Alzheimer’s Disease

By the year 2050 it has been estimated that 30 per cent of the population in Western Europe will be over the age of 65 and as many as 10 per cent will have Alzheimer’s disease (AD) [1]. Also, mild cognitive impairment (i.e. a preclinical stage of AD) has an estimated prevalence of 20–30 per cent in elderly people [2]. Dementia currently costs the UK health care system approximately £17 billion per year and is predicted to reach £35 billion within 20 years [3]. The social and economic implications of this are greatest among women because their life expectancy, and risk of developing AD, are greater than for men [4]. However, it has been calculated that if severe cognitive impairment could be reduced by 1 per cent per year, this would cancel out the estimated increases in the long-term health care costs [5], as well as reducing the significant emotional costs. Studies into the effects of HRT on AD risk and/or prevention have been conflicting.

Support for the protective effect of sex hormones on cognitive impairment and AD has come both from studies into the negative effects of early surgical oophorectomy, prior to the onset of menopause, and the positive effects of MHT prescribed postmenopause. In the former category it has been reported, for example, that oopherectomy before 49 years of age is associated with a significant increased relative risk of dementia, and this risk increased the earlier the age of oophorectomy [6]. Furthermore, the risk disappeared if women were prescribed MHT until at least 50 years of age. Case control and cohort studies have also reported a reduced relative risk of AD in postmenopausal women treated with MHT, compared with never-users. Meta-analyses of these studies suggested that the relative risk was reduced by up to 34 per cent [7]. Importantly, most of these studies were carried out in the US, at a time where typical clinical practice was to prescribe MHT from the perimenopause until around 60 years old [8]. More recently this practice has been postulated to be particularly significant. For example, a prospective observational study reported that although HRT initiated early postmenopause protected against AD, that hormonal treatment initiated after the age of 60 years did not [4]. Furthermore, in a large multicentre randomized controlled study (RCT) (the Women’s Health Initiative Memory Study), women over 65 years old randomized to MHT (with medroxyprogesterone acetate) had an increased risk of ‘all-cause’ dementia compared with the placebo group [9]. This finding is supported by a recent nationwide case control study from Finland [10], which reported a 9–17 per cent increased risk of developing AD in women prescribed, respectively, estrogen-only or estrogen plus progestin HRT. However, there was no increased risk reported in women prescribed MHT prior to the age of 60 for less than 10 years and the absolute risk in all women was small (i.e. 9–18 extra cases per 10 000 women per year).

In summary, these studies suggest that MHT prescribed to older women may have a small neutral or negative effect, particularly if it is combined with a progestin. However, MHT prescribed at a ‘critical period’ around the time of menopause (particularly postsurgical menopause in younger women) may be associated with a neutral or reduced risk of dementia in later life. Larger, appropriately powered RCTs are still needed to test the ‘critical period’ hypothesis further. The primary difficulty in designing such studies is the long delay between randomization and development of (or protection against) symptoms of dementia.

A compromise to this obstacle is to study the effects of MHT on cognitive markers for subsequent AD, including verbal episodic memory [11]. However, two recent double-blinded, placebo-controlled clinical trials designed to analyse the effects of MHT on cognitive function following prescription during the critical period did not detect any significant cognitive benefits. The first was the ancillary arm of the Kronos Early Estrogen Prevention Study (KEEPS), the Cognitive and Affective Study (KEEPS-Cog). This analysed the cognitive (and mood) effects of transdermal 17β-estradiol (50 mcg/day) or low-dose oral CEE (0.45 mg/day) + cyclical oral progestogen started within the first 3 years of natural menopause (average age 53 years). The study reported the absence of treatment-related cognitive benefits following a mean length of follow-up of 2.85 years [12]. The second study, the Early versus Late Intervention Trial with Estradiol (ELITE), analysed the effects of oral 17β-estradiol (1 mg/day) ± cyclical vaginal progesterone on cognition in women <6 years (i.e. Early Intervention, average age 55 years old) or >10 years (i.e. Late Intervention) postmenopause. The trial reported the absence of an effect of HRT on verbal memory, executive function or global cognition in either group [13].

Currently, the likelihood of an adequately powered RCT being funded to study the relationship between AD risk following prescription of MHT to younger women is low [14]. However, current evidence suggests that those requiring short-term MHT to treat significant perimenopausal symptoms should, at the very least, be reassured that there are no longer-term risks of therapy on cognitive function.

