The incidence of silent celiac disease (CD) in children with idiopathic localization epilepsies and the indications for routine CD screening were determined in a study of 72 patients (31 girls and 41 boys; mean age 12.6 years; age at onset 6.4 years) observed over a 5 year period at the Institute of Neurology and Gastroenterology, University Magna Graecia of Catanzaro, Italy. The enzyme-linked immunosorbent assay (ELISA) for antigliadin antibodies (AGA) and the immunofluorescent undirected test for antiendomysium antibodies (AEA) were used to confirm a diagnosis of CD. ELISA has >96% sensitivity and 97% specificity for IgA-AGA and IgG-AGA antibodies. AEA. is a more specific but less sensitive test than AGA. Twenty five patients had childhood partial epilepsy with occipital paroxysms (CPEO), and 47 had CPE with centrotemporal spikes (CPEC). Two patients (8%) in the CPEO group had antiendomysium immunoglobulin (Ig) A antibodies, and their jejunal biopsies showed atrophy of the villi and hyperplasia of crypts, confirming the diagnosis of CD. Brain CT was normal in one and showed occipital cortical-subcortical calcifications in the other. Treatment with a gluten-free diet was followed by seizure remission, and no calcifications developed in the patient with a normal CT after 3 year follow up. None of the patients with CPEC had positive antibody tests for CD. 
COMMENT. Celiac disease screening is recommended in patients with childhood partial epilepsy with occipital paroxysms (CPEO). Early diagnosis of CD and dietary intervention may reverse the tendency to seizures and the development of brain calcifications. Routine screening for CD is not indicated in patients with infantile extraoccipital seizures. A new diagnostic test for CD using human recombinant tissue transglutaminase (TAGA) may be more sensitive in the diagnosis of silent forms of CD.