Early pregnancy GDF15 trajectories and their association with normal gestation, pregnancy loss, and hyperemesis gravidarum.

Abstract

Growth differentiation factor 15 (GDF15) rises dramatically in early pregnancy and has been linked to nausea, vomiting, and hyperemesis gravidarum (HG). Significant challenges remain as GDF15 levels overlap between normal and complicated pregnancies, limiting its current clinical utility as a biomarker. Furthermore, early pregnancy GDF15 trajectories remain poorly characterised because most studies lack longitudinal sampling and rely on single or infrequent measurements. We analysed longitudinal serum GDF15 samples and Pregnancy Unique Quantification of Emesis 24 (PUQE-24) questionnaire scores from women enrolled in two early pregnancy clinical trials (PEP Cohort and VOMIT Study), using robust nonlinear statistical models to characterise trajectories. Analyses included uncomplicated (reference) pregnancies, early pregnancy loss (EPL), and HG. In normal pregnancies, GDF15 rose dramatically until week 7, and plateaued by week 9, with a sigmoidal model fit (P < 0.01). Pregnancies with female foetuses had approximately 19% higher GDF15 (P < 0.01), and the rise occurred one day later in male foetuses (P < 0.02). PUQE-24 scores in normal pregnancy peaked around week 8 and declined thereafter, aligning with the GDF15 plateau. In women with HG, GDF15 plateaued by the time of diagnosis, with higher levels compared to normal pregnancies in the early window (P < 0.05) but overlapping confidence intervals by week 8. EPL showed lower GDF15 levels and a slower rise, but also heterogeneity between types of EPL. GDF15 trajectories in early pregnancy diverge by outcome and foetal sex. These patterns clarify GDF15 dynamics in normal pregnancy and complications such as HG and EPL and define optimal sampling windows and tentative reference intervals. These findings support further evaluation of GDF15 as a biomarker to improve diagnosis and guide treatment. Independent Research Fund Denmark. The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing.