Ozawa · Journal of the American Geriatrics Society 2012 · prospective cohort study · n=1081

Self-reported dietary intake of potassium, calcium, and magnesium and risk of dementia in the Japanese: the Hisayama Study.

Cited 126 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective cohort study

PubMed 22860881 · doi:10.1111/j.1532-5415.2012.04061.x · record verified 2026-08-31

What was done

In the prospective Hisayama Study in Japan, 1,081 community-dwelling participants aged 60 and older without dementia were followed for 17 years. Dietary intake of potassium, calcium, and magnesium was evaluated using a 70-item semiquantitative food frequency questionnaire. Hazard ratios for incident all-cause dementia, vascular dementia, and Alzheimer's disease were estimated across intake quartiles using multivariable-adjusted Cox proportional hazards models.

What was found

Over 17 years, 303 participants developed all-cause dementia, including 98 with vascular dementia and 166 with Alzheimer's disease. Comparing the highest with the lowest intake quartiles, multivariable-adjusted hazard ratios for all-cause dementia were 0.52 (95% CI = 0.30-0.91) for potassium, 0.64 (95% CI = 0.41-1.00) for calcium, and 0.63 (95% CI = 0.40-1.01) for magnesium. For vascular dementia, hazard ratios were 0.20 (95% CI = 0.07-0.56) for potassium, 0.24 (95% CI = 0.11-0.53) for calcium, and 0.26 (95% CI = 0.11-0.61) for magnesium. There was no evidence of a linear association between intake of these minerals and risk of Alzheimer's disease.

Why it matters

These results link higher intake of potassium, calcium, and magnesium to a lower risk of vascular dementia rather than Alzheimer's disease. This supports the potential role of dietary mineral intake in maintaining vascular cognitive health.

Limits

Dietary intake was self-reported using a food frequency questionnaire, creating potential for misclassification and recall error. As an observational study, residual confounding cannot be excluded, and causality cannot be determined. Findings in this Japanese cohort may not generalize to other demographic or dietary populations.

Cited by