Estrogen and cognitive function

INTRODUCTION — The role of estrogen and menopausal hormone therapy (MHT) in cognitive function and Alzheimer disease (the major cause of dementia) has attracted considerable scientific and public interest. Finding a means to delay the onset of dementia is of substantial public health significance because approximately 5 percent of United States females over 60 years of age have dementia; this rises to 12 percent after age 75 years and 28 percent over age 85 years. In addition, the prevalence of dementia will increase markedly in coming years since older adults are the fastest-growing segment of the population.

This topic review will discuss the biologic basis for the hypothesis that estrogen is related to cognitive function and/or Alzheimer disease. Individual areas that will be explored are the role of estrogen in the following:

●Maintenance of cognitive function in females without dementia

●Prevention of Alzheimer disease

●Treatment of females with Alzheimer disease

Specific recommendations concerning the overall risks and benefits of postmenopausal estrogen therapy are reviewed separately. (See "Menopausal hormone therapy: Benefits and risks" and "Treatment of menopausal symptoms with hormone therapy".)

ESTROGEN AND COGNITIVE FUNCTION

Biologic basis — Substantial biologic evidence supports the importance of estrogen to cognitive function. Estrogen receptors have been identified throughout the brain and appear particularly concentrated in the basal forebrain. The basal forebrain is of special interest since it is the major source of cholinergic innervation to the hippocampus. The cholinergic system is a neurotransmitter system important for regulation of memory and learning, while the hippocampus is the primary region of the brain mediating cognitive function. In experiments using animal models and cell lines, several mechanisms have been identified whereby estrogen may influence cognitive function:

●The cholinergic system is enhanced by estrogen [1-3]. As an example, estrogen increases synthesis of acetylcholine (the chemical cholinergic neurons use to communicate with other nerve cells) by stimulating choline acetyltransferase activity and raises the concentration of hypothalamic nicotinic acetylcholine receptors.

●The glutamate system, a second neurotransmitter system involved in learning and memory, is also influenced by estrogen [4,5]. Estrogen increases the expression of proteins from the N-methyl-D-aspartate (NMDA) receptor, which is involved in glutamate activation and enhances long-term potentiation (the process by which we learn new things) in conditions that favor activation of NMDA receptors.

●Estrogen stimulates neurons and their ability to communicate with each other [6,7] and may contribute to regulation of genes that influence neuron survival, differentiation, regeneration, and plasticity. As an example, in the hippocampus, dendritic spine density is higher and synaptic properties of neurons are augmented in the presence of estrogen.

●Estrogen may protect nerve cells from excitotoxins and may act as an antioxidant to shield nerve cells from free radical damage [8,9].

●The brain is a site of high energy requirements and is reliant on efficient mitochondrial function. Estrogen may improve mitochondrial function in the brain [10]. Mitochondrial estrogen receptors appear to be present in neurons, including hippocampal neurons [11], and estrogen benefits the bioenergetics system, including enhancing ATP production and increasing mitochondrial respiration [12-14].

●Estrogen may influence DNA repair. For example, in ovariectomized monkeys, estrogen supplementation increases transcription of DNA repair enzymes in the brain [15].

●Nonetheless, estrogen also increases levels of C-reactive protein, an inflammatory marker; inflammation has been linked to diminished cognitive function and to dementia. In older postmenopausal females, oral estrogen may increase the risk of stroke, a risk factor for cognitive impairment and dementia. (See "Menopausal hormone therapy: Benefits and risks", section on 'Stroke'.)

Epidemiologic evidence

Effect of endogenous estrogen — Females appear to be at greater risk than males for eventual age-related dementia and memory impairment, but little is known about sex-based memory differences earlier in life. During the menopausal transition, some females describe bothersome memory symptoms, such as forgetfulness, difficulties with word retrieval, and "brain fog" [16,17].

However, in a cross-sectional study of over 200 males and females ages 45 to 55, females in their late reproductive or perimenopausal years (determined by menstrual status and serum estradiol and follicle-stimulating hormone concentrations) outperformed age-matched males in detailed memory tasks [18]. These apparent sex differences were attenuated in the postmenopausal years. Higher serum estradiol concentrations in females, a reflection of ovarian activity, were associated with better performance. Thus, some evidence suggests that in early midlife, females appear to outperform males on memory tasks, but with the onset of menopause and decline in serum estradiol, any memory advantage diminishes.

