Frequency and predictors of stroke death in 5,888 participants in the Cardiovascular Health Study

W. T. Longstreth, C. Bernick, A. Fitzpatrick, M. Cushman, L. Knepper, J. Lima, C. D. Furberg

Research output: Contribution to journalArticlepeer-review

166 Scopus citations


Background: Few population-based studies have examined in detail issues of stroke-related deaths in elderly people. Methods: Participants in the Cardiovascular Health Study (CHS) are 65 years of age or older, have had extensive baseline evaluations, and have been followed-up for fatal and nonfatal cardiovascular and cerebrovascular disease outcomes. Investigators adjudicated these outcomes and classified strokes by types and subtypes. Results: Over 7 years, 1,310 (22.2%) of 5,888 participants died, and 455 (7.7%) experienced incident stroke. For the 5,888, stroke mortality was 3.2 per 1,000 person-years. For the 455, it was 36.1 per 1,000 person-years, with the most lethal type being hemorrhagic and the ischemic subtype being cardioembolic. After controlling for age and stroke type, the only other independent predictor of death after any stroke was poor performance on a timed walk measured before the incident stroke. Considering only ischemic stroke, the independent predictors of death were African American race and poor performance on timed walk. Conclusion: In CHS, death attributable to stroke is common. As in other studies, the most lethal stroke type was hemorrhagic, and ischemic stroke subtype, cardioembolic. Slow walking, possibly a measure of frailty, was associated with an increased risk of death of stroke. Finally, African Americans faced a greater risk of death than others after an ischemic stroke.

Original languageEnglish (US)
Pages (from-to)368-375
Number of pages8
Issue number3
StatePublished - Feb 13 2001

ASJC Scopus subject areas

  • Clinical Neurology


Dive into the research topics of 'Frequency and predictors of stroke death in 5,888 participants in the Cardiovascular Health Study'. Together they form a unique fingerprint.

Cite this