Progressive cognitive impairment after stroke.

OBJECTIVE: We examined the putative relationship between stroke and cognitive function in the population-based prospective cohort of the Cardiovascular Health Study (CHS).
METHODS: Of the 5888 participants of the CHS aged 65 years or older, there were 5364 with more than one modified mini-mental (3MS) examination between 1992 and 1998. To determine the effect of baseline stroke before first and subsequent (stroke between two consecutive examinations) 3MS examination on cognitive function, linear regression models were computed with potential confounders entered as additional independent variables. Stroke was divided into right and left hemispheres or posterior circulation on the basis of the clinical and/or imaging information by the hospital that treated the event and subsequent adjudication by CHS committee.
RESULTS: Participants with baseline stroke had an average 3MS decline of 1.2 (95% confidence interval [CI]: -0.7 - -1.7) points per year more than those without one. Those with a history of subsequent stroke had an average first year 3MS decline of 6.2 (CI -8.7 - -3.7) for left hemisphere, 3.5 (CI -5.3 - -1.8) for right hemisphere, and 1.1 (CI -3.9 - 1.6) for posterior circulation more than those without stroke. The effect of stroke on the rate of cognitive decline appeared to ameliorate after the first year (test for linear trend among those with stroke, P = .003).
CONCLUSION: Results from this prospective population-based data study show that stroke in the left hemisphere results in a more pronounced decline in cognition than that in the right hemisphere and that cognitive loss because of stroke appears to attenuate over time, perhaps as a result of relearning.