The Relationship of Selenium Status to Respiratory Outcome in the Very Low Birth Weight Infant

Abstract

Objective. To examine the relationship between plasma and erythrocyte selenium and glutathione peroxidase (GPx) levels in premature infants and outcome measures.

Design. Prospective observational longitudinal study.

Setting. Two regional neonatal intensive care units in the South Island of New Zealand, an area with low soil selenium.

Patients. Seventy-nine infants with birth weights less than 1500 g or gestation less than 32 weeks admitted within 48 hours of birth from November 1992 through November 1993.

Main Outcome Measures. Oxygen requirement at 28 days (chronic lung disease), or 36 weeks postmenstrual age and for all or most of the time from birth (bronchopulmonary dysplasia), total days in oxygen, retinopathy of prematurity, periventricular hemorrhage, or ventricular dilatation.

Results. Initial infant plasma selenium and GPx levels were about two thirds of maternal levels and fell a further 30% in 28 days. In contrast to adults, there was a poor correlation in infant plasma between selenium and GPx at birth and 28 days. Plasma selenium at 28 days was significantly lower in infants with chronic lung disease and bronchopulmonary dysplasia. After controlling for gestational age and age when fully fed orally, 28-day plasma selenium was significantly associated with the log of total days of oxygen requirement, each drop of 0.1 µmol/L in 28-day selenium being associated with a 58% increase in days of oxygen dependency. No significant associations of other parameters of selenium status and respiratory outcome were found, and there were no significant associations of any parameters of selenium status with other outcome measures.

Conclusions. This study demonstrates for the first time in human infants that low plasma selenium levels are significantly associated with an increased respiratory morbidity. Whether selenium deficiency is etiologically important in determining the respiratory outcome or the result of sickness in the infant should be investigated in a randomized, controlled trial.