GLP-1 exposure in early pregnancy: new study
Why this study matters
Use of glucagon-like peptide-1 receptor agonists has expanded among women of reproductive age, while evidence about exposure during pregnancy remains limited. A study published in Annals of Internal Medicine examined whether continuing GLP-1 receptor agonist dispensing into early pregnancy was associated with non-live birth, abnormal fetal growth or major congenital malformations.
The researchers used US Merative MarketScan insurance claims from 2011 to 2024. They identified 3,572 pregnancies among women aged 16 to 55 who had a GLP-1 medicine dispensed in the 90 days before their last menstrual period. Of these pregnancies, 41.1% were among women with type 2 diabetes.
What the researchers reported
The study emulated a trial comparing two dispensing strategies: at least one further GLP-1 dispensation during the first trimester, or no continuation after the pre-pregnancy period. The weighted risk of non-live birth was 29.7% with continuation and 27.1% with noncontinuation. The adjusted risk ratio was 1.09, with a 95% confidence interval from 0.98 to 1.23.
Among live-birth pregnancies linked to infant records, the weighted prevalence ratio for major congenital malformations was 1.21 (95% confidence interval 0.83 to 1.82). It was 1.29 (0.82 to 2.06) for babies small for gestational age and 1.08 (0.84 to 1.40) for babies large for gestational age.
The authors concluded that these outcomes were not definitively more frequent with continuation into early pregnancy. That wording matters: it is not the same as establishing safety. The confidence intervals for major congenital malformations and small-for-gestational-age births were wide enough to remain compatible with no increase or with clinically relevant differences.
What the study cannot establish
This was an observational analysis of US claims, not a randomised trial or a UK cohort. A dispensation records access to medicine, not whether or when every dose was taken. The comparison was also between continuation and noncontinuation among women dispensed a GLP-1 medicine shortly before pregnancy; it was not a comparison with an otherwise equivalent group with no recent exposure.
The authors identified potential residual confounding from prior glycaemic control. Underlying diabetes, obesity, medicine indication and other differences can affect pregnancy outcomes independently of GLP-1 exposure. Infant linkage was available for only part of the live-birth cohort, adding further limits to interpretation.
UK regulatory and governance implications
The study provides useful evidence for an area with sparse human data, but it does not override UK product information or regulatory advice. The MHRA states that GLP-1 medicines should not be taken during pregnancy, while trying to become pregnant or during breastfeeding. Its published guidance also addresses contraception and medicine-specific waiting periods before a planned pregnancy.
For UK weight-management and prescribing services, the paper is therefore relevant to governance rather than a basis for changing treatment. It reinforces the importance of applying current product information, recording pregnancy-status discussions and maintaining clear routes for patients to report a suspected or confirmed pregnancy. The MHRA's Yellow Card guidance provides the official route for reporting suspected medicine side effects and safety concerns. Any clinical decision remains for the patient and an appropriately qualified healthcare professional.
Key takeaway
In this claims-based study, continuation of GLP-1 dispensing into the first trimester was not associated with a definitive increase in non-live birth, abnormal fetal growth or major congenital malformations. However, imprecision, possible confounding and the limits of dispensing data mean the findings cannot establish safety. Current UK restrictions remain the relevant standard for clinics.