Comparison of population-based models of risk factors for TIA and ischemic stroke

Abstract
Objective: To determine whether there is a difference in the risk factors for ischemic stroke and for TIA. Background: TIA is associated with a high risk for ischemic stroke, but some have considered TIA as mild ischemic stroke. Prevention of disabling stroke is sufficient reason to label TIA as a precursor for stroke, but some risk factors may be more or less associated with TIA than with ischemic stroke, suggesting differences in mechanism. Methods: The medical records linkage system for the Rochester Epidemiology Project provided the means of identifying first episodes of TIA in the Rochester, MN population among those who had not had ischemic stroke. Control subjects were selected from an enumeration of the population through the medical records. The exposure to various risk factors was ascertained. The conditional likelihood approach to estimate the parameters of a multiple logistic model permitted estimation of the OR for TIA for each risk factor while adjusting for confounding variables. Results: The multivariable logistic regression model for TIA shows that the estimates of the ORs for ischemic heart disease, hypertension, persistent atrial fibrillation, diabetes mellitus, and cigarette smoking are similar to the ORs for those variables in the ischemic stroke model. However, the OR for mitral valve disease in the TIA model is 0.4, suggesting that mitral valve disease is unlikely to be associated with cerebral ischemic episodes that are brief enough to be called TIA.