Oophorectomy, estrogen, and dementia: a 2014 update

Abstract

Current evidence suggests that estrogen may have beneficial, neutral, or detrimental effects on the brain depending on age, type of menopause (natural versus induced), or stage of menopause (early versus late), consistent with the timing hypothesis. Three studies have now compared women who underwent bilateral oophorectomy before menopause with referent women and consistently showed an increased risk of cognitive decline and dementia. These studies suggest a sizeable neuroprotective effect of estrogen naturally produced by the ovaries before age 50 years. In this article, we focus on neuroprotection as related to cognitive decline and dementia. Several case-control studies and cohort studies also showed neuroprotective effects in women who received estrogen treatment (ET) in the early postmenopausal stage (most commonly at ages 50-60 years). The majority of women in those observational studies had undergone natural menopause and were treated for the relief of menopausal symptoms. However, the clinical trials by the Women's Health Initiative showed that women who initiated ET alone or in combination with a progestin in the late postmenopausal stage (ages 65-79 years) experienced an increased risk of dementia and cognitive decline regardless of the type of menopause. Three observational studies have now formally tested the timing hypothesis, and showed that the neuroprotective or harmful effects of estrogen depend on age at the time of initiation of treatment and on stage of menopause. Therefore, women who undergo bilateral oophorectomy before the onset of menopause or women who experience premature or early natural menopause should be considered for hormonal treatment until the average age of natural menopause (around age 50 years). Recommendations for the use of ET by women who experience natural menopause at typical ages remain less certain, and more research is needed.

Keywords: Cognitive impairment; Dementia; Estrogen; Menopause; Oophorectomy; Timing hypothesis.

Fig. 1

Proposed chain of causality linking bilateral oophorectomy to the increased risk of cognitive decline and dementia. In our etiologic hypothesis, genetic variants or non-genetic factors are effect modifiers (or interaction variables) rather than confounders. APOE = apolipoprotein E; ESR1 = estrogen receptor 1; ESR2 = estrogen receptor 2.

Fig. 2

An update on the timing hypothesis. The effect of estrogen on the risk of cognitive decline or dementia varies with the age at which estrogen is removed (via bilateral oophorectomy) or added as pharmacological treatment. The figure shows the results of 13 studies expressed as relative risks (RR), odds ratios (OR), or hazard ratios (HR) and 95% confidence intervals (plotted on a logarithmic scale). Women with ovarian conservation have a reduced long-term risk of cognitive decline or dementia compared to women who underwent bilateral oophorectomy before menopause (most commonly before age 50 years). * One additional study confirmed the protective effect of ovarian conservation; however, the results were not reported using comparable relative risk estimates and could not be displayed (Bove et al., 2014). Treatment with estrogen in the early postmenopausal stage (most commonly at ages 50–60 years) is associated with a reduced long-term risk of cognitive decline or dementia in three meta-analyses. However, initiation of estrogen treatment in the late postmenopausal stage (ages 65–79 years) is associated with an increased risk of cognitive impairment or dementia. Three studies tested the timing hypothesis by conducting stratified analyses by age at the time of hormonal treatment. The two strata for the MIRAGE study are shown in red, for the Kaiser study are shown in blue, and for the Cache County study are shown in green. CEE = conjugated equine estrogen; MIRAGE = Multi-Institutional Research on Alzheimer Genetic Epidemiology study; MPA = medroxyprogesterone acetate; WHIMS= Women’s Health Initiative Memory Study (Bove, 2014; Rocca, 2007; LeBlanc, 2001; Shumaker, 2004; Shumaker, 2003; Yaffe, 1998; Hogervorst, 2000; Phung, 2010; Bove, 2013; Whitmer, 2011; Henderson, 2005; Shao, 2012). [Modified from Rocca et al., 2011(Rocca et al., 2011)].