Potassium, Calcium, and Magnesium Intakes and Risk of Stroke in Women

Susanna C. Larsson; Jarmo Virtamo; Alicja Wolk


Am J Epidemiol. 2011;174(1):35-43. 

In This Article

Materials and Methods

Study Population

Details of the Swedish Mammography Cohort have been reported elsewhere.[28] In brief, the cohort was established in 1987–1990, when all women born between 1914 and 1948 and living in central Sweden (Västmanland and Uppsala counties) received a questionnaire on diet via the mail. In the autumn of 1997, the 56,030 participants who were still alive and residing in the study area received a new expanded questionnaire that included approximately 350 items concerning diet and other lifestyle factors; 39,227 women (70%) completed the 1997 questionnaire. Compared with women who completed the baseline (1987–1990) questionnaire but not the 1997 questionnaire and who were still alive in 1997, those who completed both questionnaires were on average younger and had a slightly lower incidence of all types of stroke (total stroke). The age-adjusted yearly incidence rates of total stroke per 10,000 persons between 1987 and 1997 among women who completed the 1997 questionnaire and those who did not were 9 and 10, respectively.

Because information on several potential confounders (e.g., cigarette smoking, diabetes, hypertension, and physical activity) was first obtained in 1997, only women who completed the 1997 questionnaire were included in the present study. We excluded women with an erroneous or a missing national identification number, those with a history of stroke, coronary heart disease, or cancer before the start of follow-up, and those with implausible values for total energy intake (i.e., 3 standard deviations from the loge-transformed mean energy intake). This left 34,670 women aged 49–83 years for the present analysis, which used data from 1998–2008. The study was approved by the Regional Ethical Review Board at the Karolinska Institutet, Stockholm, Sweden.

Baseline Data Collection

The 1997 questionnaire included questions on educational level, weight, height, cigarette smoking, physical activity level, aspirin use, medical history, family history of myocardial infarction before 60 years of age, alcohol consumption, and diet. Body mass index was calculated by dividing the weight in kilograms by the square of height in meters. Pack-years of smoking history were calculated as the number of packs of cigarettes smoked per day multiplied by the number of years of smoking. Participants reported their level of activity at work, home/housework, walking/bicycling, and exercise in the year before study enrollment. The questionnaire also included questions on inactivity (watching television or reading) and hours per day of sleeping and sitting or lying down. The reported time spent at each activity per day was multiplied by its typical energy expenditure requirements, expressed in metabolic equivalents, and amounts for all activities were added together to create a metabolic equivalent-hours per day (24 hours) score.[29]

Dietary Assessment

A 96-item food frequency questionnaire was used to assess diet in 1997. On this questionnaire, participants indicated how often, on average, they had consumed various foods over the past year, by using 8 predefined frequency categories ranging from never to ≥3 times per day. For commonly consumed foods such as milk, cheese, and bread, participants could indicate how often per day or week they consumed these foods (open question). Nutrient intakes were calculated by multiplying the frequency of consumption by the nutrient content of age-specific (<53, 53–65, and ≥66 years of age) portion sizes by using composition values from the Swedish Food Administration Database.[30] All nutrients, except alcohol, were adjusted for total energy intake through the use of the residual method.[31] The questionnaire also asked about use of dietary supplements, including multivitamins with minerals, and some specific vitamin and mineral supplements, including calcium and magnesium. The food frequency questionnaire has been validated,[32] and the Spearman coefficients for the correlations between estimates (intake in mg/day) from the dietary questionnaire and the mean of 14 24-hour recall interviews were 0.77 for calcium and 0.73 for magnesium; potassium intake has not been validated.

Case Ascertainment and Follow-up

Incident cases of first stroke that occurred between January 1, 1998, and December 31, 2008, were ascertained by linkage of the study cohort with the Swedish Hospital Discharge Registry, which provides virtually complete coverage of the discharges. The International Classification of Diseases, Tenth Revision, was used to identify stroke events. Strokes were classified as cerebral infarction (code I63), intracerebral hemorrhage (code I61), subarachnoid hemorrhage (code I60), and unspecified stroke (code I64). Dates of death were obtained from the Swedish Death Registry.

Statistical Analysis

Participants were followed from January 1, 1998, until the date of first stroke event, death, or December 31, 2008, whichever came first. We used Cox proportional hazards regression models with age as the time scale to estimate the relative risks of stroke by category of potassium, calcium, and magnesium intake. Participants were categorized into quintiles (for total stroke and cerebral infarction) or tertiles (for hemorrhage strokes because of the smaller number of cases) of potassium, calcium, and magnesium intakes based on the distribution in the whole cohort. Entry time was defined as a subject's age in months at start of follow-up, and exit time was defined as a subject's age in months at stroke event or censoring. The proportional hazards assumption was tested and was found to be met for all variables except diabetes. We adjusted for diabetes by stratification in the Cox model. The multivariable model included the following variables: smoking status (never, past, or current smokers), pack-years of smoking (<20, 20–39, or ≥40 pack-years), educational level (less than high school, high school, or university), body mass index (<20, 20–24.9, 25–29.9, or ≥30), total physical activity level (quartiles), self-reported history of hypertension (yes or no), aspirin use (yes or no), family history of myocardial infarction before 60 years of age (yes or no), and intakes of total energy (in kcal/day, as a continuous variable), alcohol consumption (nondrinkers or <3.4, 3.4–9.9, or ≥10.0 g/day), and quintiles of protein, cholesterol, total fiber, and folate. The selection of variables for inclusion in the multivariable model was based on the association between the variable and potassium, calcium, or magnesium intake, as well as the association between the variable and risk of stroke observed in the present cohort study or reported in the literature.

Tests for linear trends were conducted by modeling the minerals as continuous variables by using the median value of each category. We conducted stratified analyses by history of hypertension (yes or no) to assess possible effect modification by this variable. Tests for interaction were performed using the likelihood ratio test. The statistical analyses were performed using SAS, version 9.1 (SAS Institute Inc., Cary, North Carolina). All P values were 2-sided.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: