1. Objectives

To evaluate the effectiveness and safety of vitamin A supplementation for preventing acute lower respiratory tract infections (LRTI) in children up to seven years of age

2. How studies were identified

The following databases were searched between February and July 2010:

• Cochrane Acute Respiratory Infection Group’s Specialized Register
• CENTRAL (The Cochrane Library 2010, Issue 1)
• MEDLINE
• EMBASE
• CNKI
• VIP
• WHO ICTRP

Reference lists and relevant journals, theses and conference proceedings were also hand-searched

3. Criteria for including studies in the review

3.1 Study type

Randomized controlled trials, excluding quasi-randomized trials

3.2 Study participants

Children aged up to seven years, without measles-related pneumonia or HIV infection

3.3 Interventions

Vitamin A with or without other micronutrients, versus placebo or the same micronutrients without vitamin A

3.4 Primary outcomes

• Incidence or prevalence of acute LRTI (confirmed by doctors on the basis of pre-defined criteria, including fever, tachypnoea, remission with or without cough, and chest or radiological signs)

(Acute LRTI refers to severe infection that affects the airways below the epiglottis)

Secondary outcomes included the incidence or prevalence of signs and symptoms of acute LRTI, including cough with or without fever, increased respiratory rate, increased sputum production, and specific radiological changes of the lung; and adverse events associated with the administration of vitamin A, including raised intracranial pressure, vomiting, nausea, and enlargement of the liver

4. Main results

4.1 Included studies

Ten randomized controlled trials, enrolling 33,179 children, were included in this review

• Participants’ ages ranged from zero to 83 months and sample sizes ranged from 613 to 15,419 children
• One trial involved children hospitalized for diarrhoea, and another involved children hospitalized predominantly for protein-energy malnutrition
• Seven trials administered bolus mega doses of vitamin A (100,000 or 200,000 IU), and four trials used low-dose schedules of vitamin A supplementation, including 5000 IU/day, 8333 IU/week, 10,000 IU/week, and 45,000 or 20,000 IU every two months
• Other supplemental micronutrients included vitamin E in one trial and zinc in two trials

4.2 Study settings

• Bangladesh, Brazil, Canada, the Democratic Republic of the Congo, Ecuador, Ghana, India, Indonesia, Mexico, and the United States of America
• Two trials were hospital-based, and eight trials were community-based
• Most trials were conducted in areas with a high prevalence of malnutrition, vitamin A insufficiency, or conditions reducing vitamin A absorption

4.3 Study settings

How the data were analysed
One comparison was made: vitamin A versus placebo for the prevention of acute LRTI in children up to seven years of age. Data were summarized using the generic inverse variance method to produce rate ratios (RR) with corresponding 95% confidence intervals (CI). Random effects models were used in all analyses. Overall pooled analyses were not performed; instead, data were pooled into the following subgroups for analysis:

• By dosage of vitamin A: mega doses versus low doses
• By state of nutrition, development or weight of the participants
• By age of the participants

Results
Vitamin A versus placebo
Incidence of acute lower respiratory tract infection: Mega-dose vitamin A (100,000 or 200,000 IU)
Pooled analysis of two community-based trials in which mega doses of vitamin A were used showed no statistically significant difference between treatment and control groups in the incidence of acute LRTI (1.13, 95% CI [0.80 to 1.60]). However, one of the trials contributing to the pooled analysis reported a statistically significant increase in the risk of acute LRTI with vitamin A treatment (RR 1.39, 95% CI [1.03 to 1.88]). In one hospital-based trial using mega-dose vitamin A, no treatment effect was found on the incidence of acute LRTI (RR 1.07, 95% CI [0.92 to 1.26]). Two other trials did not provide data in a suitable format for pooling, but also reported no effect of vitamin A on incidence rates.

Incidence of acute lower respiratory tract infection: Low-dose vitamin A (up to 5000 IU/day)
Among community-based trials, one study found no effect of low-dose vitamin A on the incidence of acute LRTI (RR 1.16, 95% CI [0.77 to 1.76]), and one found no effect on the age-adjusted incidence of acute LRTI (RR 1.01, 95% CI [0.73 to 1.40]). One hospital-based trial reported no treatment effect of low-dose vitamin A on the incidence of acute LRTI.

Incidence of acute lower respiratory tract infection: State of nutrition, development, or weight
In one high-dose trial, underweight children receiving vitamin A had a statistically significantly lower incidence of acute LRTI in comparison to controls (RR 0.38, 95% CI [0.17 to 0.85], p=0.019), while normal-weight children had a statistically significantly increased incidence (RR 2.22, 95% [1.25 to 3.95], p=0.0067). Another trial using both high- and low-dose schedules reported similar findings, with a significant increase in the incidence of acute LRTI in normal-sized children (RR 1.83, 95% CI [1.26 to 2.67]), and a non-significant reduction among stunted children (RR 0.48, 95% CI [0.21 to 1.12]). A further low-dose trial reported no differences in treatment effect between underweight and normal weight groups.

Incidence of acute lower respiratory tract infection: Age
Pooled analysis of three trials demonstrated no significant effect of vitamin A on the incidence of acute LRTI in children up to five years of age (RR 1.13, 95% CI [0.88 to 1.43]). In one high-dose trial, children ≤23 months had a borderline statistically significantly higher incidence of acute LRTI with vitamin A treatment (RR 1.19, 95% CI [0.99 to 1.43]), while those >23 months did not (RR 0.98, 95% CI [0.75 to 1.26]). One further trial reported no differences in incidence rates between age groups.

Additional outcomes
In three trials, the prevalence of symptoms of acute LRTI, including cough, cough plus fever, cough plus instantaneous respiratory rate, tired ribs, daytime cough, difficulty breathing, rapid respiratory rate, cough plus rapid respiratory rate, did not differ between treatment and control groups. However, in one high-dose trial, the two-week prevalence of cough (RR 1.18, 95% CI [1.09 to 1.27]) and rapid breathing (RR 1.18, 95% CI [1.09 to 1.27]) were statistically significantly increased with vitamin A supplementation. No trials reported on adverse effects of vitamin A supplementation.

5. Additional author observations*

Four of the 10 included trials were at low risk of bias for allocation concealment, and all trials were blinded, with six of the 10 studies double-blinded. Incomplete outcome data was high in one trial (29.6% of participants excluded from analysis), and one trial did not report on loss to follow-up. Using the GRADE approach, evidence for the outcome incidence of acute LRTI (all subgroups) was rated as moderate quality, while evidence for the outcomes cough, cough with difficulty breathing, and cough with fever was rated as low to very low quality.

The majority of evidence reviewed here suggests no overall difference between vitamin A and placebo on the incidence or clinical course of acute LRTI, with some trials demonstrating harm with vitamin A supplementation, others benefit, and most no effect. Evidence of a benefit of vitamin A among underweight children was demonstrated in one trial, while two trials showed a negative effect among children with adequate nutritional status. Low-dose vitamin A appeared safer and of similar benefit to high-dose vitamin A.

Further large randomized controlled trials evaluating the effects of different doses of vitamin A on acute LRTI are warranted. The effect of vitamin A on specific acute LRTI and the effect of various modes of vitamin A delivery on acute LRTI also require further investigation.