Does single mega-dose vitamin A in the early neonatal period reduce mortality during infancy in low- and middle-income countries?

MANUSCRIPT CITATION

Mazumder, S. Taneja, S. Bhatia, K. Yoshida, S. et al. Neovita India Study Group. Efficacy of early neonatal supplementation with vitamin A to reduce mortality in infancy in Haryana, India (Neovita): a randomised, double-blind, placebo-controlled trial. Lancet. 2015: 4; 385 (9975): 1333-42. doi: 10.1016/S0140-6736 (14) 60891-6. PMID:25499546

REVIEWED BY

Sindhu Sivanandan, MD
Division of Neonatology
WHO Collaborating Center for Training and Research in Newborn Care
All India Institute of Medical Sciences (AIIMS), Delhi, India

M Jeeva Sankar, MD, DM
Division of Neonatology
WHO Collaborating Center for Training and Research in Newborn Care
All India Institute of Medical Sciences (AIIMS), Delhi, India

QUESTION

P= In neonates who are able to feed orally at the time of enrolment and whose parents were likely to stay in the study area until at least 6 months of age,
I= does single dose of oral vitamin A (retinol palmitate 50000 IU plus vitamin E 9·5–12·6 IU) administered within 72 hours of birth
C= compared to single dose of placebo (vitamin E 9·5–12·6 IU)
O= reduce the mortality between vitamin A administration and 6 months of age?

METHODS

  • Design: Randomized, double-blind, placebo-controlled trial
  • Allocation: The randomization list was prepared offsite at WHO (Geneva, Switzerland) using a block randomization scheme with a block size of 20 by a statistician not otherwise involved with the trial. Based on randomization list, labels with a unique infant number were printed at the WHO and affixed on the appropriate blister packs of capsule containing vitamin A or placebo, which were then sent to the study site.
  • Blinding: The vitamin A and placebo capsules were identical in color, shape, and size. Each blister pack contained 2 capsules (of either vitamin A or placebo) one for the dose and the other as a backup and was labeled with a unique number. The research team, parents, data analysis team, and safety board were unaware of the content of the capsules in the blister pack.
  • Follow-up period: Primary outcome – evaluated between Vitamin A administration and 6 months of age; secondary outcomes – evaluated between Vitamin A administration and 12 months of age
  • Setting: Community setting. Two districts, namely Faridabad and Palwal districts (population roughly 2·1 million people) in the state of Haryana, India between June 24, 2010, and July 1, 2012. More than half of births take place at home, mostly assisted by traditional birth attendants.
  • Patients: Eligible participants were neonates whose parents consented to participate, were likely to stay in the study area until at least 6 months of age, and were able to feed orally at the time of enrolment. Both singleton and multiple births were eligible for enrolment.
  • Intervention: Randomization to vitamin A (retinol palmitate 50000 IU) and minute amounts of vitamin E (9·5–12·6 IU) or placebo, i.e. vitamin E alone (9·5–12·6 IU) administered as a single dose orally by squeezing the contents of the capsule in to the mouth of the neonate within the first 72 hours after birth.
  • Outcomes:
    • Primary outcome: The primary outcome was death between Vitamin A administration and 6 months of age.
    • Secondary outcomes: A number of secondary and safety outcomes were evaluated. They include death between Vitamin A administration and 28 days of age (neonatal period), deaths between administration and 12 months of age, admission of an infant to hospital because of any illness between administration and 6 months of age, potential adverse events in the 3-day period after supplementation, and vitamin A status in a subsample of infants at 2 weeks and 3 months of age.
      • Hospital admission was defined as an admission to hospital as an inpatient, as reported by the mother.
      • Adverse event was defined as any harmful or undesired effect as reported by the family that might have resulted from the supplement
  • Data collection: Study worker collected information through home visits on days 1 and 3, post supplementation. Subsequently, information about mortality and hospital admissions was collected through home visits on day 29 and at 3, 6, and 12 months of age.
  • Analysis and sample size: Based on a prior published study, the authors assumed the control group mortality to be 40 per 1000 live births up to 6 months of age. The authors had initially calculated a total sample size of 40,200 infants to detect a 15% reduced mortality rate (40/1000 live births in control vs. 34/1000 live births in experimental group) between administration and 6 months of age in the vitamin A group with 85% power, 95% confidence, and taking into account 10% attrition. After an interim analysis after 20,046 infants had been recruited, the data safety monitoring board recommended an increase in sample size to 45, 000 because the recorded mortality at 6 months in both groups was lower than originally estimated. Analysis was by intention-to-treat. For infant serum vitamin A level, the authors estimated a sample size of 720 infants randomly selected at 2 weeks of age and at 3 months of age (360 from each study group) to be adequate to detect an increase of 0·25 SD in mean serum concentrations of retinol in the vitamin A group, with 80% power, taking into account up to 30% attrition. Blood specimens were also taken from the mothers of the 720 infants sampled at 3 months. This number would be sufficient to have an absolute precision of 3·5% for an assumed 20% prevalence of maternal subclinical deficiency of vitamin A (serum retinol <0·70 μmol/L).
  • Patient follow-up: Of the 47,777 live births screened, 44,984 neonates were randomized (22,493 received vitamin A and 22, 491, placebo). Of the 44,984 neonates randomized, the supplement was delivered to all (>99·9%) except one neonate in the vitamin A group. The follow up rates were high with only a few neonates – 28 (0.12%) in vitamin A and 19 (0.08%) in the control groups – lost to follow up. All the randomized neonates were analyzed for the primary outcome and various secondary outcomes in an intention-to-treat analysis. For the evaluation of adverse events between vitamin A administration and 3 days after, data for 13 infants in vitamin A group and 10 in placebo group were not available.

