How preclinical models of menopause can inform clinical care: A focus on midlife and reciprocal communication between clinical and preclinical science

Abstract

Midlife in women typically includes the menopausal transition, a time of hormonal transformation, adaptation, and reorganization. Coincident with this dynamic period of physiological change, there are putatively modifiable factors that influence disease, short-term and long-term health outcomes, symptom emergence, and longevity. The menopause transition could be considered a window of vulnerability; however, it is also a window of opportunity for intervention. Thus, the menopause transition is a critical sensitive window whereby there is opportunity for turning points for healthy aging trajectories. Preclinical research can aid in this pursuit of scientific discovery for modifiable factors and treatments, and their particular parameters. Rodent menopause models include surgical and transitional variations, allowing detection of precise determinants impacting menopause-related outcomes. These models permit systematic manipulation of endogenous and exogenous hormone exposures across the lifespan, with infinite outcome measurements ranging from molecular to behavioral. This research is uniquely poised to address complex, interactive hypotheses with extensive control in a relatively short timeframe, including dissociation of age and menopause effects. To understand the many dynamic changes with menopause, iterative and reciprocal communication between clinical and preclinical domains of science is key.

Midlife is a time of hormonal transformation and the beginning of the menopausal transition in women, coinciding with emerging symptoms and health-associated changes that can influence health outcomes as aging ensues. An understanding of health determinants during midlife provides scientists the ability to target particular vulnerabilities, as well as opportunities for their amelioration, during this life stage [1]. Gaining insight into variables impacting the quality of life, risk factors for disease, and emerging symptom presentation in midlife and the menopausal transition can lead to positive consequences toward healthy aging. Indeed, it is known that reproductive characteristics, such as age of menarche or menopause, have been related to longevity in women [2]. The lifespan of the woman, currently about 81 years in the United States, surpasses the reproductive life stage by almost three decades [3,4]; menopause occurs around the average age of 52, and is clinically defined retrospectively after amenorrhea for one year [5]. Therefore, a better understanding of the experience of menopause, and the decades that follow, is key to optimizing a large portion of the female lifespan, and healthy aging overall.

Menopause is often a dynamic and transitional event, but it can occur as a more abrupt event as well, secondary to surgery removing some or all of the reproductive tract. Scientific discovery and clinical insights thus far have supported the hypothesis that just as the trajectory of menopause is not uniform across aging women, the optimal therapeutic tactic will not be uniform across aging women either. Indeed, researchers participating in the Stages of Reproductive Aging Workshop (STRAW), and its follow-up STRAW+10, characterized menopause stages by predominant symptoms and hormone profiles to aid in quantification of the menopause experience into discrete phases, while acknowledging that the time of each stage, and time across stages, can vary greatly across individuals [6,7].

During each of the stages, the menopause transition can yield an array of symptoms that impact quality of life and increase the risk for life-altering health outcomes. The Study of Women’s Health Across the Nation (SWAN), which was initiated in 1996 in a racially diverse sample of women, has provided researchers with a better understanding of permanent and transient changes that women undergo throughout menopause and across aging [8]. Menopause can impact genitourinary, cardiovascular, metabolic, and sexual health, resulting in night sweats/hot flashes, sleep disturbances, and central and visceral adiposity as well as altered risk factors for osteoporosis, diabetes, and cardiovascular disease [8,9]. Midlife is a particular life phase in women where chronic pain begins to increase [10], including escalations in rheumatoid arthritis and migraines. There are also negative impacts on mood, anxiety, and cognition with menopause, although there is evidence that these effects are most prominent during the transition to menopause and in early postmenopause, and that they can attenuate later in postmenopause [11–16]. There are likely individual differences in vulnerability and resilience to menopause-related outcomes on psychosocial variables. For example, individual fluctuations of estrogens early in perimenopause have been related to psychosocial menopausal symptoms, as demonstrated in a study assaying a urinary metabolite of estradiol showing that its fluctuations were associated with depressive symptoms, negative affect, and elevated levels of sadness, rejection, and anger [17]. Another example is demonstrated by a recent longitudinal study evaluating a sample of low-income women with and without HIV, which included a large proportion of women belonging to racial groups traditionally under-represented in health research [18]. This study reported clinically significant cognitive impairments in learning and memory from premenopause (menses in the past 3 months with no regularity changes) to early perimenopause (menses in the past 3 months with a change in regularity), with impairments persisting into postmenopause (no menses within the past 12 months) [18], supporting the tenet that certain populations of women may have an increased vulnerability to long-lasting menopause-induced cognitive decline. Additional work has demonstrated that psychosocial and health-associated risk factors in women have been related to the trajectory of depressive symptoms in midlife when followed over 15 years [19]. Research similar to that described here, identifying complex health-related factors that impact long-term health outcomes, is critical to ultimately yield discoveries that translate to the clinic thereby initiating a real impact on the quality of life, health outcomes, and longevity for women during aging. The importance of research targeting the menopause transition and beyond, especially those utilizing longitudinal approaches evaluating outcomes into postmenopausal timeframes, is especially poignant when considering the fact that one-third of the female lifespan occurs, on average, after the final menstrual period.

