Clinical determinants of poststroke dementia.

BACKGROUND AND PURPOSE Frequency of poststroke dementia is high, and stroke considerably increases the risk of dementia. The risk factors for dementia related to stroke are still incompletely understood. We sought to examine clinical determinants of poststroke dementia in a large well-defined stroke cohort. METHODS The study group comprised 337 of 486 consecutive patients aged 55 to 85 years who 3 months after ischemic stroke completed a comprehensive neuropsychological test battery and MRI, including structured medical, neurological, and laboratory evaluations; clinical mental status examination; interview of a knowledgeable informant; detailed history of risk factors; and evaluation of stroke type, localization, and syndrome. The DSM-III definition for dementia was used. RESULTS Frequency of any poststroke dementia was 31.8% (107/337), that of stroke-related dementia (mixed Alzheimer's disease plus vascular dementia excluded) was 28.4% (87/306), and that of dementia after first-ever stroke was 28.9% (79/273). The patients with poststroke dementia were older and more often had a low level of education, history of prior cerebrovascular disease and stroke, left hemispheric stroke (reflecting laterality), major dominant stroke syndrome (reflecting laterality and size), dysphasia, gait impairment, and urinary incontinence. The demented patients were also more frequently current smokers, had lower arterial blood pressure values, and more frequently had an orthostatic reaction compared with the nondemented stroke patients. The correlates of dementia in logistic regression analysis were dysphasia (odds ratio [OR], 5.6), major dominant stroke syndrome (OR, 5.0), history of prior cerebrovascular disease (OR, 2.0), and low educational level (OR, 1.1). When we excluded those with cerebrovascular disease plus Alzheimer's disease or those with recurrent stroke, the order of correlates remained the same. When the patients with dysphasia (n=30) were excluded, the correlates were major dominant syndrome (OR, 4.6) and low educational level (OR, 1.1). CONCLUSIONS Our data suggest that a single explanation for poststroke dementia is not adequate; rather, multiple factors including stroke features (dysphasia, major dominant stroke syndrome), host characteristics (educational level), and prior cerebrovascular disease each independently contribute to the risk.