Serum potassium and stroke risk among treated hypertensive adults

BACKGROUND: In prospective studies, hypokalemia has been shown to be associated with a subsequent increase in stroke risk in treated hypertensive adults after 4 to 16 years of follow-up. Stroke risk associated with more recent assessments of hypokalemia has not been examined. METHODS: We used data from an on-going, population-based, case-control study of incident stroke at Group Health Cooperative (GHC). Cases were treated hypertensive adults, 30 to 79 years old, who sustained an incident ischemic or hemorrhagic stroke between July 1989 and December 2000. Controls were GHC members without a history of stroke who met same inclusion criteria as cases. Serum potassium (K(+)) levels were collected from GHC laboratory database. Hypokalemia (< or =3.4 mM/L) was defined using the most recent serum K(+) measure in the year before the index date (event date for cases; random date within calendar year of identification for controls). Multivariate logistic regression estimated the relative risk of ischemic and hemorrhagic stroke. RESULTS: Among 593 ischemic and 125 hemorrhagic stroke cases, and 2397 controls, few subjects were hypokalemic: 3%, 6%, and 2%, respectively. Using the normal range of serum K(+) as a reference (3.5 to 5.0 mM/L), hypokalemia was associated with an elevated risk of ischemic (odds ratio OR: 2.04; 95% confidence interval CI: 1.14-3.64) and hemorrhagic stroke (OR: 3.29; 95% CI: 1.45-7.48) in adjusted analyses. Associations were not modified by diuretic use. CONCLUSIONS: In adults with treated hypertension, hypokalemia in the year before a stroke was associated with an increased risk of incident ischemic and hemorrhagic stroke independent of diuretic use when compared to normal serum K(+) levels.