While there have been great strides in research discovery to understand women’s health, especially in the last two decades, we do not yet have a concise understanding of the biological drivers of variations in menopause, the effects of hormone therapies, and the distinctions impacting outcomes from each of these. We have come a long way in understanding the science of menopause, but more work needs to be done. Animal models of menopause can aid in this pursuit. As risk factors get identified, the array of currently validated animal models of menopause is poised to increasingly contribute impactful scientific knowledge. Systematic evaluation in these models could pinpoint, for example, risk factors that are modifiable or dependent on age, menopause history, or depression, anxiety and cognitive profiles; in fact, it may be that the sensitive window of modifiable risk factors varies depending on which profile is pinpointed and manipulated [20,21]. Furthermore, while it is true that the menopause transition is likely a time of vulnerability for the development and emergence of symptoms and disease in women, it is also likely that during the menopause transition there is an opportunity to espouse health-promoting behaviors and disease prevention strategies [22]. Such a supposition has been supported by work with rodents [23,24] and humans [8,9,16].

In order for preclinical animal models of menopause to usefully inform clinical outcomes and care, one must recognize the importance of interdisciplinary collaboration and reciprocal insights between basic and clinical scientists. For a truly translational approach to the study of menopause, it is important to establish preclinical models for the most common menopause etiologies. To institute a sound menopause model, the basic scientist must understand the clinical presentation as well as its physiological and anatomical profiles, if known. This involves learning from, consulting, and collaborating with scientists with clinical expertise, as well as the clinicians who treat this population. Along these lines, another necessity is becoming proficient in knowledge set forth by organizations focusing on the promotion of women’s health during aging, such as the North American Menopause Society (NAMS), and groups of researchers performing longitudinal research in women, including but not limited to the SWAN, Women’s Health Initiative (WHI), Seattle Midlife Women’s Health Study, Biodemographic Models of Reproductive Aging (BIMORA) Project, and Kronos Early Estrogen Prevention Study (KEEPS) [8,25–28]. There are also recent longitudinal studies, such as the Swiss Perimenopause Study, which assesses perimenopausal women across 13 months, evaluating a considerable range of health-related factors including genetic, epigenetic, endocrinological, physiological, and psychosocial foci [29].

Clinical insights will be enriched through continued research in menopausal women, and preclinical models can be developed and improved across time in response to these increased data about the clinical experience. It is optimal that this feed of information is reciprocal. That is, when we obtain greater detail about the clinical experience, we can use this greater detail to enrich preclinical models so that systematic experimental manipulations and subsequent fine-tuned data can be obtained. This will, in turn, achieve clearer scientific mechanisms and explanations of outcomes, and can be translated to the clinic so research in women can be done in a more targeted manner based on the preclinical outcomes. The considerable benefits of reciprocal, iterative communication across clinical and preclinical research domains is especially poignant as we recognize that systematic and direct manipulation of many variables that are invasive, difficult to frequently test, or hard to control in humans can be accomplished in the animal model work.