MAIN RESULTS

The infant and maternal characteristics of vitamin A and placebo groups were comparable. The primary and secondary outcomes are listed in the table and discussed below.

Primary outcome
Outcome Vitamin A

N = 22,493;

n (per 1000 enrolled infants)

Placebo

N= 22,491

n (per 1000 enrolled infants

Risk difference (95%CI) Relative risk (95% CI) P
Mortality after vitamin A administration up to 6 months of age 656 (29.2) 726 (32.3) -3.1

(-6.3 to 0.1)

0.9

(0.81 to 1.0)

0.056
Secondary outcomes                                                                      
Mortality after vitamin A administration until 28 days of age 281 (12.5) 298 (13.2) -0.7

(-2.8 to 1.3)

0.94

(0.80to 1.11)

0.477
Mortality after vitamin A administration up to 12 months of age 879 (39.1) 939 (39.1) -2.7

(-6.3 to 1.0)

0.94

(0.86 to 1.02)

0.151
Hospital admissions after vitamin A administration up to 6 months of age 1348 (59.9) 1399 (62.2) -2.3

(-6.7 to 2.2)

0.96

(0.90 to 1.04)

0.315
Adverse events Vitamin A

N= 22, 480;

n (%)

Placebo

N= 22,481;

n (%)

Relative risk

(95% CI)

P
Deaths within 3 days of vitamin A administration 104 (0.5%) 101 (0.5%) 1.03 (0.78 to 1.35) 0.834
Bulging fontanelle

(reported by study worker)

261 (1.2%) 140 (0.6%) 1.86 (1.52 to 2.29) <0.0001
Bulging fontanelle

(physician confirmed)

205 (0.9%) 80 (0.4%) 2.56 (1.98 to 3.32) <0.0001
Serum retinol Vitamin A Placebo Mean difference

(95% CI)

P
2-week old infants N=341 N=323
Mean (SD), μmol/L 0.68 (0.23) 0.55 (0.21) 0.13 (0.09 to 0.16) <0.0001
3-month old infants N=281 N=278
Mean (SD), μmol/L 0.74 (0.29) 0.70 (0.28) 0.04 (-0.01 to 0.08) 0.125
Mothers of 3-month old infants N=286 N=276
Mean (SD), μmol/L 1.18 (0.41) 1.17 (0.41) 0.01 (-0.006 to 0.008) 0.819

Efficacy: Vitamin A administration, compared to placebo, did not have an effect on mortality between administration and 6 months of age. The relative risk of death was similar when deaths due to accidents were excluded from the analysis. There was no significant interaction between sex and mortality between Vitamin A administration and 6 months of age. There was no effect of the intervention on mortality at 12 months of age.

Safety and adverse effects of intervention: There was no difference in the risk of hospital admission in the first 6 months of life between the vitamin A and placebo groups. Bulging fontanelle diagnosed by a health worker (with or without confirmation by a physician) was significantly higher among neonates who received vitamin A. In all but 4 cases, the bulging fontanelle resolved within 72 hours. Other adverse events like, vomiting, fever, diarrhea, inability to suck or feed, and convulsions were not different between the two groups. The number of infants who died within 72 h of Vitamin A administration was also comparable in the two groups.

Effect on serum retinol levels and sub clinical vitamin A deficiency: Infants administered Vitamin A at birth had higher serum retinol levels and lesser prevalence of sub-clinical vitamin A deficiency (defined as serum retinol concentrations less than 0·70 μmol/L) at 2 weeks. However at 3 months of age, the serum retinol levels and prevalence of sub clinical vitamin A deficiency was comparable to placebo group.

