Does Ozempic cause birth defects? Separating fear from evidence

What did @realdrbae actually say?

The core claim is this: a JAMA study found that women who stopped GLP-1 medications within the first trimester showed "no statistically significant increase in major congenital malformations or birth defects." The creator also said the standard recommendation to stop two months before conception comes from animal studies by Novo Nordisk, not human data. And they added that for type 2 diabetes, continuing these medications through pregnancy may be appropriate if benefits outweigh risks.

That is a lot of ground to cover in under two minutes. To the creator's credit, they flagged the animal-study origin of the precautionary guidance, which many providers and patients don't realize. They also correctly told viewers not to try to lose weight during pregnancy. But there are real gaps in what got said, and one framing choice that deserves scrutiny.

Does the science back this up?

The JAMA study the creator references is almost certainly Patorno et al., 2024, published in JAMA Internal Medicine, which used U.S. insurance claims data on pregnancies with first-trimester GLP-1 exposure. The study did find no statistically significant increase in major congenital malformations compared to unexposed pregnancies, which aligns with what the creator said. However, that same study and related research flagged elevated risks for other outcomes.

Specifically, Cesta et al., 2024, in JAMA Internal Medicine using Scandinavian registry data, found associations between first-trimester semaglutide exposure and increased risk of fetal growth restriction and preterm birth, though not malformations. The FDA's labeling for semaglutide still carries a warning against use in pregnancy based on animal reproductive toxicity data showing embryo-fetal deaths and structural abnormalities at doses relevant to human exposure. The animal data is not nothing. Citing it as the sole reason for caution and then essentially dismissing it oversimplifies the picture.

What did they get wrong (or right)?

They got the malformation finding right. On that specific outcome, the human observational data so far is reassuring. Credit where it is due.

What they got wrong, or at least incomplete: framing the two-month washout recommendation as purely animal-study-driven caution that the JAMA data now counters. That is a misleading contrast. The JAMA studies are retrospective and observational. They cannot establish causation, and the sample sizes for specific defect subtypes are not large enough to rule out smaller risk signals. The creator did not mention the fetal growth and preterm birth concerns from the same body of literature.

The claim that these medications can be continued through pregnancy for type 2 diabetes if benefits outweigh risks is also not standard guidance right now. Current clinical consensus, including from the American Diabetes Association, does not recommend GLP-1 receptor agonists during pregnancy for type 2 diabetes. Insulin remains the preferred option. That is not a minor caveat. Saying providers may "reasonably recommend" continued use without that context could mislead a pregnant patient with diabetes into thinking this is common practice.

What should you actually know?

If you are on a GLP-1 medication and become pregnant or are trying to conceive, the current standard recommendation is to stop the medication. The two-month pre-conception washout is based on semaglutide's long half-life, roughly one week, meaning it takes several weeks to clear the body. That pharmacokinetic reasoning is valid independent of whether the animal studies prove human harm.

The human data on birth defects specifically is somewhat reassuring so far, but the research is new, the datasets are not complete, and other pregnancy outcomes like growth restriction deserve attention. The creator's summary is not wrong on the malformation point, but presenting it as a clean reassurance is premature. Anyone on these medications who is pregnant or planning pregnancy should be talking to an OB and their prescribing provider, not making decisions based on a TikTok summary of one study, however well-intentioned.