Biomarkers — Examining biomarkers of estrogen exposure in postmenopausal females is another approach to understanding whether estrogen may influence cognitive function. These markers may provide a broader representation of overall estrogen status than simple questions on postmenopausal hormone use.

Two studies have investigated whether plasma estrogen concentration may be related to cognitive function. Neither found any relation between estradiol or estrone and cognition or cognitive decline over time [19,20]. However, one of these studies reported that females with the highest levels of free estradiol and bioavailable estradiol (perhaps representing estradiol available to the brain) showed less cognitive decline on the Mini-Mental Status Examination (MMSE) over time (figure 1), although this was based upon a small number of females who had evidence of cognitive impairment [19].

The length of a female's reproductive period (ie, menarche until menopause) is one indicator of total exposure to endogenous estrogen. In an observational study of 3602 females over age 55 years, a longer reproductive period was not associated with a reduced risk of dementia [21].

Effect of exogenous estrogen — Numerous epidemiologic studies have examined whether postmenopausal estrogen use affects cognitive function in older females without dementia. Many of these reports included large numbers of females over age 65 years, collected detailed information on postmenopausal hormone use, or conducted extensive cognitive testing. Data are conflicting, with some epidemiologic studies showing evidence of mild benefit [22-25] and other larger studies reporting no benefit or possible harm.

The largest series is the Nurses' Health Study, a cohort of over 13,000 females aged 70 years and over, where no relationship was seen between current or long-term hormone use (at least 5 to 10 years of use) and results on six different cognitive tests (figure 2) [26]. Females who were in the menopausal hormone therapy (MHT) group who also had the apolipoprotein E (APOE) E4 allele were more likely to experience a worse cognitive decline [27].

Clinical trial evidence — Although epidemiologic evidence is conflicting, clinical trial evidence has generally ruled out any global (but not domain-specific) cognitive benefits of unopposed estrogen or combined estrogen-progestin therapy administered to females over age 65 years without underlying dementia. Thus, we suggest that MHT should not be prescribed for preservation of cognitive function in older postmenopausal females.

Strong evidence of cognitive benefits for females taking MHT at younger ages (eg, near menopause) is also lacking, and thus, MHT should not be prescribed for preservation of cognitive function in younger females either.

In contrast, strong support for the lack of global cognitive benefits with combined estrogen-progestin therapy and even the suggestion of some harm in older postmenopausal females comes from an ancillary study of the Women's Health Initiative (WHI), the WHI Memory Study (WHIMS), which assessed annual MMSE scores in 4532 postmenopausal females enrolled in the larger WHI study who were over age 65 years and free of probable dementia at baseline [28]. After a mean follow-up of 4.2 years, no significant improvement in global cognitive function was seen with combined estrogen-progestin therapy compared with placebo [28]. In addition, more females in the MHT group had substantial and clinically important declines in MMSE scores compared with placebo (6.7 versus 4.8 percent, respectively).

Similar results were reported in the WHIMS estrogen-alone trial of 2947 females over age 65 years who had previously undergone hysterectomy. After a mean follow-up of 5.4 years, mean MMSE scores were significantly lower in the estrogen group compared with placebo, particularly in females who had lower cognitive function at baseline [29]. The overall difference between the estrogen and placebo groups (0.26±0.13 units) was not clinically relevant, but the females in the estrogen group were at greater risk of having a clinically significant decline (≥10 units) on the MMSE (relative risk [RR] 1.47, 95% CI 1.04-2.07). Of note, the baseline cognitive function was significantly lower in the unopposed estrogen WHIMS population when compared with the combined estrogen-progestin population, but results were similar when data from the two studies were pooled.

In a much smaller randomized trial (n = 461) with shorter follow-up (mean three years), females with cerebrovascular disease were randomly assigned to receive unopposed estradiol or placebo [30]. After excluding females who had had a recurrent stroke, no significant relation was found between estrogen and substantial decline on the MMSE (RR 0.74, 95% CI 0.49-1.13) or on five tests of specific cognitive domains (RRs ranging from 0.81 to 1.33). In secondary subgroup analyses, there was a significantly decreased risk of substantial cognitive decline on the MMSE among females with normal cognition at baseline (n = 268; RR 0.46, 95% CI 0.24-0.87); however, no significant relations were observed for any of the tests measuring specific cognitive domains (RRs ranging from 0.68 to 0.93). These subgroup analyses should be interpreted with caution.

Similarly, another small trial (n = 115) of oral estradiol therapy in females aged 70 years and older reported no difference in cognitive performance after 20 weeks of follow-up in those assigned to estradiol versus placebo. These null results were consistent across tests of a large variety of cognitive domains [31].