Maternal vitamin A status: About 39% of mothers of study infants aged 3 months had low vitamin A status (serum retinol less than 1·05 μmol/L) and about 12% had evidence of vitamin A deficiency (serum retinol less than 0·70 μmol/L). As mothers did not receive any vitamin A supplements as part of this trial, these proportions represent the background maternal vitamin A status in the study population.

CONCLUSION

  1. The authors concluded that among newborn infants, a single dose of 50,000 IU of vitamin A compared to placebo did not result in a statistically significant decrease in mortality between vitamin A administration till 6 months of age. The authors argued that a 10% mortality reduction, although statistically insignificant, may be clinically important because it could lead to a 6% reduction in deaths in the first half of infancy – about 40% of deaths occur within 72 hours of life before Vitamin A administration is possible; hence 10% reduction in mortality between administration and 6 months would translate to a 6% reduction in deaths between birth and 6 months of age. A total of 322 neonates would need to be given vitamin A to prevent one infant death in the first 6 months of life.
  2. The authors also concluded that vitamin A administration within the first 72 h of life is safe and well tolerated, with the exception of a small excess risk of transient bulging fontanelle.
  3. Administration of vitamin A to neonates significantly improved serum retinol concentrations at 2 weeks of age but this improvement was short-lived and was no longer detectable at 3 months of age.
  4. However, formulation of public health policy for neonatal Vitamin A administration is not simple and the government needs to consider the prevalence of vitamin A deficiency at National and Sub National levels.

Visit Acta to access a pdf copy of this EBNEO commentary!

COMMENTARY

Vitamin A deficiency (VAD) is a major public health problem in low and middle-income countries, where its deficiency is associated with increased mortality. In order to prevent VAD, WHO recommends high-dose vitamin A, 100,000 to 200,000 IU every 4–6 months for children aged between 6 months and 5 years of age in countries at risk of VAD. The evidence for this recommendation came from systematic reviews and meta-analyses of randomized trials.1The Cochrane review1 included 43 studies that enrolled 230,354 children – a huge number – trying to answer whether VAS is beneficial. From the analysis of 17 trials that reported on all-cause mortality, the review concluded that vitamin A administration led to a 24% reduction in mortality (RR = 0.76; 95% CI 0.69, 0.83) when compared to no treatment.

It was, however, not known if the same mortality benefit would be extended to infants below 6 months of age by giving vitamin A at birth. Studies in young infants done in developing countries yielded conflicting results. A meta-analysis2 by Gogia et al included 6 RCTs (4 from Asia and 2 from Africa) that used ≤ 50, 000 IU of vitamin A in the neonatal period. They found no convincing evidence that neonatal VAS reduced risk of all-cause mortality during infancy or increased the risk of early adverse effects, including bulging fontanelle. In contrast, a meta-regressional analysis suggested that in regions where at least 22% of pregnant women have vitamin A deficiency, high dose vitamin A administration may have a protective effect against infant mortality.3 Hence the international community embarked on three randomized trials in three different regions – India, Tanzania and Ghana to better inform global policy makers about neonatal vitamin A administration. All the 3 trials were similar in all respects expect the geographical regions in which they were conducted. In India about 12% of mothers were deficient in vitamin A as compared to 2.5 % in Ghana and 6.5% in Tanzania.

The Indian trial reviewed herein4 did not show a conventionally statistically significant effect of vitamin A administration on mortality at 6 months (RR 0.90; 95% CI 0.81 to 1.00) and at 12 months of age (RR 0.94; 95% CI 0.86 to 1.02). The authors argued that the confidence interval, though reached the null value of 1, also included a potential 6% survival benefit in the first half of infancy. In contrast, the ‘sister’ studies from Ghana5 and Tanzania6, while showing no benefit on mortality at 6 and 12 months, could not rule out a small risk of increased mortality at 6 months of age. The differences in efficacy between these trials could be due to differences in the characteristics of the study population including the background rate of maternal vitamin A deficiency state. The Indian population had a higher maternal vitamin A deficiency status (12.1%) as compared to Ghana (2.5%) and Tanzania (6.5%). Such a survival benefit had been earlier reported by two previous trials in India 7 and Bangladesh.8

