Is there an association between type 1 diabetes and epilepsy in childhood?

Background The relationship between type 1 diabetes and epilepsy in the paediatric population is unclear. There is evidence to suggest that antibodies to glutamic acid decarboxylase, an auto-antigen implicated in type 1 diabetes, are also associated with the development of epilepsy. However, seizures occurring in children with diabetes may be attributed to hypoglycaemia rather than an underlying seizure disorder. Objectives To identify the number of children with both type 1 diabetes and epilepsy in the region, and to elucidate associated management issues. Methods A data collection form was designed following a literature search. Paediatric consultants and nurse specialists caring for children with diabetes in the Northern Deanery were contacted to identify patients who might also have epilepsy. Information was then collected from case notes by two reviewers and analysed using Microsoft Excel. Results Data from 10 regional hospital clinics was reviewed: 16 children out of 1258 seen in diabetic clinics were being managed for both conditions (1.3%). Of the 12 sets of notes accessed (75% ascertainment), there was an equal sex distribution. In six patients diabetes was diagnosed before epilepsy (50%). Six children had at least one documented hypoglycaemic seizure (50%). Three children had a family history of diabetes (25%), and three children had a family history of epilepsy (25%). One child had a family history of both diabetes and epilepsy (8%). Three children had positive pancreatic islet cell antibodies (25%), one had positive anti-GAD antibodies (8%), one child had primary hypothyroidism (8%) and one had coexisting coeliac disease (8%). Diabetic control was thought to be good in two patients (17%); six patients achieved good seizure control (50%). Five children had special educational needs (42%), three of whom had poorly controlled epilepsy and diabetes. Conclusion There is no evidence that epilepsy occurs more frequently in our regional population of children with type 1 diabetes; however, the numbers are small and a more comprehensive national prospective study may be indicated. From a practical perspective, it should not be assumed that seizures—with or without low associated low blood glucose levels—are due to diabetes in isolation, as this may delay the diagnosis of coexisting epilepsy.