Female aging: when translational models don’t translate

Abstract

For many pathologies associated with aging, female patients present with higher morbidity and more frequent adverse events from treatments compared to male patients. While preclinical models are the foundation of our mechanistic understanding of age-related diseases, the most common models fail to recapitulate archetypical female aging trajectories. For example, while over 70% of the top age-related diseases are influenced by the systemic effects of reproductive senescence, we found that preclinical studies including a menopausal phenotype make up < 1% of published aging biology research. The long-term impacts of pregnancy, birthing, and breastfeeding are also typically omitted from preclinical work. In this perspective, we summarize limitations in the most commonly used aging models, and we provide recommendations for better incorporating menopause, pregnancy, and other considerations of sex in vivo and in vitro. Lastly, we outline action items for aging biology researchers, journals, funding agencies, and animal providers to address this gap.

Motivation

In her 1949 novel The Second Sex, French existentialist philosopher Simone de Beauvoir contemplated the origins of perceived female inferiority, outlined inequities that exist between men and women, and offered a profound argument for the dismantlement of this partisanship.¹ Over 70 years later, inequities related to sex and gender persist on multiple scales and across diverse domains. Here, we consider the basic biology of aging. Specifically, we comment on limitations in the most commonly utilized preclinical models in aging biology research and the resulting obstacles encountered in our mechanistic investigation of female aging.

Our understanding of the basic biology of aging has flourished in recent decades, owing in large part to the use of model organisms including Drosophila melanogaster (fruit flies), Caenorhabditis elegans (roundworms), and Murinae (rodents).² Such model systems have allowed for stratification of the diverse impacts of aging on organismal health into a smaller set of underlying qualities (i.e., the Hallmarks of Aging).³ Information gained from preclinical studies have demonstrated that the effects of time’s arrow can be manipulated by interventions such as caloric restriction,⁴ exposure to youthful circulatory factors,⁵ and removal of senescent cells.⁶ The animal models employed have provided and continue to provide unique opportunities to monitor, measure, and manipulate aging phenotypes over a condensed lifespan, offering novel insights into human physiology and pathophysiology.

A central tenet of animal models is that the system used shares essential physiological characteristics with humans. Unfortunately, this is all-too-often not the case when it comes to female aging. For example, female patients present with osteoporosis four times more frequently than male patients, resulting in a higher incidence of hip fractures.⁷ Yet, aging female rodents do not naturally present with a decline in bone mass.⁸ Female individuals also present with sarcopenia earlier in their lifespan than male indvidualss,⁹ though sex differences in rodents are minimal.¹⁰ The incidence of Alzheimer’s disease (AD) and non-AD dementia is higher in female individuals when compared to age-matched male humans.^11,12 Yet, recapitulating this difference has proved challenging in animal models, and several studies have demonstrated that female rodents better retain memory-related functions over time in comparison to age-matched male rodents.^13,14 Female individuals are twice as likely to present with knee and hand osteoarthritis in the clinic,¹⁵ but in our recent study, we found that male mice presented with more severe cartilage degeneration than female mice.^15,16 Although these are just some reported examples of the discrepancies between clinical and preclinical observations, many more disconnects likely exist but have not yet been identified due to the predominant use of male models in aging biology studies.¹⁷

This gap in our understanding of how aging affects the onset and progression of diseases in female patients has likely contributed to worsened health outcomes. Meta-analyses on disability status over the last 20 years from France, Spain, and the US have consistently demonstrated that people who are female live with higher morbidity than age-matched male counterparts.^18-20 Female patients are 50% more likely to have a heart attack misdiagnosis²¹ and 33% more likely to have a stroke misdiagnosed than male patients.²² Between 1997 and 2000, 80% of drugs removed from the market were due to adverse events caused in female consumers,²³ with most of these drugs intended to treat age-related diseases. Even as of 2020, female patients still reported more side effects due to prescription drugs for age-related diseases when compared to male patients.^24,25 Although psychological, social, and economic factors clearly contribute to these disparities, our laggard understanding of aging female physiology represents a major barrier in our ability to prevent, diagnose, and treat diseases in older female individuals.

In an attempt to better identify, understand, and ultimately resolve these preclinical and clinical discrepancies, the purpose of this perspective is to (1) outline current issues confounding the study of sex as a biological variable within aging biology research, (2) identify advantages and disadvantages in current models used to study female aging, and (3) define action items to increase the translatability of preclinical aging studies for older female people.