Depression

Major depressive disorder is the leading cause of disease-related disability among women worldwide [15]. Furthermore, later life depression in women increases cardiovascular mortality by 50 per cent [16]. Also, there is increasing evidence that fluctuation in reproductive hormones, such as at the time of the perimenopause, increases the risk of a major depressive episode in some women [17–26]. The biological basis of this risk is probably multifactorial. However, most of the evidence, albeit not all, suggests that it is predominantly driven by effects of estrogen on the brain. Recently, a double-blind RCT involving euthymic women in perimenopause, or early postmenopause, found that only 17.3 per cent of those randomized to transdermal 17β-estradiol (100 µg), and intermittent micronized progesterone, developed clinically significant depressive symptoms, compared with 32.3 per cent of women in the placebo group [27]. In addition to a protective effect, some studies have also reported that estrogen may have an antidepressant effect in perimenopausal women. For example, a 6-week RCT of 34 women with perimenopausal depression (major and minor) reported that 17β-estradiol (50 µg daily) was associated with a significant improvement in mood compared with placebo [28]. These findings were replicated in a 12-week double-blind RCT of 17β-estradiol (100 µg daily) in 50 women with perimenopausal depression (major/minor) or dysthymic disorder [29]. Furthermore, these results were still significant at the 4-week follow-up. Consequently, some doctors in the UK prescribe sex hormones as the first line of treatment in women with reproductive depression [30–34]. Nevertheless, the biological evidence to support this approach is limited and prescription for clinical depression (i.e. as defined by the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM–V) or the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD–10)) in this patient group remains controversial. Studies are required that directly compare the effects of estradiol, preferably in the form of an implant, patch or gel (i.e. to avoid the effects of first-pass metabolism) with antidepressant medication and/or cognitive behaviour therapy (CBT).

It is outside the scope of this chapter to comprehensively review earlier basic science and animal studies into these mechanisms, but these can be found elsewhere [35]. Instead the remainder of the chapter will focus on some of the recent in vivo techniques that have been used to study the neurobiological mechanisms that underpin estrogen’s effects on the brain over the menopause.

Neurobiological Mechanisms

Estrogen is a steroid hormone synthesized by aromatization of androgenic precursors (i.e. androstenedione and testosterone). Naturally occurring estrogens include, in order of potency, 17β-estradiol, estrone and estriol. Estradiol is mainly produced by ovarian granulosa and theca cells and is the predominant form of estrogen found in premenopausal women. Research in ovarectomized rhesus monkeys has recently demonstrated that estradiol can also be produced directly by neurons [36]. This finding suggests a novel role for ‘neuroestrogen’ as a neurotransmitter.

Estradiol mainly exerts its effects by binding to two intracellular estrogen receptors (ERs), ERα and ERβ [37]. However, subsequent modulation of brain function includes an interaction with a number of neurotransmitter systems [38, 39]. This interaction has been studied indirectly using neuroendocrine challenge tests and, more recently, more directly using in vivo brain imaging.

Neuroendocrine Studies

This early in vivo technique provided an indirect method to explore the effects of estrogen on neurotransmitter systems important to memory, AD and mood in the menopause. Studies have reported modulatory effects on MHT on the cholinergic, serotonergic and dopaminergic systems.

In Vivo Brain Imaging Studies

Early studies into structural integrity of neural tissue in postmenopausal women using structural magnetic resonance imaging (sMRI) reported that MHT had a neutral effect on grey and white matter volumes [53–55]. Others, including our group, have found that compared with never-users, MHT has positive effects on regional grey and white matter concentrations [56–58]. This has included modulation of brain regions that are known to be important in memory and mood. For example, studies have reported atrophy of the hippocampus, prefrontal cortex and medial temporal lobe regions (e.g. amygdala) following chronic depression and/or stress [59]. Furthermore, atrophy of these same regions is also associated with memory impairment. A consistent finding reported in sMRI studies postmenopause has been that HRT users had regional sparing of grey matter in prefrontal regions [56, 60] and the hippocampus [56, 58]. It has also been reported that MHT users had sparing of white matter in medial temporal lobe regions [56]. Furthermore, white matter hyperintensities (a putative marker of brain aging) have been reported to be less extensive in MHT users than non-users in cross-sectional [61] and longitudinal [62] studies. In summary, studies in postmenopausal women suggest that MHT either improves or has a neutral effect on the structural integrity of neural tissue in brain regions important to memory and mood.

