Enhanced cognitive performance with estrogen use in nondemented community-dwelling older women

To examine the association between history of postmenopausal estrogen use and cognitive function in a large sample of nondemented community-dwelling older women.

A community of older residents in Cache County, Utah.

A total of 2338 nondemented women aged 65 and older.

All subjects were administered the Modified Mini-Mental State Examination (3MSE). Self-reported information on current and past use of estrogen after menopause was also obtained using a structured interview. Estrogen use was trichotomized as: no use, past use, and current use. Apolipoprotein E (APOE) genotype was determined and was dichotomized by the presence of an epsilon4 allele. A series of variance/covariance models was conducted with the 3MSE score as the dependent variable, first considering estrogen use alone, then adding, sequentially as covariates, education, age, health status, APOE genotype, current depression status, and history of head injury.

In the simplest bivariate model, the 3MSE means (and confidence intervals) were 92.1 (91.7-92.4), 93.5 (93.1-93.9), and 94.4 (94.0-94.7) for never-, past-, and current users, respectively. In the final model (R2 = 0.28), no use of estrogen replacement therapy (P = .006), lower education (P < .001), poorer perceived health status (P = .035), current depression (P = .014), and presence of at least one APOE epsilon4 allele (P < .001) each independently predicted lower 3MSE score. Both current and past estrogen users had significantly higher 3MSE scores than never-users (P = .0063 and P = .0096, respectively).

In this large community study, women who had used estrogen after menopause scored higher on the 3MSE. This finding remained, even after controlling for the effects of age, education, APOE genotype, and other variables that may affect cognition. These data support studies reporting a beneficial role of estrogen on cognition in postmenopausal women, particularly among current estrogen users.