Key Takeaways

• A large meta-analysis shows no causal link between antidepressant exposure and neurodevelopmental disorders, including autism and ADHD.
• Initial risk signals disappear after adjusting for confounding factors, including maternal psychiatric illness.
• Paternal antidepressant use shows similar associations, supporting genetic/environmental contributions.
• No increased risk was identified for specific SSRIs or SNRIs.
• Untreated maternal psychiatric illness carries significant risks for both mother and child.

Recently, there have been several high-profile reports questioning the use of SSRI and SNRI antidepressants during pregnancy. These debates have not been spurred by new research demonstrating the detrimental effects of antidepressant use during pregnancy but rather reflect and amplify the limitations of research on medication use in pregnancy. It is ethically not possible to conduct randomized controlled trials in pregnant women, where one group of pregnant women receives medication and the other receives placebo.  Furthermore, registration studies (those required by the FDA to establish efficacy) of new drugs specifically exclude pregnant women.

While pregnancy registries provide the most detailed and accurate information regarding the reproductive safety of medications, they are not available for all medications. Furthermore, they are costly to establish and maintain and are generally small, so they may not detect small differences in rare outcomes or fully account for confounding variables. Most of our data now comes from large medical databases, where researchers can assess outcomes in large populations and estimate the impact of confounding factors.

Last week, Chan and colleagues published a study in Lancet Psychiatry examining the risk of neurodevelopmental disorders (NDDs) in children prenatally exposed to SSRIs and SNRIs. First, the good news: this large-scale study found no evidence supporting a link between exposure to antidepressants and risk for neurodevelopmental disorders, including autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder (ADHD), in exposed offspring.

This is not the first, nor the only, report on this topic. While some studies have suggested a link between SRI exposure and increased risk of neurodevelopmental disorders, particularly autism, a growing body of carefully conducted studies controlling for confounding factors does not support an association between prenatal SRI exposure and elevated risk.

The challenge is that each new study tends to increase anxiety and renew the search for a definitive answer. There is no absolute certainty in this field. Especially when studying long-term outcomes, many factors influence the results. These studies are complex and often difficult to interpret, and the clinical significance of findings can be hard to extract from dense statistical language, leaving readers wondering if they fully understand the results.

Study Design

In this systematic review and meta-analysis, Chan and colleagues identified studies that included (1) mothers or fathers with antidepressant use before or during pregnancy and (2) reported data on neurodevelopmental disorders in offspring. The analysis included 37 studies, with 648,626 antidepressant-exposed pregnancies and 24,967,806 unexposed pregnancies.

Relative risks (RRs) were pooled using random-effects meta-analyses. The research team assessed publication bias, conducted subgroup analyses, and evaluated study quality using the Newcastle-Ottawa scale.

Study Results

In unadjusted analyses (comparing exposed and unexposed pregnancies), antidepressant use during pregnancy was associated with a modestly increased risk of neurodevelopmental disorders in offspring (RR 1.13, 95% CI 1.08–1.18), including ADHD (1.35, 1.24–1.47) and ASD (1.69, 1.24–2.30). Antidepressant exposure had no impact on risk for intellectual disabilities, motor disorders, or speech and language disorders.

However, these associations were attenuated or became non-significant in sensitivity analyses accounting for confounding factors such as maternal psychiatric illness, familial or genetic influences, and misclassification. This suggests that the increased risk observed in unadjusted analyses is unlikely to be due to direct medication exposure and more likely reflects underlying psychiatric illness or shared genetic and environmental factors.

Both SSRI and non-SSRI antidepressants showed similar risk patterns for ADHD and ASD. No significant difference in ASD risk was observed between high- and low-dose exposure.

To better account for genetic and environmental contributions, the researchers conducted additional analyses.

Pre-Conception Exposure to Antidepressants and Risk of NDDs

In a subgroup analysis, researchers focused on women who had used antidepressants prior to pregnancy and compared those who continued treatment during pregnancy with those who discontinued before conception. These groups are more comparable with respect to underlying psychiatric illness than comparisons including women without psychiatric histories. The risk of ASD and ADHD was similar in both groups, a finding that does not support a causal link between antidepressant exposure and increased risk of neurodevelopmental disorders.

Antidepressant Use in Fathers and Risk of NDDs

The researchers also examined outcomes in children whose fathers used antidepressants during the pregnancy. Paternal antidepressant use was associated with increased ADHD risk (1.46, 1.38–1.56) and ASD risk (1.28, 1.16–1.40). This finding further supports the idea that observed risks are not due to direct fetal exposure to medication but are more likely related to shared genetic or familial environmental factors.

Subgroup Analysis of Specific Antidepressants

In adjusted analyses controlling for confounding by indication, only amitriptyline and nortriptyline were associated with increased risk of ADHD (amitriptyline 1.74, 1.00–3.03) or ASD (amitriptyline and nortriptyline 2.02, 1.32–3.10). No significant associations were found for specific SSRIs or SNRIs. The authors hypothesize that these associations may reflect the fact that these tricyclic antidepressants are typically not first-line choices and may be used to treat more complex or refractory neuropsychiatric disorders.

The Bottom Line

The findings of this meta-analysis are consistent with prior well-designed studies examining neurodevelopmental outcomes after prenatal antidepressant exposure. As reviewed in detail by Andrade (2025), unadjusted analyses often show a small but statistically significant association with ASD and ADHD.

However, when studies account for underlying psychiatric illness, genetic vulnerability, and environmental factors, these associations diminish or disappear. This indicates that the observed risks are most likely not due to antidepressant exposure during pregnancy but rather to confounding factors.

For me, the most clinically compelling data come from analyses comparing women who continue antidepressants during pregnancy with those who discontinue treatment. Both groups demonstrate a similarly small but statistically significant increased risk of having a child with ASD or ADHD. This pattern suggests that the observed risk is not driven by medication exposure but rather by factors related to the underlying psychiatric illness or shared genetic vulnerability.

While no medication can be considered completely risk-free, the risks of discontinuing antidepressants must also be considered. Women with mood and anxiety disorders who discontinue treatment during pregnancy are at increased risk:

• Relapse of psychiatric illness
• Increased risk of postpartum depression
• Higher rates of pregnancy complications, including small for gestational age and preterm birth

There is also a growing body of literature indicating that untreated maternal psychiatric illness functions as an early adverse exposure that can shape developmental trajectories and increase vulnerability in children. These studies suggest that maternal depression and related adversities are associated with alterations in stress physiology, epigenetic marks, and early emotional and behavioral regulation, which in turn elevate risk for psychiatric symptoms.  Genetic vulnerability to psychiatric illness may begin to influence fetal brain development in utero and that the prenatal environment—including factors such as maternal stress, inflammation, and health behaviors—can amplify or mitigate this risk. 

Clinical decisions should always be made on a case-by-case basis. While some women may choose to discontinue antidepressants prior to conception, this is not the best option for all.  For many families, the greater concern may not be the use of antidepressants during pregnancy but rather the consequences of ongoing, untreated psychiatric illness during critical periods of brain development.

—Ruta Nonacs, MD PhD