In women, natural menopause is due to a transitional follicular depletion in the ovaries, which initially leads to erratic estrogen levels from irregular follicular growth and development, and a lack of sufficient opposing progesterone levels [30,31]. It is notable that during the menopause transition, the fluctuation in estrogen and progesterone levels is highly variable across women [32]. After this transition, when ovulation and cycling cease due to lack of follicular development, circulating levels of estrogen and progesterone drop to very low, often undetectable, levels; circulating (albeit comparatively low) levels of estrogens that are present during postmenopause in women are thought to be primarily from adipose tissue [33]. In response to these low gonadal hormone levels, the pituitary gonadotropins luteinizing hormone and follicle stimulating hormone increase [5,30,34]. In women, the typical process of transitional menopause via follicular depletion is comorbid with age, making it exceptionally difficult to tease apart whether symptomology and risk factors are due to aging or menopause, or potential interactions thereof. While menopause, aging, and the relationships amongst them are not uniformly experienced across all women, certainly highlighting the enthusiasm as of late for personalized medicine, there are commonalities in symptoms and risk factors that can, and should be, studied.

There are many commonalities and differences in the woman and female rat reproductive cycles across their lifespans. While the young adult women’s typical menstrual cycle is 28 days, the young adult female rat’s typical estrous cycle is 4–5 days. In both women and rats, the path to reproductive senescence occurs across time unless a situation such as surgical intervention or pathology ensues. Depending on the specific rat strain, typically between 10 and 14 months of age, estrous cyclicity becomes irregular. The hypothalamus and pituitary are the nexus of reproductive-related changes in the rat, and the ovary responds to these changes with stable or elevated circulating estrogen levels and sometimes elevated progesterone levels, depending on estropause status [31,35]. Unlike in the woman, the female rat does not undergo extensive follicular depletion in its lifetime. Because the rat does not undergo marked follicular and hormone depletion with aging, with the appropriate methodology we can experimentally induce marked follicular depletion when we so choose. In fact, experimenters can leverage the timing of this experimental manipulation to occur in young adulthood, middle age, or old age. This allows preclinical researchers to systematically dissociate the effects of aging from the effects of follicular depletion or other ovarian status factors in question.

In the domain of menopause modeling, ovariectomy (Ovx), or surgical removal of the rodent ovaries, has been traditionally used to analyze effects pertaining to ovarian hormone loss. The Ovx model is also important in providing a clean hormonal slate for further assessment of various hormone or drug treatments that may interact with circulating ovarian hormones. While Ovx effects on various cognitive and physiological systems have been well-characterized in the preclinical literature [36,37], providing insight into one form of surgical menopause, other surgical menopause etiologies are less characterized. This includes hysterectomy, or surgical removal of the uterus, which is the second most common gynecological surgical procedure, and has a lifetime risk of 25% in the US [38,39]. There was a period of time wherein the non-pregnant uterus was thought to play minimal to no role in broader cognitive and physiological functioning; one excerpt provides an example of the staple and traditional thinking in the field: “… without the appropriate hormonal stimulation the uterus is a quiescent and, to a large degree, useless organ.” [40].

In some human menopause studies, the effects of hysterectomy have been obscured by grouping together women with intact ovaries, regardless of uterine status. However, increasing clinical evidence is yielding new considerations for the long-term effects of hysterectomy. Hysterectomy alone has been found to increase early-onset dementia risk by 38% [41], and detrimental effects were increased with a younger age at surgery. Like that of clinical research, a novel preclinical model of hysterectomy has been found to induce cognitive deficits [42]. Notably, it is debated whether ovarian failure accelerates following hysterectomy; this could be critical to the cognitive outcomes observed in the decades following surgery. There is evidence that women who undergo hysterectomy with ovarian conservation during reproductive years transition to menopause earlier than women with an intact uterus and ovaries [43–45]. With animal models, we can systematically address these targeted questions by evaluating brain, behavioral, hormonal, and ovarian outcomes simultaneously whilst independently modulating variables of interest, such as age and ovarian status. Indeed, with the development of the 4-vinylcyclohexene diepoxide (VCD) model of menopause, we are able to observe changes in these critical variables across a period of ovarian follicular depletion as a model of the menopause transition [46–48]. In order to be translationally relevant to a broad number of menopause experiences in women, we must study the vast reality of surgical practices performed in women, including how they might interact with aging. This is an important component to deciphering the effects of variations in menopause on any physiological system of interest, and additional questions such as effects of hormone therapy and pharmaceutical applications can be built on to this foundational knowledge.

