Neonatal outcomes in gestational diabetes trials: What clinicians should consider when interpreting treatment effects.

Abstract

Determining whether treatment of gestational diabetes mellitus benefits the newborn is central to clinical guidelines and patient counseling. However, the conclusions drawn from major trials may depend not only on the biological effect of treatment, but also on the neonatal outcomes and fetal growth standards selected. Composite endpoints can obscure or exaggerate treatment effects when they combine events that differ in clinical importance, frequency, or responsiveness to treatment. Similarly, different definitions of fetal overgrowth can substantially alter the reported prevalence of large- and small-for-gestational-age infants and therefore the apparent balance between treatment benefit and possible overtreatment. This review examines landmark and illustrative studies, including ACHOIS, MFMU, HAPO, TOBOGM, DipGluMo, Big Baby, and GRACE, together with methodological literature on composite outcomes and birth weight classification. Across these studies, treatment effects were sometimes driven by a single frequent component, masked by biochemical or rare outcomes, or altered by the growth chart applied. Absolute birth weight thresholds are transparent but do not account for gestational age or infant sex, whereas percentile-based definitions provide adjustment but yield different results across GROW, INTERGROWTH-21st, WHO, and other standards. Clinicians and investigators should therefore assess neonatal benefit using clinically coherent outcomes reported individually, together with prespecified measures of fetal growth and potential treatment-related harm. Absolute birth weight thresholds and percentile-based measures, including both large- and small-for-gestational-age outcomes, should be reported together, with explicit specification of the growth chart and sensitivity analyses using alternative standards.