COMMENTARY

Title, Date of Publication & Journal

The prevalence of neurodevelopmental disorders in children prenatally exposed to antiepileptic drugs.

January 2013

Journal of Neurology Neurosurgery and Psychiatry.

Claim supported by evidence?

The press release states: “Children whose mothers take the antiepileptic drug sodium valproate while pregnant are at significantly increased risk of autism and other neurodevelopmental disorders.”  But it is important to note that while this is the case compared to the non–epileptic population, it is not proven by this paper compared with epileptic women on other types of drug or on no medication at all.

So the paper does not demonstrate that epilepsy medication is responsible for Neurodevelopmental disorders (NDs).  It could be that all epileptic mothers are pre-disposed towards having children with NDs (see below).

The paper does not strongly support a claim that children of mothers with epilepsy would have had lower risk of neurodevelopmental disorders had their mothers switched from monotherapy sodium valproate to other therapy during pregnancy, although this might be true.

Summary

• A small prospective study comparing 201 children from mothers with epilepsy to partially-matched 214 control mothers.
• The study is multi-purpose and not specifically designed to address the effects of epilepsy medication on NDs in children
• (NDs) assessed at age 6 years were more common in the epilepsy group.
• Within the epilepsy group NDs were slightly more common for mothers using monotherapy sodium valproate (VPA), but it’s not statistically significant.

Study Conclusions

• In epileptic mothers taking monotherapy VPA 6/50 (12%) children have NDs.
• In epileptic mothers not taking monotherapy VPA 9/151 (6%) children have NDs.
• But those numbers are not statistically significantly different and could be down to random error
• By contrast, in non-epileptic mothers, 4/214 (2%) children have NDs.
• Therefore, as the authors themselves state, genetic influences due to epilepsy rather than medication might be driving the apparent association.
• Further research is warranted on the possible impact of sodium valproate and other antiepileptic drugs used during pregnancy on offspring neurodevelopment.

Strengths/Limitations

• Major limitation is that a specific subgroup of epileptic mothers is compared to the whole control group. This confuses mother’s epilepsy with the form of therapy. 
• Epilepsy rate is already known to be very high in those with autism, the main ND described here. This may be driving the observation of increased ND rate in children of epileptic mothers (20-30% of children with autism spectrum disorders develop epilepsy by adulthood).
• Prospective study with clear protocol reduces potential bias; however only 50% of those approached for the control group were recruited, introducing some bias due to self-selection.
• Control subjects were well-matched for parity, age, location and date of recruitment.

‘Before the headlines’ is a service provided to the SMC by volunteer statisticians: members of the Royal Statistical Society (RSS), Statisticians in the Pharmaceutical Industry (PSI) and experienced statisticians in academia and research.  A list of contributors, including affiliations, is available here.