OBJECTIVE:

To test the hypothesis that bone mineral density (BMD), a marker of cumulative estrogen exposure, is associated with cognitive function in nondemented older women.

DESIGN:

A prospective cohort study.

SETTING:

Clinical centers in Baltimore, Maryland, Minneapolis, Minnesota, the Monongahela Valley near Pittsburgh, Pennsylvania, and Portland, Oregon.

PARTICIPANTS:

We evaluated 8333 older community-dwelling women enrolled in the Study of Osteoporotic Fractures who were not taking estrogen replacement.

MEASUREMENTS:

Calcaneal and hip BMD were measured at baseline and at follow-up (4-6 years later); vertebral fractures were ascertained radiologically at year 6. Women were administered a modified Mini-Mental State Exam, Trails B, and Digit Symbol at baseline and at follow-up.

RESULTS:

Compared with women with higher bone mineral density, women with low baseline BMD had up to 8% worse baseline cognitive scores (P = .001) and up to 6% worse repeat cognitive scores (P = .001), even after multivariate adjustments. For 1 SD decrease in baseline hip BMD or calcaneal BMD, women had a 32% (95% CI, 19-47%) or a 33% (95% CI, 20-48%) greater odds of cognitive deterioration (worst 10th percentile of change). Women with vertebral fractures had lower cognitive test scores and a greater odds of cognitive deterioration than those without fractures (OR = 1.29; 95% CI, 1.03-1.60).

CONCLUSIONS:

Women with osteoporosis, whether measured by baseline BMD, reductions in BMD, or vertebral fractures, have poorer cognitive function and greater risk of cognitive deterioration. Our findings suggest a link between two of the most common conditions affecting older women. Further understanding of this association may be important for new treatment and prevention directions.