Headache in idiopathic/genetic epilepsy: Cluster analysis in a large cohort

Atalar A. C., TÜRK B. G., Ekizoglu E., Gok D. K., BAYKAL B., ÖZGE A., ...More

EPILEPSIA, vol.63, no.6, pp.1516-1529, 2022 (SCI-Expanded) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 63 Issue: 6
  • Publication Date: 2022
  • Doi Number: 10.1111/epi.17205
  • Journal Name: EPILEPSIA
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, BIOSIS, CAB Abstracts, EMBASE, International Pharmaceutical Abstracts, MEDLINE, Veterinary Science Database
  • Page Numbers: pp.1516-1529
  • Keywords: cluster analysis, genetic epilepsy, headache, headache phenotypes, idiopathic epilepsy, myoclonic seizures, JUVENILE MYOCLONIC EPILEPSY, UNITED-STATES, MIGRAINE, CHILDHOOD, CHILDREN, BURDEN, CLASSIFICATION, PREVALENCE, MIGRALEPSY, MECHANISMS
  • Hacettepe University Affiliated: Yes


Objective The link between headache and epilepsy is more prominent in patients with idiopathic/genetic epilepsy (I/GE). We aimed to investigate the prevalence of headache and to cluster patients with regard to their headache and epilepsy features. Methods Patients aged 6-40 years, with a definite diagnosis of I/GE, were consecutively enrolled. The patients were interviewed using standardized epilepsy and headache questionnaires, and their headache characteristics were investigated by experts in headache. Demographic and clinical variables were analyzed, and patients were clustered according to their epilepsy and headache characteristics using an unsupervised K-means algorithm. Results Among 809 patients, 508 (62.8%) reported having any type of headache; 87.4% had interictal headache, and 41.2% had migraine. Cluster analysis revealed two distinct groups for both adults and children/adolescents. In adults, subjects having a family history of headache, >= 5 headache attacks, duration of headache >= 24 months, headaches lasting >= 1 h, and visual analog scale scores > 5 were grouped in one cluster, and subjects with juvenile myoclonic epilepsy (JME), myoclonic seizures, and generalized tonic-clonic seizures (GTCS) were clustered in this group (Cluster 1). Self-limited epilepsy with centrotemporal spikes and epilepsy with GTCS alone were clustered in Cluster 2 with the opposite characteristics. For children/adolescents, the same features as in adult Cluster 1 were clustered in a separate group, except for the presence of JME syndrome and GTCS alone as a seizure type. Focal seizures were clustered in another group with the opposite characteristics. In the entire group, the model revealed an additional cluster, including patients with the syndrome of GTCS alone (50.51%), with >= 5 attacks, headache lasting >4 h, and throbbing headache; 65.66% of patients had a family history of headache in this third cluster (n = 99). Significance Patients with I/GE can be clustered into distinct groups according to headache features along with seizures. Our findings may help in management and planning for future studies.