Risk of Alzheimer's disease among elderly patients with anemia: population-based investigations in Olmsted County, Minnesota.

PURPOSE: To evaluate the association, if any, between Alzheimer's disease (AD) and anemia among the elderly.
METHODS: Both case-control and cohort methodologies were used to evaluate this association. The case-control study included all incident cases of AD whose onset occurred during 1980-1984 in Rochester, MN (n = 302). An age- and gender-matched control was selected from among Rochester residents seen for care at Mayo Clinic during the year of onset of the case. All community medical records for cases and controls were reviewed to identify women with hemoglobin levels of < 12 g/dL and men with hemoglobin levels of < 13 g/dL during the year of onset and the prior year for cases and during the index year and year prior for controls. Logistic regression was used to estimate odds ratios (OR). The retrospective cohort study used the 618 residents of Olmsted County, MN, who were > or = 65 years of age and whose anemia, as defined above, was newly recognized in 1986. Standardized incidence ratios (SIR) for AD were estimated by use of the person-years of follow-up in the cohort and the incidence rates of AD for this community.
RESULTS: In the case-control study, an almost twofold increase in occurrence of AD was associated with anemia (OR, 1.88; 95% CI, 1.17-3.03). For men and women, respectively, the ORs were 1.81 (95% CI, 0.75-4.39) and 1.96 (95% CI, 1.11-3.47). The cohort study showed no overall increase in risk of AD (SIR, 0.98; 95% CI, 0.67-1.37). For men, the SIR was a comparable 1.49 (95% CI, 0.79-2.56), whereas for women the SIR was only 0.79 (95% CI, 0.49-1.23).
CONCLUSIONS: Although anemia may be a risk factor for AD in the elderly, the mechanism by which anemia contributes to the pathogenesis of AD or to the unmasking of AD symptoms is unknown.