At the outset, it is important to note the distinction between sex and gender and to define a priori the terms used throughout this perspective. Sex is “a multi-dimensional biological construct based on anatomy, physiology, genetics, and hormones,” while gender is “a multi-dimensional construct that encompasses gender identity and expression, as well as social and cultural expectations about status, characteristics, and behavior as they are associated with certain sex traits.”²⁶ Although gender is clearly an important contributor to the disparities observed within our population, in this perspective we focus on differences as they pertain to sex. For simplicity and consistent language, throughout this perspective, we use the term “female” to refer to people or animals sexed as female at birth, as typically defined according to appearance of the genitalia.

Shortcomings in the study of female aging at the bench

Female aging in humans is inextricably intertwined with reproductive senescence and menopause, leading to systemic endocrine alterations that affect tissues and organs throughout the body.²⁷ Other features commonly associated with female aging include the impacts of pregnancy and breastfeeding on health later in life.^28,29 The ubiquity of these features highlights their critical importance in understanding female aging trajectories.

Female aging post-menopause:

With current lifespan estimates, female individuals spend on average over one third of their life post-menopause, with an increased disease incidence corresponding with this phase.³⁰ Clinically, menopause onset is defined as the time when menstrual cycles have ceased for at least 12 months with a corresponding loss of ovarian follicular function (i.e., ovarian senescence).³¹

While the field of reproductive aging has long recognized the impacts of menopause-related disruption to the hypothalamic-pituitary-gonad axis,³⁰ the role of reproductive senescence in the etiology of many other age-related diseases has been largely overlooked in preclinical studies. Likely representing a major contributor to this gap, the most used models to study aging do not exhibit menopause. Neither fruit flies nor roundworms demonstrate menstrual (shedding of the uterine lining resulting from changes in sex hormone circulation) or estrous cycling (cyclical changes in sex hormones that do not result in uterine lining shedding), and thus, do not undergo menopause. The most commonly used rodent models have an estrous, but not a menstrual, cycle, and the estrous cycle shows sex hormone fluctuations similar to those observed in female humans.^32,33 At 9-12 months of age (the equivalent of ~30-38 years in humans³⁴), the estrous cycle of rodents becomes irregular and the estrus period prolonged, a phase referred to as ‘estropause’ or, in mice, ‘mouseopause’.³⁵ However, approximately three-quarters of aged rodents spontaneously rejuvenate their ovarian follicles in middle-age and re-establish the circulation of sex hormones comparable to that of a pre-menopause state, despite being unable to reproduce.^27,35,36 The small percentage of rodents that fail to rejuvenate their ovarian follicles transition to an anestrous state of low ovarian sex hormone levels, akin to a perimenopause state in humans (i.e., the transition from regular menstrual cyclicity to irregularity that eventually leads to cessation of menstrual cycles).^27,35,36 Therefore, approximately three-quarters of the most widely used aging female rodents reflect a pre-menopause state and one quarter reflect a perimenopause state. Most importantly, none of the readily available rodent models mimic a menopause phenotype.^27,36 When considering other animal models commonly used in aging research, some studies have suggested that nonhuman primates display a menopause phenotype in the last 1-2 months of their lifespan.^37,38 In fact, it has been reported that the only known organisms that live a substantial portion of their lifespan with altered levels of circulating sex hormones following reproductive senescence are killer whales, pilot whales, and humans.³⁹

To better understand the potential impact of the gap in menopausal phenotypes amongst commonly used model organisms, we performed a literature search to determine the number of clinical trials, mammalian studies, and mammalian studies that considered menopause in aging research (see Supplementary Note for detailed search information). We focused on the top 22 most prevalent diseases in older adults (aged 50 years or older), as outlined in a recent systematic analysis of the Global Burden of Disease Study.⁴⁰ We then classified these top diseases into 5 categories: cardiovascular, metabolism, orthopedic, cancer, and cognitive/neurological (Figure 1A, Table S1). Amongst adults ages 50-74 years old, male individuals displayed a higher incidence of cardiovascular diseases, such as stroke and heart disease, while female individuals displayed a higher incidence of musculoskeletal disorders, such as low back pain and osteoarthritis. Sensory deficits such as loss of hearing and vision were also more prevalent in female individuals at this age range. In adults 75 years and older, female individuals continued to display an increased incidence of musculoskeletal disorders and sensory deficits, in addition to an increased incidence of stroke, heart disease, and diabetes when compared to male counterparts (Table S2).

