The Clinical Dilemma of Prenatal Pharmacotherapy
For clinicians managing pregnant patients, the decision to prescribe or maintain medication involves a delicate risk-benefit calculus. Historically, observational studies have frequently identified associations between prenatal exposure to various drug classes—including acid-suppressive medications and antipsychotics—and an increased risk of neurodevelopmental disorders (NDDs) like Attention-Deficit/Hyperactivity Disorder (ADHD) and Autism Spectrum Disorder (ASD). However, a significant question persists: is the medication itself the causative agent, or are these associations artifacts of maternal underlying conditions and shared familial factors?
Recent high-quality evidence from massive population-based cohorts in South Korea, Hong Kong, the United States, and Sweden is now providing much-needed clarity. By utilizing advanced statistical techniques such as sibling-matched analyses and propensity score overlap weighting, researchers are effectively ‘filtering out’ the noise of confounding variables, leading to a paradigm shift in how we view prenatal drug safety.
Acid-Suppressive Medications: Reassurance from Sibling Controls
Acid-suppressive medications, including Proton Pump Inhibitors (PPIs) and H2 receptor antagonists (H2RAs), are among the most commonly used drugs during pregnancy. A landmark retrospective cohort study by Hong et al. (2026), published in JAMA, examined over 2.7 million mother-child pairs from the South Korean National Health Insurance Service database. The study aimed to resolve the ambiguity surrounding PPI/H2RA exposure and offspring neuropsychiatric risk.
Initial findings using an overlap-weighted cohort suggested a modest but statistically significant increase in risk. Exposed children had an adjusted hazard ratio (HR) of 1.14 for ADHD and 1.07 for ASD. However, the study’s strength lay in its sibling-matched analysis, which compares children born to the same mother but with different exposure statuses. In this model, which controls for shared genetic and environmental factors, the associations vanished. The adjusted HR for ADHD dropped to 0.98 (95% CI, 0.95-1.02), and for ASD, it was 0.98 (95% CI, 0.92-1.04).
These results indicate that the slight risks observed in traditional models likely reflect ‘confounding by shared familial factors’—essentially, the characteristics of the families who require these medications rather than the medications themselves. For clinicians, this provides significant reassurance that managing severe gastroesophageal reflux in pregnancy with standard acid-suppressants does not inherently elevate NDD risk.
Antipsychotics: Separating the Drug from the Disease
The management of psychosis or severe mood disorders during pregnancy is even more complex. Two major studies, one from Hong Kong (Wang et al., 2021) and one from the United States (Straub et al., 2022), have addressed the neurodevelopmental impact of prenatal antipsychotic exposure.
The Hong Kong Cohort
Wang and colleagues analyzed over 330,000 mother-child pairs. While primary analyses showed a small increased risk of preterm birth (wOR 1.40), the neurodevelopmental outcomes told a different story. The weighted hazard ratio for ADHD was 1.16 and for ASD was 1.06, neither of which reached statistical significance after robust adjustment. Furthermore, when comparing gestationally exposed children to siblings who were not exposed, the risks remained non-significant. The researchers concluded that the findings do not support a recommendation to discontinue regular antipsychotic treatment, given the high risks associated with untreated maternal psychosis.
The MAX and MarketScan Data
Similarly, Straub et al. utilized two massive US databases (MAX and MarketScan) covering over 3 million pregnancies. While unadjusted data showed a nearly two-fold increase in NDD risk for exposed offspring, this risk largely disappeared after adjusting for maternal characteristics (pooled adjusted HR 1.08). The only potential outlier was aripiprazole (HR 1.36), which the authors noted requires further replication. The consensus remains that maternal mental illness, rather than the antipsychotic drugs used to treat it, is the primary driver of observed neurodevelopmental differences.
Maternal Eating Disorders: A Persistent Risk Factor
While medication risks often dissipate under sibling-controlled scrutiny, the impact of maternal disease itself can be more profound. A Swedish population-based study by Mantel et al. (2022) examined the offspring of mothers with eating disorders (ED). Unlike the medication studies, this analysis found that the risk remained elevated even after accounting for many confounders.
Maternal anorexia nervosa (AN) was associated with an HR of 1.42 for ADHD and 2.04 for ASD. Bulimia nervosa and unspecified eating disorders showed even higher correlations. Crucially, the risk was significantly higher when the eating disorder was active during pregnancy (Ongoing AN: ADHD HR 2.52; ASD HR 3.98). Even after adjusting for parental psychiatric comorbidities and using family clusters (cousin comparators) to account for shared genetics, the associations for AN and BN remained significant. This suggests a potential biological or environmental impact of the ED environment—such as nutritional deficiencies or metabolic stress—on the developing fetal brain.
Expert Commentary and Clinical Implications
The collective weight of this evidence suggests a transition in clinical thinking. We must move away from the ‘teratogenic anxiety’ associated with common medications and toward a more holistic focus on maternal health and disease stability. The ‘Healthy User Bias’ and ‘Confounding by Indication’ are powerful forces in observational data; when a mother is sick enough to require medication, her underlying health status is the most significant variable for the child’s outcome.
Key takeaways for practice include:
- Medication Reassurance: For acid-suppressants and most antipsychotics, the risk of NDDs is not causally linked to the drug exposure. Maintenance of maternal stability is paramount.
- Focus on Disease Activity: The Swedish ED data highlights that active disease during pregnancy carries a higher risk than a history of disease. This underscores the need for pre-conception optimization and intensive prenatal support for women with chronic conditions.
- Monitoring and Support: Offspring of women with severe mental health or eating disorders should be closely monitored for neurodevelopmental milestones to ensure early intervention, regardless of their medication exposure.
Conclusion
The evolution of epidemiological methodology—specifically the use of sibling-matched designs—has significantly refined our understanding of prenatal risks. As clinicians, our focus should remain on the effective management of maternal health. By ensuring that conditions like gastroesophageal reflux, psychosis, and eating disorders are appropriately managed, we provide the best possible environment for fetal neurodevelopment. The evidence suggests that, in most cases, the disease is a greater threat than the treatment.
References
1. Hong S, Lee S, Kim H, et al. Prenatal Exposure to Acid-Suppressive Medications and Risk of Neuropsychiatric Disorders in Children. JAMA. 2026; doi:10.1001/jama.2025.23956.
2. Wang Z, Chan AYL, Coghill D, et al. Association Between Prenatal Exposure to Antipsychotics and Attention-Deficit/Hyperactivity Disorder, Autism Spectrum Disorder, Preterm Birth, and Small for Gestational Age. JAMA Intern Med. 2021;181(10):1332-1340.
3. Straub L, Hernández-Díaz S, Bateman BT, et al. Association of Antipsychotic Drug Exposure in Pregnancy With Risk of Neurodevelopmental Disorders: A National Birth Cohort Study. JAMA Intern Med. 2022;182(5):522-533.
4. Mantel Ä, Örtqvist AK, Hirschberg AL, Stephansson O. Analysis of Neurodevelopmental Disorders in Offspring of Mothers With Eating Disorders in Sweden. JAMA Netw Open. 2022;5(1):e2143947.