Estrogen and the brain: does estrogen treatment improve cognitive function?

In this paper we describe potential reasons for the discrepancies between data from basic sciences and observational studies and those of large treatment studies investigating the association between brain function and sex steroids. Observational studies which often showed positive associations between hormone use and cognition can be affected by 'recall bias' and 'healthy user bias', while outcomes of treatment studies were hypothesized to be modified by age at treatment, age at or type of menopause, health status, addition of a progestogen or type of estrogen treatment. However, meta-analyses of data from treatment studies negate many of these hypotheses showing at best mainly short-term (up to 6 months) positive effects of estrogen treatment on cognition regardless of age. This positive effect may reverse, particularly in older women with prolonged hormone treatment, which was predominantly seen after addition of progestogen. Medroxyprogesterone acetate seemed to have worse effects on cognition than other types of progestogen in these long-term studies. Estradiol with or without a progestogen was three times more likely to have positive effects on cognition than conjugated equine estrogens. However, two-thirds of studies showed no associations at all which may be an underestimate given the possibility of publication bias. We briefly review alternative treatments, such as testosterone and soy-derived supplements, but currently insufficient data are available for conclusive comments. Women who have undergone surgical menopause or who undergo natural menopause before age 47 may benefit most from hormone treatment and a special case may need to be made for this group. Long-term safety studies for this group are urgently needed.