Breastfeeding appears safe with antiseizure medications

Breastfeeding while taking common antiseizure medications was not linked to worse child development at age 6 years.
*Prospective cohort study; Level 2 (OCEBM).

Citation

Li Y, Loring DW, Cohen MJ, et al. Breastfeeding and Neurodevelopmental Outcomes in Children of Women Taking Antiseizure Medications. JAMA Network Open. 2026;9(9):e2637040. doi:10.1001/jamanetworkopen.2026.37040

Background

Many women with epilepsy avoid breastfeeding because of concern that antiseizure medicines enter breast milk. This study addressed whether breastfeeding during modern treatment is associated with child development at age 6 years.

Patients

273 children of women with epilepsy taking antiseizure medications and 84 children of healthy women. Key exclusions included nonepileptic seizures, maternal intelligence score below 70, and medical conditions likely to affect pregnancy or child development.

Intervention

Breastfeeding while taking antiseizure medications, mainly levetiracetam or lamotrigine; duration was none, less than 6 months, or 6 months or longer.

Control

No breastfeeding, shorter breastfeeding, and breastfed children of healthy women.

Outcome

Verbal language and adaptive behavior using standard scores; exact score ranges were not reported. Higher scores indicate better function.

Follow-up Period

Birth to child age 6 years.

Results

Primary language and adaptive behavior outcomes were similar between breastfed children of women with epilepsy and healthy women.

Comparison Outcome Significant finding
Breastfeeding ≥6 months vs none, among women with epilepsy Adaptive behavior (primary) Mean difference 4.68 points (95% CI 1.34 to 8.03)
Breastfeeding ≥6 months vs none Practical skills Mean difference 5.27 points (95% CI 1.84 to 8.70)
Breastfeeding ≥6 months vs none Conceptual skills Mean difference 3.50 points (95% CI 0.28 to 6.71)

CI, confidence interval.

Shorter breastfeeding showed no significant benefit. Results were based on completers. Benefits appeared strongest in monotherapy; polytherapy subgroups were small.

Limitations

Observational design cannot prove causation. Residual confounding, especially family environment, is possible. Breastfeeding exclusivity was not measured. Medication subgroup analyses were underpowered. Specialty-center recruitment may limit generalizability. The clinical importance of 3- to 6-point score differences is uncertain.

Funding

U.S. National Institutes of Health; no funder role reported.

Clinical Application

Encourage eligible women with epilepsy taking modern antiseizure medicines to breastfeed, aiming for at least 6 months when feasible.