The case for precision language regarding menopause

We most often think of menopause following on from decreased ovarian activity, resulting in lower production of 17β-oestradiol (E2) and progesterone, which happens as women age. At an average age of 51 years, women typically report cognitive issues, mood changes, sleep problems, hot flashes and night sweats. ^{Reference Harlow, Gass, Hall, Lobo, Maki and Rebar1} While this is the menopause that comes to mind when we use the term, there are ‘many menopauses’, each with their own causes, ages and physiological, cognitive and mood repercussions. Understanding each of their long-term effects on women’s health is imperative for precision treatment.

Previous guidelines have been published attempting to specify the types of menopause, as well as the staging of the most common type, spontaneous menopause. The 2022 position statement of the North American Menopause Society (NAMS) recognises the following types: (a) natural or spontaneous menopause, diagnosed retrospectively following 12 months of amenorrhea and characterised by a gradual decrement of E2 due to ageing at approximately 51 years of age; (b) induced menopause (bilateral oophorectomy, including or not including the fallopian tubes, BSO), on average 10 years earlier than spontaneous menopause and marked by a complete and abrupt loss of E2; (c) primary ovarian insufficiency (POI), a loss of ovarian function before the age of 40 years; and (d) early menopause, which occurs in women aged 40–45 years. ^{Reference Harlow, Gass, Hall, Lobo, Maki and Rebar1}

Within spontaneous menopause, the Stages of Reproductive Aging Workshop (STRAW+10) identified different phases of ovarian ageing. ^{Reference Harlow, Gass, Hall, Lobo, Maki and Rebar1} Spontaneous menopause, itself, is the main phase and corresponds to stage zero (0), five stages precede spontaneous menopause (−5 to −1) and two follow it (+1 to +2). Stages −5 to −3 correspond to the general reproductive interval, with stage −3 signalling the late reproductive stage, which is subdivided into two substages (−3b and −3a). Stage −3b is characterised by regular menstrual cycles with no change in length or follicular phase, but anti-müllerian hormone (AMH) and antral follicle count (AFC) are low. Stage −3a is characterised by shorter cycles and early follicular phase follicle-stimulating hormone (FSH) levels. Stages −2 to −1 correspond to the spontaneous menopause transition, which begins with perimenopause at stage −2 (increased FSH and decreased oestradiol and progesterone levels), continuing with early-stage spontaneous menopause. The latter is characterised by variability in menstrual cycle length, elevated but variable FSH levels and low AMH and AFC levels. Stage −1 signals late spontaneous menopause and is marked by amenorrhoea of 60 days or longer, increased variability in cycle length and high fluctuations in hormonal levels. Stage +1 is defined as early post-menopause (post-spontaneous menopause), and this is further subdivided into stages +1a, +1b and +1c. Stage +1a is the end of perimenopause and signals the end of the 12-month period of amenorrhoea − an indication that spontaneous menopause has occurred. Stage +1b lasts 1 year, and there are rapid changes in average FSH and oestradiol levels; this stage is also characterised by vasomotor symptoms. Stage +1c is the period in which there is a stabilisation of high FSH levels and low oestradiol levels, lasting from 3 to 6 years. Finally, stage +2 corresponds to late post-spontaneous menopause, which is characterised by symptoms of vaginal dryness and urogenital atrophy (somatic ageing).

However, STRAW describes only the stages of spontaneous menopause transition: there is no such transition for BSO. While the initial symptoms of both are similar, symptoms of BSO are more severe and more often longer lasting, unless mediated by hormone therapy. BSO has been linked to more severe vasomotor symptoms, hot flashes, cardiovascular disease, dementia, metabolic syndrome, low bone mineral density and sleep disturbance. ^{Reference Nappi2}

Most literature has focused on ‘postmenopausal’ women and thus has not specified which type of menopause their results refer to or which types were included in their study. Indeed, a review from our group showed that only 18% of clinical papers published in 2018 differentiated menopause type. ^{Reference Edwards, Duchesne, Au and Einstein3} Importantly, those studies that specified the type of menopause or included only one type of menopause showed the clearest cognitive outcomes. In looking at how oestradiol levels affected cognitive changes, those that included all menopause types led to mixed results, with some showing null results and others positive effects, leaving a muddied literature to decipher.

The importance of designating the different menopauses cannot be sufficiently emphasised. For instance, abundant research is focusing on the effects of ‘menopause’ on cognition and mood to try to understand why Alzheimer’s disease is more prevalent in women than in men. In this sense, the relationship between cognition and mood is important because depression is a risk factor for Alzheimer’s disease. Psychiatric epidemiological research has shown that women have almost three times the odds of mood disorders than men, and a psychiatric episode at one point in their reproductive cycle makes them more vulnerable to other psychiatric episodes later in life. ^{Reference Epperson, Wisner and Yamamoto4} A recent review has highlighted this point and suggested that women are at risk for Alzheimer’s disease specifically during and after the spontaneous menopause transition, and that this risk might depend on women’s depressive history. ^{Reference Barth, Crestol, de Lange and Galea5} However, the evidence regarding mood disorders in any type of menopause is limited. In general, the literature shows a low prevalence of mood disorders for women with spontaneous menopause ^{Reference Epperson, Wisner and Yamamoto4} but a high risk for early middle-aged women following ovarian surgery. ^{Reference Nappi2}

When the literature focuses on ‘postmenopausal’ women, describing long-term health effects such as increased risk of osteoporosis, cardiovascular disease and Alzheimer’s disease, it can often include women with multiple menopause types. This conflation of menopause types can lead to more severe effects on cognition and mood with BSO, as well as increased risk of all health risks, ^{Reference Rocca, Shuster, Grossardt, Maraganore, Gostout and Geda6} driving worse results for what we interpret as spontaneous menopause. The effect of this is that we lose the importance of changes with BSO and risk a subsequent pathologisation of spontaneous menopause.

In this sense, research from our group ^{Reference Edwards, Duchesne, Au and Einstein3,Reference Gervais, Gravelsins, Brown, Reuben, Karkaby and Baker-Sullivan7,Reference Calvo, McFall, Ramana, Galper, Fuller-Thomson and Dixon8} and others has shown that early-life BSO might be the riskiest type of menopause, and this is consistent with epidemiological studies showing that, without oestradiol therapy, early-life BSO is linked to a higher risk of late-life Alzheimer’s disease. Taken together, findings like these highlight how critical it is to know the type and age of menopause, especially when studying or treating ‘postmenopausal’ women.

Overall, in our future studies and treatments it is critical to know the age and cause of ovarian cessation, referring to them by precisely as ‘early-life BSO’ or spontaneous menopause. When we do this, we begin to understand that the cognitive and mood effects of other types of menopause – POI, the female athlete triad, radiation and chemotherapy-induced ovarian failure – are woefully understudied and, importantly, that each of these menopauses may require their own hormone treatments.

The case for precision language regarding hormone therapy

Hormone therapy can involve oestradiol (E2) alone, or a combination of synthetic or bioidentical oestrogens and progestogens. There are three endogenous oestrogens: E2, oestrone and oestriol. The most typically prescribed oestrogens are E2 or conjugated equine oestrogens (CEE), while for progestogens the most common are either micronised progesterone (P4) or progesterone- or testosterone-derived synthetics, such as medroxyprogesterone acetate (MPA). ^{Reference Nappi2}

Thus, synthetic and naturally occurring E2 and oestrones are key ingredients in menopausal hormone therapy. Too long neglected as anything but a ‘sex’ steroid, E2 increases the numbers of neurites, synapses and neural activity. It is mainly produced in the gonads, but it is also synthesised in the brain from the conversion of testosterone via aromatase, and within the brain in areas critical for memory function and mood regulation. For example, E2 is synthesised in the prefrontal cortex, which supports executive function; and in the limbic system, which supports both memory and emotion regulation. Within the limbic system, neurons responsive to E2 are located in brain areas important for memory consolidation (hippocampus), sensory integration (thalamus), emotion processing (amygdala) and general cognitive functions (basal forebrain). Moreover, E2 has interactions with brain-derived neurotrophic factor (BDNF), a growth factor that enhances dendritic spines and memory functioning and modulates emotion regulation. Thus, E2 – and not just any type of oestrogen – plays a key role shaping memory, affective processing and, in general, brain plasticity.

Irrespective of what we know about E2’s beneficial brain effects, most hormone therapy studies encompass a broad range of oestrogens with multiple properties. For instance, a study specifically designed to test the efficacy of hormone therapy in preventing dementia, the Women’s Health Initiative Memory Study (WHIMS), included CEE plus MPA and found the combination to carry a risk for heart disease, stroke, blood clots, breast cancer and cognitive decline. Consequently, rather than simply stopping the use of CEE, the prescription of all types of hormone therapy (including those containing E2) was dramatically reduced. However, when different types of hormone therapy are compared, they are not all created equal; for example, E2-containing hormone therapy is more effective than CCE in improving verbal memory performance. ^{Reference Wroolie, Kenna, Williams, Powers, Holcomb and Khaylis9} Thus, linking CEE outcomes to the effects of oestradiol therapy would be a mistake.

