Development and internal validation of early- and mid-pregnancy prediction models for macrosomia in nulliparous singleton pregnancies

In 6,371 nulliparous singleton pregnancies in the nuMoM2b cohort, visit-updated models predict birthweight >4000 g with AUC 0.73 and >4500 g with AUC 0.82.

• Why it matters: Macrosomia is linked to adverse outcomes, but most tools depend on later ultrasound, limiting earlier counseling and monitoring.
• What they did: The study built visit-updated multivariable logistic regression models using maternal characteristics, anthropometry, periconceptional diet, and serial ultrasound, then assessed performance via repeated stratified cross-validation and compared against WHO fetal growth chart ultrasound percentiles.
• The result: Early-pregnancy maternal-data-only models showed moderate discrimination (AUC 0.63 for >4000 g; 0.72 for >4500 g), while later ultrasound improved accuracy; predictions for >4500 g are cautious due to 77 events and need external validation.

AI summary of the abstract · may contain errors · not medical advice · read the paper.