Objective: Recent studies have suggested that vascular dementia in older persons is more common than previously hypothesized. A substantial proportion of dementia in old age may be an early manifestation of cerebrovascular disease (CVD), that eventually becomes clinically evident as an acute cerebrovascular accident. This study was aimed at assessing whether cognitive impairment and cognitive decline in older persons free of stroke are associated with higher risk of future stroke, independently of other risk factors.
Design: Population-based prospective study.
Participants: A total of 5024 subjects from the Established Populations for Epidemiologic Studies of the Elderly, who were alive and had no history of previous stroke at the sixth follow-up visit. Subjects who had reported a stroke in a previous interview or with a diagnosis of cerebrovascular disease in a hospitalization record during the previous 3 years were excluded.
Measurements: Cognitive function was assessed by the Short Portable Mental Status Questionnaire (SPMSQ). Occurrence of a stroke was prospectively assessed by examining hospital discharge diagnoses and death certificates.
Results: During 19,533 person-years of follow-up, 259 strokes were recorded (13.3/1000 person-years). Stroke incidence was lowest in those with normal SPMSQ score (12.1/1000 person-years), intermediate in those with moderate impairment (16.3/1000 person-years), and highest in those with severe impairment (30.9/1000 person-years). Adjusting for age, education, smoking, history of hypertension, blood pressure, heart attack, diabetes, and disability, the relative risks of stroke for moderate and severe cognitive impairment were 1.2 (0.9-1.6) and 2.2 (1.2-3.8), respectively. The association between cognitive impairment and incident stroke was not mediated by hypertension or diabetes. Compared with subjects with stable or improved SPMSQ score in the previous 3 years, those who declined had higher risk of stroke.
Conclusions: The elevated risk of subsequent strokes in older persons with cognitive impairment suggests that CVD may play larger role in causing cognitive impairment then previously suspected. It remains to be demonstrated whether reducing modifiable risk factors for CVD decreases the burden of cognitive impairment in older persons without stroke.