SSRIs During Pregnancy: Weighing the Real Risks and Benefits

The cultural framing of this decision assumes that stopping medication is the “safe” choice and staying on medication is the “risky” one. This framing is clinically inaccurate. Untreated or inadequately treated perinatal depression and anxiety carry real, measurable risks to both the pregnant person and the developing fetus — risks that are often larger in absolute terms than the medication risks that receive more attention.

The decision framework is not medication vs. safety. It is medication risks vs. untreated illness risks, weighted against each individual's psychiatric history, severity of prior episodes, relapse speed, and support system.

The evidence on untreated perinatal mood and anxiety disorders includes the following well-documented risks:

• Preterm birth — depression during pregnancy is associated with a significantly elevated risk of preterm delivery through HPA axis activation and inflammatory pathways
• Low birth weight — elevated cortisol from untreated depression affects fetal growth and uterine blood flow
• Inadequate prenatal care — depression impairs the energy and motivation required to attend appointments, take prenatal vitamins, maintain nutrition, and engage with prenatal healthcare
• Substance use — untreated depression and anxiety increase the likelihood of substance use as self-medication, which carries far greater fetal risk than most psychiatric medications
• Amplified postpartum depression — perinatal depression is the strongest predictor of severe postpartum depression; untreated antenatal illness increases the postpartum period's risk substantially
• Impaired bonding — depression during pregnancy and postpartum impairs the mother-infant attachment process, with developmental consequences for the child
• Maternal suicide risk — suicide and accidental overdose are leading causes of maternal mortality in the United States; undertreated severe depression during pregnancy is a serious mortality risk

SSRIs are among the most-studied medications in pregnancy. We have more than 30 years of data, including large population-based registry studies, prospective cohort studies, and systematic reviews. The overall picture from this body of evidence:

• Most SSRIs were classified as Category C under the old FDA system (animal studies show risk; adequate human data limited but clinical benefit may outweigh risk)
• Under the current FDA labeling system (which replaced A/B/C/D/X categories), prescribers are required to weigh individual risk-benefit and consult available data — which for most SSRIs is substantial
• Sertraline (Zoloft) and escitalopram (Lexapro) have the most data in pregnancy and are generally considered preferred first-line options

Sertraline: The Preferred Pregnancy SSRI

Sertraline has the largest safety database of any SSRI in pregnancy. It crosses the placenta minimally compared to other SSRIs, produces relatively low levels in breast milk, and has not been consistently associated with major structural malformations in large, well-controlled studies. For a pregnant patient starting antidepressant treatment, sertraline is usually the first-choice SSRI.

Persistent Pulmonary Hypertension of the Newborn (PPHN)

In 2006, the FDA issued a warning about a potential association between SSRI use in late pregnancy and PPHN — a serious condition in which the newborn's blood vessels do not dilate properly after birth. This warning generated significant anxiety among pregnant patients and providers.

The updated picture from subsequent large studies (2012 and later) is more reassuring. The absolute risk of PPHN in the general population is approximately 1–2 per 1,000 births. In SSRI-exposed pregnancies, multiple large studies have found no significant elevation, and those that found a small increase found absolute risks in the 2–3 per 1,000 range — a very small elevation in an already rare condition.

Cardiac Defects: Why Paroxetine Is Avoided

Paroxetine (Paxil) is generally avoided during pregnancy because of consistent evidence of a small increased risk of cardiac septal defects. This association is more consistently replicated across studies than the PPHN data and represents a genuine reason to avoid paroxetine specifically in pregnancy — not all SSRIs. If a patient is on paroxetine and becomes pregnant, a switch to sertraline or escitalopram is usually recommended after careful risk-benefit discussion.

Neonatal Adaptation Syndrome

When SSRIs are used close to delivery, some newborns experience a transient period of adjustment that includes jitteriness, feeding difficulty, mild respiratory changes, and irritability. This is called neonatal adaptation syndrome (sometimes called poor neonatal adaptation syndrome or neonatal abstinence syndrome in this context). Critically:

• It is transient — typically resolving within 48–72 hours
• It is not permanent harm — long-term neurodevelopmental outcomes are not affected
• It is manageable with monitoring in the hospital, not a medical emergency in the vast majority of cases

Some providers consider a gradual dose reduction near the due date to minimize this; others do not. This is a conversation to have with your obstetric and psychiatric providers together.

