In a multi-national epidemiological study in Sweden and the United Kingdom, we aim to study if maternal epilepsy and use of anti-epileptic drugs (AEDs) during pregnancy is associated with increased risk of autism spectrum disorders (ASD). In addition, we will examine maternal health outcomes associated with AED use to provide data regarding the risk-benefit ratio of continuing medication use during pregnancy. Epilepsy is the most common serious neurological disorder affecting women during pregnancy, and is linked with a 10-fold increased risk of death during delivery. Antiepileptic drugs are necessary for seizure control, but effects of AED use during pregnancy remain understudied. Although there is good evidence linking use of AEDs during pregnancy to congenital malformations, there is less high-quality evidence concerning adverse neurodevelopmental outcomes. Two studies have found an association between AED exposure during pregnancy and ASD, but both studies had fewer than 10 ASD cases exposed to AEDs. Moreover, many studies of AEDs and neurodevelopmental outcomes have methodological issues. Two review of past studies of neurodevelopmental outcomes in children exposed to AEDs found substantial problems that prevented definitive conclusions regarding adverse effects of AED exposure. The chief problems included: confounding, selection bias, and small sample sizes with limited statistical power. We propose an epidemiological study that addresses each of these issues, in a combined sample of nearly 3 million children born in Sweden and the United Kingdom. All exposure, confounder, and outcome data are derived from clinical and population databases in Sweden and the UK.
In Aim 1 we will estimate associations of maternal epilepsy and child ASD.
In Aim 2, we evaluate the safety of AED use and of specific AEDs during pregnancy with regard to child ASD.
In Aim 3, we will estimate the risk or risk reduction of pregnancy and delivery outcomes associated with AED use during pregnancy, such as seizure, self-harm, or death. Throughout all analyses, causal inference techniques will be employed to determine the validity of findings. Methods include propensity score matching, sibling and negative controls, and simulation of unobserved confounders.
We propose an epidemiological study of maternal epilepsy, use of anti-epileptic drugs during pregnancy and risk of autism spectrum disorders. The study will feature nearly 3 million children born in Sweden and the UK using computerized health data registers.