Figure 1: Menopause affects the majority of age-related diseases, but few preclinical studies incorporated a menopausal phenotype.

A) The top 22 age-related diseases⁴⁰ were grouped into 5 disease categories. On March 4, 2023, we performed a PubMed literature search for known menopause-related effects in humans when considering these top 22 age-related diseases (Figure 2B, Table S1). On July 4, 2023, another PubMed literature search was performed to determine the number of (a) clinical trials, (b) mammalian animal studies, and (c) menopause-inclusive mammalian, animal studies on each disease category (Figure 2C). Summary of disease categories and specific search terms are listed in Table S1. B) The percentage of the top age-related diseases that are known to be associated with onset of menopause. References used to determine the association between each disease and menopause are listed in Table S1. C) The number of articles on PubMed for clinical trials, mammalian studies, and menopause-inclusive mammalian studies for each disease category. Created in BioRender.

When considering articles that supported or refuted an association between menopause and age-related pathologies, over 70% of the most prevalent age-related diseases are associated with and/or potentially impacted by menopause (Figure 1B and Table S1).⁴¹ Despite this, few studies have factored menopause into basic biology of aging studies. As of July 4, 2023, we identified 595,251 published clinical trials and 2,771,411 published preclinical mammalian studies relating to these five disease categories (Figure 1C). Less than 1% (17,678) of the preclinical studies relating to these diseases considered menopause (Figure 1C).

Pregnancy, birthing, and breastfeeding:

Other critical features typically missing from the most utilized mammalian models of female aging relate to pregnancy, birthing, and breastfeeding. In the US, 86% of female individuals give birth at some point in their lifetime,⁴² and epidemiological studies have revealed that childbirth can affect both immediate health as well as long-term health of the childbearer.²⁸ People with a history of complications during pregnancy have an increased risk of cardiovascular and metabolic diseases later in life.²⁸ Breast cancer risk is lower in people who were younger at the time of their first full-term pregnancy,⁴³ had a higher number of childbirths,²⁹ had a history of preeclampsia,⁴⁴ or who breast fed for longer after giving birth.⁴⁵ Breastfeeding is also associated with decreased risk of ovarian cancer,⁴⁶ type II diabetes,⁴⁷ and high blood pressure later in life.⁴⁸ Not considering these prevalent biological factors in aging biology research can impede the ability of preclinical studies to most effectively elucidate phenotypes and outcomes in older female individuals.

Opportunities for more representative models of female aging

In 2013, a now widely referenced work reported that mice fail to mimic the same genomic responses of inflammatory diseases seen in humans, thereby highlighting the translational inadequacy of the models most commonly used.⁴⁹ This work led to increased adoption of ‘dirty’ (i.e., pet store) or ‘humanized’ (i.e., engineered with human genes or engrafted with human organs) mice, both of which better recapitulate the intricacies of immunity in humans.⁵⁰ The aging biology community can similarly benefit from re-evaluation of the animal models used to study female aging with the long term goal of making preclinical studies more translatable.

Ovariectomy model of menopause:

Of available menopause models, the simplest is arguably ovariectomy (OVX) (Table 1). OVX surgically excises ovaries from the animal, representing a quick and reproducible procedure that recapitulates the loss of sex hormones that is typical of menopause. Given that OVX eliminates all ovarian sex hormones, it also provides a unique opportunity to study signaling cascades and phenotypes resulting from individual sex hormones. For example, OVX followed by estradiol versus progesterone treatment can help disentangle the distinct role of each on downstream phenotypes.²⁷ However, OVX as a model of menopause has several limitations. First and most obviously, natural menopause retains ovaries intact while OVX does not. As a result, OVX creates an abrupt cessation in ovarian function, which contrasts the progressive loss of ovarian function observed with natural menopause over time.²⁷ This gradual decline in function is of interest given that many menopause-related symptoms first present during this transition period (i.e., perimenopause).⁵¹ To address this limitation, some researchers have excised the ovaries at the conclusion of estropause, thereby allowing for a pseudo-perimenopause phase followed by a post-menopausal phase.²⁷ Another limitation is that OVX also results in depletion of all ovarian-produced sex hormones, including those that do not change with menopause, such as testosterone.²⁷

Table 1. Strengths and limitations of current rodent models of menopause for aging biology research.