A meta-analysis of clinical trials reported that neither unopposed estrogen nor combined estrogen-progestin therapy protected against a decline in overall cognitive function in normal, older (>60 years) postmenopausal females [32].

None of the initial WHI results described above [28,29] address the "critical window hypothesis," which states that early, rather than later, initiation of MHT is necessary to derive later cognitive benefits (similar to what has been observed for coronary heart disease). (See "Menopausal hormone therapy and cardiovascular risk", section on 'Timing of exposure'.)

While limited clinical trials suggest that early exposure to MHT may provide protection against later cognitive impairment, data from the WHI on early postmenopausal females found no beneficial effect:

●In a study of 343 females who had participated in randomized trials of two to three years of early postmenopausal hormone use for prevention of bone loss, follow-up cognitive testing (performed at a mean age of 65 years) suggested that females who had received MHT were at lower risk of cognitive impairment (odds ratio [OR] 0.36, 95% CI 0.15-0.90) than females who had been assigned to placebo [33]. However, these results were based on only eight cases, and the confidence interval was quite wide; in addition, 50 percent of subjects who had originally participated in the trials were not included in analyses, as they either did not participate in the cognitive testing (30 percent) or were noncompliant during the trials (20 percent).

●No beneficial effect of MHT was seen in an addition to the WHI called the "WHIMS-Young (WHIMSY)." This follow-up study included 1272 participants who had been ages 50 to 55 years when originally assigned to MHT or placebo [34]. Cognitive assessment administered via telephone interview (on average seven years after the trial ended when the mean age of participants was 67 years) showed similar global cognitive function scores in females who had received MHT near menopause (unopposed conjugated estrogen or combined with medroxyprogesterone acetate) when compared with placebo.

Some investigators maintain that female advantages in verbal abilities support the hypothesis that verbal memory specifically should be better maintained among postmenopausal hormone users, although general cognition may not be affected. Data to support this hypothesis are limited. In several small studies, females taking hormones seemed to perform better on verbal memory tasks [35-37], while in three of four large studies, hormone use was not related to verbal memory [26,38-40].

In a follow-up report from WHIMS, combined estrogen-progestin therapy had a negative impact on verbal memory [41]. Other specific cognitive domains have also been examined. In a few studies, verbal fluency (word generation rather than memory) in females taking estrogen was better than in nonusers [26,38,40], and in WHIMS, there was a suggestion of better figural memory for females assigned to combined MHT than placebo [41].

Trials in younger menopausal females — Experimental data in females have also contributed to the hypothesis that estrogen is involved in cognitive function. A randomized, clinical trial of 46 postmenopausal females on oral estrogen or placebo therapy found that estrogen-treated females demonstrated increased brain activation (on magnetic resonance imaging [MRI]) during cognitive tasks [42]. In another study, cerebral blood flow was increased during cognitive tasks in females taking postmenopausal estrogen [43].

The issue of low-dose estrogen administration in the early years was addressed in the Kronos Early Estrogen Prevention Study (KEEPS), a trial of MHT in 693 younger menopausal females ages 45 to 54 years who underwent cognitive testing [44]. KEEPS tested two hormone regimens: 0.45 mg of oral conjugated equine estrogen with 200 mg micronized progesterone, and 50 mcg transdermal estrogen with 200 mg micronized progesterone, versus placebo. Up to four years of MHT had no overall effect on cognition when compared with placebo.

In the Early versus Late Intervention Trial with Estradiol (ELITE), 643 postmenopausal females, stratified according to time since menopause (<6 or >10 years; early versus late, respectively), received oral estradiol (with vaginal progesterone for females with a uterus) or placebo for a median of five years [45]. When compared with placebo, estradiol (whether it was started early or late) had no effect on verbal memory, executive function, or global cognition. These findings suggest that early exposure to estrogen after menopause does not have a differential beneficial effect on cognition when compared with later exposure, or when compared with no exposure.

ESTROGEN AND DEMENTIA

Pathophysiology — Amyloid plaques are one of the pathologic hallmarks of Alzheimer disease. Many investigators believe that excessive deposition of these plaques is critical to the development of this disorder. Preliminary data suggest that estrogen may be important in preventing amyloid deposition, in addition to any role it plays in cognitive function [46]. (See "Prevention of dementia".)

