NEW research has heightened the need for GPs’ and other prescribers’ awareness of the potential risks associated with specific antiseizure medicines when weighing the benefits of prescribing to reproductive-aged women.

Among pregnant women, five in 1000 take antiseizure medicines and women with epilepsy who do not take the medicine while pregnant are at risk of seizures. Previous studies (here and here) have shown that pregnant women with epilepsy have a higher mortality than women without epilepsy, and some of this risk may be attributed to seizures. Therefore, it is essential for prescribers to know the risk for the child and which antiseizure medicines can be used to ensure the health of both mother and child.

The SCAN-AED Study, a Nordic register-based study of antiepileptic drugs in pregnancy, reported a substantially increased risk of adverse neurodevelopment in children whose mothers were treated with antiseizure medicines during pregnancy.

The study (we are co-authors) leverages real-world data on over 4 million mother–child pairs in five countries followed for up to 20 years, with 25 000 children prenatally exposed to antiseizure medicines, 16 000 born to mothers with epilepsy.

The results show that prenatal exposure to topiramate and valproate was associated with a two- to fourfold increased risk of autism spectrum disorder and intellectual disability. They also suggest elevated risks of child neurodevelopmental disorders with common antiseizure duotherapies.

Among 21 634 unexposed children of mothers with epilepsy, 1.5% had a diagnosis of autism and 0.8% intellectual disability by age 8 years. In same-aged children of mothers with epilepsy exposed to topiramate and valproate monotherapy, 4.3% and 2.7%, respectively, had autism, and 3.1% and 2.4% had intellectual disability. After adjusting for a wide range of potential confounders, the analysis yielded hazard ratios of 2.8 (95% confidence interval [CI], 1.4–5.7) for autism and 3.5 (95% CI, 1.4–8.6) for intellectual disability after topiramate exposure and 2.4 (95% CI, 1.7–3.3) and 2.5 (95% CI, 1.7–3.7), respectively, after valproate exposure. These associations were dose-dependent, demonstrating elevated risks with higher daily doses of valproate (³ 750 mg) and with topiramate (³ 100 mg).

While previous research has found an association of similar magnitude between exposure to valproate during pregnancy and child neurodevelopmental disorders, the current findings on topiramate are novel. Topiramate is no longer recommended for first line use in most guidelines because it carries an increased risk of congenital malformations in babies exposed prenatally, but safety evidence on long term outcomes has been sparse.

Reassuringly, our latest findings confirm previous results that monotherapy with lamotrigine and levetiracetam is not associated with increased risks of autism and intellectual disability in children of women with epilepsy. These antiseizure medicines are widely recommended as first line therapy in women of reproductive age and considered effective for both focal and generalised epilepsies, without carrying the risk of aggravating seizure frequency.

Worryingly, though, our data suggest that some common duotherapies are associated with increased risk of child neurodevelopmental disorders within the same range as topiramate and valproate exposure, even without these being one of the duo-medicines. Children whose mothers had used a combination of levetiracetam and carbamazepine as well as a combination of lamotrigine and topiramate had increased risks of 2.5–3.5-fold. However, this did not apply to children born to mothers who had used a combination of levetiracetam and lamotrigine.

The SCAN-AED Study is an ongoing study funded by the Nordic governments and led by researchers in Norway, Denmark, Finland, Iceland, Sweden and Australia, with the aim to optimise the choice of medicines and folic acid treatment in pregnant women with epilepsy. It is one of the first studies large enough to investigate long term risks of both monotherapy and combination therapies with more than one antiseizure medicine, which may be necessary in some women with epilepsy to ensure they are seizure-free during pregnancy.

The evidence provided in this longitudinal, multiregister study is important to both doctors and women with epilepsy. Notably, our findings suggest that topiramate may not a be a safe alternative to valproate, showing a clear association with adverse neurodevelopment in children exposed to topiramate, particularly at doses of 100 mg per day.

The findings are highly relevant for prescribers in Australia. They underscore the importance of clear alignment of regulatory warnings, prescribing restrictions, and clinical guidelines to facilitate informed treatment decisions and optimal prescribing practices. Currently, valproate is available on the Australian Government Pharmaceutical Benefits Scheme (PBS) for reimbursement without restriction, despite widespread regulatory warnings cautioning against its use in women of reproductive age (here and here).

PBS restrictions have recently been updated to allow use of levetiracetam and lamotrigine as first line treatment in women of childbearing potential. Topiramate is listed on the PBS for seizures―if other antiseizure medicines have failed satisfactory control―and as prophylaxis to treat migraines in case of contraindication or intolerance to other guideline-indicated medicines (eg, b-blockers, pizotifen).

GPs and other prescribers need to be aware of the potential risks associated with specific antiseizure medicines when weighing the benefits of prescribing these to reproductive-aged women.

Associate Professor Helga Zoega is a pharmaco-epidemiologist at the School of Population Health, Faculty of Medicine and Health, University of New South Wales. She is also a Professor of Public Health at the University of Iceland in Reykjavík.

Associate Professor Marte-Helene Bjørk is with the Department of Clinical Medicine at the University of Bergen, and the Department of Neurology at Haukeland University Hospital in Bergen, Norway.



The statements or opinions expressed in this article reflect the views of the authors and do not necessarily represent the official policy of the AMA, the MJA or InSight+ unless so stated.

Subscribe to the free InSight+ weekly newsletter here. It is available to all readers, not just registered medical practitioners.

If you would like to submit an article for consideration, send a Word version to


Abortion care must be part of the core curriculum for GP and ob/gyn training
  • Strongly agree (63%, 133 Votes)
  • Strongly disagree (18%, 37 Votes)
  • Agree (11%, 24 Votes)
  • Neutral (4%, 9 Votes)
  • Disagree (3%, 7 Votes)

Total Voters: 210

Loading ... Loading ...

One thought on “Antiseizure drugs in pregnancy: child risk of autism

  1. Michael King says:

    Am I allowed to take issue with the authors including Autism Spectrum Disorders as “adverse neurodevelopment”? While I’m at it, I’ll take issue with calling those on the autism spectrum (myself included) disordered. Are Macintosh or Linux “disordered” Windows? Perhaps it’s the other way around and it is the neurotypical operating system that is deficient in some way. Or maybe we are all just part of the natural variety of the human species and the “adverse” or “disordered” are related solely to very narrow societal definitions.

Leave a Reply

Your email address will not be published. Required fields are marked *