According to an editorial, this research adds knowledge about the links between the use of metformin, glibenclamide, and insulin for women with gestational diabetes and important clinical outcomes. It notes however, that though the researchers took great care to collect and control for confounders that may affect their outcomes of interest, meta-analysis has its limitations. In this case, five different sets of criteria were used to define gestational diabetes in the 15 studies; and although the targeted glucose values were similar among studies, they were not identical, and timing of postprandial glucose tests also varied. In addition, tests for heterogeneity cannot fully adjust for effects on outcomes driven by individual differences in patient adherence to study protocols or effects related to socioeconomic differences of subjects within each study. Furthermore, long term follow-up of both mothers and their offspring are needed to fully assess the safety as well as the benefits of these drugs when taken during pregnancy.