Psychiatric Medication During Pregnancy: What the Evidence Actually Shows

The most important thing to understand: untreated mental illness in pregnancy is also a risk. The choice is never “medication risk vs. no risk.” It is always “medication risk vs. illness risk.” Untreated depression in pregnancy is associated with preterm birth, low birth weight, impaired prenatal care adherence, and postpartum complications. Untreated anxiety can drive behaviors (substance use, restriction of prenatal care) that are more harmful than the medications being avoided. The risk-risk framework is the only honest way to approach this decision.

The FDA's old A/B/C/D/X pregnancy category system — a simple letter grade — has been replaced by narrative labeling that provides actual evidence summaries and is far more useful clinically. The old categories persisted in popular culture long after they were retired and continue to mislead patients and even some providers. A medication rated “C” under the old system did not mean it was dangerous — it usually meant there was simply limited data.

An important population-level reality: approximately 50% of pregnancies in the United States are unplanned. This means many people are exposed to psychiatric medications during the critical first trimester before they know they are pregnant. The research on accidental first-trimester exposure is a significant part of what we know about medication safety — and it is generally reassuring for most medication classes.

SSRIs and SNRIs are the most commonly used psychiatric medications in pregnancy — and they have more safety data than any other psychiatric medication class in this context.

The overall picture: SSRIs are broadly considered low-risk in pregnancy. The two most studied — sertraline (Zoloft) and escitalopram (Lexapro) — have the most evidence and are generally preferred as first-line choices when starting or maintaining an antidepressant during pregnancy. Fluoxetine has more data than most, though a longer half-life is a consideration.

What to Know About Specific Risks

• Neonatal Adaptation Syndrome (NAS) — some newborns exposed to SSRIs near delivery show temporary symptoms: jitteriness, feeding difficulties, irritability. This resolves on its own, typically within days, and is not associated with long-term harm. It is not a reason to stop medication during pregnancy; it is a reason to inform your delivery team.
• Persistent Pulmonary Hypertension of the Newborn (PPHN) — early studies suggested a possible association with late third-trimester SSRI use; subsequent research has found the absolute risk to be very small (approximately 3 per 1,000 births, compared to approximately 2 per 1,000 in the general population), and causality has not been firmly established. This risk does not override the benefit of treating significant depression or anxiety in most cases.

For context on starting or managing psychiatric medication more broadly: Starting Psychiatric Medication: What to Expect. And for more on anxiety specifically during pregnancy: Anxiety & Pregnancy.

This category requires more nuance because the medications vary significantly in their risk profiles.

Benzodiazepines

Benzodiazepines (lorazepam, clonazepam, diazepam) are generally not first-line during pregnancy and are avoided in the first trimester when possible. Earlier studies suggested cardiac malformation risk; subsequent evidence has not consistently confirmed a significant causal effect at low doses. However, benzodiazepine use near delivery is associated with neonatal sedation and withdrawal. Short-term use for acute anxiety during pregnancy may be appropriate in specific circumstances — this is a risk-benefit decision to make with your prescriber.

Mood Stabilizers

• Lithium — earlier studies suggested a significant cardiac malformation risk (Ebstein's anomaly); more recent, larger studies suggest the actual risk is much lower than originally thought, though not zero. Lithium remains a consideration for bipolar disorder in pregnancy when the illness severity warrants it, with cardiac screening (fetal echocardiogram at 16–20 weeks) and careful monitoring of lithium levels throughout pregnancy. Levels can fluctuate significantly as blood volume changes.
• Valproate (Depakote) — avoid during pregnancy. Valproate is a significant teratogen associated with neural tube defects, cardiac malformations, and neurodevelopmental effects including lower IQ in exposed children. It is contraindicated in pregnancy unless there is no effective alternative and the severity of illness makes it necessary. Women of reproductive age on valproate should use effective contraception and discuss transition planning with their prescriber.
• Lamotrigine — generally considered the preferred mood stabilizer in pregnancy for people with bipolar disorder. Reasonable evidence for relative safety; no consistent signal for major malformations at typical doses. Important caveat: lamotrigine levels drop significantly during pregnancy due to increased metabolic clearance, meaning doses often need to be increased during pregnancy and then reduced postpartum to avoid toxicity.
• Atypical antipsychotics (quetiapine, olanzapine, aripiprazole) — mixed data; no consistent signal for major malformations, but metabolic effects (gestational diabetes risk with olanzapine) are relevant. Used when needed for bipolar disorder or psychosis when the benefit outweighs the risk.

ADHD medication in pregnancy is an area where data is limited but growing, and where individual decision-making matters most.