Bearing in mind the various animal models of menopause (Figure 1), we now have fortuitous opportunities abound to experimentally induce follicular depletion, or surgically manipulate parts of the reproductive tract, at any age. In addition, we can systematically combine experimental methodologies in animal models. This permits us to ask how particular reproductive structures interact. For example, we can determine the impact of simultaneous surgical removal of structures in combination, as compared to no surgical structure removal, or removal of only one of the structures alone. We also have the opportunity to ask questions systematically combining transitional and surgical menopause, and can furthermore methodically alter which of these menopause parameters comes first; for instance, it can be determined whether the outcome of surgical ovarian removal changes depending on whether transitional menopause occurred before surgery. This could be assessed by experimentally inducing follicular depletion or not in the rats, and then subsequently performing Ovx on all subjects. If Ovx has a different outcome depending on whether follicular depletion was induced before surgery, we could interpret the findings in a translational context, and suggest that for studies in women consideration of transitional menopause status before surgical removal of the ovaries could yield notably different impacts. In animal models, methodical, incremental, and stepwise assessments of putative influences can be implemented in individual or combined experimental protocols of menopause induction. A feasible goal of preclinical work is elucidating precise determinants that impact symptoms and risk for menopause-induced health outcomes.

Figure 1.

A depiction and comparison of various rodent menopause models with representative reproductive tract structure alterations.

Any animal model has its strengths and weaknesses. It is accepted that no animal model can exactly recapitulate human health conditions; this is true for preclinical menopause models as well. These confines include, but are not limited to, the following truths of rodent models: 1) They have a short lifespan, 2) Things are simplified in many ways, including diet, light/dark cycle, and pharmaceutical exposures, 3) The etiology of menopause does not parallel women. Notably, however, these limitations of the model also bear footing to its strengths. Specifically: 1) They have a short lifespan, which means we can study across the entire trajectory of aging, birth to old age, in about 2 years; 2) Things are simplified in many ways, including the diet, the light/dark cycle, and pharmaceutical exposures, which means we have abundant experimental control over environmental and extrinsic variables and exposures that we cannot realistically or ethically control in humans, and; 3) The etiology of menopause does not parallel women, which means that we have the opportunity to control when it occurs thereby allowing us to dissociate menopause from aging if we so choose. These examples highlight the tenet that our depth of knowledge about a model is what yields its ‘strengths’ and ‘weaknesses,’ and that this depth of knowledge overall is in actuality an asset. Indeed, we can capitalize on what we know about the detailed characteristics of the model to optimize our approach to asking a scientific question; in fact, the weaknesses of a particular model could be leveraged to also be its strengths. Moreover, interpreting the data from an animal study in the context of such strengths and weaknesses is critical to translating its results to the clinic. The extensive control over, and the ability to systematically manipulate, variables in an animal model yields preclinical scientists uniquely poised to address many complex, interactive questions in a straightforward and simplified manner.

