Abstract
BACKGROUND: Triptan medications are proposed as an association of cerebral ischemic stroke in patients with migraine, but population-based data on this are lacking. METHODS: A retrospective, observational, cohort study was performed using computerized claims data from a proprietary, insurance-based registry-Marketscan (by Merative). Patients with at least 1 claim related to a migraine diagnosis were included. The primary exposure was initiation of any triptan medication defined by at least 1 prescription fill. The primary end point was time to cerebral ischemic stroke. Secondary end points included time to retinal stroke (central retinal artery occlusion, other retinal artery occlusion), intracranial hemorrhage, and myocardial infarction. The association between triptan initiation and study end points was modeled using a propensity score-overlap weighted Cox model. Adjusted hazard ratios (HR) and corresponding 95% CIs were computed. RESULTS: In total, 869 092 patients (median age 40.0 years [Q1-Q3, 30.0-50.0]; 77.5% female) met study selection criteria of whom 287 629 initiated a triptan and 581 463 did not. Median follow-up was 2555 days in the entire cohort. Initiation of a triptan was associated with higher hazard of cerebral ischemic stroke (adjusted HR, 2.37 [95% CI, 1.65-3.51]; absolute risk difference of 0.17% per year). Of 4 secondary end points, 2 were associated with triptan use: intracranial hemorrhage (adjusted HR, 2.37 [95% CI, 1.65-3.41]) and myocardial infarction (adjusted HR, 2.06 [95% CI, 1.60-2.66]). CONCLUSIONS: In a population-based cohort study, initiation of a triptan was independently associated with risk of subsequent cerebral ischemic stroke.