RESUMEN
OBJECTIVE: We constructed epilepsy multimorbidity networks to study associations with chronic conditions, and co-prescriptions and drug-disease networks to assess potential interactions. We conducted a population-based study in Catalonia, Spain, with electronic files of 3,135,948 adult patients with multimorbidity, 32,625 of them with epilepsy (active diagnosis any time during 2006-2017). We constructed epilepsy comorbidity networks using logistic regression models from odds ratio estimates adjusted by age, sex, and comorbidities with R software and generated trajectories to study the progression of epilepsy. We constructed drug-disease and co-prescription networks using mixed models with repeated measures adjusting by age, sex, and period with chronic prescription invoiced data. Comorbidity more frequently preceding epilepsy included cerebrovascular accident (OR: 3.59), congenital anomalies (2.18), and multiple sclerosis (1.33); and following epilepsy: dementia (1.91), personality disorder (1.59), alcohol abuse (1.22), and Parkinson (1.21). Mental retardation (13.08), neurological cancer (8.49), benign neoplasm (4.69), infections (3.14), and psychosis (1.58) might precede or not epilepsy. A common progression was to schizophrenia, dementia, and other neurological diseases (mainly cerebral palsy and other degenerative diseases of nervous system). Co-prescription associations with major-moderate potential interactions were 54% for carbamazepine, 61% phenytoin, 53% phenobarbital, and 32% valproate. Major potential interactions were with antipsychotic, anxiolytic, opioid, cardiovascular, and other anti-seizure medications (ASMs). The most frequent comorbidities of epilepsy were congenital, cerebrovascular, and neurological and psychiatric conditions. High comorbidity and co-prescription with potential interactions can increase the complexity of care of patients with epilepsy.