GENETIC EPIDEMIOLOGY OF SEIZURE DISORDERS IN ROCHESTER

The objective is to test genetic and nongenetic models to explain the familial distribution of seizure disorders. The data to be used in the analysis were collected in the Rochester-Olmsted County Record Linkage Project of the Minnesota Comprehensive Epilepsy Program. Offspring of women with epilepsy have consistently been found to have higher risk of siezures than do offspring of men with epilepsy. In addition, fertility rates in individuals with epilepsy have been found to be lower than population rates, and the fertility reductions are more extreme for epileptic men than for epileptic women. If the fertility reduction were selective, occurring mainly in individuals with the most familial forms of epilepsy, then a lower proportion of the epileptic fathers than of the epileptic mothers would be expected to have familial forms of epilepsy. Therefore, selective fertility could explain the higher risk of seizures in offspring of women than in offspring of men with epilepsy. First, fertility rates in epileptic men and women in Rochester will be compared with sex-specific population rates, using age- and calendar year-specific person-years in the analysis to control for age, secular period, and length of follow-up. Fertility rates in epilepsy cases with different ages at onset (0-19 vs 20+), seizure types (generalized vs partial) and etiologies (idiopathic vs symptomatic) will be compared, to determine whether or not any observed fertility reductions are selective for particular epilepsy diagnoses. Second, seizure risks in offspring of spileptic men and women will be compared, controlling for selective fertility effects by stratifying by age at onset, seizure type, and etiology of epilepsy in the parents. Third, the consistency of the familial distribution of seizure disorders with various genetic models will be evaluated, using segregation analysis and controlling for age at onset, fertility effects, and ascertainment. The models to be tested include autosomal and X-linked single locus models, polygenic threshold models, and two models of maternal transmission. The information to be obtained in this study can be used for improved genetic counseling for seizure disorders, and may also lead to development of improved anticonvulsant medications and/or methods to prevent onset of seizures in some individuals.