Created in BioRender.

Menopause Models	Strengths	Limitations
Ovariectomy (OVX)	Short time to onset of menopause Cost-effective and highly reproducible Can be used to systematically determine the influence of single sex hormone replacement following OVX	Does not retain ovaries intact as is the case with natural menopause Abrupt menopause transition with no perimenopause phase Depletes all sex hormones, including those unaffected by natural menopause
4-vinylcyclohexene diepoxide (VCD)	Slow and progressive menopause onset Retains ovaries intact Recapitulates perimenopause and menopause stages	Time- and labor-intensive Potential for investigator toxicity, therefore requiring additional safety precautions
Spiny Mouse (Acomys Cahirinus)	Naturally occurring menopause Non-invasive and least harmful to the animal	Absence of age- and sex-matched non-menopausal mice Display a unique regenerative capacity and how this feature affects the presentation of menopause-related phenotypes is unclear Least characterized of murine menopause models Long lifespan relative to other commonly used rodents and a long time to menopause onset Not commercially available

From an implementation perspective, most studies incorporating OVX models to date have used young animals, despite the fact that menopause in humans manifests in the setting of aging cells and tissues.²⁷ Age is likely an important consideration, as the systemic effects of menopause in a young organism can vary greatly when compared to the effects in aged counterparts.⁵² For example, young female patients who undergo oophorectomy have clinically distinct presentations compared to older female patients undergoing natural menopause.^53,54 While the reasons for this difference are unclear, the manifestation of age-related disease following natural menopause presumably occurs through a combination of cell-intrinsic (e.g., genetic vulnerability of aged cells) and extrinsic (e.g., a change in the circulating sex hormone profile) factors, rather than a single cause alone. As such, OVX performed in young animals likely misses key mechanistic insights into diseases that manifest in older female humans.

4-vinylcyclohexene diepoxide (VCD) model of menopause:

In contrast to surgical ablation, the VCD model of menopause allows for a slow and progressive loss of ovarian-based sex hormones while preserving the ovaries intact (Table 1). VCD is an ovarian-specific toxin that causes primordial and primary ovarian follicle apoptosis.^36,55 This chemically-induced model of menopause requires daily intraperitoneal injections of VCD in an oil-based vehicle to deplete follicles, thereby allowing for evaluation of both perimenopause and menopause phases.³⁶

While the VCD model may better recapitulate the menopause transition and is less invasive than OVX, it is more time-, labor-, and cost-intensive. When considering possible toxic effects to animals receiving VCD, no secondary toxicity to organs beyond the ovaries in young⁵⁶ or middle-aged female mice⁵⁷ have been reported. In rats, while the model was minimally toxic in prepubescent animals, 100% of sexually mature rats suffered from peritonitis following VCD injections.⁵⁸ To our knowledge, the VCD model has not yet been used or validated in middle-aged or aged rats. One alternative to mitigate the presentation of peritonitis is oral administration of VCD and triptolide (an herb from a woody vine natively found in China), which induces infertility in rodents.⁵⁹ While we have not found evidence of this dual treatment approach as a model of menopause, it has the potential to attenuate some of the toxic side effects of intraperitoneal VCD administration. Another limitation of this model is that, like OVX, VCD has typically been performed in young animals, thereby potentially masking systemic cell and tissue responses that are age-dependent.

Despite limitations, the potential impact of the OVX and VCD approach for understanding menopause mechanisms is considerable. To illustrate this point, we revisit here two of the aforementioned examples of the disconnect between preclinical and clinical findings, osteoarthritis and osteoporosis. As noted above, the incidence of knee osteoarthritis is higher in age-matched female patients than in male patients over the age of 50,¹⁵ with female patients presenting with more severe cartilage degeneration at the time of joint replacement than male patients.⁶⁰ Yet, male mice have more severe age-related cartilage degeneration than non-menopausal female mice.¹⁶ Menopause induction by both OVX and VCD more closely recapitulates the clinical presentation, as evidenced by an accelerated and more severe cartilage degeneration compared to non-menopause mice.^57,61 Likewise, female patients demonstrate a precipitous drop in bone mass,⁷ a phenotype that is not recapitulated by mice undergoing natural aging.⁸ In contrast, the onset of menopause in mice following OVX or injection with VCD results in significant loss of bone mass.⁵⁶ These examples illustrate how use of a menopause model can increase the translational relevance of preclinical findings.