The authors’ selection of mortality at 6 months as the primary outcome is reasonable because mortality benefit is a very important advantage for implementing large-scale supplementation programmes in areas with high infant mortality rates. In the Indian study, Vitamin A administration significantly improved serum concentrations of retinol at 2 weeks but not at 3 months of age. In contrast, there was an improved serum vitamin A level at 3 months but not at 2 weeks in Tanzanian neonates and there was no improvement in vitamin A status at either time points in Ghanaian neonates. How could one explain these differences? Does vitamin A act by mechanisms other than increasing the serum levels of retinol or could serum retinol be a poor marker of body stores? In children, where VAS is unquestionably beneficial, vitamin A may act through immune modulation in addition to prevention of vitamin A deficiency9. It is possible that the effect of VAS is modified by factors unrelated to deficiency, such as sex of the child and the receipt of vaccines.9

Vitamin A supplementation was generally safe and well tolerated except for a small increase in bulging fontanelle (physician confirmed) 0.9% in vitamin A group compared to 0.4% in the placebo group. However this side effect was equal in Ghana (0.3%)and Tanzania (0.14%). The administration of excessive amounts of vitamin A can lead to toxicity; the manifestations depend on the infant’s age, hepatic function, dose and duration of supplementation. In one randomized study,10 11.5% of infants who received 3 doses of 50,000 IU of vitamin A at 6, 10 and 14 weeks of age developed bulging fontanelle as compared to 1% of those who received placebo. The bulging fontanelle is usually transient and resolves within 24-72 hours and is not associated with long term developmental consequences.11 The strengths of the study are its huge sample size, robust methodology, and high rates of compliance to the intervention and very high follow up rates. The study population is representative of large parts of South-East Asia. The two similar companion studies from different LMIC settings is another notable feature. In all the 3 studies, the mortality rate noted in the study population was lower than the initially presumed mortality rate on which sample size was calculated. The lower mortality rates could have been due to improvement in health care and health care related spending or increased awareness in the study population.

The companion trials from African sub-continent confidently state that the trial results do not support universal neonatal Vitamin A administration as a child survival intervention in either Ghana or Tanzania. However in the Indian scenario, where the prevalence of maternal vitamin A deficiency is high, a possibility of a small mortality benefit cannot be ruled out completely. A way forward would be a meta-analysis of the 3 Neo-vita trials and analysis of trials grouped based on the vitamin A deficiency status of its population. Until such evidence is available, one cannot recommend universal vitamin A supplementation to the neonates at birth.

References

  1. Imdad A, Herzer K, Mayo-Wilson E, Yakoob MY, Bhutta ZA. Vitamin A supplementation for preventing morbidity and mortality in children from 6 months to 5 years of age. The Cochrane database of systematic reviews 2010:CD008524.
  2. Gogia S, Sachdev HS. Neonatal vitamin A supplementation for prevention of mortality and morbidity in infancy: systematic review of randomised controlled trials. Bmj 2009;338:b919.
  3. Rotondi MA, Khobzi N. Vitamin A supplementation and neonatal mortality in the developing world: a meta-regression of cluster-randomized trials. Bull World Health Organ 2010;88:697-702.
  4. Mazumder S, Taneja S, Bhatia K, et al. Efficacy of early neonatal supplementation with vitamin A to reduce mortality in infancy in Haryana, India (Neovita): a randomised, double-blind, placebo-controlled trial. Lancet 2015;385:1333-42.
  5. Edmond KM, Newton S, Shannon C, et al. Effect of early neonatal vitamin A supplementation on mortality during infancy in Ghana (Neovita): a randomised, double-blind, placebo-controlled trial. Lancet 2015;385:1315-23.
  6. Masanja H, Smith ER, Muhihi A, et al. Effect of neonatal vitamin A supplementation on mortality in infants in Tanzania (Neovita): a randomised, double-blind, placebo-controlled trial. Lancet 2015;385:1324-32.
  7. Rahmathullah L, Tielsch JM, Thulasiraj RD, et al. Impact of supplementing newborn infants with vitamin A on early infant mortality: community based randomised trial in southern India. Bmj 2003;327:254.
  8. Klemm RD, Labrique AB, Christian P, et al. Newborn vitamin A supplementation reduced infant mortality in rural Bangladesh. Pediatrics 2008;122:e242-50.
  9. Benn CS, Aaby P, Arts RJ, Jensen KJ, Netea MG, Fisker AB. An enigma: why vitamin A supplementation does not always reduce mortality even though vitamin A deficiency is associated with increased mortality. International journal of epidemiology 2015;44:906-18.
  10. de Francisco A, Chakraborty J, Chowdhury HR, et al. Acute toxicity of vitamin A given with vaccines in infancy. Lancet 1993;342:526-7.
  11. Adverse events following the administration of vitamin A. (Accessed 21/11/2015, at http://www.who.int/immunization/programmes_systems/interventions/Adverse_events_vitA.pdf.)