The standard advice to anybody on one of these drugs who wants to become pregnant is to stop first — and what is known about preventing pregnancy on them is thinner than that advice assumes. This study is the first thing this desk has read suggesting that stopping earlier may not remove the risk it is meant to remove — and it is not an argument for doing anything else.
Three groups
Using a national dataset linking medical records with pharmacy dispensing, women with prepregnancy overweight or obesity delivering between January 2022 and January 2026 were sorted into three groups: exposed before and into pregnancy, former users who stopped before conceiving, and nonusers. [1] Groups were matched one to one on maternal age, race and ethnicity, prepregnancy BMI, prepregnancy diabetes and prepregnancy hypertension, then adjusted further in regression.
429 women were exposed into pregnancy, with a median total exposure of 44 days — mostly early, often before a pregnancy was known. 801 had stopped beforehand. 2,203 had never taken it.
What was found
Against nonusers, the exposed group had higher adjusted odds of excessive gestational weight gain, aOR 2.88, 95% CI 2.04 to 4.05; gestational diabetes, 1.59, 95% CI 1.03 to 2.44; excessive fetal growth, 1.78, 95% CI 1.03 to 3.08; and cesarean delivery, 3.35, 95% CI 2.15 to 5.22.
The women who had stopped before conceiving looked much the same: excessive gestational weight gain 1.98, 95% CI 1.56 to 2.51; gestational diabetes 1.43, 95% CI 1.05 to 1.93; excessive fetal growth 1.54, 95% CI 1.01 to 2.36; and cesarean delivery 3.92, 95% CI 2.81 to 5.47. There was no statistically significant difference between the two drug groups on any outcome.
The rebound reading
The authors interpret the pattern as rebound after stopping rather than as harm from exposure in the womb, and the logic is sound: if intrauterine exposure were driving the outcomes, the group exposed during pregnancy should have fared worse than the group that stopped. They did not.
That fits what is known about what happens after these drugs are discontinued — weight returns at roughly a kilogram a month, and a pregnancy is nine months long. Rapid regain during pregnancy is a plausible route to excessive gestational weight gain, gestational diabetes and a larger baby. Plausible is not demonstrated, and this study did not measure the rebound directly.
What it does not say
That anybody should keep taking a GLP-1 while pregnant. These drugs are not recommended in pregnancy, no group in this study continued them deliberately, and nothing here tests that option. The comparison is between two ways of stopping.
It also does not identify what should be done instead. Slower tapering, earlier preconception planning, structured support during the transition — all plausible, none tested here, and all of them decisions for an obstetrician rather than a website. The space between continuing and stopping has more options than two, and none has been studied in this situation.
Why it matters for people buying online
Because unintended pregnancy on these drugs is not rare. A median exposure of 44 days describes women who found out they were pregnant and stopped — which is the common case, not an unusual one.
Anybody who could become pregnant while taking one of these drugs should be having this conversation before they start, not after. It is also a real cost that never appears in a monthly price: a pregnancy after stopping may carry a higher chance of a cesarean delivery and gestational diabetes, neither of which is in anybody’s arithmetic.