Abstracts

RATIONALE: Previous studies have consistently reported higher risk of epilepsy in offspring of affected women than in those of affected men. The cause of this "maternal effect" is unknown. We tested genetic and nongenetic explanations for the maternal effect in the families of 1560 probands with idiopathic/cryptogenic epilepsy ascertained from voluntary organizations. METHODS: Histories of seizure disorders in offspring were obtained by interviewing the probands. Male probands had lived separately from their offspring more often than female probands, and did not report epilepsy in any of the offspring from whom they had been separated. To reduce reporting bias, therefore, offspring who had lived separately from the probands for ?1 year were excluded. RESULTS: The maternal effect differed markedly by birth year of the offspring. Among offspring born prior to 1965, cumulative incidence of idiopathic/cryptogenic epilepsy to age 20 was significantly higher in offspring of female vs. male probands (6.5% [17/266] vs. 0 [0/71], p=0.03, log rank test). In contrast, among offspring born in 1965 or later, no maternal effect was apparent (risk to age 20: 5.8% [25/603] among offspring of female probands vs. 7.1% [8/242] among offspring of male probands [p=0.77]). To evaluate the roles of reporting and selection bias in this difference across offspring birth cohorts, we examined population-based data on offspring of parents with idiopathic/cryptogenic epilepsy in Rochester, MN, which were obtained by record linkage rather than by interview. A very similar pattern was observed in the Rochester data. Risk was higher in the offspring of female vs. male probands born before 1965 (5.2% [7/162] vs. 1.8% [2/138], p=0.04) but not in those born in 1965 or later (4.3% [4/149] vs. 3.3% [2/90], p=0.81). Moreover, the reduced maternal effect in recent birth cohorts was primarily due to increased risk in offspring of male probands, rather than reduced risk in offspring of females. CONCLUSIONS: These findings add an intriguing piece to the puzzle presented by the maternal effect in epilepsy, and suggest that it is unlikely to be explained by a genetic mechanism.