SSRI safety in pregnancy: what the new evidence actually says

The conversation happens regularly in general practice. A woman on an antidepressant discovers she is pregnant. She wants to know if she should stop. She has heard things — that SSRIs affect foetal development, that they are in the “not safe” category, full stop. And she is frightened.

That framing — “not safe” — carries more certainty than the evidence has ever supported. New research published in JAMA Network Open is making the nuance more visible, and it matters to how GPs navigate this question.

A team at Vanderbilt University Medical Center used a target trial emulation design — a methodology that mimics a randomised controlled trial from observational data — applied to linked maternal and neonatal electronic health records from over 1,000 pregnancies between 2006 and 2022. It is among the more methodologically rigorous approaches applied to this question in the perinatal literature.

---

What the study found

Continuing SSRIs during pregnancy was associated with two measurable effects at birth: a small reduction in Apgar scores at one and five minutes (0.39 and 0.28 points respectively) and a 73% increased odds of meconium staining (OR 1.73, 95% CI 1.22–2.45).

Both of these fall under what is known as neonatal adaptation syndrome — a pattern of transient effects observed in SSRI-exposed newborns in the days following birth. The symptoms are real: some infants show jitteriness, feeding difficulties, or mild respiratory changes. They typically resolve within a few days and rarely require treatment beyond monitoring and supportive care. They are not trivial to experience in the room. They are not, on the current evidence, indicators of lasting harm.

What the study did not find is clinically significant. There was no meaningful increase in NICU admission (OR 1.23, 95% CI 0.82–1.83). No significant increase in preterm birth. No significant increase in congenital anomalies overall (OR 1.09, 95% CI 0.63–1.90). The absolute risk of the serious outcomes most patients ask about — “will my baby have a birth defect, will my baby be early, will my baby be in intensive care” — was not meaningfully elevated.

The study carries real limitations. Sample size, while substantial, limited the power to detect rare malformations. Medication adherence during pregnancy could not be confirmed from prescription records alone. Depression severity data was available for only approximately 10% of participants, leaving residual confounding possible. The findings need to be held with appropriate uncertainty.

---

Both sides of this

The message here is not “SSRIs are safe in pregnancy.” Every pharmacological decision in pregnancy involves balancing risk on both sides of the ledger. The message is that the risk profile of SSRIs during pregnancy appears considerably more nuanced than the “not safe” shorthand that often reaches patients — and that the counterfactual matters.

Untreated depression during pregnancy carries significant risks of its own: preterm birth, low birthweight, impaired maternal-infant bonding, and worse postpartum outcomes. The woman who stops her antidepressant abruptly because she is frightened — without a structured clinical conversation about what she is trading — is not removing risk from the picture. She is shifting from one risk profile to another, often without knowing it.

There is a harder pattern embedded in this. Many patients, and some clinical systems, treat “removing medication” as the cautious default position in pregnancy. For women with stable moderate-to-severe depression on a treatment that is working, abrupt discontinuation is not cautious — it is a clinical risk that has not received the same systematic attention as the medication’s effects.

The counterfactual is invisible in a way that the medication is not. Neonatal adaptation syndrome at birth has a name and is attributable. Relapsed depression in the second trimester is attributed to the illness, not to the decision to stop. That asymmetry shapes how risk is communicated to patients, and it needs to be named.

---

What this means this week

If you are pregnant and currently on an antidepressant, this research is not a green light and it is not an alarm. It is context for a conversation with your GP or psychiatrist — one that should happen before decisions are made, not after.

For GPs, several practical points follow from the evidence:

Preconception conversations are the right moment. For women of reproductive age on SSRIs for depression, a proactive discussion before conception — covering what to expect, what the evidence shows, how to monitor for neonatal adaptation — allows an informed decision rather than a reactive one. It also makes the conversation far less fraught when a pregnancy is confirmed.

Births of SSRI-exposed infants should occur at a facility equipped for neonatal resuscitation — not because serious complications are likely, but because the team should be prepared to monitor for neonatal adaptation effects and not be surprised by them.

The ‘stop everything’ assumption needs active challenge. For mild depression with good psychosocial supports and other management options in place, a tapering plan before conception may be appropriate. For moderate-to-severe depression, the evidence for continuing a treatment that is working is robust. These are different clinical scenarios and should be discussed individually.

Depression in the perinatal period is common and undertreated. The fear that drives women to stop their medication without clinical guidance may, on the available evidence, be doing more harm than the medication itself.

---

Verdict: yes — this changes how GPs should frame the SSRI-in-pregnancy conversation.

More nuance, not less medication. The difference matters.

---

Sources cited

1. SSRI pregnancy fears may be overstated — Medical Republic, September 2026
2. JAMA Network Open — AMA peer-reviewed open-access journal