Bangladesh, Ghana, Guinea-Bissau (3 trials), India (2 trials), Indonesia, Nepal, Pakistan, the United Republic of Tanzania, and Zimbabwe
• All studies were set in LMIC, with varying degrees of vitamin A deficiency and infant mortality

4.3 Study settings

How the data were analysed
Vitamin A supplementation was compared with placebo or no treatment for the prevention of mortality and morbidity in neonates. Dichotomous data were summarized using the generic inverse variance method to produce risk ratios (RR) based on cumulative risk and rate ratios per years of follow-up, both with corresponding 95% confidence intervals (CI). Data from cluster-randomized trials were adjusted for clustering. Data for term neonates were presented separately in only one trial for the outcome all-cause mortality. All-cause mortality data from two other trials were also considered as being from term neonates as the inclusion criterion of a normal birthweight (≥2500 g) was specified. Due to the paucity of term neonate data available, analyses including all infants, regardless of gestational age at birth, were also conducted. For all secondary outcomes, data were not reported separately for term neonates and were therefore analysed across all infants. In both trials with maternal vitamin A supplementation, no significant interaction was reported between maternal and neonatal supplementation arms, and thus all data from neonates, regardless of maternal supplementation, were included in analyses.

Provided sample size was sufficient, the following subgroup analyses were planned to investigate heterogeneity:

• Maternal vitamin A supplementation
• Birthweight
• HIV status of mother and infant
• Dose and frequency of vitamin A supplementation
• High baseline infant mortality
• Co-morbidities
• Timing of vitamin A supplementation (within 48 to 72 hours of birth versus later)
• Infant sex
• Geographic region

Results
All-cause infant mortality at six months of age
Risk ratios
In three studies including 22,721 term neonates, a non-significant 20% reduction in the risk of death from any cause was found among those supplemented with vitamin A (RR 0.80, 95% CI [0.54 to 1.18]; I²=63%). Pooled analysis of all infants regardless of gestational age at birth was also non-significant (RR 0.98, 95% CI [0.89 to 1.07], 11 trials/154,634 infants; I²=47%). In subgroup analysis of data from all infants, the effect did not differ meaningfully by sex. However, when subgrouped by region, infants from Asia benefited from vitamin A supplementation (RR 0.89, 95% CI [0.80 to 0.99], 6 trials/86,391 infants; I²=29%), whereas infants from Africa were at borderline increased risk of all-cause mortality (RR 1.10, 95% CI [1.00 to 1.21], 5 trials/68,243 infants; I²=0%).

Rate ratios (per years of follow-up)
Supplementation with vitamin A did not significantly reduce the rate of death from any cause at six months of age among term infants (rate ratio 0.93, 95% CI [0.67 to 1.30], 2 studies; I²=51%) or among data from all infants (rate ratio 0.99, 95% CI [0.84 to 1.17], 5 studies; I²=46%). In subgroup analyses by sex and region, no statistically significant results were found.

All-cause infant mortality at 12 months of age
Risk ratios
There was no significant difference between groups in the risk of death from any cause at 12 months of age (RR 1.04, 95% CI [0.94 to 1.15], 8 studies/118,376 infants of any gestational age; I²=47%). In subgroup analysis by sex, no statistically significant results were found. Infants from Africa receiving vitamin A were at borderline increased risk of all-cause mortality at 12 months of age (RR 1.08, 95% CI [1.00 to 1.17], 6 trials/71,553 infants of any gestational age; I²=0%).

Rate ratios (per years of follow-up)
Vitamin A supplementation did not significantly reduce the rate of death from any cause at 12 months of age among term infants (rate ratio 0.94, 95% CI [0.59 to 1.50], 3 studies; I²=72%) or among all infants (rate ratio 1.06, 95% CI [0.92 to 1.22], 6 studies; I²=31%). Supplemented female infants of any gestational age had a greater rate of all-cause mortality at 12 months (rate ratio 1.21, 95% CI [1.05 to 1.41], 5 trials; I²=0%; p=0.06 for subgroup differences). In subgroup analysis by region, no statistically significant results were found.

Secondary outcomes
Cause-specific mortality related to diarrhoea and acute respiratory infections at six and at 12 months and cause-specific morbidity related to diarrhoea and acute respiratory infections at six months did not differ significantly between treatment and control groups. Vitamin A deficiency (serum retinol <0.70 μmol/L) and anaemia were each reported in one study only, with no significant differences between groups. Pooled estimates of adverse events in the first 48 to 72 hours post-supplementation showed an increased risk of a bulging fontanelle with vitamin A supplementation (RR 1.53, 95% CI [1.11 to 2.11], 5 trials/100,256 infants; I²=71%); no other significant differences were found. One study reported on adverse events during the first month, and found no significant difference between groups. No data on other secondary outcomes were reported in the included trials.

5. Additional author observations*

Using the GRADE approach, evidence for the primary outcome all-cause mortality at six months of age among term neonates was rated as being of very low quality due to risk of bias in one of the three included trials, imprecision, and statistical heterogeneity. Other outcomes assessed using GRADE were rated as high quality, with the exception of the adverse effect diarrhoea in the first 48 to 72 hours post-supplementation, which was downgraded to low quality due to imprecision and statistical heterogeneity.

Overall, no reduction in the risk of all-cause mortality was found with neonatal vitamin A supplementation. While infants in Asia-and therefore at greatest risk of vitamin A deficiency-appeared to benefit from supplementation, those in Africa did not, and were at increased risk of mortality in some analyses. Limited data were available on other outcome measures such as cause-specific morbidity and mortality, adverse effects, and vitamin A deficiency.

Future trials should investigate the biological mechanisms mediating the effect of vitamin A on infant mortality, and determine the reasons underlying the differential effect by geographical region. Longer follow-up periods, and stratifying effects by sex, age of administration, prematurity, and intrauterine growth restriction would be valuable in future trials.