The beneficial impact of clarifying type of menopause and hormone therapy

In considering the effects of hormone therapy on cognition and mood, it is critical to be specific about the type of hormone therapy and the type of menopause. Importantly, the effects of even one type of hormone therapy, oestradiol therapy (ET), may vary depending on type of menopause, timing of administration and age of the patient.

Studies using large cohorts such as UK Biobank have reported mixed results, with some finding cognitive benefits of hormone therapy for ‘menopausal women’ and others not finding any benefit at all. However, UK Biobank data collection does not specify the type of hormone therapy; in addition, these studies did not clearly identify the type of menopause. Thus, this amalgam of hormone therapy and types of menopause may have led to mixed results. ^{Reference Calvo, McFall, Ramana, Galper, Fuller-Thomson and Dixon8} On the other hand, when the KEEPS trial focused on spontaneous menopause and oestradiol therapy (ET_SM) over the course of 4 years, it found no significant effects of oestradiol therapy on cognitive decline in spontaneous menopause at average age 54 years. In addition, the Women’s Health Initiative Memory Study of Younger Women (WHIMSY, an ancillary study to WHIMS) focused on CEE-based therapies and also found no benefit or risk to cognitive function for women in spontaneous menopause aged 50–55 years. Thus, NAMS does not recommend hormone therapy as treatment for cognitive decline or Alzheimer’s disease for women in spontaneous menopause. ^{Reference Nappi2}

In contrast, other numerous large studies of women with surgical menopause prior to spontaneous menopause have shown cognitive and mood benefits of oestradiol therapy. Thus, the NAMS position statement does recommend oestradiol therapy following ovarian removal prior to spontaneous menopause and initiated immediately after surgery (ET_BSO). Consistent with the NAMS position statement, our Canadian cohort of 366 early middle-aged women taking (ET_BSO) has shown generally positive effects of oestradiol therapy for spatial working, associative memory and maintaining hippocampal DGCA23 volume, ^{Reference Gervais, Gravelsins, Brown, Reuben, Karkaby and Baker-Sullivan7} all of which are of utmost importance because decrements in all three of these have been suggested as early biomarkers for late-life Alzheimer’s disease. Thus, the mixed results of the studies from UK Biobank, as well as others, are a cautionary tale of the poor evidence that ensues when neither menopause type nor hormone therapy is specified. Conversely, studies that specify hormone therapy and menopause type have identified successful treatments for maintaining cognition.

Regarding the use of oestradiol therapy to treat mood disorders for spontaneous menopause, the evidence is also limited and very mixed. Some studies have focused on spontaneous menopause and found small to medium benefits of CEE use (but not oestradiol therapy) in treating mood disorders, while others have reported an increased risk. To date, no type of hormone therapy has been approved to treat depression during perimenopause or spontaneous menopause. Further research is needed to understand whether oestradiol therapy might be beneficial for mood changes related to early-life BSO.

Each type of menopause has unique health consequences; each type of hormone therapy has different effects on each type of menopause. We conflate them at our peril. The few studies that have specified the type of menopause and hormone treatment indicate that early-life BSO might be the most severe menopause type. In addition, they show that, while oestradiol therapy is not necessarily effective as a cognitive or mood treatment in spontaneous menopause, it may be effective following surgical menopause. Future research should carefully specify the type of menopause and hormone therapy. Indeed, although a radical suggestion, we might get rid of the terms ‘menopause’ and ‘hormone therapy’ by instead simply using the terms that apply to the type we are discussing. As we have argued, blanket terms lead to muddied results. Adding precise language to studies leads to precision treatments, upholding the integrity of scientific research and the quality of treatment.

Therefore, to advance rigorous language for rigorous results, we propose:

1. (a) limiting studies to one menopause type or comparing different types;

2. (b) studying different hormone therapies separately or, in comparing results, being clear about precisely which types are being compared – CEE, oestradiol therapy or E2 (for unopposed E2);

3. (c) carefully clarifying both type of menopause and type of hormone therapy by:

   1. (i) designating menopause type by its age: BSO_age, POI_age, SM_eage, etc.;

   2. (ii) indicating type of hormone therapy for each type of menopause: ET_SM, ET_BSO, CEE_SM, CEE_BSO, CEE_POI, etc.

By clarifying our language, we will begin to get results that point to the appropriate treatments for each type of menopause.