The inconsistencies between findings in these studies are probably due to a variety of methodological factors. First, the effects of MHT may be limited to grey or white matter compartments within brain regions that are vulnerable to the effects of aging [63–65]. Thus studies that specifically focused on other brain regions [54] or whole brain grey and white matter volumes [53] may not detect significant between-group differences. Second, some studies failed to explicitly exclude ex-users of MHT among ‘non-users’ [53, 54] and/or may have included women that did not receive MHT around the time of menopause with current users [53, 54, 61]. The ‘critical period’ hypothesis, however, predicts that these factors would reduce between-group differences. Third, as described above, it is probable that factors such as the type of MHT used (e.g. CEE versus estradiol), whether MHT is opposed or unopposed by progestogens, and the mode of administration (e.g. oral or implant) may modulate the putative protective effects of MHT on brain aging [66]. Finally, the methodological technique applied to analysing brain structure may be important. Thus although brain morphometry of bulk regions (a combined measure of both white and grey matter of whole brain or lobes) can be examined using hand-tracing methods, subtle regional differences in grey and white matter may occur that are undetectable using this approach. This difficulty has largely been overcome by studies using voxel-based morphometry (VBM), which generate statistical parametric maps of the significant between-group differences in grey matter concentration [56, 60]. In summary, studies suggest that initiation of MHT at the time of menopause may modulate age-related differences in regional grey and white matter concentration in regions that are important to memory and mood. Further studies, using functional imaging, suggest that MHT may also modulate brain activity in these regions.

Functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) techniques have permitted analysis of the effects of estrogen on brain function in peri- and postmenopausal women. Several observational studies have reported a significant effect of MHT on brain function in postmenopausal women. An early PET study [53] that compared long-term MHT users to non-MHT users reported significant differences in the relative regional cerebral blood flow (rCBF) in brain regions that included the frontal and parahippocampal gyri during verbal memory, and parietal and parahippocampal cortices during visual memory. Furthermore, these changes were associated with improved memory performance – suggesting that these functional differences had significant behavioural consequences. A subsequent PET study in postmenopausal long-term MHT+ users, MHT non-users and women with AD reported that MHT– women had metabolic ratios that were intermediate to HRT+ women and AD patients in brain regions characteristically hypometabolic in AD (including the dorsolateral prefrontal, middle temporal and parietal cortices) [67]. Although the behavioural performance did not differ between MHT+ and MHT women, it was suggested that the relative hypometabolism in MHT women might be an early indicator of future cognitive decline.

The above studies of long-term MHT use have been supported by research into the short-term effects of MHT in women closer to the ‘critical period’. For example, a placebo-controlled, crossover fMRI study reported that MHT+ was associated with increased activation of frontal, prefrontal and inferior parietal regions during a verbal memory task [68].

In summary, studies of brain function in older women report that long-term MHT, if prescribed since around the time of menopause, and short-term MHT, if prescribed close to the time of menopause, probably have positive modulatory effects on brain function in brain regions that are important to memory. These findings have been supported by studies in younger, premenopausal women.

Pseudo-menopause

A useful model to research the effects of reduced ovarian function on the brain is to study the consequences of acute ovarian steroid suppression (i.e. post-GnRHa) in premenopausal women. One advantage of this approach is that it avoids confounding effects of the ‘healthy user’ bias. Studies using this approach by our group and others have reported a significant reduction in brain activity in similar regions reported postmenopause, including the dorsolateral prefrontal and parietal cortex [69, 70]. Furthermore, these changes have been reported to be reversed following estrogen (and progesterone) add-back [69] or the return of ovarian function [71].

In summary, studies into the effects of MHT on brain function in younger and older women suggest that it modulates brain function in regions critical to memory and mood, including the prefrontal cortex and hippocampus. However, these studies do not provide insight into putative biological mechanisms that might underpin these actions. One mechanism indirectly suggested by neuroendocrine challenge studies is that HRT might modulate specific neurotransmitter systems in these regions.

To study this we extended our earlier findings and analysed the interaction between GnRHa and a muscarinic antagonist, scopolamine. We used a verbal memory task to probe specific brain regions important to memory and mood. We reported that scopolamine reduced activation in the left inferior frontal gyrus (LIFG) during encoding, which was attenuated further by GnRHa [39]. Furthermore, using a visual working memory task we also found an interaction at the parahippocampus [72]. These findings were also associated with significant behavioural effects. Following pseudo-menopause with GnRHa, cholinergic antagonism produced a more significant deficit in response accuracy and response time, respectively. In summary, these findings suggest that one mechanism via which acute loss of ovarian hormones might modulate brain function is via an interaction with the cholinergic system in frontotemporal brain regions. Another putative mechanism via which sex hormones could modulate brain function in these regions is by a direct effect on the neuronal and glial function. This effect can be studied using proton magnetic resonance spectroscopy (1H-MRS).

Stay updated, free articles. Join our Telegram channel

Join