First vs. Second/Third Trimester Risk Windows

Structural malformation risk (cardiac defects, neural tube defects) is concentrated in the first trimester, when organ systems are forming. Second and third trimester exposure carries different risk profiles — primarily neonatal adaptation syndrome and the PPHN data (which is specifically associated with third-trimester exposure). The timing of pregnancy when discussing medication changes matters significantly.

Several early studies reported associations between prenatal SSRI exposure and autism spectrum disorder, generating significant concern. The critical methodological issue in these studies is confounding by indication: maternal depression itself is associated with elevated autism risk (likely through shared genetic pathways and inflammatory mechanisms), making it statistically very difficult to separate the effect of the SSRI from the effect of the condition the SSRI is treating.

The most rigorous studies — including large sibling-controlled and propensity-matched analyses that control for maternal psychiatric history — have found no independent SSRI effect on autism risk after accounting for the mother's depression or anxiety. The apparent association in earlier studies appears to be explained by the underlying condition, not the medication.

Venlafaxine (Effexor) — SNRI

Venlafaxine, an SNRI (serotonin-norepinephrine reuptake inhibitor), has a pregnancy safety profile broadly similar to SSRIs. It is not a first-line choice simply because sertraline has more data, but for patients who are doing well on venlafaxine, continuation is a reasonable discussion. Venlafaxine does have a more significant discontinuation syndrome than most SSRIs, which is relevant if dose adjustments near delivery are considered.

Bupropion (Wellbutrin)

Bupropion's safety data in pregnancy is more mixed. Some studies have raised concerns about a possible cardiac septal defect association — similar to paroxetine, though less consistently replicated. Bupropion is generally not a first-line choice in pregnancy, particularly in the first trimester, and this is a specific conversation to have with your prescriber if you are currently taking it.

The decision about continuing, stopping, switching, or adjusting psychiatric medication in pregnancy should be individualized and shared between patient and provider. The key variables:

• Severity of prior episodes — a patient with mild, reactive depression responds differently to this question than a patient with severe MDD who was hospitalized without medication
• Speed of relapse — some patients relapse within days to weeks of stopping medication; others have longer windows. This is critical information for the decision.
• Current trimester — the risk calculus differs by trimester (structural risks in the first, neonatal adaptation syndrome more relevant near delivery)
• Support system — a patient with a strong support network and access to intensive therapy may have more flexibility than one who is isolated and has not responded to therapy alone
• History of suicidality or severe impairment — significantly elevates the risk calculation of stopping medication

Why Abrupt Stopping Is Almost Always Wrong

“Cold turkey” discontinuation of SSRIs is not recommended at any time, and especially not during pregnancy. Abrupt stopping carries significant discontinuation syndrome risk (flu-like symptoms, “brain zaps,” severe mood dysregulation), increases the risk of rapid relapse, and removes the gradual taper that allows monitoring. If a decision is made to stop, it should involve a supervised taper with your prescriber.

The breastfeeding picture differs from the pregnancy picture in one notable way. During pregnancy, the primary concern is the medication crossing the placenta. For breastfeeding, the concern is the amount transferred to the infant through breast milk.

Sertraline is the preferred SSRI for breastfeeding: it consistently shows the lowest levels in breast milk relative to maternal dose, and studies of nursing infants of sertraline-taking mothers show negligible measurable levels in infant serum.

An important exception to note: Paroxetine — which is generally avoided during pregnancy due to cardiac concerns — actually has among the lowest breast milk transfer rates of any SSRI. For a patient who was on paroxetine before pregnancy and needs to restart postpartum while breastfeeding, it is one of the preferred options from a breastfeeding safety standpoint. This apparent contradiction often surprises patients; the pregnancy risk and the breastfeeding risk profile can differ significantly.

Fluoxetine has a longer half-life and its active metabolite norfluoxetine accumulates more in breast milk than sertraline. It is generally not a first choice for breastfeeding mothers, though the absolute levels in most studies are still low.

A productive shared-decision conversation about psychiatric medication in pregnancy should cover:

• Your psychiatric history — specifically, how quickly you relapsed when you stopped medication in the past, and how severe the episode was
• Specific medication data — ask about the specific SSRI you are on (not “antidepressants in general”), its specific pregnancy data, and whether a switch to sertraline or escitalopram would be appropriate
• Absolute risk numbers — ask for absolute risk percentages, not just relative risk ratios. “What is the actual probability of this outcome, and how does it compare to the untreated baseline?”
• The untreated illness risks — explicitly ask your provider to walk through the risks of untreated depression or anxiety on your pregnancy, not just the medication risks
• Timing and trimester considerations — what adjustments, if any, are recommended for different stages of the pregnancy