A patient planning a pregnancy recently asked me which antidepressant is proven to be the safest. It is a completely fair question and one I hear regularly. Yet “proven safe” sets a standard that perinatal pharmacology simply cannot meet.
Randomized controlled trials designed to establish antidepressant safety in pregnancy have never been conducted, and they likely never will be. Randomizing pregnant patients with significant depression to a placebo would raise profound ethical concerns: Treatment works, and untreated maternal illness carries real risks, including relapse, preterm birth, impaired bonding, and severe postpartum depression. Practically, detecting a doubling of a rare birth defect would require tens of thousands of participants and years of pediatric follow-up.
Instead, clinicians rely on observational studies and preclinical animal research. While animal models help screen for major teratogens, notorious exceptions like thalidomide remind us that species differ. Human decisions depend on large observational cohorts.
With those limits in mind, sertraline and escitalopram have the most consistently reassuring safety profiles. More importantly, the most useful clinical question is not whether a medication carries zero risk, but how its potential risks compare with the well-established harms of leaving maternal depression untreated.
Real physician voices, twice a week
Free, and one click to unsubscribe.
The hidden trap of confounding
A central challenge in interpreting pregnancy data is that women who take antidepressants differ from women who do not, largely because of the underlying illness itself. Depression is closely associated with factors like smoking, substance use, obesity, and reduced access to prenatal care, all of which independently affect pregnancy outcomes. When researchers compare medicated patients against healthy, unmedicated controls, the illness’s footprint can easily be mistaken for a drug effect.
The scale of this bias is significant. In one large meta-analysis, 16 of 20 adverse outcomes initially appeared elevated in exposed pregnancies; once researchers accounted for maternal depression, only two, preterm birth and NICU admission, remained statistically significant. Similarly, adjusting for depression and maternal health factors drops the apparent relative risk for cardiac malformations from 1.25 to a nearly flat 1.06.
Even newer study designs cannot eliminate every limitation. Pharmacy records show that a prescription was filled, not taken, and exposed newborns are often examined more closely, inflating detection rates. Isolated statistical associations, such as a reported link between citalopram and diaphragmatic hernia, must be treated as early signals rather than proof.
What the data actually show
“Safety” is not a single metric. It spans congenital malformations, pregnancy complications, newborn adaptation, and long-term neurodevelopment, each with its own baseline rate and evidence quality.
Birth defects: Major malformations occur in roughly 2 to 4 percent of all pregnancies. Combined data on first-trimester sertraline exposure show no increase in overall malformations or heart defects, and large cohorts show similarly reassuring results for escitalopram. Most historical concerns have clustered around paroxetine (particularly cardiac defects) and fluoxetine, alongside some persistent signals with venlafaxine.
Newborn complications: Up to 30 percent of exposed newborns experience transient neonatal adaptation symptoms such as jitteriness, irritability, or feeding difficulties, which typically resolve within two weeks. More serious complications, like persistent pulmonary hypertension of the newborn, remain rare. While exposure after 20 weeks roughly doubles the relative risk, the baseline rate is so low that the number needed to harm is about 1,000, and sertraline ranks lowest among studied antidepressants.
Bringing the evidence back to the exam room
In practice, sertraline is often a first-line choice when starting treatment, not because it is superior on every single metric, but because its safety profile is remarkably consistent across outcomes and backed by decades of clinical experience.
For a patient already doing well on an antidepressant before pregnancy, however, staying the course is usually the safest path. Switching medications mid-stream exposes the fetus to two drugs during cross-tapering and significantly raises the risk of maternal relapse, often without a clear safety advantage. Paroxetine is the main exception worth re-evaluating.
When we do prescribe, the goal is the lowest effective dose, not the lowest possible dose. Underdosing creates a lose-lose scenario, exposing the fetus to both the medication and inadequately treated maternal illness. Because pregnancy alters drug metabolism, some patients will even require a dose increase in the second or third trimester to stay well.
When I sit down with a patient planning a pregnancy, we move past the impossible standard of “proven safe.” Instead, we focus on shared decision-making: weighing small, often uncertain medication risks against the very real consequences of untreated illness, and building a plan that keeps both mother and baby healthy. Ultimately, our aim is not to eliminate all risk, but to empower patients with knowledge and compassion so they can make the best choices for themselves and their families.
Timothy Lesaca is a psychiatrist in private practice at New Directions Mental Health in Pittsburgh, Pennsylvania, with more than forty years of experience treating children, adolescents, and adults across outpatient, inpatient, and community mental health settings. He has published in peer-reviewed and professional venues including the Patient Experience Journal, Psychiatric Times, the Allegheny County Medical Society Bulletin, and other clinical journals, with work addressing topics such as open-access scheduling, Landau-Kleffner syndrome, physician suicide, and the dynamics of contemporary medical practice. His recent writing examines issues of identity, ethical complexity, and patient-clinician relationships in modern health care. Additional information about his clinical practice and professional work is available on his website, timothylesacamd.com. His professional profile also appears on his ResearchGate profile, where further publications and details may be found.
