The FDA's SSRI pregnancy panel: what was claimed, what pushed back, what's undecided

On July 21, 2025, the FDA convened an expert roundtable on SSRIs and pregnancy [1]. Some panelists proposed adding warnings to SSRI labels — up to a boxed warning, the FDA's strongest. The American College of Obstetricians and Gynecologists called the discussion "alarmingly unbalanced." More than a year later, no label has changed and no decision has been announced [4].

Roughly one in eight American adults takes an antidepressant, most of them women, many of childbearing age — which means this unresolved fight sits directly over one of the most consequential medication decisions a person can face. Nearly everything written about it takes a side. This page's job is different: lay out what each side actually claims, what each claim stands on, and what remains genuinely undecided.

This is information to bring to your prescriber and your obstetric provider — not medical advice, and never a reason to stop a medication on your own, least of all during pregnancy.

What happened

The panel brought together ten experts — perinatal psychiatrists, epidemiologists, developmental researchers [1][3]. Several argued that pregnant patients aren't adequately warned about potential fetal risks of SSRI exposure and proposed labeling changes; a boxed warning was floated [4]. The panel reached no consensus, and the FDA has since said it is considering the question without announcing anything [4].

The composition became the story almost immediately: nine of the ten panelists had previously and publicly questioned SSRI safety or antidepressant use in general [3]. That fact doesn't make their claims wrong — but a panel assembled overwhelmingly from one side of a live scientific dispute will produce a one-sided discussion, and by most accounts it did.

The case the panel made, fairly stated

Steelmanned rather than caricatured: SSRIs cross the placenta; serotonin plays roles in fetal development; some studies have reported associations between prenatal SSRI exposure and outcomes including neonatal adaptation syndrome (real and generally transient), persistent pulmonary hypertension of the newborn (a small absolute risk with a regulatory history of its own), and contested signals on other outcomes. Panelists argued patients deserve those uncertainties on the label rather than discovering them in a journal. "Tell patients more" is not, on its face, an anti-patient position.

The pushback, fairly stated

The response from perinatal psychiatry and obstetrics [2][3] rests on two pillars.

First, the large-study record is more reassuring than the panel implied. The major meta-analyses and registry studies — the kind with tens or hundreds of thousands of pregnancies — have not found consistent overall increases in major birth defects, autism, or long-term neurodevelopmental harm once confounding by the underlying illness is accounted for [2]. Confounding is the crux: depressed and non-depressed pregnancies differ in many ways besides medication, and studies that compare siblings or control for maternal illness see most signals shrink or vanish.

Second, untreated depression is not a neutral baseline. It carries its own documented risks — to prenatal care, to birth outcomes, to postpartum illness, and at the extreme to maternal suicide, a leading cause of maternal death. A warning that frightens people out of treatment doesn't move them to safety; it moves them to a different set of risks that the panel, by most accounts, barely discussed [2][3]. That omission is what "alarmingly unbalanced" referred to.

There is also a precedent both sides invoke: the 2004 boxed warning on antidepressants and youth suicidality, which was followed by measurable drops in prescribing and — in some analyses — unintended harms from untreated depression. Whether that history argues for or against a new warning depends entirely on which risk you weight, which is rather the point.

What this means if the decision is yours

The panel and its critics agree on more than the coverage suggests: this is a risk-risk decision, individual to your history, and the worst outcome is not choosing either arm — it's an unplanned, abrupt discontinuation in the first trimester driven by fear.

• If you're planning a pregnancy: this conversation belongs before conception, with your prescriber and an obstetric provider in the same loop — severity of your history, relapse pattern, prior discontinuation attempts, and which specific drug (they are not interchangeable in pregnancy data).
• If you're pregnant and taking an SSRI: do not stop on your own. Withdrawal and relapse risks are real, and stability matters most now. Book the conversation instead.
• If you're weighing starting one while pregnant: ask the checklist questions plus the pregnancy-specific one — "what does the data say for this specific drug, and what does my depression severity say about the other side of the scale?" The severity question matters doubly here: benefit concentrates in more severe illness, which shifts the risk-risk arithmetic.

Where this stands

No labeling decision as of August 31, 2026. The FDA has the panel transcript, the public comments, and a regulatory precedent it knows cuts both ways. We update this page when the docket moves — including if the agency announces it will change nothing, which is a result the loudest voices on both sides won't advertise, and we will.

Every drug this decision touches has its own page in the medication library — approval history, verbatim label warnings, prescription volumes, and adverse-event reports with their limits stated first.