Dementia after stroke is a predictor of long-term survival.

Abstract

The aim of this study was to determine whether dementia after stroke adversely influences long-term survival. Subjects were 251 patients > or = 60 years of age with ischemic stroke who were given neurological, neuropsychological, and functional examinations 3 months after hospitalization and were followed up prospectively. Using criteria modified from the Diagnostic and Statistical Manual of Mental Disorders-III-R, dementia was found in 66 (26.3%) patients at the 3-month baseline examination. Life-table methods were used to estimate mortality rates in the groups with and without dementia after 1 to 5 years of follow-up, Kaplan-Meier curves to estimate the cumulative proportion surviving with and without dementia, and Cox proportional-hazards analysis to compute the relative risk of mortality associated with dementia at baseline, after adjusting for other potential predictors of stroke mortality. The mortality rate was 19.8 deaths per 100 person-years with dementia compared with 6.9 deaths per 100 person-years without dementia. The cumulative proportion surviving after a median follow-up of 58.6 months was 38.9 +/- 0.08% for those with dementia and 74.5 +/- 0.04% for those without dementia. The relative risk associated with dementia was 3.11 (95% confidence interval, 1.79 to 5.41) after adjusting for the effects of demographic factors, cardiac disease, severity of stroke (Barthel Index), stroke type (lacunar versus nonlacunar), and recurrent stroke (examined as a time-dependent variable). When the Mini-Mental State Examination score at baseline was examined instead of the diagnosis of dementia, the results of the model were similar. Our study is the first to demonstrate that dementia or cognitive impairment adversely influences long-term survival after stroke, even after adjusting for other commonly accepted predictors of stroke mortality. Impairment in intellectual function after stroke, independent of physical disability, has a significant impact on prognosis. Both cognitive and physical functions should be assessed in clinical studies of stroke outcome.