Epidemiologic studies — The relationship between menopausal estrogen use and dementia has been extensively examined in observational studies [47-57]. In an early meta-analysis of observational studies, the relative risk (RR) of dementia was significantly reduced (RR 0.66) [48]. However, there were methodological concerns, including variability in study design and the possibility of confounding, which may have influenced the conclusion. A 2014 meta-analysis of 15 prospective, observational studies of menopausal hormone therapy (MHT) and Alzheimer disease reported no significant association between MHT and risk of Alzheimer disease (summary RR 0.88, 95% CI 0.66-1.16) [58].

Further, in a study of more than 160,000 postmenopausal females, use of systemic hormone therapy was associated with a small increased risk of Alzheimer disease. Thus, overall, the epidemiologic studies do not support a meaningful benefit of hormone therapy for dementia.

Timing of exposure — Some epidemiologic studies suggest that use of estrogen in the early postmenopausal years is associated with a lower risk of dementia, while later use is not [54,59-61]. However, other reports have not confirmed these findings [23,57,62], so this area requires further investigation.

In one report of 1889 females (mean age 74.5 years), prior use but not current use of estrogen (unless >10 years) was associated with a lower risk of Alzheimer dementia over three years of follow-up (hazard ratio [HR] for prior use 0.33, 95% CI 0.15-0.65; and HR for current use >10 years 0.55, 95% CI 0.21-1.23) [54]. In a subsequent report based upon seven additional years of follow-up of the same cohort, females who had initiated MHT within five years of menopause had a decreased risk of later dementia, while females who started MHT five or more years after menopause did not [60].

A second, larger study among more than 5000 females also found that use of MHT only in midlife was related to a 26 percent decreased risk of dementia (RR 0.74, 95% CI 0.58-0.94), but those taking hormones only in later life had an increased risk (RR 1.48, 95% CI 1.10-1.98) [59]. This observation is consistent with the clinical trial data (the Women's Health Initiative [WHI], performed in older females) that suggest either no benefit or an increased risk of dementia. (See 'Clinical trial data' below.)

In contrast, two other studies reported that use of MHT around the time of menopause was not statistically associated with a reduction in dementia risk (both showed a nonsignificant decrease in RR) [23,62].

Clinical trial data — As noted above, the WHI Memory Study (WHIMS), an ancillary study of the WHI, reported that combined estrogen-progestin (conjugated estrogen 0.625 mg with medroxyprogesterone acetate 2.5 mg daily) therapy does not have global cognitive benefits in older, nondemented, postmenopausal females. In addition, the WHIMS investigated the effects of combined therapy on all-cause dementia risk with the following results [63]:

●After a mean follow-up of four years, MHT was associated with an increased risk of dementia (40 cases in 2229 females in the MHT group versus 21 of 2303 females in the placebo group; HR 2.05, 95% CI 1.21-3.48). This translates to an additional 23 cases of dementia per 10,000 females per year attributed to MHT (figure 3).

●Combined estrogen-progestin therapy neither increased nor decreased the risk of mild cognitive impairment (HR 1.07, 95% CI 0.74-1.55; 63 versus 59 cases per 10,000 person-years for MHT and placebo, respectively).

Similar results were reported in the WHIMS estrogen-alone trial of 2947 females over age 65 years who had previously undergone hysterectomy [64]:

●A nonsignificant increase in dementia risk was seen with estrogen compared with placebo (HR 1.49, 95% CI 0.8-2.7)

●When the unopposed estrogen data were pooled with the combined estrogen-progestin data, the increased risk of dementia was significant (HR 1.76, 95% CI 1.19-2.60)

●A nonsignificant increase in the risk of developing mild cognitive impairment was seen in the estrogen compared with placebo group (HR 1.34, 95% CI 0.95-1.60)

●The distribution of the types of dementia diagnosed in the estrogen and placebo groups was similar

The WHIMS has also examined intermediate endpoints measured on MRI, the total ischemic lesion volume [65] and the total brain volume [66], both of which are associated with dementia. For total ischemic lesion volume, although there was a larger burden for those assigned to either estrogen-progestin or estrogen alone, compared with placebo, the difference was modest and not statistically significant. This indicates that vascular disease is not likely the primary mechanism by which MHT increased risk of dementia in the trial.

In the second report [66], females assigned to MHT had slightly but significantly lower brain volume on MRI when compared with placebo. The adverse effects were most evident in females who were experiencing cognitive deficits before starting MHT. This was the first evidence of possible lower brain volumes associated with hormone use and requires further confirmation in additional studies. However, these data are consistent with the detrimental effects of MHT on cognitive function observed in the WHIMS trial of older females, and research is necessary to better understand how hormones might lead to such damage in those who begin therapy many years after menopause.

