Is routine vitamin E administration justified in very low-birthweight infants?

Analysis of nine randomised controlled trials of prophylactic vitamin E supplementation in very low-birthweight infants (less than 1500g) showed no statistically significant reduction in the incidence of acute retinopathy of prematurity. There was a significant reduction (49 per cent) in the incidence of intraventricular haemorrhage, but no clear evidence for a corresponding reduction in intracerebral haemorrhage and no reduction in the incidence of haemorrhage confined to the germinal matrix. By combining the estimated reduction with the known incidence of long-term neurological disability associated with intracranial haemorrhage alone, it was shown that only 1.5 per cent (point estimate) and not more than about 4 per cent of all very low-birthweight infants are likely to benefit from routine vitamin E supplementation. In view of this, and data suggesting toxicity of vitamin E at concentrations close to those considered therapeutic, the routine use of vitamin E in very low-birthweight infants is not justified on present evidence. Only large randomised trials can establish whether currently used vitamin E preparations do more good than harm.