Abstract Study question Does periconceptionally exposure to GLP-1 receptor agonist (GLP-1 RA) treatment increase the risk of pregnancy complications? Summary answer Perconceptionally exposure to GLP-1 RA had no association with pregnancy complications when compared to a control group with balanced covariates. What is known already The prevalence of obesity continues to increase globally, affecting half the adult population in Western countries, increasing the risk of infertility and obstetric and neonatal complications. Weight loss might improve risks, leading to an interest in effective weight loss methods such as GLP-1 RAs. However, none of the GLP-1 RAs are approved for use during pregnancy and at least two months before to avoid fetal toxicity. The current recommendations are based on a fetal safety principle and data from animal studies, clearly underscoring the need for more human evidence about the effect of GLP-1 RA use in the periconceptional period. Study design, size, duration A Danish nationwide observational cohort study of 522,209 pregnancies from October 2009 until June 2019 with data from health and socioeconomic registries at Statistics Denmark: the National Prescription Register, National Patient Register, Birth Register, Residential Register, and the National Diabetes Registry. All registries had complete information until June 2019, except the National Prescription Register which was until December 2018. The unique personal identification number used in Denmark enables direct linkage of individual-level information across registries. Participants/materials, setting, methods Exposure was defined as the redemption of GLP-1 RA prescription, liraglutide, +/- eight weeks from last menstruation. Primary outcome: any complication (pre-eclampsia, gestational diabetes (GDM), preterm birth (<37 + 0), low-birth-weight (<2500g), small-for-gestational-age, or large-for-gestational-age (LGA)). Evaluated risk of individual outcomes, and effect on gestational age, birth weight, and placenta weight. Outcomes were compared to an unexposed control group, weighted using fast generalized matching on propensity score estimated from relevant confounders. Risk ratio, 95% confidence interval (CI). Main results and the role of chance Within the study population of 522,209 pregnant women, we identified 93 pregnancies exposed to liraglutide in early pregnancy. The exposed women were on average older, had a higher BMI (on average 40% higher), and had a sixty times higher frequency of diabetes (DM1 or DM2) compared to unexposed women giving birth in the same period (p < 0.001). We found that liraglutide exposure was associated with a higher risk of obstetrical complications, pre-eclampsia, GDM, and preterm birth. There was no difference in the birth weight, but those exposed were more likely to give birth to an LGA child (liraglutide, p < 0.01). However, we did not find any evidence of increased risk when utilizing a propensity score matching method (liraglutid, p > 0.1). This indicates that liraglutide, in itself, is not driving the increased risk, but the indication for a GLP-1 RA prescription (e.g. diabetes or obesity) is the causal factor. The finding did not depend on the choice of method for propensity score matching. Findings pertaining to liraglutide and semaglutide, encompassing the period from 2009 to 2023, will be showcased at the primary conference. Limitations, reasons for caution Weight loss treatment was off-label for most of the study time. Wider implications of the findings The study is important for counseling women periconceptionally exposed to GLP-1 RA, that there is no evidence of an increased risk of pregnancy complications. It calls for studies on the safety profile for GLP-1 RA treatment in women, enabling early weight management for the benefit of the mother and child. Trial registration number not applicable
Read full abstract