The review included 11 studies (16,804 participants). Of these, six were case-control studies (13,978 participants) and five were cohort studies (2,826 participants). The quality scores ranged from 2 to 9, with nine studies being classified as high quality.
Mortality: Five cohort studies (2,826 participants) showed a beneficial effect of magnesium sulphate on infant mortality (RR 0.73, 95% CI 0.61 to 0.89; Ι²=36%). No effect on the relative risk was found when one study of low quality was excluded (RR 0.73, 95% CI 0.66 to 0.88). Two case-control studies assessed the association between magnesium sulphate and mortality. One found a clear beneficial effect; the other found a non-significant detrimental effect. However, when the case-control studies were stratified by gestational age, a beneficial effect seen, but the authors reported that adjusted data could not be obtained. No meta-analysis was performed on the case-control studies.
Cerebral palsy: Five case-control studies (974 participants) showed a beneficial effect of magnesium sulphate on the incidence of cerebral palsy (OR 0.64, 95 %CI 0.47 to 0.89), with evidence of heterogeneity (Ι²=70%). Stratification by birth weight found a stronger effect in infants born weighing under 1,750g (OR 0.39, 95% CI 0.20 to 0.77; Ι²=67%) than in heavier infants. One cohort study (1,097 participants) found a protective effect of magnesium sulphate treatment on cerebral palsy (OR 0.10, 95% CI 0.02 to 0.81).