Spiny mice as a naturally-occurring menopause model:

Acomys cahirinus, the African spiny mouse, was recently reported to be the first murine species known to both menstruate and undergo a slow and gradual menopausal transition similar to humans.^62,63 Specifically, spiny mice demonstrate a menopausal phenotype at 36 months, as evidenced by a progressive decline in primordial ovarian follicles. Ovarian cyclicity became more irregular between years one and three (i.e., akin to a perimenopause state), dropping precipitously in the fourth and final year of life (i.e., menopause onset). Importantly, the authors observed a significant drop in estradiol between years 1 and 2 with no changes in circulating testosterone, similar to natural menopause seen in humans. Although initial research interest in spiny mice focused on elucidating mechanisms underlying their remarkable regenerative capacity,⁶⁴ this model could be another promising research tool for incorporating menopausal phenotypes into aging biology studies. Future studies are needed to both validate this model and evaluate whether the menopause phenotype in these mice tracks with the menopause-induced systemic changes that are observed in humans. (Table 1).

Determination of menopause status in aging rodents:

It is challenging to know which rodents exhibit a pre-menopause versus perimenopause state due to the relatively low quantity of collectable serum from rodents, the relatively low levels of circulating hormones, and the insensitivity of commercially available assays. Mass spectrometry (either gas or liquid chromatography) is the gold standard for quantifying sex hormones in humans, and recent studies have mapped the sex hormone profiles seen in mice and rats across the estrous cycle using this approach.⁶⁵ However, mass spectrometry is expensive and requires a relatively large sample (~500 μl of serum per sex hormone). With the goal of increasing the translation and feasibility of using natural aging models, the development of more reproducible, reliable, and cost-efficient assays that quantify female sex hormones in rodents are much needed. Until then, as a proxy for circulating sex hormone levels, vaginal lavage and cytology can be used to track the estrous cycle and estropause status.^35,66 Post-mortem, histological evaluation of ovarian follicle numbers can also be used to determine estropause status.⁶⁷ Uterine weights also correlate with estrogen levels, offering a practical alternative for determining the hormone status of naturally aging and/or menopause models.⁶⁸

Pregnancy, birthing, and breastfeeding:

Despite numerous clinical studies showing that pregnancy, childbirth, and breastfeeding impact longevity and healthspan,^28,43 these variables remain largely absent in preclinical work that does not focus on reproductive aging. Female breeders (i.e., rodents used to generate rodent colonies) are typically excluded from research studies. As such, unless specifically requested, female rodents ordered through animal vendors are nulliparous (i.e., have never produced a litter) and have never breastfed. The use of breeders offers an opportunity for researchers to develop mechanistic understandings of how physiological processes related to birthing may influence age-related diseases. Indeed, recent preclinical studies have identified pregnancy-related changes in immunity, microbiota composition, neural signaling, and muscle regeneration.^69-72Another study demonstrated that breastfeeding alters the circadian rhythm within maternal mice.⁷³ These changes may translate to significant alterations in aging trajectories and age-related diseases, though further studies are needed. Consideration of variables such as number of pregnancies, litter sizes in rodents, complications (e.g., stillbirth), and other related outcomes (e.g., health conditions directly resulting from pregnancy such as preeclampsia) will provide valuable insight into how child-bearing features affect female aging trajectories.

Cell culture considerations:

Cell culture systems have served as an indispensable reductionist approach to evaluate aging mechanisms, but consideration of sex in these systems has generally been limited. The importance of sex as a biological variable (SABV) in vitro and in vivo has been highlighted by the National Institutes of Health (NIH).⁷⁴ Yet, in the four years since implementation of SABV requirements, 50% of studies still did not report the sex of cells used, and of those that did, only 22% used female cells.⁷⁵ Cells retain a memory of their origins; this is true of cells isolated from older animals and humans as well as cells isolated from male and female organisms. ^76,77 Therefore, reporting of cell source represents an important study design consideration.

In addition, cell culture conditions have the potential to contribute confounding effects and may preclude identification of sex-dependent changes associated with aging. Cell culture studies typically use fetal bovine serum (FBS) or serum from other animals, which may mask the effects of endogenous hormone alterations in the context of aging.⁷⁸ Phenol red, commonly used in media as a pH indicator, has estrogenic activity,⁷⁹ thereby similarly contributing potential confounding effects, especially when attempting to isolate downstream cellular responses to a menopause-induced loss of sex hormones.

