Background and purpose: High intakes of calcium, potassium, and magnesium have been hypothesized to reduce risks of cardiovascular disease, but only a few prospective studies have examined intakes of these cations in relation to risk of stroke.
Methods: In 1980, 85 764 women in the Nurses' Health Study cohort, aged 34 to 59 years and free of diagnosed cardiovascular disease and cancer, completed dietary questionnaires from which we calculated intakes of calcium, potassium, and magnesium. By 1994, after 1.16 million person-years of follow-up, 690 incident strokes (129 subarachnoid hemorrhages, 74 intraparenchymal hemorrhages, 386 ischemic strokes, and 101 strokes of undetermined type) had been documented.
Results: Intakes of calcium, potassium, and magnesium were each inversely associated with age- and smoking-adjusted relative risks of ischemic stroke, excluding embolic infarction of nonatherogenic origin (n=347). Adjustment for other cardiovascular risk factors, including history of hypertension, attenuated these associations, particularly for magnesium intake. In a multivariate analysis, women in the highest quintile of calcium intake had an adjusted relative risk of ischemic stroke of 0.69 (95% CI, 0.50 to 0.95; P for trend=0.03) compared with those in the lowest quintile; for potassium intake the corresponding relative risk was 0.72 (95% CI, 0.51 to 1.01; P for trend=0.10). Further simultaneous adjustment for calcium and potassium intake suggested an independent association for calcium intake. The association of risk with calcium intake did not appear to be log linear; the increase in risk was limited to the lowest quintile of intake, and intakes > approximately 600 mg/d did not appear to reduce risk of stroke further. The inverse association with calcium intake was stronger for dairy than for nondairy calcium intake. Intakes of calcium, potassium, and magnesium were not related to risk of other stroke subtypes.
Conclusions: Low calcium intake, and perhaps low potassium intake, may contribute to increased risk of ischemic stroke in middle-aged American women. It remains possible that women in the lowest quintile of calcium intake had unknown characteristics that made them susceptible to ischemic stroke.