Objectives: To determine whether the rate of bone loss predicts subsequent cognitive decline independently of baseline bone mass and whether apolipoprotein E (ApoE) genotype explains the association.
Design: A prospective cohort study.
Setting: Clinical centers in Baltimore, Maryland; Minneapolis, Minnesota; Pittsburgh, Pennsylvania; and Portland, Oregon.
Participants: Four thousand four hundred sixty-two women aged 70 and older (mean = 75.8) participating in the Study of Osteoporotic Fractures.
Measurements: Total hipbone mineral density (BMD) was measured 2 and 6 years after enrollment (mean follow-up = 3.5 years), and expressed as annualized percentage rate of bone change. A modified Mini-Mental State Examination (mMMSE) was administered at 6 and 10 years (mean follow-up = 4.5 years) and defined cognitive decline as a decline of three or more points on repeat mMMSE score. ApoE genotype information was available on 883 women.
Results: Cognitive decline occurred in 12% of the women with the least bone loss (by quartile), 14% in the second, 16% in the third, and 20% in those with the greatest bone loss. After adjustment for age, education, stroke, functional status, estrogen use, body mass index, and smoking, the results were similar. Those who lost the most BMD were almost 40% more likely than women in the lowest quartile to develop cognitive decline in the multivariate model (odds ratio (OR) = 1.4, 95% confidence interval (CI) = 1.1-1.8). A similar association between hipbone loss and cognitive decline was observed in the multivariate model further adjusting for ApoE e4 (OR = 1.5, 95% CI = 0.8-2.7).
Conclusions: Women with more rapid hipbone loss were more likely to develop cognitive decline than those who had lower rate of loss (or who gained bone mass). Differences in functional status, estrogen use, and ApoE did not explain this association. Further investigation is needed to determine the mechanisms that link osteoporosis and cognitive decline.