Background-Animal experiments and epidemiological studies have suggest
ed that high potassium intake may reduce the risk of stroke, but the e
vidence is inconclusive, and the role of other nutrients in potassium-
rich foods remains unknown. Methods and Results-We examined the associ
ation of potassium and related nutrients with risk of stroke among 43
738 US men, 40 to 75 years old, without diagnosed cardiovascular disea
ses or diabetes, who completed a semiquantitative food frequency quest
ionnaire in 1986. During 8 years of follow-up, 328 strokes (210 ischem
ic, 70 hemorrhagic, 48 unspecified) were documented. The multivariate
relative risk of stroke of any type for men in the top fifth of potass
ium intake (median intake, 4.3 g/d) versus those in the bottom (median
, 2.4 g/d) was 0.62 (95% CI, 0.43, 0.88; P for trend=0.007). Results f
or ischemic stroke alone were similar. Intakes of cereal fiber and mag
nesium, but not of calcium, were also inversely associated with risk o
f total stroke. These inverse associations were all stranger in hypert
ensive than normotensive men and were not materially altered by adjust
ment for blood pressure levels. Use of potassium supplements was also
inversely related to risk of stroke, particularly among men taking diu
retics (relative risk, 0.36; 95% CI, 0.18, 0.72). Conclusions-Although
these data do not prove a causal relationship, they are consistent wi
th the hypothesis that diets rich in potassium, magnesium, and cereal
fiber reduce the risk of stroke, particularly among hypertensive men.
Potassium supplements may also be beneficial, but because of potential
risks, use should be carefully monitored and restricted to men taking
potassium-losing diuretics.