Reference: Hum Reprod Open. 2026 Sep 8;2026(3):hoag073
Practice Point: The body of evidence on acetaminophen use in pregnancy still supports its safety.
EBM Pearl: The use of validated surrogate markers rather than clinical outcomes automatically diminishes a study’s validity; the use of unreliable surrogate markers in a surrogate population is little more than a distraction.
Maybe you’ve heard of the children’s book series If You Give a Mouse a Cookie, which in its purest sense depicts a charming story about the absurdity of leaps in logic using the example that if you give a mouse a cookie, it’ll eventually need a broom to clean up after its haircut. Well, back from our summer hiatus, we’re jumping right into the cynicism you know and love with a recommendation for the next story in the series: a cohort study from Denmark reporting that fetal exposure to paracetamol (acetaminophen) during pregnancy is associated with reduced markers of fertility in female infants and adolescents. Mass media and clinical email alerts alike already have a head start with headlines like "Girls born to mothers who took acetaminophen in pregnancy found to have reproductive organ changes that could affect fertility later in life."
The primary analysis was a prospective cohort of 302 infant daughters evaluated during "mini-puberty" at age ~106 days. Investigators assessed the physical exam, ultrasound-measured ovarian volume, antral follicle count (AFC), uterine width, and reproductive hormones, including anti-Müllerian hormone (AMH), and then stratified the infants by prenatal paracetamol exposure based on maternal report and urine concentrations. The median dose was 2 g per use (!). Results showed a time- and dose-dependent decrease in AFC with increasing exposure. A separate "independent confirmatory cohort" included 1,210 girls studied between 1997 and 2002 for a different clinical question but who had data from physical examinations and MRI assessments performed from infancy through adolescence. Results there were "confirmatory" and showed that among those exposed to paracetamol and NSAIDs in utero, there were reduced uterine and ovarian volumes and lower inhibin B and follicle-stimulating hormone levels at puberty.
First of all, we found ourselves asking, "Why was this even studied?" Turns out this wasn’t necessarily a fishing expedition like we may have first assumed. The authors were building on several studies in mice and rats that showed subfertility in female rodent offspring exposed to acetaminophen prenatally. These studies provided an antiandrogenic/germ-cell mechanism and a biologically plausible hypothesis the authors thought was worth investigating in humans. Fair enough—study away. But in order to actually see what happens in humans in the same circumstances, you have to design a study that has a chance of offering up valid conclusions.
We’ll start our list of grievances with the fact that none of the markers used to assess future fertility of infants or adolescents are remotely validated or reliable for this purpose. AMH is dynamic in infancy and often undetectable at birth, only becoming a reliable surrogate of ovarian reserve at age 25. AFC is useful for measuring ovarian response to stimulation during infertility treatments; otherwise, it is an unreliable predictor of natural fertility. Ovarian volume and uterine width are highly variable in infancy; established utility is for detecting masses and Müllerian anomalies, not predicting fertility. Inhibin B is the least reliable marker in the group. Moreover, there is a well-documented follicular/oocyte reorganization that happens around puberty, which makes markers of ovarian reserve hard to interpret in adolescents, too.
Secondly, if you want to know if something affects fertility, you really need a population of people ready to give birth. Call us idealists, but infants are hardly the population in question here, although they are obviously quicker to study. We get that surrogate markers or populations are necessary sometimes, but it does affect how much or little we can do with the information we gain from studying them.
Third, the paracetamol (acetaminophen) dosing the pregnant patients recorded, a median of 2 g per use, is really high. We can’t extrapolate the effects of essentially overdosing a medication to what happens with recommended use.
So, if we wrote the next book in this series based on the leap in logic that is hinted at in this study, it would go something like "If you give a mouse a Tylenol, humans may no longer be able to procreate." That is absurd, of course. But a story has nonetheless been assembled from adult mouse reproductive outcomes, a biologically plausible hypothesis, and observational human infant data based on unreliable surrogate markers and has then been published in a medical journal, passed around in mass media, and consumed by our patients. The trouble is, the evidence is not there to back it up, and now we’re the ones who need a broom to clean up the mess.
For more information, see the topic Medication and Drug Exposure in Pregnancy in DynaMed.
DynaMed EBM Focus Editorial Team
This EBM Focus was written by Katharine DeGeorge, MD, MS, MSCP, Executive Editor at DynaMed and Professor of Family Medicine at the University of Virginia. Edited by Alan Ehrlich, MD, FAAFP, Executive Editor at DynaMed and Associate Professor in Family Medicine at the University of Massachusetts Medical School; McKenzie Ferguson, PharmD, BCPS, Senior Clinical Writer at DynaMed; Rich Lamkin, MPH, MPAS, PA-C, Senior Clinical Writer at DynaMed; Claire Symanski, PhD, Medical Editor and Team Lead for ENT at DynaMed; Michael Butler, PhD, Medical Writer at DynaMed; Matthew Lavoie, BA, Senior Medical Copyeditor at DynaMed; Hannah Ekeh, MA, Senior Associate Editor II at DynaMed; and Jennifer Wallace, BA, Senior Associate Editor at DynaMed.