| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2004;35:1268.)
© 2004 American Heart Association, Inc.
Original Contributions |
Editorial Comment—Dementia After Stroke
Consultant Neurologist, Department of Neurology, Royal Free Hospital, London, UK
An extract of the first 250 words of the full text is provided, because this article has no abstract. |
Stroke and dementia were implicitly associated by the advent of the concept of multi-infarct dementia, but it was Tatemichi et al1 who put this in temporal context and brought it to popular attention by observing a 26.3% dementia incidence 3 months after hospitalization for stroke. Several subsequent studies largely confirmed Tatemichi’s work but gave incidence estimates for poststroke dementia from 6% to 32%,2,3 with most studies producing estimates toward the upper end of this range. As Ivan et al4 point out, these works are potentially inaccurate for several reasons, including cross-sectional study design, the use of hospital-based cohorts with the risk of selection bias, lack of controls, and no (or imperfect) exclusion of previous dementia, to which should be added the effect of the diagnostic criteria used.
What is remarkable about the data from Ivan et al is that it took 
10 years for 19.3% of the cases of dementia to develop, a level that in most other studies would have been achieved within 3 months. Approximately 40% of apparently incident cases in most studies are caused by preexisting dementia, most of which is Alzheimer disease (AD).5–7 Nine such cases were excluded for that reason here and, if these are added to the numerator and denominator, would amount to 18% of cases, largely explaining the difference. The remainder may be accounted for other reasons among the 28 cases that were not followed-up with cognitive examination at 6 months.
Thus, Ivan et al have produced a much lower estimate for the
Related Article:
Stroke 2004 35: 1264-1268.