Stimulants (Amphetamine Salts, Methylphenidate)

Stimulants are not categorically forbidden in pregnancy, but they are not well-studied either. The available data does not show a consistent major malformation signal. However, some studies have suggested associations with preterm birth and lower birth weight at higher doses. Many women choose to pause stimulants during pregnancy, particularly in the first trimester — especially if ADHD symptoms are manageable through behavioral strategies during that period. Others, particularly those for whom untreated ADHD poses significant functional risks (driving safety, job performance, self-care), make a risk-benefit decision to continue under close monitoring.

Non-Stimulant Options

• Atomoxetine (Strattera) — generally advised to avoid during pregnancy; limited safety data and some animal study concerns. Not a preferred option.
• Behavioral and non-pharmacological strategies — organizational tools, environmental modifications, therapy (particularly CBT for ADHD), and coaching are worth maximizing during pregnancy as adjuncts or substitutes for medication when possible.
• Bupropion (Wellbutrin) — sometimes used for ADHD and has more pregnancy data than atomoxetine; also addresses depression if comorbid. The evidence on cardiac malformation risk at standard doses is reassuring but not conclusive. A reasonable option to discuss with your prescriber for mild-to-moderate ADHD.

The framework that makes this decision most navigable is called shared decision-making — a clinical process in which the prescriber provides evidence and guidance, and the patient brings their values, preferences, and life context to the table. Neither party can make this decision well without the other.

The Risk-Risk Framework

The question is never “is this medication risky?” The question is: “Is the risk of taking this medication greater or less than the risk of not taking it — given my specific diagnosis, its severity, my history, and what alternatives exist?” For someone with mild anxiety, the answer may favor pausing medication with a therapy plan in place. For someone with treatment-resistant bipolar disorder, the answer may be clearly in favor of continuing medication.

What to Bring to Your Prescriber

• Your full psychiatric history — diagnoses, prior episodes, hospitalizations
• Your current medications and doses
• Your prior history with medication changes — how did you respond when you tried to taper or stop?
• Your support system — who is available to help monitor for relapse?
• Your OB's contact information so your prescriber can coordinate

Questions to Ask Your OB and Psychiatrist

• “What are the specific risks of this medication at my dose during pregnancy?”
• “What are the risks of my untreated mental health condition during pregnancy?”
• “Are there monitoring steps that would make continuing this medication safer?”
• “If I stop, what is the probability of relapse, and what would that look like clinically?”
• “What would therapy as an alternative or adjunct look like, and can you refer me?”

Role of Therapy

For mild-to-moderate conditions, evidence-based therapy (CBT, particularly for depression and anxiety) can serve as a partial or complete substitute for medication during pregnancy, or as an adjunct that allows for dose reduction. For moderate-to-severe conditions, therapy is a critical adjunct but is unlikely to provide sufficient stabilization on its own. Your prescriber can help determine what is realistic.

For guidance on having this conversation with your provider: How to Talk to Your Doctor About Mental Health →

Postpartum planning is as important as pregnancy planning — and it is often skipped entirely. The postpartum period carries a significant spike in psychiatric risk: postpartum depression affects 1 in 5 new mothers, and the risk is substantially higher for people with a prior psychiatric history. Planning for this window before delivery is one of the highest-value clinical interventions available.

Medication After Delivery

If medication was paused or reduced during pregnancy, the postpartum period is typically when it is reintroduced — often at higher doses than pre-pregnancy, in anticipation of the PPD/anxiety spike. If medication was continued throughout pregnancy, postpartum monitoring for symptom changes is still essential.

Breastfeeding and Medication

Most psychiatric medications transfer into breast milk, but at varying levels and with varying clinical significance. The most important factors are the level of transfer and the infant's age and health:

• Sertraline and paroxetine have the lowest breast milk transfer of the SSRIs and are generally preferred for breastfeeding mothers. Infant serum levels are typically undetectable.
• Fluoxetine — longer half-life leads to higher infant exposure; less preferred during breastfeeding, though used when other options are insufficient.
• LactMed (NLM database) — a free, evidence-based resource on drug levels in breast milk. Your prescriber can use this to review specific medications.

NICU Planning

If neonatal adaptation syndrome is anticipated (third-trimester SSRI exposure), inform the delivery team in advance. Hospitals with NICUs monitor for NAS and manage it supportively. This is not an emergency — it is a predictable, manageable event that is much easier to handle when the care team is prepared.

Connecting to Postpartum Support

Postpartum Support International (PSI) offers free resources, online support groups, and a directory of providers trained in perinatal mental health. Establishing the connection before delivery makes it far easier to reach out if symptoms emerge. For more information: Postpartum Mental Health: What Every New Parent Should Know →