Some Antiseizure Medications May Be Safer than Others During Pregnancy

Why It Matters

The findings influence prescribing decisions for women of childbearing age, balancing seizure control with fetal neurodevelopmental safety. As antiseizure drugs are increasingly used for epilepsy and migraine, clearer safety profiles are critical for clinicians and patients.

Key Takeaways

• Valproate linked to higher neurodevelopmental disorder risk
• Levetiracetam, lamotrigine, phenytoin show no adverse signals
• Zonisamide and oxcarbazepine show possible risk signals
• Study uses 2000‑2021 US insurance data, 14,993 exposed
• Findings hypothesis‑generating; need replication before practice change

Pulse Analysis

Antiseizure medications have become a mainstay for managing epilepsy and migraine in women of reproductive age, yet their safety during pregnancy remains a contentious issue. Historically, valproate has been flagged for teratogenicity and neurodevelopmental harm, prompting guidelines to favor alternatives. However, most safety data have focused on birth defects, leaving a gap in knowledge about long‑term cognitive outcomes. The recent BMJ study addresses this gap by examining neurodevelopmental disorders, such as autism and intellectual disability, across a broad spectrum of commonly prescribed drugs.

Using two large U.S. insurance databases spanning 2000‑2021, the investigators identified nearly 15,000 pregnancies exposed to one of ten antiseizure agents and compared them with almost 9,000 unexposed births. Children were followed for up to eight years, allowing the team to calculate hazard ratios for various outcomes. Valproate showed a 29% increased risk for any neurodevelopmental disorder, while zonisamide exhibited a similar signal. Several drugs—including lamotrigine, oxcarbazepine, phenobarbital, and topiramate—were associated with higher odds of intellectual disability, and oxcarbazepine and carbamazepine showed modest links to ADHD. In contrast, levetiracetam and phenytoin did not demonstrate adverse neurodevelopmental signals, reinforcing their reputation as relatively safer options.

Clinicians must now weigh these nuanced findings against the imperative of seizure control during pregnancy. The study’s observational nature precludes definitive causality, and the authors stress the need for replication in other health‑system datasets. Nonetheless, the emerging risk signals for drugs previously deemed benign on a malformation basis suggest that neurodevelopmental safety warrants independent evaluation. Future research should incorporate granular cognitive testing to differentiate global intellectual impairment from specific learning deficits, ultimately guiding more precise, patient‑centered prescribing for pregnant individuals.