Risk Factors for a First Febrile Convulsion in Children: A Population Study in Southern Taiwan

PURPOSE: To identify risk factors for a first febrile convulsion among 3-year-old children by a matched case-control population study. METHODS: All 11,714 neonatal survivors born in Tainan City between October 1989 and September 1991 were enrolled. At age 3, 10,460 children were available for telephone survey for febrile convulsions, and were confirmed by home visit interviews. Those without history of seizure were randomly matched to each febrile convulsion case by age, gender, and residence district. RESULTS: Two hundred fifty six children had febrile convulsions, and 218 of them and their matched controls were available for analysis. The febrile convulsion cases had significantly more febrile episodes (four or more) per year (33.0 vs. 22.5%; p = 0.021), and cases had a higher percentage of developmental delay (3.7 vs. 0.4%; p = 0.046) and a higher percentage of febrile convulsions in their siblings (12 vs. 0.4%; p = 0.011) than controls. The other sociodemographic, environmental, and biologic variables showed no differences between cases and controls. Step-wise logistic regression showed a highly significant independent association between febrile convulsions and history of febrile convulsions in the siblings, and a moderate one between febrile convulsions and the number of febrile episodes per year. CONCLUSIONS: The presence of febrile convulsions in the siblings and the number of fever episodes per year were the independent and significant predictors of febrile convulsion for an individual case in our population-based sample.