Enduring language deficits in children of women with epilepsy and the potential role of intrauterine exposure to antiepileptic drugs.

OBJECTIVE: Exposure to certain intrauterine antiepileptic drugs (AEDs) can negatively influence the language skills and intelligence of young children. It remains unanswered whether these deficits are transient or persist as children grow up. This study aims to evaluate the language function of children of women with epilepsy (CWE) aged 9-13 years in comparison with their peers, and its relationship with intrauterine AED exposure. METHODS: We included 191 CWE in our study from the Kerala Registry of Epilepsy and Pregnancy. Children in the same age group (n = 144) and without maternal epilepsy or antenatal AED exposure served as controls. We used Clinical Examination for Language Function version IV to assess language in both groups. Relevant data related to maternal epilepsy and AED use were obtained from the registry records. RESULTS: The average Core Language Scaled Score (CLSS) was significantly lower in CWE as compared to controls (83.19 vs 90.18, P = .001). Similarly, the mean scaled scores in other language parameters were also significantly lower in CWE. In the multivariate analysis, compared to control children, the average CLSS in CWE was 4.5 units lower (95% confidence interval [CI] = -8.8 to -0.2, P = .04) with AED monotherapy exposure and 7.3 units lower with exposure to AED polytherapy (95% CI = -13.8 to -0.8, P = .03). Intrauterine exposure to phenobarbitone (n = 61) and valproate (n = 55) as either monotherapy or polytherapy showed a negative effect on CLSS in CWE as compared to control children. However, carbamazepine (n = 75) and phenytoin (n = 37) use was not associated with significant variation of CLSS. In head-to-head comparisons between AED monotherapies in CWE, phenobarbitone showed a negative effect on CLSS (-14.7, 95% CI = -23.1 to -6.4, P = .001) as compared to carbamazepine. SIGNIFICANCE: Intrauterine exposure to phenobarbitone and valproate impairs language development in CWE, with effects persisting into the second decade.

Language, intelligence, and educational outcomes of adolescents with antenatal exposure to antiseizure medications: Prospective data from the Kerala Registry of epilepsy and pregnancy.

OBJECTIVE: In a prospective cohort of children (13-21 years) of women with epilepsy (CWWE), we compared those exposed to antiseizure medications (ASM) in utero to those without exposure to ASM regarding their language proficiency and intelligence. We also compared their educational performance with state-wide averages. METHODS: Research staff blinded to the ASM exposure of CWWE administered the Clinical Evaluation of Language Fundamentals IV and the Wechsler Intelligence Scale for Children IV to test their language proficiency and intelligence. We assessed their educational performance with a questionnaire. CWWE without antenatal exposure to ASM served as comparators for language and intelligence tests. The educational performance of CWWE (regardless of ASM exposure) was compared with the state-wide averages published by the government. RESULTS: In total, 446 children (mean age 16.5 ± 2.2 years; 236 girls) participated in the study. Their ASM exposure involved monotherapy for 272 (61%), polytherapy for 133 (29.8%) and none for 41 (9.2%). The commonly used ASMs (mono & polytherapy) were carbamazepine (n = 192), valproate (n = 124), phenobarbitone (n = 95), and phenytoin (n = 73). The full-scale intelligence quotient of CWWE (n = 146) with antenatal exposure to ASM (89.2 ± 21.5) was significantly lower (p = 0.03) than that of CWWE (n = 11) unexposed to ASM (96.9 ± 8.8). The CELF core language scaled score for the exposed CWWE (n = 132) was significantly lower than that of unexposed children (n = 12; 99.2 ± 19.9). Compared with state-wide averages, CWWE (n = 386) had delayed initiation of education (2.6% vs. 0.1%), increased dropout rates (1% vs. 0.11%), increased usage of special assistance during examinations (4.4% vs. 0.1%) and a lower rate (19.9% vs. 37%) of enrolment in universities. CONCLUSION: The language and intelligence functions of CWWE with exposure to older ASMs were lower than those of unexposed CWWE. Compared to state-wide averages, a significantly higher proportion of CWWE had difficulties with education, and only a smaller proportion enrolled in higher education.