Fifty years’ follow‐up of childhood epilepsy: Medical outcome, morbidity, and medication



To describe the long‐term prognosis of childhood epilepsy, with special emphasis on seizure remission, relapse, medication, associated neurologic impairment, mortality rate, and cause of death.


A prospective longitudinal study on a population‐based total cohort of 195 children with epileptic seizures in 1962‐1964. Data were collected from medical records and a questionnaire.


Follow‐up data from 94% of the initial cohort showed the best long‐term prognosis for seizure freedom for children with no intellectual or neurologic impairment. These children had later seizure onset, shorter total duration of epilepsy, and were more often medication free. Only a few of them had isolated relapses. Generalized, rather than focal, epilepsy was associated with fewer relapses and less ongoing medication. The “true incidence” group, with onsets during the inclusion period of 1962‐1964, had the best long‐term prognosis for seizure freedom, with 90% seizure‐free after 50 years. Although only 10% of this group had ongoing seizures at follow‐up, 22% still used anticonvulsive medication, often with old drugs, that is, phenobarbital or phenytoin, as one of the anticonvulsive drugs. The standardized mortality ratio (SMR) was 2.61 for the whole group, with no difference between those with or without other neurodeficits. Those who died young either had neurologic impairment or died from epilepsy‐related conditions; later deaths often followed non–epilepsy‐related conditions. No one in the incidence group died of SUDEP (sudden unexpected death in epilepsy).


This 50‐year, long‐term follow‐up of a cohort of persons with childhood epilepsy in general demonstrates a better outcome for seizure freedom compared to our follow‐up after 12 years and to previous reports. We also report a low incidence of seizure relapses. Remission of seizures does not automatically lead to termination of medication. The mortality rate associated with SUDEP was lower than previously reported.