Prediction of poststroke dementia

Objective: To investigate prospectively the frequency and clinical determinants of poststroke dementia (PSD) in a cohort of consecutive ischemic stroke inpatients in southern Taiwan. Methods: A standard stroke evaluation protocol was conducted at admission and 3 months after an ischemic stroke. The protocol included clinical, neurologic, neurobehavioral, and functional assessments as well as neuroimaging examinations. Diagnoses were made according to the Neurologic Adaptation of the 10th edition of the International Classification of Diseases criteria for dementia. Results: Excluding patients with prestroke dementia, a total of 283 patients were surveyed at 3 months after stroke; 26 (9.2%) of them met the criteria for PSD. The correlates of PSD in logistic regression analyses were age 65 years or older (odds ratio [ OR] 6.6) vs <65 years, previous occupation as a laborer (OR 3.3), prior stroke (OR 3.1), left carotid vascular territory (OR 12.5) vs vertebrobasilar and unknown territories, moderate to severe stroke severity (OR 3.4), and cognitive impairment (OR 4.5) and poorer functional status at admission (OR 4.5). Based on the significant predictors identified, the logistic regression model correctly classified PSD in 93.4% of subjects. Conclusion: The lower frequency of PSD in this study from Taiwan compared with previous studies from Western countries may have been due to the relatively younger age of the elderly population and the use of stricter diagnostic criteria.