Several steps can be taken to overcome these limitations in standard cell culture conditions. Cells can be cultured in phenol red-free media and in the presence of serum isolated from age-matched male or female organisms to better maintain phenotypic sex differences in vitro. Moreover, for studies aimed at investigating the role of the menstrual/estrous cycle or menopause in age-related phenotypes, media can be designed to mimic the circulatory environment across these life stages. For example, prior to the onset of estropause, female mice have circulating estradiol levels of 2.7 ± 1.0 pg/mL and circulating progesterone levels of 31,323 ± 6,108 pg/mL.⁶⁵ Conversely, with OVX, female mice have undetectable levels of estradiol (< 0.3 pg/mL) and low levels of progesterone (3,940 ± 2,135 pg/mL).⁶⁵ Media composed of charcoal stripped FBS (i.e., FBS with native sex hormones removed^80,81) and supplemented with sex hormones at levels that recapitulate in vivo aging microenvironments can be valuable for better understanding the impact of menopause on female cellular aging. Similar practices have already been utilized in the literature for modeling menstrual cycles,^82,83 illustrating the feasibility of this approach.

Disentangling chromosomal versus hormonal sex-differences:

An interesting and important question in our understanding of the impact of aging on female cellular and tissue declines is whether the observed changes have predominantly genetic or hormonal origins. The “four core genotypes” model includes mice engineered to have XX chromosomes with male or female gonads or mice with XY chromosomes with male or female gonads.⁸⁴ These models offer unique opportunities to disaggregate chromosomal from hormonal sex differences in the development of age-related diseases.⁸⁴ For example, the four core genotype model revealed that the presence of XX chromosomes increases lifespan, independent of gonad.⁸⁵ As noted by the authors, an important limitation is that while this study accounted for chromosomal differences, it did not account for menopause.⁸⁵ It would be interesting to repeat these studies in menopause-induced mice to evaluate whether chromosomal effects on longevity persist. In other applications of the four core genotypes model, the XX genotype was shown to increase resistance to Alzheimer’s related pathology, regardless of the type of gonad.⁸⁶ The primary limitation of this model is that these mice are not commercially available and, thus, need to be bred in-house and aged. Additionally, this model does not consider the impact of other baseline X escapee genes, epigenetic regulation of autosomal genes by sex chromosomes,⁸⁷ reactivation of a silent X chromosome, or secondary phenotypic effects (e.g., the influence of sex chromosome gene expression on the hypothalamic pituitary gonad (HPG) regulation of sex hormones).^88,89 Despite these limitations, the four core genotype model has the potential to provide valuable insights into sex differences associated with age-related pathologies.

Parallel approaches can be used in cell culture experiments to disentangle hormonal versus chromosomal sex differences. For example, XY or XX cells from wild type animals or humans can be cultured in media that contains male or female circulating factors. Using these approaches in vitro allows for detailed interrogation into the molecular mechanisms dictating chromosomal versus hormonal sex differences in a manner easily accessible to researchers.

Closing the gap

Closing the gap in our understanding of sex-based differences in biological aging will require the engagement of stake holders at all levels of aging biology research, including researchers, peer-reviewed journals, funding agencies, and animal providers. In Figure 2, we present a list of recommendations, some of which echo those made by leaders in women’s health.^90-93

Figure 2: Specific and actionable recommendations to researchers, peer-reviewed journals, providers of animals for aging research, and funding agencies to close the gap in our understanding of female aging.

Created in BioRender.

Conclusion

To date, few aging biology studies outside the field of reproductive biology have employed models that recapitulate key features of aging female physiology, such as menopause. The result is a healthcare system that lacks mechanistic data on how to treat age-related diseases in female patients. Herein, we mapped progress and highlighted opportunities for advancing female aging biology research. We outlined the need for preclinical aging studies to incorporate more representative models of female aging, including menopause, pregnancy, and breastfeeding. We also presented considerations for the maintenance of sex-based circulatory factors in vitro as well as investigations that disentangle chromosomal versus hormonal sex differences in aging biology. Finally, we reiterated recommendations made by women’s health leaders and added suggestions for biomedical researchers, journals, animal providers, and funders. Features unique to female aging are relevant to all aging biologists, not just those studying reproductive senescence or specializing in sex hormone signaling. These topics represent valuable opportunities for the field to tackle fundamental aging biology questions to the benefit of women’s health.