Factors affecting survival and occurrence of stroke in patients with transient ischemic attacks

To assess the predictive value of a series of demographic and clinical variables for stroke and survival in a population after a first transient ischemic attack (TIA).

Cox proportional hazards regression analysis was used to determine the association of various demographic and clinical factors with survival and stroke in 330 residents of Rochester, Minnesota, who had an initial TIA with first medical attention within 120 days during the period 1955 through 1979.

We investigated several demographic, diagnostic, and treatment variables, including initial clinical manifestations (pure sensory TIA and unilateral carotid hemispheric TIA), to estimate the significant (P < or = 0.01) predictors of survival and of stroke. Follow-up was limited to 10 years.

Relative survival for patients with a first TIA was 94% at 1 year and 87% at 5 years after first medical attention. Three interactions were significant predictors of survival: (1) age at TIA and gender (young women had the best survival and older women had the worst survival), (2) systolic blood pressure and congestive heart failure (patients with low systolic blood pressure and congestive heart failure had the worst survival), and (3) calendar year of onset and diabetes mellitus (survival was worst for patients with diabetes during the early years of the study). Only age was a significant independent predictor of stroke after TIA (hazards ratio, 1.45 per 10 years).

Estimating risks of stroke and death after TIA on the basis of demographic and clinical variables without reference to the mechanism of TIA is of limited clinical utility. Age is the most significant such predictor. Interactions that reflect comorbidity, such as diabetes, blood pressure abnormalities, or heart disease, may affect survival but not the risk for occurrence of stroke.