Objective: To examine the association between plasma vitamin A levels and outcome measures in very low birthweight (VLBW) infants, including meta-analysis of all observational studies.
Design: A prospective observational longitudinal study of plasma vitamin A levels measured in the cord blood; maternal blood in the first 48 h after delivery; and the infants' blood at 48 h, 7 days and 28 days of age and correlated with antenatal and postnatal events. A meta-analysis of all published observational studies on the association of vitamin A with respiratory outcome in the VLBW infant was undertaken.
Patients: Fifty-seven infants (88% of all eligible) VLBW infants (< 1500 g) admitted from January through October 1993 to one of two regional neonatal intensive care units in the South Island of New Zealand.
Results: Exposure to antenatal steroids led to a significant increase in infant cord plasma vitamin A levels (P = 0.003), but no influence on infant plasma vitamin A levels at any other time. Exposure to postnatal steroids produced a significant rise in infant plasma vitamin A levels between 7 and 28 days (P = 0.008). After controlling for gestational age, antenatal and postnatal steroid exposure, low vitamin A levels at 48 h increased the risk of developing chronic lung disease (odds ratio for 50 microg/l decrease: 2.04, 95% CI 1.19-5.77) and bronchopulmonary dysplasia (odds ratio 1.96, 95% CI 1.14-6.87). On combining our results in meta-analysis with those of other published prospective observational studies, infants with chronic lung disease had lower plasma vitamin A levels at all times.
Conclusions: Our results support an association between low plasma vitamin A levels and adverse outcome in the VLBW infant.