The progression of comparative medicine has advanced translational science, and this effect is far reaching for many fields. Herein, we focus on women’s health during midlife and menopause, providing select illustrations of how translation can happen, progress that has been made, and progress that could be made in the future as we train the next generation of scientists who will be the cartographers of the future menopause maps. Leveraging the advancement of knowledge that has been gained about the midlife and menopause in women from clinical scientists, preclinical scientists have been able to improve the utility of related animal models. This has occurred not only in the realm of experimentation techniques, but also via expansion of improved accuracy of modeling, interpretation, and in turn, translational applicability. In fact, parallel findings have been discovered across species in this light, thereby permitting further precision of manipulation and interpretation in the next studies. It could also be stated that providing concrete exemplars illustrating such relationships is essential to convincingly show that animal model results are clinically meaningful. Thus, the following is an example of how animal work is consistent with clinical outcomes and has been used to systematically tease apart such driving factors: As aging ensues, memory often becomes impaired; in women, is this memory change due to age, or menopause, or both? Notably, causality is especially difficult to discern in women since transitional menopause is often concomitant with age. Leveraging the induced follicular depletion (VCD) ovary-intact rat model, we dissociated the effects of aging and follicular depletion, and systematically tested the hypothesis that cognitive changes were prominent early in the menopause transition, even if onset occurred in young adulthood. Specifically, this was done via a longitudinal experimental design addressing how age and ovarian status impact spatial memory across the transition to menopause [49]. The addition of a vehicle-treated, regularly aging cohort allowed further interpretation with an eye toward translation. This group permitted a direct comparison of a “gold standard” normally aging rat group to the experimentally-induced accelerated follicular depletion group across the many cognitive maze testing times as the experiment (and follicular depletion) progressed. Findings from this preclinical study were manifold, including results revealing that age at the onset of transitional menopause impacts cognition; specifically, follicular depletion at a younger versus older age had a stronger detrimental impact on spatial memory. These animal model data demonstrating that earlier follicular depletion may be especially negative for memory correspond to findings showing that: 1) younger women who undergo ovarian hormone loss via oophorectomy showed poorer memory compared to women who received oopherectomy later in life [50], 2) loss of ovarian hormones earlier in life in women, including with premature ovarian failure or bilateral oopherectomy, negatively impacts cognition later in life [51], and 3) premenopausal bilateral oopherectomy in women is associated with a higher risk of dementia [41].

In the context of the impact of animal models on reproductive system discoveries leading to real world implications on women’s health, we can peer back to some of the earliest classic literature. Much of the original sex behavior studies were focused upon males, and at that time it was noted that a major ‘complicating factor’ in deciphering mechanisms of female sex behavior was the cyclic nature of the behavior itself. Fortunately, one of the exceptions was published in 1917; Stockard and Papanicolaou [52] used guinea pigs to show that vaginal cytology changes could be closely correlated with alterations in ovarian function, and it soon became understood that cyclic sex behavior alterations could be predicted by characterizing these vaginal cell types. Papanicolaou began studying a modified staining version of vaginal smears, and soon noted cancer cells, publishing a paper, “Cancer of the uterus: The vaginal smear in its diagnosis,” and a related book with Herbert Traut. The team described that the simple assessment of vaginal cytology under the microscope was influenced by the reproductive cycle, and that cytology from the vagina was reliably impacted by hormones, and could detect malignancy. The research of Papanicolaou, started in rodents, led to revolutionary clinical procedures for early detection of cervical cancer; the Pap smear. This reciprocal nature of women’s health research, from preclinical insights to the development of clinical guidelines, continues over 100 years later.

In summary, midlife in the female often includes broad, sweeping changes in steroid and feedback hormones; the resulting impact is extensive, likely yielding a physiological period of reorganization. As such, this window of time could be a period of vulnerability, but it also could be a window of opportunity for therapeutic intervention. There are unique benefits to using rodents to ask and answer incisive questions in this domain. In fact, questions about the impact of front-line, clinically-relevant life history factors speaking to menopause etiology, including presence of the uterus and/or ovaries, and extent of ovarian follicular depletion, can be directly and systematically addressed via experimental manipulations in rodents. Data from basic science are incrementally building on the clinical literature, and the evidence underscores the idea that to understand the many dynamic changes with, and due to, menopause, iterative and reciprocal communication between clinical and preclinical domains of science is key. Identification of modifiable risk factors and their parameters for turning points of positive versus negative impacts could be experimentally targeted to the midlife timeframe, as compared to younger and older timeframes, to directly address whether midlife is a period of particular sensitivity for particular factors of interest. Preclinical research evaluating specific menopause profiles when induced at varied ages, including when distinct hormone therapies are mapped onto these different menopause profiles, will yield valuable data regarding modifiable factors leading to the beneficial and detrimental trajectories of aging through the midlife and beyond.

Data availability

No data were used for the research described in the article.