Background: At least two previous studies (Pediatrics 1995; 95: 263-269 and JAMA 1996; 276: 1805-1810) have suggested that prenatal exposure to magnesium sulfate may reduce the incidence of cerebral palsy in very low birthweight (VLBW) infants, by as much as 90%. However, these were multicenter observational reports and the infants were born before 1989, before surfactant and antenatal steroids (factors associated with improved outcome) became routinely used.

Purpose: To determine if in utero exposure to magnesium sulfate is associated with improved neuromotor outcome in VLBW infants born at a single regional perinatal center, in the current era when surfactant and antenatal steroid use are more prevalent.

Methods: Maternal and neonatal data, including maternal magnesium therapy, were obtained on all VLBW infants born at our Regional Perinatal Center in 1994. Neuromotor outcomes at 2 years corrected age were obtained from our NICU Follow-up Clinic, or by phone contact with the infant's primary care physician for those not followed up at our clinic. Data were grouped according to magnesium exposure and analyzed by chi-square or Fisher's exact test.

Results: Data from 110 infants have thus far been analyzed. Thirty-four infants (31%) were exposed prenatally to magnesium and 54 (49%) to steroids. Ninety-nine infants (90%) received surfactant. Mothers given magnesium were comparable to non-treated mothers for age, education, race, and delivery mode but were more likely to have received steroids (P<.001) and antibiotics (P<.005) before delivery and less likely to have had an illicit drug history or positive toxicology screen (P<.005). Infants exposed to maternal magnesium were comparable to non-exposed infants for birthweight, gestational age, sex, Apgar scores, surfactant treatment, indomethacin therapy, patent ductus arteriosus, necrotizing enterocolitis, intraventricular hemorrhage, and periventricular leukomalacia. Furthermore, the incidence of cerebral palsy at 2 years corrected age was similar between the groups.

Conclusion: In our VLBW infants, born and cared for at a single perinatal center in 1994, prenatal magnesium exposure was not associated with a lower incidence of cerebral palsy, despite a variety of factors favoring the exposed infants, including antenatal steroid use and absence of illicit drugs.

Speculation: Differences in prenatal and/or neonatal care between centers, or between groups, may have acted as confounders in prior studies which concluded that magnesium exposure was protective.