Overall, unopposed estrogen and combined estrogen-progestin therapy do not prevent dementia in older, nondemented, postmenopausal females, and both therapies increased risk in the only large, randomized trials to date.

Given the available WHIMS data, neither unopposed estrogen nor combined estrogen-progestin therapy should be given to females over age 65 years for preservation of cognitive function and prevention of dementia.

The WHIMS trials do not address the question of the impact of estrogen administration in the early postmenopausal years on later dementia risk. However, due to important risks of postmenopausal hormone use (eg, stroke), MHT should not generally be considered for chronic disease prevention in postmenopausal females. (See "Menopausal hormone therapy: Benefits and risks".)

Recently menopausal females — In the Kronos Early Estrogen Prevention Study (KEEPS) trial described above, transdermal 17-beta estradiol therapy was associated with a decrease in amyloid-beta deposition on neuroimaging studies, particularly in apolipoprotein E (APOE) E4 carriers [67]; it remains unclear the exact interpretation of these findings in terms of cognitive health.

Treatment of dementia — Several small, randomized, clinical trials have evaluated estrogen treatment of females with Alzheimer disease.

In the largest study, 120 hysterectomized females with mild or moderate Alzheimer disease were randomly assigned to estrogen therapy (0.625 or 1.25 mg oral conjugated estrogen) or placebo for one year [68]. There was no difference between the groups on several tests of cognitive function, motor function, activities of daily living, or mood (figure 4). A smaller, 16-week trial reported similar results [69], as did a meta-analysis of seven trials [70]. Thus, estrogen therapy alone does not appear to be an effective treatment for Alzheimer disease. However, these studies are relatively small and cannot eliminate the possibility of a modest benefit from estrogen therapy.

There are several questions that have not been addressed (eg, will regimens other than oral conjugated estrogen or medroxyprogesterone acetate be effective? What is the result of beginning MHT at an older age compared with continuously taking it through menopause?). In addition, limited data suggest that estrogen may play a role in enhancing response to tacrine and might be of value as part of a combined pharmaceutical approach [71]. (See "Treatment of Alzheimer disease", section on 'Estrogen replacement'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Menopause".)

SUMMARY AND RECOMMENDATIONS

Menopausal hormone therapy (MHT) in older, postmenopausal females:

●There is limited epidemiologic support for the hypothesis that estrogen preserves overall cognitive function in nondemented females. However, in the Women's Health Initiative (WHI), both unopposed estrogen and combined estrogen-progestin therapy had no global cognitive benefits in older, nondemented, postmenopausal females and led to accelerated cognitive decline (see 'Epidemiologic evidence' above and 'Clinical trial evidence' above). It is possible that there are domain-specific cognitive benefits (such as verbal memory) with MHT, but definitive evidence is lacking. (See 'Clinical trial evidence' above.)

●Early, observational, epidemiologic studies of estrogen for the prevention of clinical Alzheimer disease indicated possible benefits but were limited by small numbers of patients and problems with study design (see 'Epidemiologic studies' above). In contrast, the WHI demonstrated that unopposed estrogen and combined estrogen-progestin therapy could possibly increase risk in females over age 65. The impact in younger postmenopausal females is still unclear. (See 'Clinical trial data' above.)

●We suggest not using estrogen or combined estrogen-progestin MHT for either the preservation of cognitive function in nondemented older females (Grade 2C) or for the treatment of Alzheimer disease (Grade 2C). (See 'Clinical trial evidence' above and 'Timing of exposure' above.)

MHT in younger postmenopausal females:

●Although epidemiologic and some clinical trial data suggest that use of MHT in the early postmenopausal years provides protection against later cognitive impairment, the WHI (the WHI Memory Study-Young ["WHIMSY"]), the Kronos Early Estrogen Prevention Study (KEEPS) trial, and the Early versus Late Intervention Trial with Estradiol (ELITE) trial in younger females saw no beneficial effect.

●The WHI trials have not addressed the question of early hormone use and later dementia risk. (See 'Clinical trial evidence' above and 'Timing of exposure' above.)

●We suggest not using estrogen or combined estrogen-progestin MHT for the preservation of cognitive function or dementia in younger (early) postmenopausal females (Grade 2C). (See 'Clinical trial evidence' above and 'Timing of exposure' above.)