Objective: To estimate whether higher magnesium levels in umbilical cord blood at delivery are associated with increased total pediatric (fetal + neonatal + postneonatal) mortality.
Methods: During the Magnesium and Neurologic Endpoints Trial, in addition to randomizing mothers having preterm labor into arms containing magnesium sulfate, other tocolytic agents, or saline controls, we obtained biologic specimens at delivery, including umbilical cord venous blood on which was determined the serum ionized magnesium level using the AVL 988-4 analyzer (Graz, Austria). Laboratory results were then matched with the pediatric mortalities. The study power was based on the anticipated reductions in neonatal intraventricular hemorrhage related to magnesium usage from 18.9% to 4.4%. For alpha =.05, 1-beta (power)=80%, two tailed, the total number of infants needed would be 140.
Results: Of 149 mothers who gave permission for randomization, ionized magnesium levels were available for 82 children. Seven deaths occurred (one immediately before delivery, three as neonates, and three in the postneonatal period). The median level of ionized magnesium among the seven dead children was 0.76 mmol/L; among the 75 survivors, the median level of ionized magnesium was 0.55 mmol/L (Mann-Whitney U test, P =.03). Using multivariable logistic regression analysis, the association remained statistically significant when controlling for possible confounding factors (adjusted odds ratio 7.7, 95% confidence interval 1.2, 47.6, P =.03).
Conclusion: These findings of a dose response between serum ionized magnesium and deaths in children increase our concern about the improper use of tocolytic magnesium.