Severity and outcomes according to stroke etiology in patients under 50 years of age with ischemic stroke.

UNLABELLED: To analyze the association of stroke etiological subtypes with severity and outcomes at 3 and 12 months in patients ≤50 years. Observational study of patients admitted to a stroke unit (2007-2013). VARIABLES: demographic data, vascular risk factors, comorbidities, severity on admission (NIHSS), and good functional outcome (mRS ≤ 1) at 3 and 12 months. We used multivariate analyses to evaluate the influence of stroke etiology on severity and outcomes. We included 214 patients, 58.3 % men, mean age 41.4 years. General linear models showed all etiologies were more severe than lacunar strokes (P < 0.05). Atherothrombotic strokes showed greater severity than those of undetermined and uncommon etiology, whereas cardioembolic strokes were more severe than cryptogenic. Taking into account specific etiologies, atherothrombotic strokes (B = 5.860; 95 % CI 2.979-8.751), cervical artery dissection (CAD) [B = 7.485; 95 % confidence interval (CI) 4.734-10.237], and atrial fibrillation (AF) strokes (B = 5.773; 95 % CI 2.704-8.132) were more severe than other etiologies. Logistic regression models showed that strokes of uncommon etiology, especially those not related to CAD, had a lower probability of good outcome at 3 months [odds ratio (OR) = 0.197; CI 95 % 0.044-0.873], whereas atherothrombotic strokes were associated with this probability at 12 months (OR = 0.187; 95 % CI 0.037-0.951; P = 0.007). In patients ≤50 years of age, strokes of atherothrombotic, cardioembolic (particularly those due to AF), and uncommon etiology had a greater severity than the rest. Furthermore, strokes of uncommon etiology, especially those different from CAD, decreased the probability of a good outcome at 3 months, as did atherothrombotic strokes at 1 year.